Vous êtes sur la page 1sur 228


Philosophy and Medicine


H. Tristram Engelhardt, Jr., Center for Ethics, Medicine, and Public Issues, Baylor
College of Medicine, Houston, Texas and Philosophy Department, Rice
University, Houston, Texas
Stuart F. Spieker, Center for Ethics, Medicine, and Public Issues, Baylor College of
Medicine, Houston, Texas

Associate Editor
Kevin W. Wildes, SJ., Department of Philosophy, Georgetown University,
Washington, D.C.

Editorial Board
George J. Agieh, School of Medicine, Southern Illinois University, Springfield, Illinois
Edmund Erde, University of Medicine and Dentistry of New Jersey, Camden, New
Patricia A. King, J.D., Georgetown University Law Center, Washington, D.C.
E. Haavi Morreim, Department of Human Values and Ethics, College of Medicine,
University of Tennessee, Memphis, Tennessee

The titles published in this series are listed at the end of this volume.
Department ofHealth and Society,
Linköping University,
Linköping, Sweden

An Action-Theoretic Approach

Second Revised and Enlarged Edition


Library of Congress Cataloging-in-Publication Data
Nordenfelt, Lennart, 1945-
On the nature of health an action-theoretic approach / by
Lennart Nordenfelt. -- 2nd ed.
p. cm. -- (Philosephy and medicine ; v. 26)
Includes bibliographical references and index.
ISBN 978-0-7923-3470-5 ISBN 978-94-011-0241-4 (eBook)
DOI 10.1007/978-94-011-0241-4
1. Health--Philosophy. 2. Holistic medicine--Philosophy.
I. Title. II. Series.
RA776.5.N67 1995
613--dc20 95-1945

ISBN 978-0-7923-3470-5

Printed on acid-free paper

All Rights Reserved

© 1987 and 1995 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 1995
Softcover reprint ofthe hardcover 2nd edition 1995
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
To the Memory of My Father





1. Why a Philosophy of Health? 1
2. What Should Be Required of an Acceptable Theory of
Health? 4
3. Preliminaries on Conceptual Analysis 6
4. Two Perspectives on Health and Disease 11


1. The Classical Background 15
2. On the Concept of a Goal 17
3. The Biostatistical Theory: A General Presentation 18
4. The Prima-Facie Merits of the Biostatistical Theory 21
5. A Critical Assessment of the Biostatistical Theory 23


1. An Action-Theoretic Approach 35
2. On the General Concepts of Action and Ability 37
3. On Ability and Disability 46
4. On First- and Second-Order Ability 49
5. On the Concept of Vital Goals 53
6. Health as a Person's Ability to Fulfill His Basic Human
Needs 57
7. Health as a Person's Ability to Attain the Goals Set by
Himself 65
8. Towards a New Philosophy of Health: A Welfare Notion
of Health 76
9. On Some Properties of the Welfare Notion of Health 97


1. Maladies 105
2. Other Factors Compromising Health: Old Age,
Pregnancy, Grief 112


1. On the Relation Between Health and Society 119
2. On Some Consequences of the Welfare Notion of Health
for Medicine and Health Care 129


1. Answers to the Requirements of a Good Theory of
Health 145
2. Glossary 148


1. A Classical Debate - Physiologism Versus Ontologism 151
2. The Problem of Historical Change 154
3. Towards a Reconstruction of MediCal Ontology 157
4. Some Modern Definitions of Diseases 162
5. An Analysis of the Disease Concepts 167
6. Summary 172






This edition of the book contains a Postscript which relates my theory of

health to some contemporary theories of European origin, viz. the ones by
David Seedhouse, William Fulford and Ingmar Porn. In a number of extra
footnotes I have also somewhat updated the text both in relation to later
developments in my own theories of health and quality of life, but also in
relation to recent important publications in the field.

I am grateful to Mr Malcolm Forbes and Professor Stuart Spieker for

perusing the texts.

Linkoping in November 1994

Lennart Nordenfelt


This work is a part of a project entitled "Health, Disease and Cause of

Death" which has been financed by the Swedish Research Council for
the Humanities and the Social Sciences. The members of this project
have been, beside myself, Drs 0ivind Larsen and Erik Falkum, Oslo,
Norway, Dr. Ingemar Lindahl, Stockholm, Sweden, and Miss Eva Nys-
trom, Uppsala, Sweden. I wish to thank my project colleagues for their
continuous and invaluable support.
Special thanks are also due to the following persons: Professor Ingmar
Porn for his many pertinent remarks on basic philosophical questions;
Drs. Filip Cassel, Bertil Martensson, and Bo Petersson for a number of
detailed comments on both stylistic and philosophical issues; Drs. Ants
Anderson and Per Sundstrom for their corrections of some of my mis-
takes in medical matters; Mr. Brian Beattie, Dr. Paul Needham, and Dr.
Craig Dilworth for perusing the text and improving my English style;
Mrs. Maud Wolffelt and Miss Lena Hector for the patience and care with
which they have helped me to prepare this volume for the printer.
I am also grateful to the editors of the series Philosophy and Medicine,
Professors H. Tristram Engelhardt Jr. and Stuart Spicker, for their
generosity in including this book in the series.
I dedicate the book to the memory of my father, Dr. Olof Nordenfelt,
with whom I had numerous fruitful discussions about the nature of
health and illness.

Linkoping, June 1987

Lennart Nordenjelt



This study of the concept of health is an attempt to combine central ideas

in modern philosophy of medicine with certain results from analytical
action theory. What emerges from the study is a concept 'of health based
on an action-theoretic foundation. A person's health is characterized as his
ability to achieve his vital goals.
The general conception is not new. This study has been inspired by a
number of scholars, both ancient and modern. The most important
influences from the latter have been those of Georges Canguilhem, H.
Tristram Engelhardt Jr., Caroline Whitbeck and Ingmar Porn.
The novel aspect of this book consists of elaborations made to the
general conception. First, the action-theoretic platform is analysed in some
detail. The nature of the ability involved, as well as the conditions for
having that ability, are specified. Second, the vital goals of man are given
considerable attention. Some previous attempts to define such vital goals
are analysed and criticized. A new characterization is proposed, in which
the vital goals are conceptually linked to the notion of happiness. A person's
vital goals are such states of affairs as are necessary and together sufficient
for his minimal happiness. Third, a number of consequences of this con-
ception are observed and analysed. One issue which is particularly empha-
sized is that of whether the concept of health is a theoretical or a normative
The aim of this book is to elaborate a cogent and at the same time
reasonably simple theory of a general concept of health. Although most of
the examples employed are drawn from organic phenomena, the concept
is designed to cover both somatic and mental health. The distinction between
soma and psyche is important only when we discern different possible
compromiserS of health, such as diseases, impairments and defects. These
compromisers can be of either a somatic or a mental kind.
An analytical project with the scope and purpose of the present study
always involves risks. A conceptual structure which aims partly at simplifi-
cation cannot be expected to explain all the subtleties of a complex concept

of ordinary language. This book thus reflects the extreme difficulty of

analysing a concept such as health.
The book, however, is written in the belief that traditional conceptual
analysis - containing tentative definitions and an exploration of their
consequences - is an indispensable tool in the philosophy of health. Pro-
gress is possible in this field, as in all other academic fields, only if positions
are stated clearly and their logical consequences are investigated in detail.
In such a way it becomes possible to reveal the strengths, as well as the
weaknesses, of the position in question.


Chapter one commences by listing a number of reasons for analysing the

concept of health. It is argued that health has a central place both in general
anthroplogy and in the theory of medicine. Moreover, it is emphasized that
a well-reasoned conception of health is essential also for clinical practice
and health care in general. This has become even more evident in those
countries where health authorities have been entrusted with the duty not
only of treating obvious cases of illness, but also with the duty of protecting
and promoting health.
The analysis starts with a presentation of a set of adequacy conditions
for an acceptable theory of health. It is claimed that such a theory should
be able to provide a definition of the concept of health, as well as an analysis
and mapping of several other concepts which are in different ways related
to health. This holds, in particular, for the conceptual "opposites" of health,
such as 'illness', 'disease', 'impairment', 'defect' and 'disability'. The theory
should also be able to explain the position of the concept of health in both
its societal and scientific contexts.
This explication of objectives is followed by some semantic prelimi-
naries. The philosophical nature of the project is discussed in some detail.
In particular, the two following questions are considered: (1) is there only
one concept of health ? and (2) what is the procedure by which we can clarify
this (or these) concept(s)?
In providing an answer to the first question this book departs both from
an Aristotelian essentialistic view of concepts and from an extreme nomi-
nalism, which entails that there is no clearly definable concept of health at
all. The analytic program of the book is based on the belief that there is
a set of highly interrelated uses of the term "health" (and its cognates in
other languages) which indicate a core sense of the term. This core sense,

which the following discussion attempts to explicate, is to be found both

in exclusively medical language and in everyday language. The medical
concept of health can, we believe, be understood as a technical- and more
narrowly definable - version of the everyday concept.
In relation to the second question, two possible procedures are consider-
ed, one being a completely empirical sociolinguistic study, and the other a
traditional philosophical conceptual analysis. A decision is made in favour
of the latter procedure, where this is interpreted as entailing (a) the conside-
ration and criticism of some prevalent abstract characterizations of health
in the light of ordinary language, and (b) a well-reasoned proposal for an
improved abstract characterization.
The final section of Chapter one sets the stage for the main analysis of
the study. This is done by introducing the two basic perspectives from
which 'health', 'disease', and other similar concepts are normally viewed.
The two perspectives are here called the holistic and the analytic perspec-
tives. From the former one focuses on the state of the human being as a
whole, and judges whether he or she is healthy. From the latter one directs
one's attention to particular parts of the human organism, and considers
their structure and function.
It is contended that prevalent theories of health and disease can be
viewed as stemming from one or the other of the two perspectives. A theory
taking the holistic perspective is one which employs mainly social and
humanistic concepts; a theory employing the analytic perspective, on the
other hand, uses a biological or chemical set of concepts.
Chapter two is devoted to a critical assessment of an important theory
of health belonging to the analytic perspective, viz. the biostatistical theory
(BST) of Christopher Boorse. The basic idea in this theory is that diseases
of living organisms are internal states which interfere with the normal
functioning of these organisms. (What is to be considered as normal
functioning is calculated statistically with respect to an age group of a sex
of a species.) Health is defined as the absence of diseases in this sense. In
the basic characterization of these concepts no attention is thus paid to the
performance of the organism in a larger, for instance a social, context.
In assessing this very influential theory of health our discussion pays
much attention to the concept of a function, and the relation between an
organism's functional ability and its environment. Two major points are
being made here: (1) The BST is not a clear theory of health and disease
unless we know at what level of integration the biological functions are to
be identified. We get very different results if the relevant functions are

interpreted to be, on the one hand, the functions of the microscopic parts
of the body, or, on the other hand, the gross functions of the major organs.
Both of these extreme alternatives can yield counterintuitive consequences.
(2) The BST faces a number of difficulties when we consider more closely
the dynamic interaction between an organism and environmental change
pertaining to the organism.
As a consequence ofthis discussion it is argued that our ordinary (as well
as scientifically medical) conceptions of disease cannot be entirely formed
according to a biostatistical model. Considerations concerning pain and
disability are obviously crucial. This conclusion provides good reasons for
considering a theory where pain and disability play an essential role viz.
a theory constructed from a holistic perspective.
The purpose of Chapter three is to provide and defend a holistic theory
of health, which is mainly intended to apply to human beings, although
some applications to non-humans are also made. The key concept in this
theory is the concept of ability, which is therefore given a substantial
analysis within the framework of modern action-theory. The main stages
in this analysis are the following:
The traditional distinction between ability and opportunity for action is
made. A combination of ability and opportunity constitutes what is here
called practical possibility: it is practically possible for a person to perform
an action if, and only if, he is both able and has the opportunity to perform
Ability is defined as that kind of possibility for action which is determin-
ed by factors internal to the agent's body or mind. The specification of an
ability must, however, always presuppose a situational background. It is
argued that, when this background is not explicitly stated, there is a tacit
presupposition of a set of "standard circumstances".
What counts as standard circumstances will, however, vary from place
to place and from society to society. As a result, a person with a particular
physical and mental make-up may be able to perform a required action in
one environment but not in another. This implies an important relativi-
zation in the case of the notion of ability. A fortiOri, this also applies to the
notion of health.
This relativization is in one important respect reduced by the intro-
duction of the concept pair first-order ability and second-order ability. The
idea here is the following: a person may be unable (in the immediate,
first-order, sense) to perform a certain action, but still have a second-order
ability to perform it. This then means that he will, given that he undergoes

adequate training and exercise, obtain the first-order ability to perform the
action in question. It is argued that the ability involved in health is of the
second-order kind.
After these preliminaries the fundamental task of the book is formulated
in the following terms: what is the set of goals, and by whom are they set,
which define the abilities that constitute health? This required set of goals
is designated the vital goals of man. Two important proposals for defining
the vital goals are given and discussed in some detail.
(a) The vital goals of man can be deduced from his basic needs (the
(b) The vital goals of man are identical with the goals that he
himself sets during the course of his life (the subject-goal
The first proposal is found to be too weak; the second is found to be both
too weak and too strong. Some improvements of the two proposals are
The major suggestion of the whole essay is then introduced: the vital goals
of a human being are. goals whose fulfillment is necessary and jointly
sufficient for the minimal happiness of their bearer. This is the tenet of what
is here to be called the welfare theory ofhealth. (In the case of humans welfare
is identified with happiness.)
The concept of human health is thus connected with the concept of
happiness. Health is in itself, however, neither sufficient nor necessary for
happiness. Health is a person's ability, in standard circumstances, to real-
ize his minimal happiness. It is not sufficient for happiness since, if circums-
tances are not standard, for instance in cases of accident or war, health
need not result in happiness. Nor is health necessary for happiness, since
the vital goals of an ill yet happy person can to a great extent be fulfilled
by people other than the person himself, for instance relatives and others
taking care of him.
The qualifying concept of minimal happiness is introduced and defended
in the context of an analysis of happiness. Happiness is presented as a
multidimensional concept ranging from a very high degree (along some
dimensions one can even speak of complete happiness) to a very low
degree. It is argued that the vital goals of man are conceptually connected
to some minimal degree of happiness to be decided upon by evaluation.
The concept of health thus derived is not theoretically decidable in the
following sense: the analysis of the concept is not sufficient to establish an

operational procedure for determining whether a particular person is

healthy or not. The analysis leaves undetermined where exactly the level
of minimal happiness in the long run is to be placed.
It is the contention of this essay that this openness mirrors the extent
to which health is an evaluative concept. What is to be counted as "real"
minimal happiness has to be decided upon, and cannot simply be the result
of empirical investigation. Since the level of "real" minimal happiness
determines the vital goals, it also determines health.
In Chapter four there is a discussion of the major conditions which
compromise health, for instance diseases, injuries, pregnancy, and senility.
Following a recent study by Culver and Gert [27) diseases, injuries and
some other bodily phenomena are grouped together in a category called
"maladies". It is typical of maladies that they are entities internal to a
person's body or mind which tend to impair his health, but which need not
do so.
This section provides a general characterization of maladies as well as
a suggestion for making sharp distinctions among the various categories of
Chapter five attempts to reveal in what respects and in what senses the
concept of health is relative to the society employing the concept. A
fundamental distinction is made between society as a platform for action
and society as a producer of values.
In the former sense society influences the concept of health by determin-
ing the societal standard circumstances within which the human being oper-
ates. In the latter sense society influences the concept by indicating the
levels of minimal happiness.
In the same chapter the question is asked whether the welfare theory of
health would have important consequences for the science or practice of
medicine. It is argued that the consequences for the science and practice
of rehabilitation are more profound than for traditional medicine in the
sense of the art of curing diseases. The openness of the set of vital goals,
for instance, will only marginally affect the list of accepted maladies. The
main reason for this is that most acknowledged maladies strike their
subjects in such a basic and general way that they so obviously disable their
bearers whatever the choice of vital goals.
In a special study the welfare theory of health is used to determine the
status of a "controversial" disease or illness, viz. homosexuality. The fun-
damental question discussed is whether a homosexual is as a rule unable
to realize some of his vital goals. The conclusion drawn is that homosexual-

ity is a disease (or impairment) only given very special provisos: (i) repro-
duction is an indisputable vital goal, (ii) the homosexual does not merely
choose not to reproduce, but is also unable - for physical or mental reasons
- to reproduce. The latter proviso, obviously, is not generally true.
Finally, the welfare concept of health is applied to the realm of non-
humans. It is argued that the welfare concept of health is applicable also
to the non-human living world. With the higher animals both the ideas of
ability and happiness can be retained. To the lower animals and the plants
the concept can only be extended through analogy. It is, however, disputed
that health in lower animals and plants should be identified simply with
normal probability of survival and reproduction. Health, in these cases, can
also be understood in terms of usefulness: a corn plant is healthy if it
contributes, given standard circumstances, in an expected way to certain
goals and ultimately to the happiness of its cultivator.



Health has not generally been viewed as a proper object of philosophical

study. It is not well known that health and health care were important
topics for Plato and Aristotle, as well as for Descartes, Locke, and Kant.
Few people know that the dominant school of medicine in Europe until the
seventeenth century - Galenic medicine - was an application of central
themes in Aristotle's natural philosophy, or that many of the schools that
followed were highly influenced by Descartes' philosophy of man. Even
fewer would believe that philosophical analysis or speculation could make
any valuable contribution to modern medicine. Medicine has for a long time
- so many would put it - been liberating itselffrom the bonds of philosophy
in its move to become an empirical science. 1
There is thus a case for clarifying the position of a philosophy of health
and for considering its potential value. We shall here only make a brief
attempt in this,direction. This will be undertaken from three points of view:
from that of philosophical anthropology; from that of a philosophy of
medical science; and, most particularly, from the perspective of the art of
(1) Health as a subject of philosophical anthropology.
It is important to emphasize that medicine has no monopoly on the
concept of health or related concepts. Health, impairment, disease and
disability are concepts which are well embedded in ordinary thinking and
which have a long non-scientific tradition. There is an ordinary concept of
health which can be used by the layman with the same accuracy as he can
use most other central concepts characterizing man, for instance concepts
of morality, emotional concepts, or concepts of excellence.
To characterize a human being in terms of health or illness is to describe
one aspect of the 'status' of this human being, what we often call his 'state
of well-being'. This aspect is to be distinguished from his moral status and
his emotional status, as well as from his status as regards his intelligence
or talent.

These various aspects may be difficult to separate, which can entail

confusion, sometimes with societally dangerous consequences. Immoral
and illegal behaviour is sometimes confused with illness. A criminal is
labelled sick; an ill man is labelled a criminal.
The purpose of philosophical anthropology is to understand and intellec-
tually organize the qualities of man in order to construct a reasonable view
of mankind. This entails the detection of those qualities which are unique
to human beings, as well as those which are essential to them but which
they share with other living creatures. Health belongs to the latter category;
it is not unique to humans but some of its manifestations are obviously
peculiar to them. It must therefore be a proper and important task for
philosophy to analyse the concept of health and determine its relation to
other central characteristics of human beings.
(2) Health as a subject for the philosophy of science.
Having established that medicine has no monopoly on health, we must
acknowledge the obvious: there is a science, or a collection of sciences, with
a particular interest in health, viz. medicine. The medical sciences focus
their interest in health on its primary negative counterpart, disease.They
meticulously describe the various human diseases; they classify them,
investigate their etiology and look for therapies with which to combat them.
The concept of disease is thus a key-concept in medicine. Thus, unders-
tanding the concept of disease is essential to understanding medicine as a
science. The concept and its relatives must therefore be proper subjects of
that discipline which treats of the sciences and their theories and concepts.
That discipline is the philosophy of science.
(3) The importance of analysing the concept of health for clinical
medicine and health care.
But is conceptual analysis not of very limited interest to the clinician?
Are the conceptual problems of clinical practice not normally solved in a
very pragmatic and efficient way? The healthy person is simply the one who
performs his daily work without complaint; the person who is ill consults
doctors and seeks treatment.
This simple view can, however, be challenged for a number of reasons.
Consider four cases:
(a) Health authorities nowadays do not just treat people already afflicted
with disease, they also devise programs for the prevention of disease. One
important element in such a program is the detection, at a very early stage,

of pathological changes in the population. In order to make such detections

the authorities have to approach many people who consider themselves
healthy, and who do not themselves seek any health care. This procedure
is commonly known as screening.
Here, obviously, the decision concerning the existence of disease in an
individual must be made independently of any judgment on the part of the
individual as regards his own health status.
(b) A second example can be drawn from psychiatry, including forensic
psychiatry. Mental hospitals contain a great number of patients who do not
consider themselves ill, and who have not approached the medical estab-
lishment willingly. According to law, senior psychiatrists in most countries
are entitled to incarcerate individuals who are thought to suffer from
certain mental diseases. Such laws presuppose clear definitions of the
mental diseases in question. They also presuppose the existence of criteria
for the presence of those diseases, which are independent of the subjects'
own judgments?
(c) Our third example stems from the health insurance authorities. These
authorities are continually obliged to make decisions concerning health and
disease without ever meeting the suffering individuals. It is true that they
base most oftheir decisions on certificates issued by medical doctors - who
have met the patients - but they also perform a critical evaluation of the
certificates. The certificate must fulfill certain requirements, and these
requirements are not merely of a simple legal kind. Not all conceivable
diagnoses are accepted as indicating instances of disease. There is an
independent judgment on the part of the insurance authorities as to what
diagnoses should be classified as indications of disease and thereby justify
the payment of a sickness benefit. 3
(d) But clinical medicine and health care do not focus merely on disease.
The concept of health also occupies a central position. This fact has
recently been emphasized by many health authorities and health organi-
zations. In Sweden, a new Public Health Act has been instituted. 4 Accord-
ing to this act it is the duty of the health care organization not only to treat
disease, but also to promote health. It does not suffice to take care of those
who have been stricken by disease and who actively seek help. It is also
the responsibilty ofthe health-care personnel to maintain and promote the
already existing health of the population. But in order to do this in a
sensible way, we clearly require a reasonable understanding of, and some
consensus concerning, the nature of health. What is not clear is whether

such a consensus presently exists, over and above the application of the
notion to certain obvious diseases.
In sum, there seem also to be urgent practical needs - in addition to the
general philosophical and scientific ones - for correctly characterizing
health and disease.



An analytic procedure of this kind presupposes a set of adequacy con-

ditions. According to what criteria should a theory of health be judged?
What questions should the theory be able to answer? In the following we
shall provide a list of such questions.
(1) What are the logical relations between the health-concepts?5
As has already been indicated, there is a comprehensive semantical field
of interrelated concepts, which will here be called "health-concepts". These
concepts, it will be assumed, bear certain logical relations to one another.
The task is to determine the nature of these relations. The following
concepts will be called "health-concepts" (and they will all be treated to
some degree in this essay): health, vitality, illness, disease, impairment,
injury, defect, disability and handicap.
The theory should, in particular, be able to say whether a negative
health-concept, such as that of disease, can be applied simultaneously with
that of health to the same subject. Can a person be healthy and have a
disease at the same time?
(2) What are the logical relations between the concept of health
and some other central humanistic concepts?
The concept of health is obviously related to a number of other central
concepts applicable to humans, and is indeed sometimes confused with
some of them. This holds, in particular, for the concepts associated with
norms, such as 'decency', 'morality', and 'legality'.
Health is also supposed to have an intimate relationship with happiness.
What is this relationship? Is health a prerequisite of happiness? Another
important conceptual relation is the one between 'health' and 'ability'. If
health is in some way related to 'ability', how is it to be distinguished from
the concepts of excellence: talent, intelligence, strength and creativity?

(3) What is the relation between human health and the health of
other living beings?
Most health-concepts seem to be applicable to other living beings, ani-
mals as well as plants. A dog can be healthy and can acquire diseases and
injuries; so can a cauliflower. It is plausible to assume that these appli-
cations of health-concepts are not radically different from their uses in the
human case. A reasonable theory of health should be able to account for
the similarities as well as the differences between human health and the
h~alth of animals and plants.

(4) What is the relation between mental and somatic health?

Is there such a thing as mental health? Thts has been a subject of
controversy. Still, all civilized countries officially acknowledge this concept
by instituting mental hospitals and offering medical treatment to persons
who are said to be mentally ill. A theory of health should be able to take
a coherent standpoint on this issue.
(5) What is the relation between health and the environment?
Human health and human illness are not isolated phenomena. A man
who is healthy or ill lives in both a physical and a cultural environment.
These environments influence him in several ways. Some such influences
are simple causal influences: the environment may directly create illness by
hurting the subject or by putting extreme pressure on him. Other influences
are more subtle and their relations to health more indirect. Nevertheless,
as will be pointed out, in particular in Chapter five, it is quite important
to detect them.
The environment constitutes the background of an individual's life. It
defines the possibilities of his development and action. One environment
physically allows one kind of development and action; another environ-
ment allows a different kind. Thus healthy people may show different kinds
of characteristics depending upon the environment in which they live.
Another subtle influence is the one stemming from society. Society sets
goals for its members. Some of these goals become standards. according to
which the health of the members is determined. A theory of health should
be able to give an account of these indirect influences.
(6) What is the place of the health-concepts in science?

The science of medicine has as its subject matter the phenomena repre-
sented by the health-concepts.This raises a number of theoretical questions
concerning the nature of these concepts. What kind of qualities do they
refer to? Are they purely biological; or are they biostatistical or perhaps
anthropological; or do they belong to a number of different spheres?
To this can be added a more radical question: are the health-concepts
basically descriptive, scientific, concepts or are they basically evaluative?
What is the point in talking about "positive" and "negative" health-con-
cepts? Do they indicate some kind of evaluation? What is the nature of this
evaluation and to what extent is it compatible with a science of medicine?


Consider first some fundamental questions about health-concepts and

conceptual analysis in general. The most crucial question is whether there
is only one concept of health and one concept of disease. Does not the
multitude of definitions in the area of health indicate that there are many
concepts of health and disease? If so, then perhaps one particular theory
is suitable for analysing some of these concepts, while a different theory is
suitable for others.
These questions first deserve some background from the theory of se-
According to Aristotelian philosophy, defining is the process by which
the "essence" of things is revealed. The essence of an object consisted of
those of its properties in virtue of which the object belonged to a particular
species. The essence of man, for instance, was the property of being a
rational animal. To define the species of man, then, was tantamount to
presenting this property.6
According to Aristotle all things in the world - not just biological
creatures - belong to species; and all species have their essences. There-
fore, in general, the process of definition consisted in characterizing the
essence of things.
It is characteristic of this view that defining is directed towards the
world, to nature itself. Aristotle considers it possible to inspect nature,
"find" the essences of things, and, on this basis, to formulate definitions.
Such a view of definition is often called a realist view.
In contrast to this realist approach to definitions there is a nominalist
approach. A nominalist does not believe in "true essences", nor in any
natural hierarchies of phenomena. In his view, what we have is a world

which can be structured in many different ways, and a language which can
be used in many different ways. There are certain ways of structuring the
world and of using the language which are currently dominant. But nothing
in principle prevents us from changing them.
According to this view, to define concepts such as health or disease is
not to find the true nature of these phenomena (there is no such true nature)
but to determine a particular way of using language. This way may be
currently accepted or newly stipulated.
Within nominalism one can find two important but diverging tendencies.
One of these is the exact opposite of Aristotelian essentialism. It denies not
only that the world has a natural organization, but also that language has
any clearly definable uses.And, at least according to some advocates, it
even denies the desirability oflanguage's having a clearly definable use. This
position will be called the strong version of nominalism. 7
A different nominalist position is the following: It is indeed a convention-
al affair how we structure the world. It is dependent on our ways oflooking
at things and on our particular purposes. On the other hand, our use of
language is not completely arbitrary. A presupposition of efficient commu-
nication between individuals is that they structure the world and use
language in almost identical ways. Since communication seems to be rather
efficient in most societies, it is a reasonable hypothesis that there is a fairly
definite structuring of the world and use of language within them. This
structure and language use can be defined. For example, we can define how
the terms "health" and "disease" are used in the Anglo-American society.
This is, according to this' version of nominalism, one reasonable interpre-
tation of the phrase "to define the concepts of health and disease".
If it were to turn out that part of a particular language use is unclear and
difficult to define - which in fact implies that communication must be
ineffective - then a nominalist of this persuasion would not hesitate to make
an explication of the notions involved. Nor would he hesitate to recommend
making certain changes (often simplifications). In contrast to the nominal-
ist of the strong persuasion, he finds clear definitions desirable.
The philosophy to be followed and defended in this essay will be of the
latter form, and will be called weak nominalism. According to this view,
then, there is a fairly definite use (or interrelated uses) of the term "health"
and its relatives. This use (or uses) determines the concept of health to be
analysed in this essay.
But if this is our position, how should we explain the following phenom-

(i) the history of medicine, as well as the history of philosophy,

displays a great variety of definitions of health-concepts
(ii) there seems to be a great diversity in the attribution of health
and disease between different cultures and even within one and
the same culture.
(i) Consider first the mUltiplicity of definitions. It is true that there are a
great number of suggested definitions of the health-concepts. All the great
medical theoreticians have formulated such definitions. This is true of
Hippocrates, Galen, Sydenham, Boerhaave and Bernard. And nowadays,
representatives of disciplines such as psychology, sociology and anthropol-
ogy have made their contributions. They have offered new characteri-
zations of health which bring in elements from their own disciplines,
emphasizing, for instance, the societal context. 8
But to say that there is a multiplicity of definitions does not amount to
saying that there is a multiplicity of concepts. First, many of the definitions
are very similar; some of them may only appear to be distinct, for example
they may use synonymous terms. Second, a concept is not identical with
a definition. There may be a great number of definitions, all of which
constitute attempts to define a single concept. There could be several
reasons why these attempts are different. One reason may simply be that
some attempts arefailures; they are unacceptable characterizations of the
concept in question; they do not capture the logic of ordinary discourse.
Another reason is that the various attempts differ in degree of ambition.
Some only attempt to give a rough characterization; others try to give a
very detailed picture of the concept in question.
What has been said about definitions here pertains to what are normally
called lexical definitions. These are definitions aimed at characterizing a
given concept, a given mode of speech. There is another kind of definition,
which can introduce new concepts. It is the stipulative definition. By giving
such a definition one decides that a particular term is to have a certain
sense. It is indeed possible to define "health" in a stipulative way and
introduce a new concept of health. The question is only what use one is
going to make of such a concept. If it were completely unrelated to the
ordinary concept of health it would not be used in ordinary discourse, and
would therefore be of no interest to us. If it were so related, it would have
the character of a technical concept of health, to be used for particular

It is important to note that a technical concept of health need not be

explicitly introduced by a stipulative definition. There may also be a current
technical use of the term "health" in a particular subculture of that culture
which acknowledges the term. It is, for instance, a plausible hypothesis that
the professional medical subculture utilizes a technical concept of health
without having explicitly introduced it. This hypothesis will be dealt with
in Chapter four, the section on Maladies.
In acknowledging the existence of technical concepts of health we shall
to some extent sustain the popular idea that health is a 'family' concept.
According to Ludwig Wittgenstein most natural concepts are family con-
cepts, by which he meant that no concept has one singular use but a number
of different uses. These different uses are, he said, related to each other in
the sorts of ways that the members of a family are related. 9
In making this concession we should not conclude, however, that there
is no point in defining health. In this study it will be assumed that the
concept of health has fairly clear boundaries and that these can be detected
by a careful analysis.
Befcre leaving the issue of there being a multiplicity of definitions, a
further fact should be emphasized. There is an important distinction
between definitions and empirical theories. Both definitions and theories
characterize phenomena; but, from a theoretical point of view, they do it
in very distinct ways. A definition, as we have said, characterizes a concept.
An empirical theory characterizes the phenomenon represented by the
concept, for instance by specifying what its causes could be. This principal
distinction is unfortunately not always recognized in scientific discourse.
It may be unclear whether a particular author intends a certain statement
to be a definition or an empirical theory. It is, for instance, unclear whether
Galen meant that his idea of health, as the balance between the primary
elements of the body, was a definition or an empirical theory.
For the purposes of this essay it is important to note the following: Two
persons may very well have distinct empirical theories about the phenom-
ena of health and disease, and still share the same concept.
(ii) Consider now the fact that there are quite different uses of the terms
"health" and "disease" in different societies and even within one and the
same society. People can attribute the qualities health and disease to
different subjects and apparently on different grounds. This need not be due
to the fact that they have different "definitions". They may never have
formulated a definition. They simply use the terms in different ways.

There are a number of explanations of such variations which are consis-

tent with the fact that the attributors share the same concepts of health and
disease. We shall here consider two explanations. Let us assume that the
attributors share the following (simplified) concept of health: "C is in
health" means that C is able to fulfill some of his vital goals. By a disease
they mean a state of the individual's body which tends to disable him i.e.
which tends to cause him to be ill.
(a) A and B disagree in a particular attribution of health because
of different judgments of a single case.
First, this difference can have a very simple explanation. A and B may
inspect the single case with different degrees of care. A may notice features
which B overlooks.
Second, while observing the same phenomena, they may interpret them
differently. Assume that a physician A and a pentecostal pastor B are asked
to judge whether a third person C is ill. C exhibits "extreme" behaviour;
he shouts without communicating with any fellow human being; his face
is directed upwards; he seems to be in contact with God. A gives the verdict
that C is ill. B judges otherwise; he considers C to be in contact with God.
A notices C's superficial behaviour. He notices that C has lost contact
with "ordinary" reality, and predicts that he will not be able to take care
of himself. A judges C to be disabled. B interprets C's state differently. He
judges that C is far from disabled. It is merely that his attention is intensely
concentrated on a particular object; thus he is temporarily out of contact
with the world around him.
These differing judgments are clearly consistent with A 's and B's sharing
the same concept of health. The difference between them lies in the fact that
A considers C to be disabled, whereas B does not.
(b) A and B disagree on a particular attribution of health because
their attributions are given against the backgrounds of different
This explanation of a disagreement is based on the particular analysis
of health to be given in Chapter three. The general idea can be put very
briefly: ability is always measured against a certain background, natural
and cultural. A person with a certain set of internal qualities may be able
to reach a particular goal in one environment, i.e. in one society but not
in another. Let us assume that A looks upon C from the point of view of

the first society. Hence, C is judged to be in health. B looks upon him from
the other perspective. Hence, C is judged to be disabled, to be ill.
Here again, A and B may share the same concept of health, the difference
between them lying in their application of the concept in different contexts.
Our list of explanations of disagreements between people who make
judgments about health can be lengthened. A more complete discussion of
some of these aspects will be found in Chapter five, section 1: On the relation
between health and society. The purpose of this discussion was merely to
present and critically analyse a set of arguments to the effect that there is
a great multiplicity of concepts of health.
Legitimately refusing to accept such arguments, however, does not prove
that there is only one "true" notion of health. This is not the standpoint
taken in this essay. There is no Aristotelian species called "health"; what
there is, is a use (or a number of related uses) of the term "health". The
assumption is made, however, that this use (these uses) is consistent
enough to allow a characterization and an explication.
Such a characterization can be performed in slightly different ways and
with slightly different purposes. One way is to make a detailed sociolinguis-
tic study and try to pinpoint similarities and differences in various
subcultures. It would be of interest to trace these subcultures, see how
influential they are, and try to describe the technical concepts of health and
disease which they employ. But such a pursuit could not substitute the kind
of project envisaged in this essay.lO
The present project is more traditionally philosophical; its purpose is to
find a core element in prevalent uses ofthe term "health", and try to develop
it in such ways that it will become coherent and useful for scientific
purposes. The aim is not merely one of lexicography, but also of logical
reconstruction: to sharpen the borders of the concept of health. The con-
cept will thus be influenced by the process of analysis. Thus there is an
element of stipulation in the present program, though the basis is an already
existing concept of health.


In ordinary thinking about the phenomena of health and disease there is

a tendency to oscillate between two perspectives. From the first perspective
one focuses on the general state of a human being and considers whether
or not the person is healthy. This means asking questions such as the

following: How does this person feel? What is he able to do? Can he
function in a social context?
From the second perspective one directs one's attention to particular
parts of the human organism and considers their structure and function.
One asks questions such as: Is this organ normal? What is the pulse rate
of this man? What does the tissue of the liver look like? What capacity do
the lungs have?
The first perspective, which focuses on the human being as a whole, will
here be called a holistic perspective. A study pursued from this perspective
will use concepts borrowed from ordinary language, psychology, anthro-
pology or sociology. Examples of such concepts are those of well-being,
pain, depression, ability, adaptability, disability and handicap. II
The second perspective, which concentrates on the parts of the orga-
nism, will here be called an analytic perspective. A study pursued from this
perspective will use mainly biological, chemical and statistical concepts. It
will involve inspecting organs and tissues, studying their functioning and
measuring their rate of change, as well as calculating the relative frequency
of the values obtained. 12
What are the ,sources of these two perspectives and why do they both
have a prominent place in our thinking? The source of the first perspective
is obvious. The ordinary human being is primarily interested in the holistic
facts. How do I feel today? Has my pain gone? Can I go to work? What
matters to him are the realities about his whole person. The detailed
functioning of a particular organ is interesting only ifit substantially affects
his whole person in a positive or negative way. The key question for the
ordinary man is: am I healthy or not?
The source of the second perspective is the art and science of medicine.
Medicine has a task: to eliminate disease and restore the health of those
who seek its help. In order to accomplish this, medicine must acquire
knowledge about the mechanisms behind the phenomena of health and
illness. To obtain this knowledge the physician must make detailed investi-
gations involving the smallest accessible parts of the human body. As a
result, his concentration is fixed on particular internal phenomena. A key
question from this perspective is: what is the nature of this disease?
The two perspectives clearly do not exclude one another. In fact, one
cannot view the health-disease dimension from but one of these perspec-
tives alone. The ordinary man surely understands that there is an organic
(and mental) background which is responsible for his state of health. He
understands that his body is like a piece of intricate machinery which can

function well or go awry. In the first case the result is health, in the second,
illness; and the particular malfunction is a disease.
Conversely, the practising physician is highly aware of the holistic pers-
pective. The call for help comes from a person who claims that he is ill; and
the physician's task does not end until health (viewed from the holistic
perspective) is restored.
It is also clear that any serious theory of health and disease must take
both perspectives into account. The holistic phenomena of health and
illness must be properly treated, as must the analytic phenomena of dis-
eases and impairments; moreover, there must be a clear account of the
relation between the two worlds: how, for instance, is a disease related to
the general state of a person's health?
While acknowledging this, it will be argued here that most theories of
health - perhaps all plausible ones - basically stem from one of the two
perspectives. That this is a reasonable claim can be seen from the following.
It is a plausible supposition that health and disease are in some sense
conceptually related. It is not only a matter of empirical fact that diseases
affect our health. If a certain "disease" had no consequences for anybody's
health, we would stop calling it a disease. That the two concepts are
conceptually related thus means that one can be defined (at least partially)
in terms of the other.
For such a definition to be articulated, one of the concepts must be
chosen as more basic than the other. If health is chosen as the basic
concept, the concept of disease should be defined in its terms, for instance,
as a phenomenon which compromises health. But, then, health could not
itself be defined in terms of disease - that would be circular. For the
primary characterization of health we would have to find a set of concepts
not containing the concept of disease. The most natural set would be found
among those used in the holistic perspective, for instance: a man is healthy
if he feels well and can perform his social functions.
Conversely, if disease is chosen as the basic concept, it cannot be defined
in terms of health (or illness). Here, then, the most natural set of concepts
in terms of which to define it will be found among the ones used in the
analytic perspective, for instance: a disease is the abnormal functioning of
a bodily organ.
J'he main role of the perspectives in the formation of a theory of health
and disease then is to aid one in selecting one of the concepts as basic.
Moreover, the perspective chosen provides the conceptual background for
the basic concept of the theory. But a theory formed from one of the

perspectives should certainly be able to give an account of the whole

conceptual area in which 'health' and 'disease' are the key concepts.
By speaking in terms of perspectives we obtain a tool for dividing
theories in the Philosophy of Health into holistic theories and analytic
The question might now be asked: could there not be a theory of health
which operates simultaneously from both platforms? Could there not, for
example,be a theory which defines health in terms of fitness and ability, and
disease in terms of abnormal organ function?
Such a theory would, however, run counter to our supposition. If health
is defined solely in holistic terms and disease is defined solely in analytic
terms, then there is no conceptual connection between them. It then
becomes a matter of simple empirical fact that diseases affect health. In
that case it would be possible that certain diseases would have nothing to
do with anybody's health. As long as they are abnor.mal functions, they
would be diseases.
It is difficult to see how such a theory could work in the long run. It would
soon require certain additions or amendments. One natural amendment
would be to expand on the characterization of disease and say: disease is
such abnormal functioning of an organ as affects health. But if this is the
procedure chosen, then the theory in question has been transformed into
a holistic theory, since disease is being (partially) defined in terms of health.
Thus, health becomes the basic concept, and the conditions for the theory's
being holistic are fulfilled.
The fundamental issue confronting the holistic and analytic theories,
then, is whether 'health-concepts' should basically be holistic or analytic.
This book will plead for the former, and the attempt will be made to
formulate a holistic theory of health in some detail.




In classical speculation on medical matters health is conceived of as a

bodily state which is in accordance with Nature. It is a state of natural
balance in the mixture (complexio) of the primary qualities of the human
body. According to Aristotle's and later Galen's teaching, the human body
consists of the four elements: earth, air, fire, and water. These elements
derive their properties from the four primary qualities: hot and cold, wet
and dry. The four elements are said to arise out of primary matter by the
action of the active qualities (the hot or the cold) on the passive qualities
(the wet or the dry). The primary qualities form homogeneous bodies of
different kinds. The homogeneous bodies, in their turn, form through
various combinations heterogeneous bodies. The properties of the latter
are the secondary qualities, such as quantity, size and weight. The homo-
geneous parts of animals are the tissues. The heterogeneous parts are the
Health is regarded as the proper balance between the respective primary
qualities of a particular human body. If there is a disruption in the balance,
e.g. if one quality should dominate over another, there is imbalance or
disease. According to Aristotle there were four possible deviations from the
ideal temperament. These consisted in the dominance of pairs of an active
and a passive quality: cold and wet, hot and wet, cold and dry, hot and dry.
Galen added four more possibilities. There may, according to him, also be
simple distemperaments. These consist in the predominance of one quality
combined with a balance between two other qualities. A mixture can thus
be merely hot, while there is a balance between the wet and the dry. In
Galen's teaching, then, there were eight distemperaments or diseasesY
Although few of the details in the ancient natural philosophy and the
Galenic philosophy of health have survived, it is important to note that two
of the ancient ideas still influence the thoughts both of the layman and the
medical scientist. These are the idea of a balance 14 between opposing
elements or forces and, in particular, the idea of a natural or normal state
of the living organism. We shall concentrate here on the latter idea.

In many contemporary presentations human health is identified with the

normal structure and the normal functioning of the human body. Diseases,
on the other hand, are identified with particular deviations from this
norm. IS The conception of health as biological normality, however, leaves,
room for different interpretations. There is, in particular, one fundamental
clarification to be made: what theoretical status do the biological norms
have? Are they norms proper, i.e. rules in accordance with which a human
body should function if it is to function properly; or are they statistical
averages, i.e. is a norm defined by a number that the majority of the
population tends to approach?
These two different interpretations of the notion of a norm give rise to
very different philosophies of health, although they may have superficial
similarities. It is therefore unfortunate that both the ancient and many
modern statements concerning normality have oscillated between the two
interpretations. King [66] says the following about the ancient use of the
term "natural":
In the Galenic tradition, "according to nature" meant something both prevalent and desirable.
"Contrary to nature" was abnormal and undesirable. The natural implied a standard to which
things ought to conform. A certain amount of leeway was acceptable, but the limits of the
permissible deviations are not at all clear (p. 133).

Also on the modern bio-medical platform one can find the dual - and
essentially unclear - view of normality, combining descriptive and norma-
tive issues. There is today a marked drive towards using statistical
normality as the basic concept, but few presentations do not at the same
time employ evaluative terms in characterizing the normal functioning of
a human body.16
In contrast to this prevalent unclarity there is one forceful modern
attempt to formulate a conception of health and disease in terms ofbiologi-
cal norms where the interpretation is unambiguous. According to this
conception the biological norms are related to certain natural goals (for
instance the goal of survival). These goals are not attributed to the body
from the outside - they are not goals according to which the body should
function - but belong to the internal constitution of the body. Scientists can
detect what these goals are by inspecting a large sample of human beings,
by making a biostatistical analysis. These are the essential tenets of a theory
presented by the American philosopher Christopher Boorse in his signifi-
cant articles ([13]), ([14]) and ([15]). This theory will from now on be called
the biostatistical theory (BST).


The concept of a goal will playa central role in the following discussion.
It has its place both in some analytic theories and, very clearly, in holistic
theories. The term "goal" has two rather different senses, however, both of
which are important in the analysis of health.
In its m.ost general sense the term "goal" refers to a state of affairs which
is the end of some sequence of events, be they natural events or actions.
But this general interpretation can be specified in two ways. Thus one can
speak of an ideal goal set by somebody, normally a human being or some
collection of human beings. This notion is tied to such concepts as 'in-
tention' and 'desire'. If a person intends to realize a state of affairs then this
state of affairs is an ideal goal of his. Alternatively, one can consider a
factual goal, by which we mean a state of affairs that an entity has, as a
matter of fa.ct, a tendency to approach.
A quite precise illustration of this latter notion can be collected from
modern molecular biology. The general idea here is the following: in the
genes of an organism there is an encoded program, which steers the
development of the organism towards a specific phenotype. The program
is connected with steering mechanisms in such a way that, given a set of
possible environments (excluding the most extreme ones), there are causal
mechanisms to effect the evolution of the phenotype. 17
In general we shall say that an organism 0 is goal-directed in this sense
if, and only if, the following conditions hold:
o includes or is connected with a program which assigns a set of goals,
and a set of steering mechanisms, such that the steering mechanisms
constantly keep 0 oriented towards one of its goals.
The notion of goal assumed in the biostatistical theory of health (BST)
is the notion of a factual goal. 18
The realization of a goal, whether ideal or factual, involves a process or
an activity of some kind. If the goal is very general or far-reaching, its
realization may presuppose a very long process involving a number of
stages. These stages constitute subgoals of the main goal.
There may be alternative sets of subgoals for one and the same ultimate
goal. This means, on the one hand, that in a particular situation there may
be different ways of realizing the ultimate goal. One can, for instance, travel
to New York from Stockholm via either Copenhagen or Oslo.
But it means, on the other hand, that different situations may require
different subgoals for the realization of one and the same ultimate goal. For

instance, during a complicated heart operation the ultimate goal of the

patient's survival cannot be achieved by natural means only; he must be
connected to a heart-lung machine. The realization of this measure there-
fore constitutes a new subgoal given the extraordinary circumstances of a
thoracic operation.
The variability of subgoals and their dependence on situations will be an
important topic in the assessment of the BST.


In introducing his theory Boorse explicitly aligns himself with the ancient
medical tradition that culminated in Galen. Boorse here quotes the medical
historian Temkin:
Such a concept of health and disease rests on a teleologically conceived biology. All parts of
the body are built and function so as to allow man to lead a good life and to preserve his kind.
Health is a state according to Nature, disease is contrary to Nature. ([130], p. 398).

Boorse, then, readily admits that in his characterization of 'health' and

'disease' these notions are teleological. The idea of a goal-directed function
is perhaps the most central idea in his theory. There is, however, one
important respect in which Boorse distinguishes himself from his ancient
predecessors. Boorse's idea of a normal or natural function does not imply
a positive evaluation of the function.
In my view the basic notion of a function is of a contribution to a goal. Organisms are
goal-directed in a sense that Sommerhoff, Braithwaite, and Nagel have tried to characterize:
that is, they are disposed to adjust their behavior to environmental change in ways appropriate
to a constant result, the goal. ([15], pp. 555-556).

Boorse admits that the highest-level goals of organisms are somewhat

indeterminate. The behavior of organisms seems to contribute to different
and independent goals simultaneously. Among these are individual survi-
val, individual reproductive ability, survival of the species, survival of the
genes and ecological equilibrium. But, he continues:
It is only the subfield of physiology whose functions seem relevant to health. On the basis of
what appears in physiology texts, I suggest that these functions are, specifically. contributions
to individual survival and reproduction. ([15], p. 556).

Basically, then, physiological functions contribute to the survival and re-

production of the relevant organism. As Boorse stresses, however, this is
not a universal truth. When one has a disease some of one's functions are

impaired and do not contribute to these goals. Accordingly, when we speak

of physiological functions in the abstract, we are speaking about "a trait's
standard contribution in some population or reference class, e.g. a species"
([15], p. 556).
The notion of a standard, or of a species design, is thus a crucial notion
in the BST.
Boorse contends that the science of biology presupposes (and perhaps
necessarily presupposes) the idea of a species design. When a biological
textbook characterizes a species it aims at describing "the typical hierarchy
of interlocking functional systems that supports the life of organisms of that
type"([15], p. 557). This typical hierarchy is determined by statistical
means. Every detail in the hierarchy is statistically normal for the species.
(Boorse is aware of the necessity of differentiating between ages and sexes).
Accordingly Boorse's idea of a normal function is the idea of a statistically
normal function.
The species design that emerges from a statistical investigation of a great
number of members of a species is, says Boorse, an empirical ideal that
serves as the basis for judgments of health regarding any species. This
notion thus gives firm ground for applying the notions of health and disease
to all kinds of living matter, not just to human beings.
From these preliminaries we can then derive Boorse's notion of disease:
Diseases are internal states that interfere with functions in the species design.
Alternatively: diseases are internal states that interfere with normal
functions. A more precise definition of a normal function is the following:
Normalfunctioning in a member of the reference class is the performance by each internal part
of all its statistically typical functions with at least statistically typical efficiency, i.e. at
efficiency levels within or above some chosen central region of their population distribution.
([15], pp. 558-559).

Boorse also contends that an organ's inability to perform a function consti-

tutes an instance of disease even if the occasion to perform it were not to
arise. A diseased person may live and function normally in a particular kind
of situation, but what distinguishes him from the healthy person is that he
is not able to meet every kind of standard situation. The hemophiliac, who
is protected from injury, still has a disease, as does the diabetic who takes
insulin daily.
An important feature of the BST is that it aims to cover mental health
as well as somatic health. Boorse denies that there is a fundamental
difference between these two notions. The notion of mental health is not,

and should not be, any more value-laden than the concept of somatic
Mental health must be a constellation of qualities displayed in the standard functional
organization of members of our species. Only empirical enquiry can show whether normal
human beings have an even temper, engage in socially considerate behaviour, and advance
the species -or make love with 'dignity and decency' ([14], p. 70).

Boorse claims that there are a number of species-uniform mental goals. It

is the task of various mental faculties to serve these goals. Perceptual
processing, intelligence and memory provide information to guide effective
action; drives serve to motivate such actions; anxiety and pain function as
signals of danger; linguistic behaviour is a device enabling cultural cooper-
ation, etc. If these faculties function normally, then the bearer is healthy.
If, on the other hand, some internal state of his were to prevent any of them
from functioning normally, then he would be mentally ill.

Let us summarize the central theses in Boorse's philosophy of health and

(i) Health is normal functional ability; and disease is an internal
state that reduces such ability to below typical levels of efficien-
(ii) Normal functional ability is calculated statistically with respect
to an age group of a sex of a species. It constitutes a statistically
typical contribution by the members of the group to certain
factual goals. In the case of somatic health these goals are the
survival and reproduction of the individual.
Boorse, however, realizes that his view of health and disease excludes
certain states which, at least according to current linguistic usage, would
be labelled as diseases. The cases he mentions are the structural anomalies
or diseases and the widespread or even universal genetic diseases. The
reasons that these cases fall outside his system are the following: (i) Health
and disease are explicitly connected to the concept of function. Structural
anomalies are compatible with normal as well as abnormal function. (ii)
Since health and disease according to Boorse are basically statistical
concepts, health is determined by the statistically normal state of the
population. If a particular genetic defect is universal or at least statistically
normal, then that defect cannot count as a disease. 19

Boorse's account of medical concepts is not exhausted by his explication

of the theoretical notions of health and disease. There are, he says, a
concept of health and, more particularly, a concept of illness, which have
connotations that extend far beyond the idea of normal (or abnormal)
biological functioning. These concepts refer to the experiential and, in
general, molar aspects of the human being. Moreover, they are highly
evaluative. To say that a person is ill is to say explicitly that he is in a bad
state of health, that he is in a state in which he should not be.
In order to illustrate the relation between the (theoretical) notion of
disease and the notion of illness, Boorse makes a comparison with some
other pairs of concepts:
Disease and illness are related somewhat as are low intelligence and stupidity, or failure to
tell the truth and speaking dishonestly. Sometimes the presumption that intelligence is
desirable will fail, as in a discussion of qualifications for a menial job such as washing dishes
or assembling auto parts. In such a context a person of low intelligence is unlikely to be
described as stupid ... And sometimes the presumption that diseases are undesirable will fail,
as with alcoholic intoxication or mild rubella intentionally contracted. ([13], p. 61).20

It is only when we use the "evaluative" member of the concept pair that
we can infer that the designated state is undesirable.
The concise definition of illness proposed by Boorse is the following: A
disease is an illness only if it is serious enough to be incapacitating, and
therefore is (i) undesirable to its bearer (ii) a title to special treatment and
(iii) a valid excuse for normally criticizable behavior. 21 According to this
definition there can certainly be diseases which are not illnesses. Boorse
allows for "lanthanic" (hidden) diseases (to borrow a term from Alvin
Feinstein), and he allows for early stages of diseases, where these have not
yet turned into illnesses.22
On the other hand, on Boorse's conception all illnesses are diseases. A
person can be ill only if he is diseased. This shows that the BST still rests
on the analytic platform. Illness (on the level of the whole person) is
partially defined in terms of disease, while the concept of disease is not tied
to the state of health of the whole person. Rather, as we have seen, it is
tied entirely to the subnormal functioning of some bodily part.


The BST of Christopher Boorse is a well designed theory. It is clear on

certain crucial points where alternative analytic theories are unclear. It

seems to explain more of the current usage of the health-concepts than

many rival theories. Let us try to summarize its prima facie merits by
jUdging it according to our list of requirements.
(1) The B ST gives an account of some important logical relations
between health-concepts.
The primary concept is the theoretical concept of disease. The theoreti-
cal concept of health is defined in terms of disease. In addition to this a
further pair of concepts is introduced: the concept of illness and the
evaluative concept of health. Both are related to the basic theoretical
concepts. An illness is a disease which is undesirable to its bearer, therefore
it is disvalued. The concept of health (in the evaluative sense) is defined
as a bodily state of affairs lacking illness. Health (in the evaluative sense)
is thus compatible with disease (in the theoretical sense).
The BST does not explicitly cover the whole area of health-concepts.
There is no specific treatment of impairment, injury or disability. The
concept of disease used in the BST seems, however, general enough to cover
impairments and injuries. This is indicated by the fact that a disease is
ontologically characterized as a bodily state. (In order to define a sharper
concept of disease distinguishable from impairment, injury and some
others, it will be suggested in the present essay that disease should be
classified as a process. See the discussion in Chapter four, section 1:
Maladies, and in the Appendix).
(2) The BST clearly indicates the relation between human health
and the health of other living beings.
The theoretical concepts of health and disease in the BST can be applied
over the whole biological spectrum. An analysis of non-human health and
disease can run parallel to the human case. Every species of animal and
plant has its species design. According to these designs every bodily part
or organ has its functions. If all parts or organs fulfill their functions in a
normal way, then the organism is healthy; otherwise it is diseased.
(3) The BST provides a theory applicable to both somatic and
mental health.
Although a theory of mental health is only sketched by Boorse, he
indicates that this subject matter should be given essentially the same
treatment. He claims that there are a number of species-uniform mental
goals, probably however going beyond the basic goals of survival and

reproduction, which could be determined by some psycho-statistical analy-

sis. If all mental faculties contribute normally to the realization of these
goals, then the bearer is mentally healthy. Otherwise he has some mental
(4) The BST gives a clear account of the scientific status of the
By giving purely theoretical (non-evaluative) definitions of the basic
concepts of health and disease the BST clearly allows that these concepts
can be directly used in science. Whether a man is healthy or diseased (in
the theoretical sense) is a completely objective affair. The tasks are to find
the specific goals of the bodily organs and the mental faculties, to calculate
the average contribution of these organs or faculties in the attainment of
the goals, and to study whether a particular organ fulfills this average
requirement. It seems that all these tasks can in principle be undertaken
by an empirical science, for instance medicine.
Two of the problem areas to be covered by a good theory of health are
not explicitly treated in the BST. One such area concerns the relations
between 'health' and some other central humanistic concepts, such as
'morality', 'happiness' and 'ability'. Some of the answers to these questions
are, however, implicit in the theory. There are no conceptual relations,
according to the BST, between, on the one hand, the theoretical notions
of health and disease and, on the other hand, morality, happiness and
ability. That the evaluative notions of health and illness are related to some
of the other humanistic concepts is obvious, but the exact nature of these
relations has not been developed in the BST.
The relation between health and environment has received very little
attention in the BST. This is, we shall shortly argue, one of the gravest
defects of the theory. The BST has been constructed assuming a more or
less fixed environmental background. It has not given consideration to the
possible effects background variation could have on health and disease. We
shall return to this issue in the next section.


On the Concept of a Function

In using the concept of a goal-related function the BST is clearly superior
to other, more simplistic, versions of a biostatistical theory. The B ST does

not identify disease with every kind of deviance from normal values.
Disease is present only when an organ functions at a subnormal level.
The conception of subnormality is important since there are many bodily
and mental deviances which we are inclined to call supernormal and which
could hardly be called diseases. High intelligence is perhaps the most
celebrated example. But if the function of an organ (or a mental faculty)
is not connected to the attainment of a certain known goal, we will be
unable to make judgments concerning sub- or supernormality. For example
the mere observation that an organ, say a gland, "produces" more than
average will not suffice to say that it superproduces. Overproduction can
be countereffective and even prevent the attainment of the gland's particu-
lar goal, or have a generally negative influence on the life of the organism.
In the spirit of the BST we could define supernormal functional ability
in the following way: An organ functions supernormally, if and only if it is
more effective than the statistical average in attaining its particular goal,
provided that this efficiency does not have side effects which are negative
with respect to this goal.
In order to assess the BST we must look more closely at the notion of
a function. Let us consider the following questions:
(i) What is the exact nature of a function? How is the concept of
function to be distinguished fromJunctional ability andJunctional
(ii) What kinds of entities are bearers of functions? Are functions
restricted to what is commonly called organs and mental fa-
culties? Or do all parts of the body, however minute, have
(i) In Boorse's text there are a number of related concepts: function,
functioning, and functional ability. These are not all explicitly defined,
although the context often helps us to understand them. For our discussion
we need more precise definitions of these concepts.
Function can be characterized in the following way:
Organ 0 has a function Fg if, and only if, 0 is directed towards
a goal G (in the factual sense defined on p. 17 ).
For the purpose of the BST we also need the following more complicated

Organ 0 (of a particular individual A) has a function Fg if, and

only if, there is a goal G, and A belongs to a species in the
majority of whose members 0 is directed to G.

These characterizations obviously do not exclude an organ's having

more than one function.
A function is then an abstract relation between an organ (or some other
part of the body) and a goal. Now, the possession of such a function says
very little about the actual power or performance of the organ in question.
That an individual organ has a function does not entail that it has the ability
to fulfill this function.
To describe the power of an organ we need the concept of functional
ability. Consider:
o has the functional ability Fga if, and only if, 0 can fulfill its
function Fg •

Such a characterization is, however, far from complete as long as the

environmental background is not specified. Most organs cannot fulfill their
functions in all kinds of environments, whether external or internal. In fact,
all talk of diseases, in the sense of the BST, presupposes that there are
situations in which an organ loses its functional ability. Thus, we must
include the notion of circumstances.
o has the functional ability Fga in circumstances C if, and only
if, 0 can fulfill Fg in C.

The notion offulfillment requires a comment. It could mean that the work
of 0 is a causally sufficient condition in C for the realization of G. Now,
the locution normally used in Boorse's texts is that an organ contributes to
the realization of a goal (presumably one of the ultimate goals of survival
or reproduction). This fact need not raise any theoretical problem. If an
organ contributes to the realization of an ultimate goal G, then it must, in
the circumstances, be causally sufficient for some subgoal of G. That 0
fulfills its function could then mean that it provides sufficient conditions for
a subgoal of G.
Could we then not choose the subgoal as the goal of the organ? As we
shall see in the following discussion there is a difficulty connected with such
a choice. It may be that, depending on the circumstances, the organ aims
for slightly different subgoals in its causal contribution to the attaining of

the ultimate goal. The goal of the organ (understood as the kind of state
for which the organ is a causally sufficient agent) must then be described
as a set of goals related to a corresponding set of circumstances.
That a particular organ, or some other part of the body, has a function
or a functional ability says very little about the actual work performed by
the organ, i.e. its functional performance. The sort of performance required
depends on the organ's place in the internal environment, in particular on
what the other organs do and what happens to them. But it also depends
on the external situation, that is, on the external pressures placed on the
It is important to notice that the intensity of functional performance can
also vary greatly in what we might call standard circumstances. Some
organs have continuously to perform near their maximum in order for their
goals to be realized. The heart must pump continuously in order to fulfill
its function; the lungs must exchange gases continuously. But, in contrast,
the stomach and the thigh muscles are used only intermittently. Some parts
of the body are used in a goal-directed way only very rarely. The adipose
tissue of a normal man will release its triglycerides into the metabolism only
when there is a significantly low input (or low uptake) of nutrients into the
The abstract message here is the following: Certain bodily goals (or
subgoals) are such that they require for their achievement or maintenance
continuous hard work by the bodily parts responsible for these goals. Other
goals can be achieved or maintained by occasional measures. Still other
goals only very rarely require any work by a particular part of the body.
This observation has some impact on our understanding of the locution
"statistically typical efficiency". It is rather easy to understand what is
meant by statistically typical efficiency in the case of the heart and the
lungs. It means the typical continuous cardiorespiratory work (say a pulse
rate between 50 and 70 and a respiratory frequency between 10 and 30).
But some other bodily parts, like the stomach, the muscles or the adipose
tissue need not, at a particular moment, perform at all, at least not in
relation to the particular goals that we are studying. (It is not denied here
that in every living cell there is some activity going on.) The statistically
typical efficiency may manifest itself in rest.
The question is then: how could this organ be differentiated from one
which does not fulfill its goal? Such an organ could presumably also be in
a state of non-activity. The obvious and plausible answer is that we can
detect the "bad" organ by placing the organism in such a situation that

goal-achievement (or goal-maintenance) requires real work from the organ

in question. We could introduce food into the stomach; we could try to
make a person walk, or we could starve the organism in order to activate
the adipose tissues.
Our examples show, however, that "statistically normal contribution to
a certain goal" can in no sense be equated with statistically normal
performance when the performance is considered irrespective of the
internal or external situation. Observe that this also holds for those organs
which have to work continuously. The character of their work changes in
relation to the situation. After a short spell of quick running the pulse rate
of a healthy man may exceed 150 and the respiration rate 100. These
statistically abnormal values may indicate a "statistically normal contri-
bution" to the goal of keeping a constant level of oxygen in the system.These
observations show in their turn that the determination of health, on this
strict biological platform, requires a study of the organism in a great variety
of situations.
(ii) Given these preliminaries we shall now make some critical remarks
on the problem of identifying biological functions. What parts of the body
are bearers of functions? How many functions are there and how do we
determine which they are?
Brief reflection suggests that these problems must to some extent be
solved on the basis of an arbitrary decision. Goals and functions in the
human organism can be identified all the way from the level of "gross
functions" to the microscopic level - the cells or even the genes. The latter
goals could be viewed as subgoals of the former, which in their turn,
according to the BST, would be subgoals of the ultimate goals of survival
and reproduction.
But can we take all these possible goals and functions into account in
a characterization of human health? Consider first the idea that the
functions of the human body are the "gross functions", the functions of
organs such as the brain, heart, lungs, liver and kidneys. The BST-defini-
tion of disease should then be interpreted in the following way: diseases are
internal states which interfere with the functions of the larger organs.
Let us study the consequences of this interpretation in an abstract way.
We assume that organ 0 functions well in a particular situation, it
maintains or reaches its goal. Organ 0 contains a number of parts 0], ... ,
° JO' The parts have their subgoals, but these are not considered so long as

o performs its function. Assume now that 0], ... , Os are severely damaged
or completely removed. This fact mayor may not result in a change on the

level of O's activity. Experience shows that there need not be a negative
result on the level of the whole organ. There may be two explanations of
this. 0 may have an overcapacity with respect to its function; or the
remaining parts may be able to compensate for the loss through increased
An example of this phenomenon is the following. A liver has been
damaged by a long period of excessive alcohol consumption, and a
substantial portion of it is no longer functioning. Still, the liver as a whole
can fulfill its main functions, such as glycogen synthesis and purification of
the blood, in the required way. Thus, on our present interpretation of the
BST, the damage to the organ is not an indication of disease.
The damage may make itself felt in certain kinds of situations, however,
where the organ is put under stress. But we certainly know of people who
have organs which function adequately in spite of considerable damage.
Thus we cannot ascribe diseases to such people, for instance the disease
of liver cirrhosis, until this phenomenon manifests itself in a disturbance
of the organic function.
Consider now the second extreme alternative. We move very far down
on the hierarchy of biological levels. Assume that we stop at the level of
cells. Assume also (which is at present quite utopian) that we have a reliable
mapping of the functions of all individual cells. We shall then say that health
obtains in the organism when all cells fulfill their functions. Otherwise there
is at least one disease.
The consequences of such a view are extremely counterintuitive. The
existence of a single cell not fulfilling its functions is sufficient for the
existence of disease and thereby the non-health of the whole person. We
all have a great number of cells which are dying or malfunctioning. Hence,
we would all be ill.
The BST is not satisfactory under either interpretation. It seems then
that we must aim at finding an intermediary level offunction analysis which
is subtle enough to detect "obvious but silent malfunctioning", but not so
subtle as to include the odd single cell. The BST has not given an account
of how this is to be done. An answer in the spirit of the BST might be the
following. For the assessment of health we should scrutinize such func-
tional abilities the suppression of which endangers survival or the ability
to reproduce. This means that we should try to find functional abilities
which are more or less irreplaceable in their causal contribution to survival
and reproduction. This would presumably put us on a fairly high level of
biological integration, roughly on the level of gross functions.

But how could we then face the problem of the "obvious malfunctioning"
of parts of otherwise well functioning organs? A state of affairs which would
disturb a particular functional ability in most human beings, need not
disturb it in a certain individual. Still, we might want to say that that
individual has a particular disease. In order to handle this we must change
the categorical characterization of disease within the BST ("a disease is a
state which interferes") to something like the following: a disease is a state
which tends to interfere (or which with a high degree of probability inter-
feres) with an organ's functional ability. (This answer is in line with our own
positive account of the concept of disease, see Chapter four, se(!tion 1:
So far we have treated the BST in a fairly benevolent way. Our
observations have forced us to make some definitions more precise than
they are in Boorse's texts. We have also suggested some minor amend-
ments which we suspect to be in the spirit of the BST. The criticisms
advanced so far do not suggest a radically different analysis of health. We
turn now to what we consider to be a major defect of the BST.

On the Relations Between Environment, Activity and Health

It is significant that the BST says very little about the dynamic aspects of
life - for instance about changes in the environment and changes in the
activity of the body - in its characterization of health. In leaving out these
matters from the discussion the theory ignores a number of intricate prob-
lems which strike at the heart of the BST-conception. 23
Consider the following phenomena. (1) A man leaves his warm house on
an extremely cold day; he is not well dressed, and he immediately starts
to feel cold. Feeling cold involves a complicated physiological process. A
main ingredient is the adaptation of the epidermic capillaries to the external
situation. Instead ofletting blood through in order to maintain the tempera-
ture and the oxygen level of the epidermic cells, the capillaries close so that
the most vital parts of the body can maintain their regular temperature.
Another ingredient in the physiological reaction to feeling cold is the
continuous work of certain muscle systems - the shivering phenomenon -
in order to create a higher internal temperature.
(2) A man wakes up, gets out of bed and starts his morning jogging
activities. After a while his pulse and respiration are increased; his muscles
produce lactic acid; his body temperature rises and he starts sweating.

These two phenomena are considered to be the reactions of a healthy

body to changed circumstances, or adaptations to new environments. They
could certainly be expressed in the language of the BST. A strain put on
the body, from the environment or from strenuous activity, requires harder
work or perhaps a different sort of work from the organs in order for their
functions to be performed. But as long as they give their species-typical
contribution to the ultimate goals, the various organs should be considered
to be in order.
Let us now repeat that the circumstances, by changing the work of the
body, in a way change the go~l-structure of the body. They do not, presum-
ably, change the ultimate goals of survival and reproduction. But they can
in radical ways change the subgoals which the organs must attain in order
to reach the ultimate goals. This is more clearly seen as one descends the
hierarchy. But we could certainly also say that the heart and the lungs have
slightly different subgoals; their new subgoals could for instance be ex-
pressed in terms of a certain intensity of activity. A description of the
subgoals of an organ should thus be made in more complicated and relative
terms: 0 has the subgoals Gt> ... , Gn given the circumstances C)' ... , Cn,
Consider now certain extreme circumstances. We shall study a situation
in which certain external factors apparently damage parts of the human
body. We shall, in particular, study the situation where a number of mi-
crobes invade a tissue, i.e. the paradigm case of an infection.
Consider the main steps in such an infectious process. A number of
pathogenic agents enter, say, the throat. They start producing toxins, which
immediately destroy a great number of cells on the mucous membranes.
The body quickly reacts to this attack. There is a great concentration of
blood at the focal points of the infection; the body temperature rises;
certain tissues create antibodies against the viruses and the pathogenic
toxin. As a result the toxin is gradually neutralised and the microbes killed.
This outline of a description of a process of infection is at the same time
a description of a species-typical reaction to a serious attack on the body.
It may very well be that every part of the body has, during this process,
given its species-typical contribution to the ultimate goals of survival and
reproduction. In fact, the infectious disease can be seen as the species-typi-
cal reaction to the circumstance of a certain microbial invasion.
But this, then, becomes paradoxical. A typical disease can be seen, on
the BST, as a species-typical reaction, i.e. as a healthy response to a
difficult environment.

This observation does not imply that we could never identify diseases
using the BST. An individual can very well react in a way which is not
species-typical. He may become infected where other members of the
species would not. His response might also bear no relation to the ultimate
goals of survival and reproduction. Cancer seems to be a cluster of diseases,
in which the defence system of the body has broken down; certain species-
typical functional abilities no longer exist.
However, the fact that certain instances of diseases can be viewed as
species-typical responses to a particular strain creates a problem for the BST
view of disease. What has gone wrong? Can we suggest an amendment?
Let us first ask: why do we, in ordinary language and according to
medical custom, view infections as diseases? The answer is simple: in-
fections are painful; they cause fever with fatigue as a frequent conse-
quence; both fever and fatigue disable us; we are unable to do the things
we normally want to do.
Facts such as these are involved in Boorse's characterization of illness.
But for reasons of principle he cannot use them in his definition of disease.
(Boorse's concept of illness presupposes his concept of disease. See our
presentation in Chapter two, section 3.) But perhaps the viewing of in-
fections as illnesses can help us give an alternative account within the
framework of the BST. If there is illness there ought to be some subnormal
function; some organ ought not contribute in its usual way to the ultimate
goals of a particular human being. Perhaps, then, the situation should be
described as follows: an infection involves a species-normal response to a
certain kind of external attack on the body. But the infection may depress
the functional ability of other organs which are not primarily affected by
it. If so, the infection would be a disease according to the BST.
(Against this one could not argue as follows: the functional ability of the
non-infected organs can be species-typical given the difficult internal envi-
ronment; hence, there would be no disease. This move would not do,
according to the BST. A disease is precisely such an internal environment
which depresses the functional ability of some organ in relation to the
individual's external environment.)
Consider this defence.
(i) The BST argument here presupposes that diseases can be clearly
anatomically isolated. Some organs and some functions are involved in the
disease. Others are external to the disease but are affected by it. This is
certainly a plausible idea with some diseases. But is it plausible with all
diseases? What about influenza? Can we distinguish there between those

organs and tissues which are involved in the disease, and those which are
only affected by it?
(ii) The BST-argument presupposes that illness as generally understood
must be caused by some subnormal function, and moreover that the sub-
normal function which is responsible for illness must be subnormal relative
to the ultimate goals of survival and reproduction.
Let us study this important presupposition in some detail. Consider first
the pain which arises from the site of the infection. In what sense does this
pain involve subnormal functional ability? What does the pain consist or!
Its organic basis is a chemical irritation, caused by pathogenic toxins, of
certain pain-receptors. This may involve the local destruction of some
neural cells. But the "gross function" of the pain consists in the sending of
a message from the damaged locus to the brain. And does this message
indicate any subnormal functional ability? If the organ giving rise to pain
is functioning at a subnormal level, is it subnormal in relation to the goal
of survival? The answer to the latter question seems clearly to be in the
negative. If moderate pain can be related at all to survival it seems rather
to be contributory to survival. The pain can induce the individual to take
steps to prevent a dangerous development of the disease.
Consider, secondly, the disability which may result from the pain, or
from other sources, for instance from high fever. Should we say that the
organic work responsible for this disability necessarily constitutes subnor-
mal functional ability vis-a-vis the goal of survival? Is, for instance, fever
always a subnormal contribution to survival? Our evidence clearly says no.
Fever may efficiently support the defensive mechanisms of the body in
exterminating the invading microbes. Therefore, it is an adequate response,
precisely with respect to the goal of survival.
But what if we should say that the disability itself constitutes subnormal
functional ability in relation to the goal of survival? This will not do. The
disability of the whole person is not a subnormal functional ability of an
organ. Thus the general disability cannot be a candidate for disease within
the framework of the B ST.
Illness, in the ordinary sense of the word, implying pain or disability, may
be due to species-typical reactions involving the execution of normal
functional ability given a certain set of circumstances. Thus illness, as
normally understood, may be due to other things than disease as conceived
on the BST.
This observation points to what is perhaps the most serious weakness
of the BST. The BST has restricted itself, at least so far as somatic health

is concerned, to the two single goals of survival and reproduction. But it

seems clear that there are other goals involved in health, in particular the
goal of having particular abilities. As will be shown in Chapter three, this
goal cannot be recognized solely on the basis of a biological inspection of
the functional network of the bodily organs?4

Concluding Remarks

The BST constitutes an attempt to reconstruct discourse on health in

purely biological and biostatistical terms. It defines health analytically in
terms of the absence of disease. The concept of disease, in turn, is defined
as a bodily state which depresses the (statistically) normal functioning of
a part of the body. The BST claims that there is, and presumably should
be, a purely biological discourse. In order to handle certain ordinary
intuitions about 'health' and 'disease', the theory also gives an account of
'health' and 'illness'. This pair of concepts is employed in an evaluative
discourse related to human ability and well-being.
The BST thus acknowledges two different discourses on health and
disease, one theoretical and descriptive, the other practical and evaluative.
It is to be observed, however, that the evaluative notion of illness is partly
defined in terms of the theoretical notion of disease. Thus, ultimately,
according to the BST, the evaluative notions of health and illness also
belong to an analytical perspective in the sense defined on p. 12.
The present analysis has shown that this account is not satisfactory. In
order to identify a bodily state as a disease, also in the allegedly theoretical
sense, one cannot merely rely on subnormal functioning vis-a-vis survival
and reproduction. Were one to do this, many paradigm cases of diseases,
for instance infectious diseases, would not qualify as such. We have argued
in this chapter that infectious diseases as well as some other "defensive"
diseases can be identified only if their relation to illness and disability is
taken into account.
This observation indicates that the primary concept, at least in these
cases, is not disease but illness or disability. These concepts belong to the
holistic and evaluative discourse. This speaks in favour of a holistic
approach to the characterization of the various health-concepts.
The following chapter will be devoted to constructing a theory within the
holistic perspective on health.



We shall now present a very different view of health. We shall consider a

human being as a socially integrated agent who performs a great number
of daily activities and is involved in many personal and institutional
From this perspective the essential criterion of health is to be found on
the conceptual level of psychology and sociology. An everyday formulation
of this is that a person is healthy ifhe feels well and can function in his social
This idea is certainly not far-fetched. We intuitively connect health with
well-being and ability, and illness with suffering and disability, and view
these features as their essential characteristics.
This holistic view also finds support in the history of medical ideas. In
fact even Galen defined health as a "state in which we neither suffer from
evil or are prevented from the functions of daily life".25 Many latter day
theoreticians make similar characterizations. The French philosopher and
physician Canguilhem says in his book On the Normal and the Pathological
[20] that health "is a feeling of insurance in life"; "illness or the
'pathological' is the direct concrete feeling of suffering and impotence, the
feeling oflife gone wrong"([20], p. 118 and p. 77). Similarly, the American
sociologist Talcott Parsons says "Health may be defined as the state of
optimum capacity of an individual for the effective performance of the roles
and tasks for which he has been socialized" ([97]. p. 117).26
In most such characterizations two kinds of phenomena are mentioned:
first, the subjective phenomenon of a certain kind of feeling, of ease or
well-being in the case of health, and of pain or suffering in the case of illness;
second, the phenomenon of ability or disability, the former an indication
of health, the latter, of illness.
These two kinds of phenomena are in many ways interconnected. There
is first an empirical, causal, connection. A feeling of ease or well-being
contributes causally to the ability of its bearer. A feeling of pain or suffering


may directly cause some degree of disability. Conversely, a subject's

perception of his ability or disability greatly influences his emotional state.
Some would argue that the relation between the two kinds of phenomena
is even stronger, i.e. that there are conceptual links between, on the one
hand, a feeling of well-being and ability and, on the other hand, suffering
and disability. According to this idea, being in great pain, for instance,
partly means that one is disabled. Some degree of disability is here a
necessary condition for the presence of pain, so that if a person's ability is
not affected, he can be said not to be in great pain.27
The assumption of a conceptual relation between pain and disability will
be accepted in the present analysis: a man cannot experience great pain or
suffering without evincing some degree of disability.But a man may have
a disability, and even be generally disabled, without experiencing pain or
suffering. There are some paradigm cases of illness where pain and
suffering are absent. One obvious case is that of coma. Another is present
in certain mental disabilities and illnesses. When a patient cannot reflect
over his own situation, then his disabilities need not have suffering as a
consequence. In short, wherever there is great pain or suffering there is
disability, but the converse is not true.
These preliminary observations indicate that the concept of disability
has a much more central place in the characterization of illness than the
corresponding concepts of pain and suffering. If only one of these
characteristics is essential to the notion of illness, then disability is the
prime candidate. This is our main reason for founding the subsequent
analysis on the concepts of ability and disability.
To this a pragmatic argument should be added. 'Ability' has the
advantage over 'suffering' and 'pain' of being more useful as a defining
criterion for scientific and practical purposes. Ability can to a greater extent
be intersubjectively established, and in certain cases even be measured on
an ordinal scale.
This analysis, however, does not deny the extreme importance of pain
and suffering - as experiences and not just as causes of disability - in the
phenomenon of illness. An adequate description of a particular illness must
often include a description of the suffering involved.
The task of this essay, however, is not to present guidelines for making
detailed descriptions of particular instances of health and illness. The task
is rather to find criteria which, in the ideal case, can be used in general
definitions of the concepts of health and illness. For this purpose 'ability'

and 'disability' are better candidates than 'pain' and 'suffering' and their


On the Stratification of Actions

Not all bodily movements are actions. Actions are such movements or
behaviour which are under our control and influenced by our will. In
philosophical terminology it is often said that actions constitute intentional
behaviour. To shake one's hand or to nod one's head is an action only if
the agent intends to do so.
Actions are normally not performed merely for their own sake. They are
typically parts of an agent's plan to reach certain goals, instruments in the
process of his forming his life.
This aspect of actions, the fact that they are to a great extent forward-
looking and goal-related, is mirrored in our way of conceptualizing many
individual actions. It is mirrored in what will here be called the stratification
of actions. 28

(i) The concepts of basic action and action generation.29

In order to understand these notions let us consider some simple

examples. We are all familiar with the idea that one performs one action
by performing another. One travels to New York by taking a boat. One
greets someone by waving one's hand. This relation of doing-by can be
extended, and sometimes be made very long. Consider the following
The revolutionary overthrew the despotic regime in his country
by killing the tyrant
by shooting him with a gun
by pulling the trigger
by moving his finger.
The question can now be put: is this series of by-doing indefinite? Can
one always go on finding further actions by which one performs other
actions? The answer to this is no. There is always some action which
initiates the chain; this is the so-called basic action. A basic action is an
action which is not performed by the performance of some other action. In

the standard case the basic action involves just the (intentional) movement
of a part of the body. (Under special circumstances the basic action can
also be constituted by omitting to move a part of one's body).
The chain of actions indicated here may be said to be generated by the
basic action. The basic action of moving one's finger generates the action
of pulling the trigger, which in its turn generates the action of shooting the
tyrant, etc.
Is there, then, a last member of the chain of generated actions? Now this
question cannot be answered a priori. It seems one cannot draw a limit for
conceptual reasons. It depends in the single case on how much is included
in the agent's intention. If the coup d'etat actually was the final purpose of
the revolutionary (which may be a plausible hypothesis), then it is the last
element of the generated chain of actions in question.
What then is the nature of the process of generation itself? What does
the expression "by doing" signify? This has been debated and quite
thoroughly analysed in recent action theory. One of the most influential
discussions is in Goldman [47].He distinguishes between four kinds of
"level generation" as he calls it: (a) causal, (b) conventional, (c) simple, and
(d) augmentative?O
The first two kinds are the most important and exhaust the vast majority
of existing types of action-generation. Let us consider their nature in more
First, causal, generation. Here there is a causal relation, but it does not
obtain between the actions in the chain, but rather between an action of
a lower level in the chain and the endstate of an action of a higher level. So
in the case of killing the tyrant by shooting him, it is improper to say that
the shooting is a cause of the killing; the shooting is a cause of the fact that
the tyrant is dead, i.e. the ends tate of the action of killing.
To summarize: when it is true to say that a person F-s by G-ing, and the
generation involved is causal, then the endstate of F-ing is caused by the
In the case of conventional generation, the generation is effected by
conventional stipulation. There is a socially determined rule, which says
that when a certain action occurs in a particular context it should count as
some other action. Examples: lifting one's hat when meeting another
person counts as greeting him; making certain laryngal noises in appropriate
sequences can count as performing actions of speaking; signing a document
in the presence of a bank-official can count as numerous actions, for
example: making an agreement, receiving a loan, or buying a house; the

moving of a piece of wood can, given proper circumstances, count as a

move in a game of chess.
In these cases the relation (of convention) can be said to hold between
the actions themselves: a G-ing in a certain context is an F-ing. Another way
of putting it is to say that G-ing, given proper circumstances, generates the
conventional result, which is the endstate of F-ing.
Now, given the concepts of basic action and generated action we see that
human beings can do things whose results and endstates vary from being
very close to the human being to being extremely distant.
Although it is true that we say that a person does or performs all the
elements contained in a chain of generated actions, there is an interesting
difference between the levels generated. (This is of particular interest for
our main task in the essay).
The levels which are closest to the agent are more under his control than
the levels farther away. Moving his finger is more under the revolutionary's
control than his overthrowing the regime. There are many things that can
go wrong in the causal sequence from the basic action to the endstate,
whereas only a limited number of things can prevent the basic action from
being performed. Hence , very much of what we do that entails actions
beyond basic actions is dependent upon the course of the world outside
When we say that we intend to open a window, we can never know with
certainty that we shall actually succeed in doing so. Our success is
dependent upon certain states of affairs in the world (such as our being able
to manipulate a standard locking mechanism on the window) which we
normally correctly believe to obtain. Thus when we claim that there is a
causal sequence from the basic action to the endstates of the ensuing
actions in the chain, this does not mean that the basic action is a sufficient
cause of these endstates. There must be other factors in the external world
which contribute to the coming about of the endstate. That this is so does
not entail any peculiarity in our use of the term "cause" here. In practically
all uses of the term "cause", be it in ordinary language or science, it signifies
only part of a sufficient condition. 31
There is an analogous consideration to be made in the case of con-
ventional generation. The basic action itself is rarely sufficient to be con-
ventionally counted as some other action, but requires certain external
circumstances in order to be so.
Consider now some further features of action-generation. We say that
the basic action contributes causally or conventionally to a certain result.

The question now is: is the basic action also necessary for these results? The
answer to this is that it varies.
In the case of ordinary causation a cause (including a basic action) is
rarely strictly necessary for its effect. There are often alternative ways of
achieving a certain end. In order to realize a revolution the death of the
tyrant need not be necessary; in order to kill the tyrant it is not necessary
to shoot him; and so on.
In the case of human action it is not only important to determine whether
a basic action is necessary as a matter of fact. In many contexts, for
instance in explanation, it is of greater importance to know whether the
agent himself considers an action necessary for a certain result. If he does,
for him there is only one way to bring about that result.
In the case of conventional generation, a relation of necessity between
a basic action and its effects need not be a rarity. The reason is that it can
always be stipulated. We can always stipulate that the only way in which
A can make a will is by signing a particular kind of document. We can
stipulate that the only way that Sweden can declare war is by the prime
minister's putting his signature to a government decision.

(ii) The concepts of accomplishment and activity, action-chain and

action-sequence, opportunity for action. 32

In the previous section we saw how a basic action can generate actions
of higher levels. In the following such generated actions will be called
accomplishments. The chain of actions related by causal or conventional
generation, from the basic action to the final accomplishment, will be called
an action-chain.
Many of our ordinary actions turn out to be accomplishments within this
theoretical framework. Further reflection shows, however, that most of
them entail more than just a basic action plus the course of nature or
convention. This simple structure presupposes that there is an opportunity
for action. Consider the case of the revolutionary. A precondition for his
creating a revolution in the way depicted by the action-chain is that he puts
himself in a position to shoot the tyrant. He must obtain a gun, and
transport himself to a place within shooting distance of the tyrant. The
accomplishment of creating a revolution by killing the tyrant, then, pre-
supposes previous actions and these actions are themselves normally
accomplishments. Such is the case with "obtaining a gun" and "travelling
to the tyrant's home". These previous accomplishments can, however, all

be seen to be parts of the complete project that we designate "creating a

revolution". The latter is, therefore, a sequence of actions together con-
stituting what we shall call an activity.
Activities can differ in their nature, partly depending on whether or not
they are strictly goal-directed. Some activities are strictly ordered in the
following way: accomplishment a provides the opportunity for accomplish-
ment b; and b, in its turn, provides the opportunity for c. A simple example
illustrating this case is the activity of building a wall of bricks: laying the
first row of bricks provides the opportunity for laying the second.
Other activities are not ordered in this strict sense of one providing the
opportunity for the other. A child's playing with balls may be a loose
activity of this other kind. His playing with one ball does not normally
provide an opportunity for his playing with a second.
In one respect there is an important analogy between accomplishments
and activities. Just as many accomplishments can be performed by different
means, so can many activities. In other words, an accomplishment is not,
in general, defined by a particular action-chain, nor is an activity in general
defined by a particular action-sequence.
Consider now the notion of an opportunity. We have said that all accom-
plishments require opportunities. This truth is a consequence of the fact
that basic actions are not sufficient conditions for the endstates of the
actions they generate. The agent may have to be at a certain place, at a
certain time, and so on, and the causally relevant conditions in the world
may have to be normal.
The idea of an opportunity is, however, equally important in the case of
conventional generation. Here it may also be easier to describe the
opportunity exactly. This has to do with the fact that conventional genera-
tion requires for its existence a rule (often explicitly formulated) which the
agent must have learnt in order for his action to result in a conventional
accomplishment. In the rules of a game, for instance, it is explicitly stated
under what circumstances a particular move can be made (i.e. what
constitutes an opportunity for the move in question). The civil law contains,
among other things, an enormous catalogue of opportunity descriptions, i.e.
descriptions of situations in which various administrative and legal actions
can be performed. 33
On the Concepts of Practical Possibility, Opportunity and Ability
What does it mean to say that a human being can perform an action? The
term "can" i~ highly ambiguous; it admits of such diverse interpretations

as (a) logical possibility (nine can be divided by three) (b) epistemic possi-
bility (for all I know, he can be thirty years of age) (c) physical possibility
(men cannot survive without oxygen) (d) ability (John can learn Russian)
(e) authority (this university can issue Ph.D degrees) (0 opportunity (Peter
can cross the road now). 34
Ifwe limit ourselves to human beings and their relations to actions, we
could take at least the last three interpretations into account; when we say
that A can perform F we might mean either: A has the ability to perform
F, A has the authority to perform F, or A has the opportunity to perform
F - or some combinations of these. For purposes of the present discussion
it is particularly important to distinguish between a person's ability and his
opportunity to perform F. (We shall in the following pay rather little attention
to the notion of authority. First, it is of only marginal interest in the
philosophy of health. And second, for many theoretical purposes authority
can be viewed as a conventional circumstance, as a kind of opportunity.)
When a person has both the ability and the opportunity (including
authority) to perform a particular action, then he can perform it in a strong
sense of the word. This strong sense of "can" will here be called practical
How should these concepts be defined? There are at least two main ways
of characterizing concepts of possibility. According to the first - the more
traditional - possibility-concepts are defined conditionally, in analogy to
dispositional properties. For example, "It is practically possible for A to
swim", means that, if A tries to swim, then A succeds in swimming. Like-
wise, "A has the ability to swim", means that, if A tries to swim, and there
is an opportunity for him, then A succeds in swimming. On the second way,
possibility concepts are defined in terms of possible-world semantics. "A
has the ability to swim", then simply means: in some circumstances A
Both of these suggestions have their merits and shortcomings. For
substantial discussions about them, see [6] and [64].35 In the present
context we shall not take a stand on this issue. We shall not propose a
definition of practical possibility or ability. For the purposes of this essay
it suffices to provide a test for the application of the two concepts. This test
is founded on the traditional analysis of practical possibility. We can
ascertain whether it is practically possible for a person to F, by letting him
try to F. From a test where A tries to F and succeeds in F-ing, we shall
conclude that it is practically possible for A to F.36

Consider now the conditions for a person's practical possiblity to act.

These conditions differ, of course, depending on the type of action and
whether we are talking about ability or opportunity. We shall first discuss
conditions for ability and such conditions as hold for all kinds of action, in
particular for basic actions.
We shall first take into account the fact that actions are (normally)
intended. A general ability to perform an action thus presupposes an ability
to form the intention to perform the action in question. This indicates that
mental preparation is necessary for a person to act intentionally.
It is impossible for A to intend to perform F if he is completely unaware
of F. This kind of situation is not so common with basic actions as with
many generated actions. (See the discussion below.) Still, the point is also
relevant for basic actions. People are not aware of all the possible move-
ments they can make with their limbs. Hence, there are certain movements
they will never intend to make.
Also, there are mental factors which may prevent a person from per-
forming a particular action. He may find an action so revolting that he
would never intend to perform it. This again is more common with certain
complex actions; most unethical actions, for instance, are complex actions.
A further interesting case of mental prevention is the one where an agent
is continually convinced that he is not physically able to perform the action
in question. If this is so he will never form the intention to perform it.
Thus, factors such as ignorance about an action, revulsion by it, or
conviction of one's physical inability, will prevent the realization of the first
stage in acting, intending to act and setting about to act. These factors,
which are not generally acknowledged in the context of ability, have parti-
cular importance for the theory of health. Many types of mental diseases
can be located in defects among the antecedents of intending.
For the realization of the second stage of action, its actual performance
and success, there are certain obvious requirements. With basic actions
these requirements all concern the biological set-up of the agent. This
set-up can be divided into various aspects. One is that the agent must not
be paralysed. Other aspects involve such things at that the the muscle
tissues be sufficiently developed, the joints function properly, and so on.
The condition of opportunity is easy to characterize in the case of basic
actions. Here opportunity consists merely in the non-existence of external
preventive factors. A has the opportunity to raise his hand if nothing physi-
cally prevents him from doing so.

These, then, are the background conditions for the practical possibility
of performing a basic action. Consider now the complex actions,
accomplishments and activities.
By definition, the performance of an accomplishment requires the
performance of some basic action. A second obvious requirement is that
the accomplishment can in fact be generated. (Note that this may depend
partly on the agent in question.) A third requirement is that the agent know
that there is a situation which constitutes the opportunity to generate the
accomplishment in question. This entails either that he has some causal
knowledge, i.e. knows what happens, given a particular basic action in a
particular situation, or that he has some conventional knowledge, i.e.
knows of a particular action-generating rule and what it says about the
required circumstances. (In some cases both kinds of knowledge may be
presupposed ).
We shall now collect these requirements (together with the ones noted
above) into one schema. The following symbols will be used: Ace for
accomplishment, Act for activity, B for basic action, 0 for opportunity and
S for action-sequence.

It is practically possible for A to perform an accomplishment Ace if, and

only if,
(i) there is at least one action-chain, B ... Ace, given an opportunity
(ii) A believes that (i), feels no revulsion against performing Ace,
and believes that he is physically able to perform Ace;
(iii) it is practically possible for A to perform B;
(iv) 0 is present;
(v) A identifies O.
Let us now turn to the case of activities. A basic requirement is that an
activity itself be performable by the agent in question. Some activities are
directed towards a goal which has to be reached in order for the activity
to be completed. Climbing Mount Everest is an activity of this kind.
The practical possibility of carrying out an activity must involve the
practical possibility of performing each action (normally accomplishment)
which is a member of some action-sequence constituting the activity. (As
we have said, there are often alternative ways of performing the activity.)

But the practical possibility of performing each member of a set of

accomplishments does not suffice for the performance of the activity. Again
the agent must have a considerable amount of knowledge. We can summar-
ize the items that he must know:
(i) He must be aware of at least one action-sequence constituting
the activity.
(ii) He must know what constitutes the opportunity for all mem-
bers of this sequence.
(iii) He must know how these opportunities are to be identified.

A further important element in the performance of some activities is the

element of coordination. It is sometimes required that one can not only
perform each of the basic actions or accomplishments involved in the
activity, but also coordinate them into a sequence with special properties
(for instance, properties of time, force or elegance). For example, to pro-
duce a melody it is clearly not enough to produce the right notes, one at
a time. The components of the tune must be coordinated in a particular way
for the result to be music.
Let us express the requirements for performing an activity in a slightly
more formal manner:
It is practically possible for A to perform an activity Act if, and
only if,
(i) there is at least one action-sequence S: Acc1 ... Accn constituting
(ii) A believes that (i), feels no revulsion against performing Act,
and believes that he is physically able to perform Act;
(iii) it is practically possible for A to perform each of Acc1 .. Accn
given their respective opportunities;
(iv) A is able to coordinate each of Acc •... Accn in the appropriate
(v) the required opportunities actually arise;
(vi) A identifies these opportunities.

That a person lacks the practical possibility of performing a co~plex

action can thus be due to any of a variety of factors. Let us express this,
as it applies to the case of accomplishments, in our formal language: (i)

There is no action-chain X ... Ace, where X represents some basic action;

or (ii) A is not aware of any action-chain X ... Ace. On the supposition that
there is an action-chain B ... Ace known to A we have the following possi-
bilities: (iii) A feels revulsion against Ace or does not believe that he is
physically able to perform Ace; or (iv) it is not practically possible for A to
perform B; or (v) there is no opportunity for Ace by the performance of B;
or (vi) A does not identify any opportunity for Ace.
Let us illustrate these reasons for non-ability by considering the accom-
plishment of starting a machine. A may be unable to start such a machine
for the reason that (i) the starting mechanism does not respond to pressure
on the starting button; or (ii) A does not know of any way of starting the
machine; or (iii) A feels revulsion against starting the machine or does not
believe that he is physically able to start the machine; or (iv) A has broken
his hand and cannot put his finger on the starting button; or (v) A is not
in the position to press the button - he may be in a different room and hence
have no opportunity to start the machine; or (vi) although A is in the
proximate vicinity of the machine he cannot find the starting button.
The ways in which an agent may be unable to perform an activity are then
easy to see. We shall not elaborate further on them here. 37


We have collected a great number of factors necessary for the practical
possibility of somebody's performing an action F. The list of factors neces-
sary for basic actions is already long, and it becomes quite complicated
when we add accomplishments and activities. Thus, when it is not possible
for a person to perform an action, this can be due to any of a number of
Most of these reasons, but not all of them, are relevant in the context
of health. In determining whether a person is healthy or not our primary
task is not to answer the general question whether it is, in fact, practically
possible for the person to perform a certain set of actions. What we want
to know is, more specifically, whether he is able to perform the actions. This
means roughly: are his internal - bodily and mental - resources for per-
forming the actions in question sufficient?
Thus our interest in the undermining of action will from now on focus
on disabling factors - factors which are internal to the agent's body or mind.
Let us rehearse which types of factors these are. We shall consider the
different action-categories in turn.

Basic actions: Lack of bodily or mental integrity; Ignorance of the

action-type; Beliefin the impossibility of performing the action; Revulsion
against the action-type. Accomplishments: Lack of bodily or mental inte-
grity for performing the involved basic action; Ignorance of the action-
chain constituting the accomplishment; Belief in the impossibility of per-
forming the accomplishment; Revulsion against the accomplishment; Non-
identification of opportunity for action. Activities: Lack of bodily or mental
integrity for performing the involved basic actions in the way required;
Ignorance of the action-sequence constituting the activity; Belief in the
impossibility of performing the activity; Revulsion against the activity or
some of its necessary components; Non-identification of opportunities for
We shall now introduce a new dimension to the analysis of ability. Let
us first note that the practical possibility of performing a particular action
is dependent on how the action is specified, in particular regarding place
and time. It may be practically possible for\ A to perform F in a particular
situation S, but not in S\. The reason for this is that S provides the right
opportunity for A. For instance, it is now practically possible in Sweden
for a foreigner, who meets certain conditions, to vote in local government
elections. A few years ago this was not practically possible; and it is not
practically possible in several other countries.
To say that it is practically possible for a particular agent A to vote in
the local elections means, according to our suggested test procedure, that
he would actually succeed in voting ifhe tried. But what is the correspond-
ing relation regarding ability in a particular situation? The presence of
ability cannot be as easily determined as that of practical possibility. "A is
able to do F", does not imply that A actually does F, if he tries. A may be
able to do F but still fail to do F when he tries because there is no
opportunity. From "A is able to do F", we can merely conclude that A does
F if he tries and if the right opportunities obtain.
But this purely formal solution does not square completely with our
intuitions. Moreover, it does not suffice for defining a basic notion of ability
for the discourse on health, for it implies that practically everybody would
be able to do practically everything. Given a sufficient manipulation of
opportunities, a person whom we would normally call disabled with respect
to a certain action would become able. Consider the following example.
A schoolboy is taken into the cockpit of an aeroplane. He is instructed
in detail by the pilot how to handle the instruments, indeed he is sometimes
physically helped in the manipulation of many of them. Thus, given the

extraordinary opportunities created by the pilot, the schoolboy is able to fly

the aeroplane.
But this kind of situation is not what we have in mind when we say that
a schoolboy is able to fly an aeroplane. We do not mean that he can succeed
in doing so, given extremely advantageous circumstances, but rather that
he can succeed in doing so given normal or standard circumstances. 38
The distinction between extraordinary and standard circumstances is
obviously crucial in the theory of medical disabilities and handicaps. We
would certainly describe a person who has lost both his legs as disabled
with respect to the action of moving about. But if he is sitting in a
wheelchair, this is not true. Thus in order to describe the person as disabled
in the first place we have to disregard the wheelchair.
But what constitutes standard circumstances? It will be argued here that
this locution cannot be completely defined in descriptive terms, and that
this is one of the reasons why the concept of disability, and consequently
the concept of health, are relative concepts. What are counted as standard
circumstances vary from epoch to epoch, and from society to society.
Consider the differences in natural environment between Greenland and
the Congo, or the cultural differences between life in the United States and
life among the aboriginal tribes of Australia. What are counted as standard
circumstances must vary enormously between these places.
This observation provides the following insight into the logic of ability.
Ascribing an ability to a person A with regard to some action is at the same
time to describe part of the world in which A resides. Ifwe ascribe an ability
to A, then we presuppose that the standard circumstances in the society
and the natural environment in which A lives, provide the opportunities for
A to execute his ability. When we say that A is able to play football, we must
presuppose that he lives or has lived in a society where the game offootball
is recognized. When we say that he is able to repair an automobile, we must
assume that he lives or has lived in a modern, industrialized society, etc.
But let us take a particular society at a particular time. How do we
determine the standard set of circumstances in this case? There is no
algorithm for solving this problem. It is not a question of simple statistics,
although how often certain circumstances arise certainly plays a role. What
counts as an unusually advantageous or an unusually difficult situation
with regard to a particular action is basically a normative question. We
follow a convention adopted in our society when we say, for instance, that
a person who is able to read should not need continuous instruction, or that
a person who is able to walk should withstand a moderate storm of, say,

20 m/sec. A person who fails to walk in a hurricane, however, even though

he tries, may still be said to be able to walk.
How often a circumstance arises is not in itself sufficient for determining
whether it is non-standard in our sense. Assume that there is a very rare
species of rat. Assume also that when A meets a rat of this species he
becomes paralysed with fear. Assume thirdly that hardly anybody else has
become frightened by this species of rat. On these assumptions we are
inclined to ascribe some disability to A, although this disability presents
itself very infrequently.
We conclude, first, that the concept of ability, as normally understood,
is related to a set of standard circumstances. This set of standard circums-
tances is relative to some natural and cultural environment. The natural
and cultural environment of Sweden has a set of standard circumstances
which is quite different from that of Malawi. We conclude, secondly, that
the concept of a standard circumstance is a normative concept. The de-
cision as to what should count as a standard circumstance for a particular
action type is partly influenced by statistical factors, but is mainly determin-
ed by the society's profile of goals.


The relativity of ability has important consequences for the discourse on

health. For example, A may be able to perform the set of actions required
for his being in health in environment E, but not in E l' From this it would
seem to follow that A may be in health in E but not in E l' Simply by moving
from one part of the world to another A may lose a previous state of health
or acquire a state of health which he did not have. This seems to be possible
without any change in his internal bodily or mental make-up. But are we
inclined to say that a person's health is relative to the extent suggested by
this? To illustrate the problem consider the following example:
An uneducated young man leaves an underdeveloped country and moves
to Sweden. In his native country he had his own farm, which he cultivated
well enough to sustain himself and his family. When he enters Sweden, say
as a political refugee, he is no longer able to lead such a life. Where in his
home country he lived relatively well, in Sweden he is disabled.
But would we say that this man is healthy in his native country, and
becomes ill upon entering Sweden? We shall introduce a distinction
between first-order ability and second-order ability in order to analyse this

case. So far our discussion has exclusively concerned first-order ability. An

intuitive characterization of second-order ability is as follows:

A has a second-order ability with regard to an action F, if and

only if, A has the first-order ability to pursue a training-program
after the completion of which A will have the first-order ability
to do F.
Second-order ability is thus compatible with first-order disability, while
the reverse does not hold. The individual in our example lacks the first-
order ability to earn his living in Sweden. He may, however, have the
second-order ability to do so. He may be able to train himself to make a
good living in that country.
Before considering the relation between second-order ability and health,
let us note that the action of training must be given the same analysis as
other kinds of action. When we ascribe to someone the first-order ability
to follow a particular training program, we must, as in the general case
above, presuppose a set of standard circumstances, and that it persists
throughout the whole training process.
Thus, a person who enters a training program, but in the end fails to
acquire the desired first-order ability, need not lack second-order ability.
First, the training program may have been poor. This would indicate that
standard circumstances did not obtain. Second, the subject may, after a
while, no longer have intended to pursue the training in a proper way. This
being so is still consistent with his having the second-order ability. Second-
order ability need not turn into first-order ability, if the agent does not
consistently try to acquire the first-order ability.
But then what about the following case? The subject is afforded adequate
training facilities, and he tries to learn throughout the whole period. Still,
after this period, he does not know how to perform the desired action. This,
we would say, indicates that the subject does not have the second-order
ability, at least not throughout the whole period of training, to perform the
action. We could then also say that he is genuinely disabled with respect to
performing the action in question.
To summarize: A person has a second-order disability with regard to an
action F, if and only if he is disabled, given standard circumstances, from
consistently pursuing a training program to acquire a first-order ability to
perform F.

The notion of second-order ability brings us closer to the biologically

founded capabilities of man. Still, it does not and cannot completely free
us from the relativity of an action to an environment. To say that A has a
first-order ability to follow a training program successfully, presupposes a
set of standard circumstances, and the standard must be determined rela-
tive to the cultural and natural environment. It may be possible that certain
people who lack a particular first-order ability, if they were put into ex-
tremely advanced and extremely expensive training programs could
achieve this first-order ability. But if such programs have not been offered,
or if they have not even been designed, they cannot be taken into account
in ascribing second-order ability to them.
What is the impact of this distinction on our analysis of health? Should
we say that the ability involved in the ascription of health is always second-
order? It seems plausible to say that the newcomer to Sweden who has
problems in taking care of himself is ill only if he lacks the second-order
ability to manage his living in the new environment. A lack of first-order
ability is not enough.
It also seems, in general, that such disability as is solely due to lack of
training is not an indication of illness. There is reason to speak of illness
only if the act of training has in its turn been prevented by internal factors,
in which case there is a second-order disability.
But would we apply the same analysis to the standard case of illness,
which has an organic cause? Here one normally begins with a first-order
ability. An agent has, for instance, a first-order ability to perform his
professional activities. Then he becomes ill, and as a result loses his
first-order ability. But would it be true to say that he no longer has the
second-order ability to do his work?
It is easy to be misled here and identify two pairs of concepts which
should be held distinct; one pair is first- and second-order ability, and the
other is power to execute a basic competence and having a basic compe-
tence. More particularly, having a basic competence is not the same as
having a second-order ability, (whereas first-order ability and power to
execute a basic competence are identical).
We normally ascribe a basic competence to someone when he knows
how to do something. According to our previous definitions this is not at
all true about second-order ability. The immigrant to Sweden has not
previously learnt anything about Sweden and in this sense does not have
the basic competence requisite for making his living in Sweden. He may,
however, have a second-order ability with regard to the same action.

Basic competence is compatible with both second-order and first-order

ability. But for our purposes it is important to see that a person who has
a basic competence vis-a.-vis a certain action F need not even have a
second-order ability with regard to F. Consider a case of long-term illness.
A professional football player has broken both his legs. It is obvious that
during his confinement he does not have the first-order ability to play
football. Still, we would say that he has, throughout this period, the basic
competence to play football. In a perfectly understandable sense he knows
how to play football.
But does the football player have the second-order ability to play football
while he lies in bed? No, for to have the second-order ability to do Fmeans
having the first-order ability to follow a program which leads to a first-order
ability to do F. But the football player, who is bound to bed, is clearly not
in the position to follow such a program; and so we may say of him that
he is ill.
The same reasoning may be applied to all paradigm cases of illness due
to disease or impairment. During an acute phase of illness, however short
it may be, an agent has lost both the first- and second-order ability to
perform the actions with respect to which he is disabled. This holds for all
the typical cases of illness involving pain and general disabililty.
We shall, however, not exclude the possibility that an agent might lose
a particular first-order ability while retaining the corresponding second-
order ability in the sense that he may be able immediately to regain his
first-order ability by performing a series of actions. But if there is no
time-lag at all between the loss of the first-order ability and the possibility
of going through a process of rehabilitation, it is doubtful whether we would
label this a state of illness.
The fact that some states of illness last for a very short time would seem
to involve serious problems as regards testing. Assume that a person has
been struck by a disease which does not last for more than a day. In order
to decide whether the disease really has caused illness, the subject must
be tested with regard to his relevant abilities. This might be easy with
certain first-order abilities. But when the time span of the illness is so short,
it would seem virtually impossible as regards his second-order abilities.
Few training programs can be carried through within the period of a day.
The problems suggested here are in practice limited. Whatever the set
of abilities required for health is, it cannot be necessary to establish the
whole set in order to make judgments about health and illness. There are
ways of taking short-cuts. The most efficient one is to inspect a patient's

set of abilities with regard to his basic actions. If it appears that a person
cannot move out of his bed, or if it appears that he cannot talk or that he
does these things with extreme difficulty, then we can immediately con-
clude that he cannot enter a particular training program, and that he cannot
do this for reasons internal to his body or mind.
Our general conclusion is that the ability involved in health is an ability
of the second-order kind. To be healthy is to have, at least, a second-order
ability to perform a certain set of actions. To be ill is to have lost or, in
general, to lack one or more of these second-order abilities.


The most difficult task in our characterization of health still remains: how
are we to specify the set of actions that a healthy person must be able to
In approaching this problem we shall first somewhat alter our manner
of speaking. Instead of talking about a set of actions that an agent must
be able to perform, we shall assume that there is a set of goals which the
healthy person must be able to achieve. This does not involve a radical
change in our philosophy. It is merely a simplification of our mode of
speech. By concentrating on (ultimate) goals, we can avoid giving a long
enumeration of sp~cific actions. Moreover, we are not forced to make
difficult decisions about the level at which action-concepts should be
It is plausible to believe that whatever the adequate answer to the
question of health should be, it will be an answer on an abstract level, which
can be summarized in terms of certain general goals. The question to be
put should rather be formulated in the following terms: what are the goals
that a healthy person must be able to realize through his actions?
In Chapter two, section 2, a basic analysis of the concept of goal was
presented. The distinction was made there between factual goals and ideal
goals. In our discussion of an analytic theory of health the notion of a
factual goal was employed. In the discussion to follow we shall speak in
terms of ideal goals and in terms of relations between actions and goals.
Let us then look into the logical relations between actions and goals. A
first assumption is that the realization of a goal requires the performance
of at least one action. If the goal is general and abstract it normally requires
a sequence of actions, i.e. an activity.

The relation between an action and its goal may be internal or external. 39
The difference is easily seen given our analysis of actions above. We say
that the relation between an action and its goal is external if the goal is a
causal or conventional consequence of the action. The relation is internal if
it follows logically from the fact that the action has been performed, that
the goal has been reached. For example, the relation between working hard
and passing an exam is external. Passing an exam is a causal consequence
of working hard, but it is not a logical consequence of working hard. The
relation between killing Smith and the death of Smith is however internal.
If Smith has been killed it follows logically that Smith is dead.
We see then that the way human action is conceptualized is crucial for
the nature of the relation. One and the same sequence of events can be
looked upon as either an action causing a goal (stabbing Smith causing the
death of Smith) or as an action entailing a goal (kill,ing Smith entailing the
death of Smith).
It is important to be conscious of these features for the following reason.
The reaching of a goal can be conceived of as, simply, the performing of
an action. However abstract a goal we choose to consider, we can always
construct an action-concept entailing it.
It can now more easily be seen that the goal-mode of discourse is just
a variant of the action-mode of discourse. Many of the conclusions drawn
from the analysis of actions can be transferred to the analysis of goal-attain-
When goals are far-reaching or abstract the process of reaching them is
normally divided into a sequence of actions (an activity). The endstate of
each member of the activity sequence can be viewed as a subgoal. In the
extreme case, the attainment of a goal requires the realization of a very
specific set of subgoals in a particular order.
As was observed with activities, however, most goals can be realized
through a great number of alternative routes. Different circumstances may
constitute opportunities for different actions in realizing the same goal. In
circumstance C I action HI may be necessary for realizing goal G. In C2
actions H2 + HI may be necessary for realizing G.
But also one and the same circumstance may provide an opportunity for
distinct actions leading or contributing to G. In such a case neither of them
is necessary for G. Normally, one can get food for the day in many different
ways. One may go out in the woods and pick blueberries, buy food in a
shop, or visit a restaurant. Many persons have the opportunity of perform-
ing these different actions at anyone time.

Observe, however, that the number of alternatives is dependent upon the

level of abstraction in our specification of actions. When the level of
specification is very detailed the number of alternatives will be very great.
Think, for example,of all the different routes one can take in walking to the
shop in order to get food.
In general, given a certain level of specification, a goal can be reached
by a person in a particular situation through the performance of one of a
set of activities. In the limiting case the set is a unit-set and the activity is
reduced to a basic action.
Now to the crucial question. What is the set of goals that a healthy
individual must be able to reach?
In answering this question we shall consider two theories which attempt
to characterize those vital goals that determine the set of abilities required
for health. (The term "vital goal" will in the following be chosen as a
technical term for the goals related to health.) According to the first theory,
the vital goals should be defined in terms of human needs; according to the
second, the vital goals should be defined in terms ofthe goals set by the agent
Our conclusion will be that neither of these suggestions, despite their
merits, can claim to provide the sole answer to the question regarding the
vital goals of man. They can at most define special technical concepts of
health. Other such technical concepts could easily be formulated.
We can then see the dimension within which one can operate in formulat-
ing different health-concepts. This can be done by manipulating the set of
goals defining the vital goals of man. Various discourses - ordinary, admin-
istrative, as well as scientific ones - suggest some differences concerning
the vital goals. And if one wants to define the concept of health this should
rather be conceived of as a family of concepts, containing the different
technical and non-technical versions as members. 40
Admitting that the concept of health is a 'family concept' does not,
however, amount to anarchy. The family has its boundaries and it is limited
to one particular dimension. It does not, for instance, include an analytic
concept of health.
Before discussing these particular proposals for the characterization of
health, let us consider some conditions that all such serious proposals must
fulfill. First, we must be able to make some demarcations. Health, as the
ability to reach a set of goals, must be distinguishable from other concepts
definable in terms of abilities related to goals.

"Ability" is often used as a word expressing excellence. To be an able

man in this sense is to be a clever man. A clever man can reach goals which
are more advanced than the ordinary man's. In all ordinary uses this
concept is distinct from that of health. We need only point out some
obvious features to establish this. Only a minority of the population is
clever, whereas we would say that the majority of the population is healthy,
and that it is in principle possible - indeed it is a goal of society - that the
whole population be in a state of health. It is, however, conceptually
impossible for the whole population to become clever. The logic of the word
"clever" is to select the best in a population. If the general standard of
ability is raised in a population then this would just result in a sharpening
of the requirements for being clever.
Another feature distinguishing health from cleverness is that one can be
- and normally is - clever in very particular respects, whereas it is not only
in certain respects that one is healthy. One can be clever at chess without
being clever at anything else. This particularization is alien to the concept
of health. To be in health is to have some general ability. We don't say that
one is in health with regard to the function of buying food.
Other similar distinctions are those between health and morality, on the
one hand, and health and legality on the other. These distinctions are
sometimes blurred, particularly in writings concerning mental illness. It is
sometimes said that mental illness is characterized by deviant behaviour. By
this one seems to mean, simply, that the behaviour does not conform to
certain rules laid down by society.
But if such talk is not supplemented with qualifications, it would imply
that all breakers of societal rules, be they criminals (including breakers of
traffic-rules), immoral persons, unconventional persons, practical jokers,
psychopaths, as well as certain other mentally-ill persons, should be classi-
fied as being ill.
Such a claim would mean a conceptual collapse entailing a revolutionary
change in our general view of man. Criminals and immoral people should
then be treated for their illnesses, and there would be no point in our
ordinary practice of retribution, the punishment and rewarding of behav-
Further reflection shows that it is not the deviance as such which is the
ultimate criterion of mental illness. The deviance can be a sign of illness
but only insofar as it points to some fundamental disability within the
subject. If the latter is the case, then there is illness; otherwise there are
only different instances of norm-breaking.

The concepts of immorality and illegality are not concepts of disability.

To be an immoral person or a lawbreaker is to be a person who actually
performs immoral or illegal actions. Nothing is thereby said about the
person's abilities or disabilities. To be a liar means that one often tells lies,
but this is consistent with one's being able to tell the truth.
Of course, certain liars and certain criminals are for various reasons
unable to follow the relevant rules. When this is the case they have excuses
for their deviant actions. If they have a genuine disability, their actions may
be reclassified, and they may lose moral or legal responsibility. There may
be a case for illness. Therefore, an immoral or criminal person may be ill.
And his immoral or criminal behaviour may be a sign of - although not the
ultimate criterion of - his illness. But it does not follow from this that the
'person is in fact ill.
Reasoning quite analogous to this can be advanced concerning other
kinds of normbreakers, breakers of rules of games, of conventions or of
rules of prudence.
We turn now to the first option for defining the vital goals of health -
the notion of a human (or a vital) need.



The concept of need is at least as involved and loaded with ideology as the
concept of health itself. The idea of a basic human need has a central place
in political theory and political debate. It is among the key concepts in the
thinking of men such as Rousseau, Mill and Marx. It is one of the most
frequently used concepts in day-to-day political discussion; many social
reforms have been motivated by reference to the needs of the population.
These facts motivate much caution in the analysis ofthe concept of need.
Moreover, there are certain linguistic reasons for being cautious when
discussing needs. There are difficulties in making exact translations of the
term "need"; for example it does not have exact equivalents in German or
the Scandinavian languages. In some contexts, "need" can be substituted
by "want", but the German "Bediirfnis" could seldom be substituted by
"WUnsch", nor could the Swedish "behov" be substituted by "onskan".41
The analysis to follow here does not pretend to be a contribution to
comparative linguistics, nor to answer the question of what the basic

human needs are. Instead, its primary aims are the following: first, to reveal
a fundamental sense of need (viz. necessary condition for the realization
of a goal), and second, to investigate if there is' a particular interpretation
of 'need' which can serve the purpose of clarifying the concept of health.
The main conclusion of the chapter is that the most favoured interpretation
of a basic human need does not serve this purpose.

The General Sense of Need - Necessary Condition of

(i) An ontological puzzle

The ontology of needs is mysterious and elusive. What kind of thing is
a need? Is it a property of a human being? Can it be located? Is it a bodily
state? Or is it something outside the human being to which he has some
kind of relation?
Questions such as these are prompted by the fact that we have at least
two kinds of locutions: A has a need of y; and y is a need for A.
Assume now a situation where we say that A has a need of food. This
locution is typically used when A lacks food, when there is no food in A's
stomach. But if the need is something present within A, then the need
cannot be identical with food.
But in the same kind of situation it is quite proper to say that food is a
need for A. Here the need is explicitly equated with the food and must
therefore be something outside A's body.
We have here encountered an obvious ambiguity; and one of our main
tasks is to clarify it.
(ii) The relational nature of needs
It will be argued here that "need" is basically a relational term, more
precisely a four-place predicate. Perhaps this is most easily seen when
considering locutions where "need" is a verb: A needs y in order to attain
G. John needs a hammer in order to repair his house. What does such a
locution mean? Essentially it means: John's using a hammer is a necessary
condition for his repairing the house. (Or simply the hammer is a necessary
condition for John's repairing the house.)
In such locutions we can detect the most general and clearest sense of
"need". "Need" here simply stands for any necessary condition for the
attaining of a goal.

In this sense it is certainly not only humans that can have needs. All
entities to which it is sensible to ascribe goals can have needs. A plant needs
chlorophyll to survive and a machine may need oil to work.
The number of needs pertaining to an individual, in this sense, could be
very great. As long as the individual has at least one goal it is strictly
speaking infinite. The reason for this is the transitivity of the relation of
being a necessary condition for something. A necessary condition for a
necessary condition for G is, ipso facto, a necessary condition for G. IfJohn
needs a torch to find the hammer which he needs for repairing his house,
then the torch (or finding the torch) is also a need for John. And since all
series of events are infinitely divisible the number of needs must turn out
to be infinite.
To this fact we shall add a complication. Needs or necessary conditions
are dependent on background situations. It may be necessary for John to
have a hammer in a given situation in order to repair his house. But in a
different situation, where Steve does the hammering, John does not need
a hammer. Therefore, needs in this general sense may vary over time
depending on situational change.
We return now to our initial question about the ontology of needs and
the different ways of expressing the existence of needs. Given the general
sense of needs analysed above the two locutions "A has a need of y" and
"y is a need for A" should have the same analysis. They are both, according
to the analysis, elliptic formulations of the following proposition: There is
a goal G and a situation S so that y (or using y) is a necessary condition
for A in S in order for A to attain G. More simply put, y is a need for A
in S to reach G. This is the four-place expansion of the locution. But what
about ontology? Where is the need? In what sense does the need exist?
If need means "necessary condition" then that which constitutes the
necessary condition is the need. In the case of John's repairing his house,
the hammer or, more strictly put, the use of the hammer, is the need.
In the general sense, then, a need is not a bodily state of a person; it is
rather any kind of state or event, in which the person mayor may not be
involved, which is a necessary condition for the person's attaining a goal. 42

Vital Needs or Human Needs

However important the preceding general analysis may be, it cannot do

justice to, or create an understanding of, the sociological and political

discussion regarding needs. This discussion is characterized by at least the

following features:
(1) Needs are restricted to human beings;
(2) Needs are related only to goals which have some objective
biological basis.
(3) Needs are enumerable, and are often presented in short lists of
"basic" needs.
The basic idea in the philosophy of the specifically human sciences is that
needs can be discovered by an inspection of human biology and psychology.
The idea is, however, not that we should look into any particular human
being, but instead find some universal characteristics, which are the typical
biologically and psychologically based human needs.
With some theorists this human nature is considered to be more or less
stable over time (cf. e.g. Abraham Maslow), with others, notice is paid to
the continuously changing aspect of human nature, and this change is
considered to be mainly due to the evolving society (cf. e.g. Rousseau and
For the sake of simplicity let us concentrate here on just one significant
theory of human needs. We shall choose Maslow's, partly because it is a
rich and substantial theory, and partly because it has a central place in
modern psychological theory. Maslow puts forward the following theses:
(i) There is a limited set of basic human needs universal to all
human beings. He suggests the following six: physiological
needs, the need of safety, the need of belongingness and love,
the need of esteem and the need of self actualization.
(ii) The basic needs are ordered in a hierarchy where the physiolog-
ical needs are the most basic and the need for self actualization
the least basic.
This hierarchy entails that a more basic need must be fulfilled
before a less basic need can be recognized by the agent in
(iii) The basic needs are manifested in human beings as physiologi-
cal and psychological drives, i.e. as tendencies within the human
being to fulfill the needs.
The latter thesis constitutes an important addition to the general notion
of a need, meaning merely a necessary condition for realizing a goal. Here
a need is connected or even identified with a particular biological and

psychological reality; Maslow assumes that for all basic needs there are
physiological and psychological mechanisms which trigger off some be-
havior on the part of the individual to realize a certain goal. For instance,
hunger is associated with (or even identified with) a certain physiological
mechanism which triggers off food-seeking behaviour.
Given this idea one can see how it can be plausible to identify needs with
bodily states. The bodily states are the biological drives.
The idea that biological drives can be identified with needs forces us to
make a distinction between two referents of the term "need". This was in
fact evident already in our introduction where we noted the following two
kinds oflocutions, "A has a need ofy" and "y is a need for A". In our abstract
discussion about needs we let y be the referent of the term "need". Using
a hammer was a need for John in order to repair the house. But when we
say that John needs a hammer in order to repair his house, the need is not
identified with the hammer. The need is rather something located within
John's body or mind.
It is obvious that the philosophy of identifying needs with biological
drives is based on the latter mode of speech. When we identify the need
of food with a physiological drive, it is not the food which is the drive but
the hunger; similarly, when we identify the need of safety with a drive within
a person's body or mind, it is not the safety which is the drive but some
safety-seeking tendency within the person.
This important distinction between a drive and its object is obscured by
the drive-philosophers themselves, since they sometimes also refer to the
objects (such as safety, love and esteem) as needs. But even if this dis-
tinction between bodily state and object is well made, it does not seem
advisable to "identify" needs with particular bodily states. It is important
to be able to say that a person has the basic need at all times. The drives,
however, do not appear at all times. They appear only when there is a lack,
when the needs are not fulfilled. Everybody has a need of food. It is,
however, only hungry people who have a physiological drive which triggers
food-seeking behaviour.
Still, the existence of drives certainly is important for the identification
of the basic needs.The idea is the following: wherever we can find a
physiological or psychological drive, universal or almost universal to all
mankind, which triggers behaviour in the direction of a certain object, when
this object is not already present, then we are dealing with a basic need.
The problem of what the basic needs are could then, in principle, be solved
by empirical biology and empirical psychology.

The question, however, could now be asked: what is the connection

between this specific sense of human need and what we called the general
sense of need? So far, we have spoken of needs as objects of certain
physiologically and psychologically triggered behaviour. Our presentation
has not answered the fundamental question of why the objects of these
bodily or mental drives should be called needs? Why are food, safety, love
and esteem human needs? What are they necessary conditions for?
The need-theorists are normally quite silent on this point; perhaps they
consider the answer to be more or less self-evident.
There is also in Maslow's writings surprisingly little on the goals of the
needs. One could, of course, grant him and other writers that one goal is
self-evident. This is the goal of the most basic needs, the physiological ones.
The satisfaction of the hunger-need and the thirst-need is a necessary
condition for the survival of the individual. (And similarly the satisfaction of
the sex-need is a necessary condition for the survival ofthe species.) But what
about the satisfaction of the other needs? They are not necessary for
Maslow's reply here is that the fulfillment of the other basic needs (with
one notable exception) is a precondition of health. If the basic needs remain
unfulfilled, if the related drives remain "in an active and chronic sense"
([78], p. 57), then the bearer is simply an unhealthy man. (This almost looks
like a definition of illness. But for Maslow's theory to work the connection
here must be einpirical. He must mean that illness is a result of the depri-
vation of anyone of the basic needs.)
Let us then add the following to our list of Maslow's theses:
(iv) The fulfillment of the basic needs is a necessary condition for
the survival of the individual (or the species) or for the health
of the individual.44
In fact these criteria are not unique to Maslow's theory. Other important
modern writers suggest similar characterizations. Braybrooke [17] says:
... deficiency in respect to the basic needs endangers the normal functioning of the subject of
need, considered as a member of a natural species.
In the case of men, such deficiencies might also be said to endanger health and sanity. By
proximately endangering health and sanity, they ultimately endanger survival (pp. 90-91).

Knut Erik Tran0Y [131] suggests a similar, although more complex, charac-
terization of his concept of a vital need.
N is a vital need for a, if and only if the following four conditions are fulfilled.

(i) Prolonged or definitive frustration of N kills or seriously disables a.

(ii) Satisfaction of N for a does not necessarily (essentially) block any of the vital needs ...
(iii) N is irreplaceable, which is to say that frustration of a vital need N cannot be compensat-
ed by satisfaction of any other need.
(iv) Satisfaction of N brings pleasure and/or cessation/prevention of harm or pain for a ...
(p. 155).

Let us now observe that we have two kinds of sufficient criteria for singling
out the basic needs. First, we have the drives, universal, or almost univers-
al, to all mankind; secondly, we have the idea that the objects of the drives
are necessary conditions for survival and health. It might seem that these
criteria could support each other; they are, however, also in potential
conflict. There may be tendencies, almost universal to mankind, which do
not go in the direction of maintaining health. In such a case what would
be the decisive criterion? Ifwe decide that the attainment of the goal, viz.
health, is the ultimate criterion we could characterize the concept of a
human need in the following way: Any state of affairs that is necessary for
the individual's survival or health is a basic need for him (irrespective of
the existence of drives within him, be they universal or individual).45
Health and Need: On the Circle of Health
It is not our task to judge Maslow's theory of needs or any other theory
of human or vital needs. Certain basic conceptual difficulties have, howev-
er, already appeared.
Our main task is to decide if a theory of this kind can be of any use in
the endeavour of characterizing health. The initial suggestion was the
following: A person A is in health if, and only if, he has the ability, given
standard circumstances, to fulfill his basic needs.
Given the above analysis, this suggestion can be translated into: A is
healthy if, and only if, he has the ability, given standard circumstances, to
fulfill certain necessary conditions for his survival or his health. It is
immediately seen that this characterization contains the terms "health"
("healthy") in two places. First, in the analysandum; and second, in the
analysans. The conclusion, then, is that this characterization is empty or
almost empty.46
An alternative which suggests itself, in order to avoid this circularity, is
to drop the criterion of health in the definition of need. A human need is,
then, simply a requisite for the survival of the individual (with the possible
addition of the survival of the species). A healthy man can, in standard
circumstances, see to it that he satisfies his physiological needs.

This may look like a very weak suggestion. Would not also the majority
of unhealthy persons fulfill such minimal requirements of health?
But the suggestion is not so implausible when we analyse what must be
packed into the locution "in standard circumstances". The standard cir-
cumstances include the environmental and cultural background. This
background dictates what are the possible and, from a legal point of view,
proper means for maintaining survival.
In a highly regulated Western society most people cannot (without
special permission) simply go fishing and collecting berries and vegetables
in order to survive. In most circumstances the only proper way to get food
for the day is to purchase it. This in its turn presupposes funds. The
standard way of obtaining funds is to get a job.
Therefore, normally, the fulfillment ofa professional role is, in a Western
society, a requisite for satisfying one's basic physiological needs. An alter-
native is that one, as a spouse, is supported by a working man or woman.
But this could very well be seen as a contribution to the latter's professional
A consequence of this analysis is that the suggested definition of health,
in terms of the fulfillment of needs, comes closer to our intuitions. Since
professional life in a Western society entails the performance of complicat-
ed actions, one is often obliged to enter into an involved series of actions
in order merely to survive. Therefore, indirectly the goal of survival will
entail a set of other subgoals, which are in a sense more advanced.
Consider the following facts. The mere process of applying for a job
requires a number of measures. You must be able to contact an employer;
you must be able to make yourself understood; you must be able to fill in
certain forms. In addition to this you must be able to perform the actions
required by the job itself.
Understood in this sense the goal of survival requires a good deal; and
it may even square quite well with ordinary medical practice. A key-cri-
terion of health in medical practice, not least for purposes of health insu-
rance, is that the individual can perform his own professional task in an
efficient way.
We notice now how our analysis of the relation between health and the
fulfillment of needs takes us away from the simple biological platform
suggested by the modern need-theorists. Health becomes a society-related
notion, because, implicitly, society determines the subgoals to be achieved
in order to survive. Moreover, these subgoals can vary a good deal depend-
ing upon the particular society and profession.

Consider now a criticism that may be directed against the need-concept

of health on this interpretation. Although the present analysis, in terms of
professional roles, gives some plausibility to the concept, its connection to
mere survival makes it still one-sided. Almost everybody is involved in
several important activities outside his professional roles: he may be in-
volved in politics, or in artistic or athletic activities. These activities may
be very important to the individual himself, and indeed to the society.
We can well imagine that some of the activities involved in the non-prof-
essional task are specialized, and are not included in the professional task.
Are we, then, never inclined to view disabilities with regard to the former
kind of activity as indicating illness?
According to the definition of health to be presented below, this is
indeed the case. All goals which are in general important to the agent
consitute the set of goals in relation to which his abilities are to be



The idea that a subject's health is identical with his ability to realize the
goals set by himselfis a promising approach suggested recently, and inde-
pendently, by two analytic philosophers, Caroline Whitbeck and Ingmar
porn. 47 In addition to a definition along these lines, they also provide a
framework for relating the notions of disease, defect and impairment to the
notion of illness. A detailed discussion and defense of this framework is to
be found in [140].
Consider first the central passages in Porn's and Whitbeck's presen-
tations of this idea. Porn says:
Health is the state of a person which obtains exactly when his repertoire is adequate relative
to his profile of goals. A person who is healthy in this sense carries with him the intrapersonal
resources that are sufficient for what his goals require of him. This does not mean, however,
that he will realize all of his goals, for his powers to act are determined not only by his
repertoire but also by the external factors making up his opportunities for action-factors over
which he does not always have control ([991, p. 5).

And Whitbeck says:


Health is a person's psychophysiological capacity to act or respond appropriately (in a way

that is supportive of the person's goals, projects, and aspirations) in a wide variety of
situations. Health encompasses certain significant components: among them, maintaining
physical fitness, having a generally realistic view of situations, and having the ability to
discharge negative feelings ([141], p. 620).

Both Porn and Whitbeck characterize health in terms of the subject's

intrapersonal (psychophysiological) repertoire (capacity) to attain his own
goals (projects and aspirations). That health is something dependent upon
the relation between a person's ability and his own goals is the basis of the
theory to be discussed below.48
Three important differences between Porn's and Whitbeck's variants of
the theory must, however, be noted first. To begin with, Porn says that
health obtains exactly when the ability is adequate to the goals. The healthy
person must actually be able to achieve his goals given standard circums-
tances. Whitbeck's requirement is obviously weaker: the healthy person
should be able to act appropriately, in a way that is supportive of his goals.
The ability need not be adequate for the goals.
Whitbeck's requirement seems very weak. People with a very low degree
of health may be able to act in a way which is supportive of their goals. A
man who is unable to walk properly may be able to make movements which
are supportive of walking, but not sufficient for it. A mentally retarded man
who wishes to pass an examination may be able to perform actions suppor-
tive of reaching the goal, but which are not sufficient for reaching it.
Porn has chosen the stronger notion, viz. that health is the ability to
actually attain a set of vital goals. Porn is also aware of the important
proviso concerning external circumstances, although he does not provide
any information about what these circumstances are.
A second interesting difference is the following. Porn calls his theory the
equilibrium theory. According to his text, health obtains at a particular point,
exactly when a person's capacity is adequate to his goals. But what about
the case of the very able person, whose capacity supersedes the require-
ments for realizing his goals? Is he ill? Strictly speaking, this seems to be
a consequence of Porn's formulations.
Porn might argue that this is a very unusual case of disequilibrium. As
soon as the normal subject realizes his great abilities he will almost auto-
matically adjust his goals in accordance with them. In those unusual
instances when this is not being done, there may indeed be a case of illness.
Such a claim could be questioned. Many people are cautious and wish
to plan their lives in a very realistic way. They may know that they now

have great abilities, but they may also be aware of the possibility that these
may cease to exist. Therefore, they set their goals with a considerable
margin of safety. This cautious way of planning life is certainly not a
man.ifestation of illness. According to this reasoning the equilibrium thesis
will in the subsequent discussion be interpreted in the following way: health
obtains when a person has, at least, the ability required for realizing his
(It will be conceded in Section 7 that there are cases when a person sets
too low goals for himself and that this may indeed be a sign of illness. It
is, however, in our view a mistake to claim that there is illness as soon as
there is disequilibrium in Porn's sense.)
Note, thirdly, a subtle difference in the ontological characterization of
health by the two theories. For Porn health is the relation which holds when
a person's ability is adequate to his goals. To Whitbeck health is the
person's capacity to act supportively in rel~tion to his goals.
From an ontological point of view there are good reasons in support of
Porn's view. To say that health is a relation is to underline the essential
relational character of the concept of health. For most purposes this
difference of expression is, however, unimportant. As long as the theories
agree on the truth-conditions for health-statements, which in both cases
contain references also to goals and circumstances, then the difference will
have little practical import. In the following, we shall sometimes, for the
sake of simplicity, use locutions to the effect that health is the ability to
reach one's goals.

Health and Its Dimensions

Is health an absolute or a relative concept according to Porn and Whit-
Here their answers differ. For Porn health is an absolute state. Health
is a state which obtains exactly when there is an equilibrium between a
person's abilities and his goals. Strictly speaking, there can be no degrees
of health.
The complement, illness, however, is measurable in degrees. Illness is
then dimensional in two respects. First, a person can be ill to varying
degrees with respect to a particular goal. He can be more or less distant
from his goal. Secondly, a person can be ill to varying degrees with respect
to the number of goals which he can fulfill.

Given Whitbeck's conception, on the other hand, health can itself be

seen as having degrees. This must be so since health, given her view,
obtains when the subject's ability is merely supportive of his goals. If the
subject's ability is only supportive to a small degree, then there is a low
degree of health; and so on.
But is there a highest degree of health? Following our previous reasoning
there ought to be a highest degree of health which would consist in the
agent's internal resources being adequate for all his goals. In answering this
question Whitbeck, however, introduces a further dimension of health.
The term "wholeness" should not, however, suggest that there exists an upper limit, a state
of optimum health. Such a suggestion would be foreign to the account given here since,
presumably, people can always increase their ability to act appropriately in some situations.
The absence of an upper limit on health does not make that concept any more obscure than
concepts such as wealth, which also have no upper limit ([141], p. 616).

A person has the highest degree of health if he can fulfill all his vital goals
in all kinds of situations. But since one can always conceive of new and
more demanding situations, there can in practice be no such thing as a
highest degree of health.
Observe an interesting complication here. In our characterization of
ability as used in the context of health we have presupposed as a standing
background some set of standard circumstances. Now in this argument
Whitbeck does not make such a presupposition. Instead she counts on
there being indefinite variations of the circumstances and measures health
in accordance with these variations. Is it possible, given our approach, to
acknowledge this dimension of health?
Two points should be made regarding this issue. First, in one interpre-
tation, Whitbeck's dictum is not reasonable. If we say generally that a
person's health is greater the more kinds of situations he can handle in
realizing his vital goals, then the conceptual difference between health and
pure strength will be blurred. Assume, for instance, that in a particular
situation one has to be able to lift a heavy weight in order to realize a goal.
A physically very strong person may succeed in doing this. But are we
inclined to say that he is healthier for this reason? No, we are not.
But given a different interpretation we can also acknowledge a di-
mension of health in the direction envisaged by Whitbeck. We can, if we
mean by different situations different sets of standard circumstances, for
instance as provided by different cultures. A person who is adaptable in the
sense that he can reach his vital goals in a great many standard environ-

ments could be said to be healthier than one who is less adaptable. Adapta-
bility is one kind of strength, but a kind which is closer to our ordinary
notion of health.
If health is a graduated concept, what becomes of its complement:
illness? Is there such a thing as illness or are there only varying degrees of
health? Whitbeck does not explicitly discuss this matter. It is obvious,
though, that she does not identify illness with lack of health in the way Porn
does. "Illness" is a term seldom used in Whitbeck's text. The following
quotation seems to indicate that Whitbeck wishes to restrict "illness" to the
context of disease.
Ifit [a disease] precedes all symptomatic episodes (episodes of illness or acute episodes), then
the disease at that stage is alternatively termed "sub-clinical". Now it is clear that during the
subclinical phase of a disease there will be no ill (or symptomatic) person to treat ([140], p.

Whitbeck seems here to adhere to Boorse's use of the term "illness". A

person who is affected and disabled by disease is ill. There is no indication
in the text that a person affected and disabled by something else, for
instance impairment, injury or motivational disturbance, should be called
Health and Value
Whitbeck, in particular, stresses the fact that her notion of health is
value-laden. Health is, she says, evaluative in the strongest sense of the
word; to use such a concept is to say that that to which it is applied is good
or bad. Disease, on the other hand, according to Whitbeck, is value-laden
only in a "capability sense" ([ 141], p. 614).
Porn's position in this respect is less clear. It is obvious that he would
say that health is a value-related concept. The assessment of health is
dependent on what Porn calls a system ojideality. A person's profile of goals
is an instance of a system of ideality. A person is healthy if his abilities are
adequate to his goals, if they are good according to his system of ideality.
Correspondingly, diseases are "evaluated as abnormal (poor, weak, etc.)
because of their causal tendency to restrict repertoires and thereby com-
promise health" ([99], p. 7.) Porn does not make a distinction between
health and disease from the point of view of values.
Is there an important difference here between Porn and Whitbeck? This
is a case where the choice of ontology for health may be important. To say
that something serves an end is to say that something is,primajacie, good.

It is certainly not to say that it is ultimately good or that it is good in any

important respect. However, from the point of view of the particular goal
it is good. To say that X promotes the attaining of a goal therefore is to
make a positive evaluation of X.
A person's abilities may serve the goals set by the person himself. His
abilities may therefore be said to be good. They are good in relation to his
According to Whitbeck, health is a person's ability to serve his goals.
Since his ability is good (from the point of view of his goals), his health must
also be good.
But what is to be said given Porn's basic position? For Porn health is
not identified with ability, but is a relation between ability, goal and cir-
cumstance. Porn may concede that saying that P is healthy involves an
evaluation of the person's ability vis-a-vis his goal( s); but it does not
necessarily involve an evaluation of the entity to which "health" itself is
applicable. It need not involve an evaluation of that relation which consti-
tutes health.
If there is an evaluation involved here it must be an evaluation distinct
from the instrumental kind mentioned above. We shall return to consider
this problem in Section 8. Let us for the present make the following
observation. The evaluative character of the concept of health (either as
conceived by Whitbeck or, possibly, by Porn) does not seem to prevent us
from assessing and measuring health by purely theoretical methods. One
can, at least in principle, assess a person's goals by observation and
theoretical analysis. The task of seeing whether an ability is adequate to
the subject's goals is also of an empirical, causal, nature. It consists in
studying whether, as a matter offact, a person, when he is trying, succeeds
in doing what he has set himself to do.
Assessment and Critical Review
(i) Preliminary considerations
Let us now make a preliminary assessment of this theory of health which
will be called the subject goal theory (the SG-theory). We shall disregard the
differences noted between Porn and Whitbeck and consider a unified
theory. This theory can be formulated in the following concise way: A is in
health if, and only if, A, given standard circumstances, has the ability to
realize at least those goals set by himself. A is not healthy if, and only if,
he is disabled from realizing any of the goals set by himself. (We shall

assume that the clause "under standard circumstances" is understood in

the same way as we have discussed above. Let us postpone a discussion
of the term "illness".)
Consider the following intuitive advantages of this theory.
(a) It provides a distinction between health and the other ability-
From the SG-theory it does not follow that all intelligent, talented and
strong people are healthy. Nor does it follow that all weak, untalented and
stupid people are not healthy.
There is, for instance, a great difference between the athlete who can
clear 2.30 m in the high jump and the ordinary man who can only jump 1.15
m. But both can be equally healthy (according to the SG-theory). The
reason for this is that the athlete and the ordinary man have very different
ambitions with regard to the highjump. The athlete's ambition more or less
equals his talent. What counts in the determination of health is not sheer
ability, but the relation between the subject's ability and his goals.
Let us note that this feature of the SG-notion distinguishes it sharply from
the need-notion of health. The need-notion could not account for individual
variation (over and above what is implied in the requirements of different
professions). By relating health to the subject's own goals individual var-
iation is maximally accounted for. (Though this involves problems to be
discussed below.)
(b) It provides a clear distinction between health and morality and
health and legality.
It is obvious that the SG-characterization of health does not say any-
thing about the moral nature of the goals set by the subject. The goals of
the healthy man need not be the morally (or legally) right ones; nor need
his abilities correspond to the morally (or legally) right goals.
The SG-notion thus seems to fulfill some minimal criteria of adequacy.
At the same time, however, it faces a number of substantial problems. We
shall consider here the following list of problematic cases:
(a) The case of unrealistic persons with too high ambitions
(b) The case of the person with extremely low ambitions
(c) The case of counterproductive goals
(d) The restricted applicability - the SG-notion is only a notion of
adult human health.

Before being able to analyse these problems properly we must give a more
precise interpretation to the notion of setting a goal for oneself.

(ii) On the notion of a goal set by the agent himself

Let us first observe that the two authors under discussion abstain from
giving precise interpretations. Porn discusses abstractly in terms of an
agent's profile of goals. There is only a short passage in which he indicates
that the goals are determined by the agent's faculty of will ([99], pp. 7-8).
Whitbeck uses quite different expressions in her characterization. We
have previously quoted the terms "goals, projects, and aspirations". In
another context we encounter "Therefore, the notion of health is closely
associated with the notions of autonomy, with the ability to act to achieve
one's purposes ... " ([141], p. 616).
The common and quite vague platform here is simply the following: A
goal set by a person is a state of affairs which the person wants to become
the case. In order to give a plausible interpretation of this idea we must,
however, make a fundamental distinction. We shall distinguish between
wants in a wider sense and wants - or intentions - in a narrower sense.
Most people have wants (we sometimes call them idle wishes) which are
completely unrealistic, and these people are also aware that they are
unrealistic. Somebody might wish that he were the world champion in a
branch of sports or that he could travel round the world; but he knows that
he cannot realize these wishes. As a result these goals do not affect his
conduct; he does not make any preparations in order to realize them. Such
goals are not goals in the narrow sense which we have in mind here.
Goals in the narrow sense are goals which w~ act upon unless prevented
from doing so. They are goals which we have decided or intended to realize.
(The two expressions "deciding" and "forming an intention" will here be
regarded as synonymous.)49
It may be contended that there is an intermediate position between
wants and intentions. A person may decide to try to perform an action
without really believing that he can succeed. (The minimal epistemic re-
quirement here is something weaker, viz.: it is not the case that he believes
that he cannot perform the action (or reach the goal).) A decision to try
is, indeed, a way of setting a goal which has a direct influence on conduct.
The important distinction to be proposed here then is the one between
wanting, on the one hand, and deciding (intending), or deciding (intending)

to try, on the other hand. We shall say that a person has set a goal for
himself only if he has decided (intends) to reach a goal, or has decided
(intends) to try to reach a goal. A goal set in this sense will be called an
intended goal. 50

(iii) A review of some problems with the SG-theory

(a) The case of the unrealistic person. Consider now the first item in our list
of problems. Should all normally-equipped persons, who happen to set a
number of unrealistic goals for themselves, be considered to have a low
degree of health?
Prima facie, on the SG-theory, the person who aims for the stars is not
healthy. Given our present conceptual background, we can, however, qual-
ify this judgment. A person who aims for the stars may only want to reach
the stars. He may be perfectly aware that the project is unrealistic and not
act in accordance with it. His goals are therefore not real goals in our sense.
Accordingly, this case of aiming for the stars is compatible with health.
It may also be the case that a person aims for the stars under the explicit
assumption that certain favourable opportunities should arise. Such a
person has set a hypothetical goal. If the favourable opportunities never
arise and the agent fails to reach his goals, this does not indicate lack of
The case of non-health to be described must presuppose the following:
(1) A has an intended goal G (i.e. A has set the goal G) (2) A has set the
goal G under the assumption that situation S obtains (3) S obtains and A
tries to bring about G (4) A fails to bring about G in S.
Do all cases fulfilling conditions (1 )-( 4) fulfil a set of sufficient conditions
for our intuitive notion of non-health? Are all people with ambitions that
are too high, ill? A positive answer to such a question would still seem
paradoxical if the ambitious man is an extremely vital person, who
succeeds in accomplishing a good deal.
Before accepting the ambitious and vital but unrealistic person as a
counterinstance to the SG-theory, we might want to introduce a new
qualification. It seems to be essential to assess the extent to which the agent
is justified in believing that a particular goal is attainable. If he is justified
in such a belief, and if he is justified in his particular beliefs concerning
means to ends, then we would not consider his goals to be unrealistic. Nor
would he be a candidate for being ill.

If the agent is unjustified, on the other hand, either concerning the whole
project, or concerning particular parts of the project, then he may be
considered to lack realism and judgment. The crucial question is, then,
whether such a lack is a sufficient criterion of illness.
A balanced answer to such a question must also take the following into
account: in giving the general assessment that a person is ill, we certainly
do not just consider one particular goal and the person's adequacy or
inadequacy vis-a.-vis that goal. We must find out whether or not equilibrium
obtains with respect to the whole range of goals that a person has set for
himself. In such a general assessment, the outcome concerning one or a few
parameters will have little influence.
A person whom we, on intuitive grounds, consider to be vital and
ambitious will most probably perform well with regard to many of his goals
(perhaps most of his goals in his private life), although he may be unrealistic
concerning some other projects (e.g. his professional ones).
This qualification, however, does not explain away the following conse-
quence of the SG-theory. A person who has set high goals for himself and
is in general unrealistic in judgments about his ability to reach these goals
has a low degree of health. This consequence indicates in a salient way that
illness on this theory may have little to do with disease as normally un-
derstood. A combination of a low degree of intelligence and an unfavour-
able character trait may be sufficient for considering the person ill.
Let us summarize our discussion about the unrealistic person. A person
who is not justified in believing that a particular state of affairs is attainable
and still makes this state an intended goal of his, is unrealistic with respect
to this goal. If such lack of realism obtains with regard to most of his
intended goals then he has, according to the SG-theory, a low degree of
This conclusion is probably controversial. In our opinion it is, however,
not obviously counterintuitive. In fact, we shall incorporate this element of
realism in our own positive account of health.

The following three arguments, however, seem to be fatal for the SG-
theory, unless the theory is considerably qualified.

(b) The case of the person with very low ambition. Does a person with very
low ambition and very few goals - which he can easily achieve - automati-
cally have a high degree of health? Does the man who is dying, and who
is - intuitively speaking - a very sick man, but who has become resigned

to his fate, have a high degree of health? Is adequacy of resources relative

to one's own goals always a sufficient indicator of health?
Consider a lazy person who chooses to lie on his sofa as much as he can.
His ambitions are minimal and restricted to necessities. His psychophysical
abilities are more than sufficient for realizing these goals in a wide variety
of situations. Indeed, this person may break his legs (and become what we
intuitively consider to be handicapped) without any change in his ability-
goal equilibrium.
Very seriously diseased persons, for instance persons in chronic coma,
do not seem to have any ambitions at all. Therefore, although they can do
nothing or very little, the adequacy of their actions is maintained.
Similar considerations can be offered concerning mentally-handicapped
persons whose faculty of will has been disturbed. Severely apathetic people
and people who suffer from katatonic schizophrenia are, intuitively, very
ill. But as long as their ambitions are minimal, the SG-theory cannot detect
their illness.
The cases referred to seem paradoxical since our ordinary intuitions
about health imply a minimal degree of vitality. The dying person and the
person in a coma are as far from being vital as is possible.
Can the SG-theory then be saved? Perhaps by an argument of the following
kind: the lazy person and the dying man do not really have such minimal
goals as they appear to have. They deceive themselves or they temporarily
conceal their real goals from themselves.
The proponent of such an argument has, however, the burden ofcharac-
terizing the notion of a real goal. He cannot, then, as we have done, identify
intended goals with real goals. According to our interpretation a real goal
is a goal which is, in fact, acted upon, unless there is an external hindrance.
The notion of a real goal that could be used to rebut our argument must
be quite different. We shall in what follows consider and accept such a
different notion in our positive account of vital goals. (See Section 8 in this
Let us now briefly deal with the remaining two arguments against the
SG-theory. Their contents will be further explored in the presentation of
the welfare theory of health.

(c) The case of counterproductive goals. People sometimes form intentions

which are contrary to some higher valued preferences of theirs or which
are even damaging to themselves. When a man decides to consume a great
amount of alcohol, it is likely that this decision is in conflict with some other

important aim of his, for instance that of performing his professional role
properly. If his consumption is excessive and enduring the result of his
decision will even damage his body or mind.
It seems counterintuitive to say that all kinds of irrational and counter-
productive goals, just because they are intended, should have the status of
vital goals. Must we not have some minimal requirement concerning the
"quality" of the goal?
This question will be given a detailed answer within the framework of
the welfare theory of health. (See Section 8 in this chapter.)

(d) The restricted applicability of the SG-notion of health. The SG-notion

presupposes in its subjects a capacity to set goals for themselves. Such a
capacity is a property peculiar to humans with the possible exception of
some of the higher animals. The effect of this observation is that the
SG-theory cannot be valid for all applications of the terms "health" and
"illness". This fact is particularly obvious in the case of botanic uses of the
terms. A plant which is healthy cannot be identified with one which is able
to attain the goals which it has set for itself.
Note, however, that this argument also holds for some humans. New-
born babies do not form intentions. When the concept of health is applied
to a newborn baby some other concept must be used. (For a treatment of
these cases, see Section 9 in this chapter and Chapter five, section 2.)



The General Idea

Let us recollect our main conclusions from the analyses of the two propos-
als on the vital goals of man.
The vital goals as determined by our basic needs: Prima facie merits
(i) This concept is a candidate for being a universal concept of
health. It is not necessarily restricted to human beings. If there
is some sense in talking about basic needs of humans there
ought to be sense in ascribing basic needs to animals and plants
as well.
(ii) It seems to admit a uniform application to all human beings. If
the basic needs are general, it seems not to be necessary to

make an investigation into any particular needs of an individual

in order to determine his health.
(iii) The notion of a basic need is, if at all defined, tied to the
concepts of survival and health. But the latter concept cannot
be used in a characterization of health itself.
(iv) Ifneed is conceptually connected merely to survival it seems to
cover too little. Man has vital goals far beyond survival.
We have observed that standard circumstances normally require one's
participating in some professional activity in order to survive.Therefore, as
a standard result of this conception, a person is healthy if, and only if, he
can fulfill his professional duties. Then the result is not completely unre-
asonable. It gives us a technical notion of health roughly equivalent to the
notion used by health insurance authorities.
Still, this notion covers too little. There are many uses of the term
"health" which go far beyond this technical use.
This observation led us to scrutinize the subject-goal conception of health.
Prima facie merits
(i) This concept is extremely sensitive to the multiplicity of human
goals and the variations in goal-setting among different persons.
It therefore answers well to the basic complaints concerning the
need-concept of health.lt can thus explain that old people can
be healthy and that some people who indeed manage to fulfill
their professional duties may still be ill. The determination of
health is completely dependent on the individual person's set-
ting of goals.
(ii) The SG-concept of health has a restricted application. It does
not go further than the realm of human beings; it cannot explain
the health and illness of animals and plants.
(iii) The sensitivity of the SG-concept to individual goal-setting
seems to be too great. We obtain counter-intuitive results both
in the case where the subject sets too low and limited goals for
himself, and in the case where he sets counter-productive or
damaging goals.
We shall now suggest a conception of health which is different from that
suggested by the need-theory and the SG-theory, out which retains impor-

tant elements from both of them. The key notion in this conception is
welfare; in the case of humans welfare will be equated with happiness.
Happiness should then be understood as a technical ,?oncept covering more
ground than does the ordinary concept of happiness.
The general idea is the following: The vital goals of man are those whose
fulfillment is necessary and jointly sufficient for a minimal degree of wel-
fare, i.e. happiness. To be healthy, then, is to have the ability to fulfill those
goals which are necessary and jointly sufficient for a minimal degree of
happiness. Observe, however, that this implies neither that health is suf-
ficient for minimal happiness, nor that it is necessary. Health is not suf-
ficient, since the ability to fulfill one's vital goals does not imply that one
actually fulfills them. And health is not necessary, since the vital goals can
be fulfilled by other means, for instance by the actions of someone else.
To relate health to happiness is to give credit both to the need-theory
and the SG-theory. In a way it could be seen as a new way of specifying
the goals of needs. Instead of relating needs to survival and health we now
propose to relate them to happiness. Our suggestion also includes essential
elements from the SG-theory. As we shall see in the subsequent analysis,
happiness is conceptually related to the fulfillment of the agent's goals. It
is important to note, however, that not all intended goals fulfill the require-
ments of being goals in the welfare sense. Moreover, there are some goals
in the welfare sense which are not goals aimed at by the agent himself.
Is, then, the question of what the vital goals of man are a theoretically
decidable question? Can any empirical investigation completely determine
this? Superficially, it may seem so. If the fulfillment of the vital goals is
necessary and jointly sufficient for minimal happiness, then our task must
be to inquire what happiness requires. What does Jones need to be happy,
and what does Smith need? The investigation may involve very difficult
methodological problems, but it seems, in principle, to be a theoretically
decidable question.
Our reply to this will be the following: the question of what the vital goals
are is only partially theoretically decidable. Roughly, it is decidable to the
extent that a modified SG-theory is correct. We can determine the vital
goals by theoretical means to the extent that they are identical with the
goals set by the agent himself.
The welfare-theory of health to be proposed here means, however,
precisely that this cannot be the whole story. The question of what consti-
tutes "real" happiness, as well as a minimal degree of such, can only be
answered by a primary evaluation; it is not a question of science. Consider

the following problem: a person with very low intelligence and a low degree
of vitality sets very few and primitive goals. Assume that he is able to fulfill
these goals and that the situation actually permits their fulfillment. From
the point of vie.w of mere goal-satisfaction this person is then a happy
It may, however, be contended that this is not enough. "Real" happiness
presupposes a minimum of complexity and subtlety. A person who fulfills
a very small and primitive set of goals does not fulfill the requirements of
minimal human welfare, minimal human happiness. Thus, ifhis abilities do
not supersede the ability to fulfill his primitive goals, then he cannot be
The problem here cannot, we believe, be solved by scientific means. It
can only be solved by a decision founded on an evaluation of what should
constitute minimal human happiness. Let us call this an evaluation of
welfare. It is, as far as we can see, an evaluation sui generis. It should, for
instance, be kept distinct from moral evaluation. An evaluation of welfare
does not require that the vital goals have any particular moral status. The
vital goals are primarily egocentric. They are goals the fulfillment of which
gives the agent himself satisfaction. This is certainly compatible with the
goals, being moral goals; but it is also compatible with their not being so. 51
The tr.eory which we propose has profound consequences for the philo-
sophy of health. A conceptual analysis cannot once and for all settle what
the vital goals of health are. It can bring us precisely to the point we have
reached in our discussion. It can tell us that the ultimate specification of
the vital goals must be left to an evaluation of welfare.
The basic conception of health offered by the welfare-theory is the
A is healthy if, and only if, A is able, given standard circums-
tances in his environment, to fulfill those goals which are neces-
sary and jointly sufficient for his minimal happiness.
Depending on one's platform of values these vital goals can be specified
in a number of ways. In this sense we can get a family of concepts of health
consistent with our basic theory. The theory itself cannot decide which
concept is to be preferred.
The important question, then, is: can anybody make this primary eval-
uation of welfare? Can we live with a great number of characterizations of
happiness and thus a great number of characterizations ofhealth? Such an
idea might seem disastrous to the science of medicine and the general

enterprise of health care. We shall say more about this later but let us now
just offer the following words of caution.
Evaluation is not a fortuitous affair. It should be made with the same care
and rigour as the making of a scientific investigation. Moreover, once the
basic evaluations of what constitutes a minimally happy life have been
made, then this has strictly logical implications for the particular appli-
cations of the terms "health" and "illness" in individual cases. Once a
speaker has made an evaluation, he cannot rationally change criteria from
one application of "health" to another.
Secondly, evaluations do not simply appear from nowhere; they are
formed in social settings. People within the same culture will tend to make
the same basic evaluations concerning the good life from a welfare point
of view. This will in most cases result in a great deal of intersubjectivity in
judgment. There will automatically be some common ground for "consen-
sus"-discussions about minimal and desirable degrees of welfare.
Thirdly, and as a continuation of the last statement, welfare evaluations
can be made explicit, and as in political affairs, be decided upon. In fact,
this is partly being done - although normally implicitly - in the social
policies of different countries. By legislating on social and medical matters
a government declares what it considers to be desirable and what minimal
levels of welfare it would tolerate without intervention. This amounts to,
among other things, specifying what it considers to be the border between
health and illness. (From now on the term "illness" will be used consistently
as the term contradictory of "health". This means that we follow the
proposals of Porn's in this respect. For a further discussion of the notion
of illness, see Chapter four, section 1).
Fourthly, there is a secondary use of the terms "health" and "illness"
which is not evaluative. This is when we ascribe health to somebody against
the background of a standard previously decided upon. Perhaps this is what
most of us do most of the time. We may presuppose some technical concept
of health or we may base our statement on what we take to be society's
judgment about the minimal limits of happiness. If we do the latter we are
not the evaluators; we make theoretical applications of standards given by
others who are the primary evaluators.
Considerations such as these will enable us to see that the enterprise of
health care is not really in danger. In order to see that medicine can remain
a science we shall have to make some further reflections based on an
analysis of the concepts of disease and impairment and similar concepts,
as will be undertaken below.

On the Concept of Happiness

(i) Introduction
The concept of happiness is as deeply rooted in the ordinary understand-
ing as the concept of health and rivals it in complexity. An important
difference between the two concepts, however, is that happiness has at-
tracted more scholarly attention than health. Happiness has a central place
in philosophical - mainly ethical - thinking. Most of the greatest thinkers
from antiquity to modern times have constructed theories or at least
well-articulated ideas about the nature of happiness. Within the framework
of this study there will be no attempt at really contributing to the theory
of happiness. For our purposes, however, we have made certain choices
between possible alternative characterizations of happiness. In making
these choices we have had substantial help from Anthony Kenny [63], V.J.
McGill [81], Georg Henrik von Wright [150] and Wladyslaw Tatarkiewicz
What kind of entity is happiness? When one says that a person A is
happy, what does one mean? Does one primarily mean that A is placed in
a certain collection of "happy" circumstances, or does one (primarily) mean
that A is in a certain mental state of "happiness"? To answer this question
is to take a position in the classical issue of whether happiness is essentially
an objective or a subjective state of affairs.53
During antiquity the predominant idea was that happiness was an objec-
tive state in the following sense: happiness is the possession of goods which
make possible the best life for man. The difficult question to answer here
concerned which goods the happy man should possess. Philosophers like
Aristotle and Plato had in mind primarily certain internal goods, such as
virtues and intellectual gifts; but a large number of external goods, such as
having a family, having friends and having some wealth, were also consider-
ed necessary for happiness. 54
The feeling of happiness - that which is often but perhaps not always
identified with pleasure - was in most ancient systems not conceptually
required for happiness. It was, of course, obvious - and pointed out by, for
instance, Aristotle and Plato - that the possession of goods normally
resulted in some satisfaction or pleasure for the possessor. But happiness
to them did not consist in satisfaction or pleasure. The "subjective" feelings
were only regular concomitants of happiness.
In contrast to this ancient notion Tatarkiewicz notes the following:

In modern times the concept of happiness was radically refashioned by being given a subjective
colouring. It ceased to be described as possession of goods but as a subjective sense of
gratification. A life is a happy one if we are satisfied with it, and it is the man satisfied with
his life who is qualified as happy. Whether or not he possesses goods and of what kind they
are makes no difference. As long as he is content he is happy ([127], p 33).

This modern subjective concept of happiness has oscillated between an

extreme hedonistic view - where happiness is identified with any kind of
pleasure - to a view where the pleasure of happiness requires some depth
and quality: happiness must be morally justified and the pleasures of
happiness should be intellectual and aesthetic rather than purely sensuous.
For the purpose of analysing a modern notion of health it is appropriate
to specify our concepts within the modern tradition of thought. Thus
'happiness' will here be understood as a subjective concept which necessar-
ily implies a feeling of happiness or pleasure.
The feeling of happiness, however, is not the whole story. Moreover, the
concept offeeling can be explicated in a number of ways. As will be shown
in the next section we can at least distinguish among feelings as sensations,
emotions and moods.

(ii) On the notions of sensation, emotion and mood

In modern philosophy of mind it is common to distinguish between the

above mentioned categories of mental phenomena. The main characteristic
of sensations is that they have a definite location in a person's body.
Therefore pain is a sensation because pain is in principle locatable to some
particular bodily part. Itchings, pangs and throbs are other kinds of locat-
able sensations. A mood on the other hand has an indefinite location.
Moreover, a mood concerns the whole person; it "colours" the whole
consciousness of the person. Such is the case, for instance, with the moods
of anguish and exhilaration. 55
Like moods, emotion')' cannot be bodily located. What is characteristic of
emotions, however, is that they have a direction. They are, as one often
says, directed towards objects, which are typically outside the person him-
self. This holds, for instance, for the emotions of love, hatred, envy and
grief. To be in love one has to be in love with somebody; to be in grief one
has to mourn something. Neither sensations nor moods proper are directed
in this way towards objects.
The fact that emotions are directed towards objects has proved to be
useful for the purpose of characterizing and classifying emotions. The

relation between an emotion and its objects is not a purely contingent fact.
A person loved or hated must, for conceptual reasons, have certain charac-
teristics or have a certain relation to the subject in order to be at all loveable
or hatable.
Let us consider the case of hatred more closely. Jones is said to hate
Brown. The following is true about Jones: he knows that Brown has helped
him in major matters throughout his whole life. He also believes that Brown
has sympathetic feelings towards him. He is not irritated by any particular
fact about Brown. He does not feel inferior to him. Given these facts we
must doubt the statement that Jones hates Brown. Brown is not the kind
of man that could be hated by Jones. Brown is not a proper object of hatred.
Our doubt about ascribing, or even our refusal to ascribe, hatred to Jones
does not depend on any external inspection of Jones involving, say, his
emotional behaviour. We draw our conclusion solely on premises describ-
ing the alleged object of a particular emotion of hatred. Brown is an object
of hatred from the point of view of Jones only if Jones believes that Brown
has contributed, or is about to contribute, to creating a state of affairs
which is highly negative for Jones. Without this prerequisite a case of
hatred cannot occur. (Observe that Jones' belief can, of course, be false.
Still, there must be a belief of this kind.)
The kind of abstract object, defined by a set of characteristics and
relations, delimiting the class of possible objects of hatred will in the
following be called the formal object of hatred. For all emotions there is a
corresponding formal object. 56
The distinction between moods and emotions is somewhat complicated
by the fact that certain mental concepts seem to occupy an ambiguous
position between the two. A typical example is depression. According to
authoritative judges there are depressions which are object-less, i.e. not
directed at any particular fact in the external or internal situation. On the
other hand certain instances of depression have an obvious object, a fact
which the subject is depressed about. An important question here is wheth-
er this really indicates that there are two completely different kinds of states
called depression, or, whether they have something important in common
motivating the identical label. If they have something in common, which
it is plausible to assume, the only candidate seems to be the experiential
state itself (or parts of it), the feeling of being depressed. The mood of
depression is characterized by a feeling of suffering; the emotion of de-
pression is partly characterized by an identical, or similar, feeling of suffer-

The preliminary assumption made here about depression is important to

bear in mind when we approach the difficult concept of happiness. There
are good arguments for maintaining that happiness is ambivalent between
being a mood and being an emotion.
We shall thus dismiss the hypothesis that happiness is a sensation. For
most uses of the word "happiness" this is quite obvious. When one is in
general happy or when one is happy about the course of events, this is not
a state of affairs which can be assigned to a part of the bearer's body.
The only candidates for being sensations are the so-called sensuous
pleasures derived e.g. from eating, drinking, smelling and perhaps from
sexual activity. The idea would then be that when one eats, the pleasure
is located in the mouth, in analogy with the way pain is located in a finger
when the finger is pricked by a needle. According to most writers on the
subject, however, not even sensuous pleasures have the status of sen-
sations. Although the cause of a sensuous pleasure can have a very definite
location, for instance on the tongue of a particular human being, the
pleasure itself has no such location. (For a discussion' of this topic, see
Consider now the hypothesis that happiness is an emotion. That happi-
ness has, or can have, an object is obvious. A man who is happy is normally
happy about, or satisfied with something. Tatarkiewicz writes that the man
who is genuinely happy must be satisfied with his whole life-situation. 58
Under this interpretation the formal object of happiness should be the
person's existence as a whole.
What conceptual requirements should then be laid upon this 'object'?
What must be true about a person's life-situation in order for it to be the
object of happiness? In order to simplify this analysis somewhat we shall
recognize that - in addition to happiness about one's whole life-situation
- there is a "molecular" use of happiness according to which one can be
happy about a single/act. (In this use "happy" is often synonymous to "glad"
and "pleased".) One can be happy about some achievement, or about the
weather, or about receiving a gift.
It is of course not true to say that a person who is happy about a single
fact is thereby generally happy. But there is a dependence relation here. A
person who is generally happy (in the emotional sense) is also happy about
a number of, to him, important facts. And the understanding of the former
must go through an understanding of the latter. What then is the formal
object of "molecular" happiness? First, a rough approximation: the formal
object of happiness is a fact which, as the subject believes, substantially

contributes to the realization of one of his wants or is identical with such

a realization.
Thus one may be happy about the weather because it makes possible a
number of things which one wants to do or have. One may be happy about
an action performed, because it means the achievement of a goal or a
substantial step towards such an achievement. Note that the object of
happiness need not be external to the subject's mind or body. For example,
one may be happy about one's own health, for the simple reason that one's
health, by definition, would contribute to the realization of many of one's
Let us now go into a fuller specification of the formal object of happiness.
Strictly speaking, it is not enough that the fact, which is the object of
happiness, make a substantial contribution to the realization of the sub-
ject's wants. It must also be the case that the realization of the want is
believed to be possible. If a desired goal is, for various reasons, believed
to be completely unattainable to the subject, in spite of a particular sub-
stantial contribution towards achieving it, then no happiness can arise.
In summary, then, our characterization says the following: X is an object
of the (molecular) happiness of A only if there is a desired goal G of A's
and A believes either that X constitutes the realization of G or that the
occurrence of X substantially contributes to the realization of G, and A also
believes that it is at least possible that G will be realized.
From this analysis we can then proceed to the molar and general notion
of happiness. If a man is happy about his whole life-situation then he
believes either that his whole life-situation constitutes a realization of his
wants or that his life-situation contributes substantially to such a reali-
The above treats of happiness as an emotion - as an intentional object-
directed notion. Does this exhaust the analysis of happiness? Must happi-
ness always be happiness about something? Test cases for such a con-
tention are those in which a man derives pleasure from some experience
or some sequence of experiences without that pleasure being the satis-
faction of a preconceived want. The simplest cases are those involving
sensuous pleasure. The pleasures of tasting (for the first time) a particular
dish or a particular wine can be completely unexpected. Similarly with the
pleasures of listening (for the first time) to a Mozart symphony or ex-
periencing a particularly beautiful sunset on a summer day. In cases like
this there has been no preconceived want which the subject satisfies by

taking part in a particular activity. Still, we could very well label these
mental states states of happiness.
There is an argument for the emotion-hypothesis here too. It runs as
follows: when for instance you are listening to a tune which makes you
happy there are two things which occur simultaneously. You realize that
there is here a desirable goal, viz. listening to this tune, and this goal is
immediately achieved. (For an idea in this direction, see [81], pp. 5-6.)
This analysis may be a plausible one for certain cases of very salient and
normally quite short-term pleasures such as the ones we have described.
The analysis is much more doubtful, however, when we consider the mild
pleasures of being active and continuously having new experiences, i.e., of
living a rich life. The subject "feels well" - arid this may be an enduring state
of mind - but he may not be conscious of the sources of his happiness,
which may be quite complex. Thus he may not be able to form any object
the materialization of which is desirable to him. Indeed, there need not be
such an object. (That there are cases of objectless feelings of well-being
becomes even clearer when we incorporate infants and the higher animals
among the subjects of happiness.) In the theory to be proposed here it will
therefore be assumed that there are objectless states of happiness, viz.
moods of happiness. Like emotions these can be both short- and long-term.
Observe that there may be important relations between a mood and an
emotion of happiness. First, the fact that a subject is in a happy mood can
be the object of the same person's emotion of happiness. When one is happy
about one's whole life-situation, the life-situation would include an enduring
mood of happiness. An enduring mood of happiness certainly contributes
to the realization of many wants, not least to the appreciation of this
realization. Contrariwise, a mood of happiness is dependent on some
degree of emotional happiness; a mood of happiness would be disturbed
or even disrupted by emotional unhappiness.

(iii) On the qualities and quantities of happiness

In analyzing happiness as an emotion - and in distinguishing between

molecular and molar happiness - we immediately realized that happiness
is a dimensional or graduated concept. There are various degrees of happi-
ness. In fact, happiness is multidimensional; the grading of happiness can
run along a number of dimensions. We shall first consider the dimension
of completeness.

(a) Completeness. Happiness - in the emotional sense - will be said to be

complete if the subject is happy with his total life-situation, in contradis-
tinction to merely some elements of it. In the latter case happiness is partial.
Partial happiness is obviously compatible with partial unhappiness or
misery. (Unhappiness is thus the emotion whose typical object is the
absence of factors contributing to the realization of one's desired goals.) In
order to ascribe some general happiness to a person, to say that he is happy
- although not ideally happy - we must require some balance in favour of
the satisfaction side. This need not mean some simple mathematical bal-
ance. It is not just a matter of counting the number of facts contributing
to one's goals. There can be happiness about a few important matters which
well outweighs the unhappiness over a great number of petty things.
The few matters can be important in two different senses. Either they
contribute to some intrinsically important goals, or they have a great causal
strength in contributing to some goals; they may even be identical with
some goals. This outweighing of unhappiness by happiness can be un-
derstood in two ways. According to a strong interpretation, the subject
should be aware of the outweighing, and consciously embrace the holistic
emotion describable as 'being happy about the total situation'. According
to a weaker interpretation the subject need not be conscious of the
outweighing. It may be merely that, as a matter of fact, he is - in the
completeness sense - more happy than unhappy. For the purposes of this
essay we shall not require the strong interpretation (Contrast [134]).59
(b) Intensity and frequency. Intensity and frequency are dimensions ap-
plicable to both the emotion and the mood of happiness. This is so since
they pertain entirely to the experiential side of happiness, the pleasure. The
pleasure can be more or less strong, and more or less frequentover a period
oftime. 60 This is - in the emotion case -quite probably connected with the
course of events, how lucky one is or how successful one is, but not
completely so. The intensity and frequency of certain feelings is also de-
pendent on the subject's internal disposition.
(c) Richness. The richness of happiness can be interpreted in at least two
different ways. One interpretation concerns the variety of pleasures, the
variety of experiences. The other, which is restricted to the emotion of
happiness, concerns the variety in the subject's profile of goals. The rich-
ness then consists in the multiplicity of goals being realized or promising
to be realized. 61
(d) Duration. Happiness - in both the emotional and the mood sense -
can be a more or less durable state. According to some treatments it is

conceptually required that happiness last for a long period. Thus Aristotle
says: "One swallow does not make spring, nor does one fine day; and
similarly one day or a brief period of happiness -does not make a man
supremely blessed and happy" ([3], 1.7.).
What are the conclusions to be drawn from this list of dimensions of
happiness? Does happiness then have an upper limit? Is there anything to
be called complete or total happiness? Theoretically a man can be com-
pletely happy or satisfied in the sense that all his wants have been satisfied.
In a sense completeness can also pertain to the dimension of duration. If
happiness lasts all through life, then happiness is complete from a temporal
point of view. (This completeness can, however, be somewhat spurious in
the case where life is very short.) Perhaps also frequency can be complete,
in the sense that there are no intervals, the feeling of happiness is contin-
uous. For dimensions such as richness and intensity there do not seem to
be any theoretical limitations. The limitations that there are, are psycholog-
ical, individual ones. A particular individual cannot appreciate more than
a certain amount of richness and intensity in experience.
A second and for the purposes of this essay more important conclusion
is the following: along all dimensions there is, or can be judged to be, a
lowest degree of happiness, below which there is no happiness at all. If the
pleasure is poor in variety and intensity, if there is a negative balance, so
that suffering dominates, then the person in question is not to be described
as happy.
On the Relation Between Health and Happiness
We have noticed in our analysis that the formal object of happiness-as-an-
emotion is the realization of the subject's goals. This fact is important in
the following two ways. First, it immediately shows an important link
between health and happiness. Health is, roughly, one's ability to fulfill
one's goals. Happiness (as an emotion) is a state which arises as a conse-
quence of goal-fulfillment. Thus, health must be an important contributor
to happiness. Second, this discovery constitutes, however, a theoretical
problem. We are still searching for a definition of a vital goal, and conjec-
turing that the concept of happiness could provide a basis for such a
definition. Now, in the analysis of the emotion of happiness, in terms of its
formal object, we come up with the notion of a goal. Have we then come
full circle?
We shall argue that there is no vicious circle here. Let us first specify
what kind of goal is presupposed in the characterization of the formal

object of happiness. "Goal" does not mean the same as "vital goal". It is
instead a state of affairs in general wanted by the subject. It need not be
an intended goal. Remember that that goal-realization which contribute.s
to happiness need not be performed by the agent himself. One may be
happy about receiving a gift which one has wanted to have for some time.
Or one may be happy about the good weather which one has been longing
for. It is impossible to intend to attain such states of affairs. Still, when they
occur one may be happy about them and they may contribute to one's
general state of happiness.
We can draw one important conclusion from this reasoning. Happiness
depends on much more than health. A high degree of happiness presup-
poses, in addition to health, good fortune. The course of events, be they
political, social or natural, must be in general supportive of the subject's
plans and purposes. But how then should the relation between health and
happiness be characterized? Our answer is that health is related to some
minimal degree of happiness on the part of the subject. 62 A person's vital
goals constitute the set of goals necessary and together sufficient for a
person's minimal degree of happiness. The vital goals thus constitute a
subset of a person's goals. A further characteristic of this subset is that it
must contain goals which are all in principle attainable through the person's
actions. The vital goals could therefore all be intended by the agent, but they
need not be, since the agent need not know - or need not consider - what
is required for his minimal happiness.
Let us illustrate this contention by considering the dimension of time. A
reasonable claim is that a momentary pleasure is not ipso facto an instance
of minimal happiness. Moreover, the fact which creates the momentary
pleasure may block the realization of long-term minimal happiness. Still,
many agents decide to opt for the momentary pleasure and - consciously
or unconsciously - abstain from realizing a necessary condition for long-
term happiness.
We shall consider this case of counterproductivity more fully in the next
section. Let us here confine ourselves to a simple illustration. A person aims
to acquire a bag of his favourite sweets. He achieves this and as a result
he derives some temporary pleasure. A later consequence of this, however,
is that his stomach starts aching and his general well-being is reduced for
a period which is much longer than the duration of the pleasure connected
with eating the sweets. In a case like this the aim of acquiring the sweets
is not a vital goal' of the person in question. It is not a necessary condition
of minimal happiness, since minimal happiness requires some duration.

More fundamentally, it is far from being a vital goal since it is a cause of

long-term unhappiness for the person whose goal it is. Still, it can very well
be an intended goal.

We shall now attempt to summarize the general relations between the

concepts of health and happiness.
A is in health if, and only if, A has the ability, given standard
circumstances, to realize his vital goals, i.e. the set of goals
which are necessary and together sufficient for his minimal
Observe that this basic conceptual relation allows for the following two
possibilities: A person who is healthy may be less than minimally happy;
a person who is ill may be very happy.
The first possibility is obvious from the fact that health implies only a
certain ability and not the corresponding action. A person in complete
health may choose not to realize his happiness. Another possible reason
is that the man in health may be in severe circumstances. Non-standard
circumstances, such as war or imprisonment, may prevent the realization
of his goals.
On the other hand, a sick person may well pass the limits of minimal
happiness. This is so mainly because such a person's vital goals can be
achieved through somebody else's actions. It is part of the purpose of health
care to compensate for the patient's loss of abilities. In health care one tries
to provide those facilities which are not only necessary for the patient's
survival but also for his general well-being.
But a sick person's happiness may also be due to other fortunate external
circumstances. The climate may be favourable during his sickness, the
political situation may be favourable, and there may be other features in
the course of events about which he feels pleased. 63
In sum, health and happiness need not go together. The existence of
health in a subject, however, considerably raises the likelihood of his
happiness. 64
Towards a Specification of the Vital Goals of Man
In specifying the vital goals of a person we must know what his minimal
happiness requires. But how is the level of minimal happiness to be deter-
mined? What exactly is welfare evaluation an evaluation of?

The evaluation in question should in principle be an evaluation of the

happiness of a particular person. The criterion of whether a certain state
of affairs is a vital goal of A's is whether this state is necessary for the
minimal intensity, frequency, richness and duration of A's feelings of hap pi-
ness. If the question concerns A's health or A's vital goals the evaluation
should concern A's happiness and nothing else.
It is important to keep this touchstone in mind while admitting the many
practical difficulties in assessing a particular person's feelings of happiness.
It is difficult to find adequate methods of measurement; what exactly
should be measured? Should we only measure behavioural characteristics
or should we also take personal avowals into account? Moreover, some
unlucky people may hardly ever be happy in the sense required by a
particular welfare evaluation. How could we then determine at all what
states of affairs are necessary for their "real" happiness?
Considerations such as these force us to take some pragmatic shortcuts.
In determining a person's basic vital goals, i.e. the kind of goals which are
vital irrespective of the person's own intentions, we can rarely rely on
individual studies. We would do better if we were to start by considering
some universal facts oflife. As a kernel we have those states of affairs which
entail the absolute necessities. Being alive is a necessary condition of being
happy, according to whatever evaluation. Hence, all the necessary con-
ditions for the maintenance of life must be included among every person's
vital goals, for instance, having food, having a sheltered home and having
some economic security.
Secondly, there are the states of affairs which have proved to add
considerably to people's happiness in a great many cases (irrespective of
their being the goals of the subjects themselves). Central among these
states are simple additions to the necessities: having some surplus food,
having greater economic and social security. The main considerations
behind labelling a state of affairs a "basic vital goal"are thus the following:
is it absolutely necessary for all happiness? Does it with a high degree of
probability, contribute to everyone's happiness?
But so far we have only spoken of empirical judgments. Where does
evaluation proper enter into the formation of the set of basic vital goals?
Evaluation proper comes in at at least two places. First, it comes in when
we put a limit on the amount of "necessities"; how much food and what
kind offood is necessary for minimal happiness; what kind of housing, and
what kind of economic and social security is necessary for minimal happi-

ness? This is obviously an evaluation, since minimal happiness is not an

objective state; it is a state to be decided upon.
Secondly, evaluation enters when we have the necessities for life (and
their quantitative extensions) and we put particular requirements on speci-
fically human ingredients in happiness. When we consider a certain degree
of education to be a basic vital goal, we claim that "real" human happiness
must involve some intellectual pleasures. Analogously, when we consider
some physical exercise to be a basic vital goal we claim that "real" human
happiness must involve some pleasures of bodily locomotion.

What are the vital goals over and above the basic ones? Here the
SG-theory gives the fundamentally correct answer. There are, however,
some important exceptions to it and we shall consider the most important
of these in the following.
(i) The case of intended goals which have been formed under
A subject is sometimes not free in setting one of his goals. He has been
forced or compelled to do so. Is a goal which is formed in such a way
automatically one of his vital goals? In answering this question we shall
consider some different cases of compulsion: first, compulsion in the sense
of internal prevention of deliberation.
This kind may have a number of varieties. The most important of these
is perhaps the case of the irresistible desire. We mean here a desire such that
the subject cannot think of anything else; his ability to weigh between
possible competing goals is blocked. The drug-addict and the alcoholic are
typical examples of subjects with, as we believe, irresistible desires. We
view them, at least in the extreme cases, as being compelled to consume
drugs or alcohol.
A quite different kind of compulsion, which is also relevant for the
theory, can be illustrated by the example of the gunman. A person threaten-
ed with a gun can be forced to perform a wide variety of actions; he can
be forced to perform things highly damaging to himself. And when he is thus
forced he is also forced to decide to perform these actions. This is com-
pulsion in the form of an external threat. What is the logical mechanism
behind this kind of compulsion? The person who is forced is placed in an
extremely grave situation. Unless he does what the gunman tells him he
risks losing his life. This means that his most central vital goal is threatened.
If the gunman is efficient there is, given the situation, only one way of

maintaining survival; the gunman must be obeyed; the victim must decide
to follow his order.
This is then a highly rational decision; obeying the gunman is acting in
the light of one's most central vital goal. But, as soon as the threat disap-
pears, when a normal situation obtains, the decision and the action may
seem highly counterproductive. It does not reflect what the agent basically
wants and therefore does not create any happiness on his part.
Our conclusion concerning these cases is obvious: a decision which has
been compelled, either in the internal or the external sense, does not, ipso
facto, reflect a vital goal. The purchase of drugs is not a vital goal of the
drug-addict simply because he actually intends to do so. Nor is the burning
down of a person's house a vital goal of his simply because a gunman has
forced him to do so.
It is, on the other hand, theoretically possible that one is forced to
perform an action which is in accordance with one's vital goals. An intend-
ed goal formed under compulsion may be identical with a vital goal. What
makes a goal a vital goal here is not whether it has been intended or not.
The touchstone is whether minimal happiness in the long run will be
(ii) The case of counterproductive goals
It is possible to intend to realize a particular goal, which does not create
happiness, but rather unhappiness, in the long run, without having been
compelled to form this intention. For such a goal we shall, in general, use
the label counterproductive goal. The counterproductivity can, however, be
of different kinds, from which we must draw different conclusions. Consid-
er first the case of accidental counterproductivity.
A person decides to build up a business, for instance to open a booksell-
er's shop. He succeeds in doing so, but after a while the business fails. It
has to be closed down and the owner goes bankrupt. The situation has
brought him unhappiness in the long run. Still, he has succeeded in realiz-
ing one of his intended goals, and this goal was not formed under com-
pulsion. Should we then say that opening the bookseller's shop was not a
vital goal for this man? No, we cannot draw this conclusion. One cannot
require from a vital goal that it must according to causal laws lead to
happiness in the long run. Other events may always occur to destroy the
effects of the realization of the vital goals.Thus the relation between a vital
goal and happiness must be formulated in a more careful way. A vital goal
of A is not a state of affairs which necessarily contributes to the minimal

happiness of A but only a goal which contributes, according to some

reasonable judgment, to his minimal happiness. We shall return to the
question of who makes this judgment.
Consider now the case of internal counterproductivity. A person conscious-
ly forms the intention, without any compulsion, to go out on a pub-crawl.
The result of satisfying this goal is that he is not able to do any work the
next day. He misses an examination and has to postpone one of his
important goals for a long time. The realization of the first goal leads in fact
to long-term unhappiness.
Here there is no question of an accident. Had the agent given the
situation a moment's thought he would have become aware of the conflict.
Maybe he was indeed aware of the conflict. He could, before making the
detrimental decision, have realized that the intended goal would not contri-
bute to his happiness. In a situation like this the realization of the pub-crawl
is not one of the person's vital goals. A more abstract explication of this
fact would run as follows. Most people have hierarchies of goals in the
following sense: Goal G 1 of A is considered by A to be more important than
goal G2 if and only if, in a case of conflict between the two, A would upon
undisturbed reflection prefer G 1 to G2 • (According to our interpretation A's
preference would be founded upon his judgment that G 1 contributes more
to his happiness than does G2 .) If in such a case of conflict A decides to
realize goal G2 , then he sets a goal for himself which is not one of his vital
goals. G 1, on the other hand, is a candidate for being a vital goal of A's,
unless there is a further goal G which is in conflict with G 1 and which upon
undisturbed reflection would be preferred by A.
There is a further case of irrational decision making but which does not
rely on any internal preference conflict. We can conceive of a man who aims
for an obviously detrimental goal- the realization of which would bring him
long-term unhappiness - but who does not, and cannot possibly, unders-
tand this fact. In this case there is no consistent internal hierarchy of goals
of such a kind that, if we were to point out their obvious consequences, the
agent would become aware of the conflict.
According to the welfare theory this kind of irrational goal must also be
excluded from the vital goals. This is then an important case when the
agent's own judgments must be disregarded and we have to involve some
external rational judges in order to determine what the agent's real vital
goals are.

(iii) The case of trivial goals

Should we then say that all intended goals which do not conflict with
higher priority goals or which are for other reasons not obviously damaging
are vital goals? This calls for some discussion of what we might call the
trivial goals of man.
During the course of their lives people quite frequently decide to do
things which seem fairly inessential. They decide to scratch their noses;
they decide to move a piece of paper from one part of a desk to another;
they decide to take a quick walk in the garden. These decisions may be
executed without any conflict with high priority goals, and they typically
do not cause any other damage. Moreover, their execution may create some
satisfaction. But are they vital goals?
The answer to this question must again be given through our interpre-
tation of minimal "real" happiness. Does scratching one's nose or the
moving ofa piece of paper contribute to the agent's happiness? No, typically
the satisfaction given is momentary. After a few seconds the whole proce-
dure has become unimportant to the subject. In such a case the goals
mentioned do not have the status of vital goals.
Let us, however, put in a word of caution here. Many seemingly trivial
goals are subgoals of major projects which have the status of vital goals.
A person's decision to make himself a cup of tea may seem to be an utterly
trivial decision. Still, having a cup of tea is part of his vital project of getting
his daily dose of nourishment. It is not strictly necessary that he has tea
in order to fulfill his vital goal. But it is necessary that he gets something
to drink. In this particular case, his having tea plays this role. Under the
description "having something to drink" his having tea is a necessary
condition for his vital goal of surviving.
Let us then propose the following criterion for distinguishing between
seemingly trivial intended goals: a goal is only spuriously trivial if there is
an adequate description of it under which it is a necessary condition for
satisfying a vital goal. The necessary subgoal is in that case also a vital goal.
Consider now the different cases from a more general perspective. We
have observed a number of instances where the realization of one (or a set
of) intended goal(s) does not lead to minimal happiness.These are then
counter-instances to a simple SG-theory of vital goals.
We have, however, also acknowledged that the criterion cannot be a
simple factual relation between the realization of the goal and the creation
of minimal happiness. Such a relation can never be foreseen with certainty

and we would therefore never be able to judge what the vital goals of man
The question of whether something is a vital goal must therefore depend
on some judgment about the probable relation between the realization of
the goal and the creation of happiness. This judgment can be made by the
agent himself, but it can also be made by external observers. Whose
judgment is to be followed?
Let us here separate the empirical problem from the evaluative problem.
Relevant empirical issues are, for instance, the following: is a certain state
of affairs stable or is it likely to disappear soon (example: does the business
aimed for have good prospects or is it likely to go bankrupt soon); does an
alleged vital goal stand in conflict with another high priority goal of the
agent; has the agent been forced to make a particular decision?
In these empirical matters anybody - the agent, as well as the external
observer - can be both right and wrong. In particular, the agent can be
wrong and the external observer be right. It is therefore quite possible for
an agent not to be aware of what his vital goals are.
In the standard cases the agent is more likely to be right than is an
external observer. The reason for this is that the agent has direct access
to more relevant information than the external observer has. Unless the
agent is highly irrational he will know much more about his own hierarchy
of goals than the external observer does; he is normally also a better judge
as to whether a decision has been formed under compulsion or not.
But as we have claimed, the judgments in this area are not only empirical
in nature. In particular, the question of whether a person's state of happi-
ness satisfies the conditions of being minimal "real" happiness is basically
not an empirical question. A subject may accept a certain condition and
say that he is happy enough; an external observer may question this,
perhaps on the ground that the conditions for happiness are so poor or on
the ground that the behaviour displayed by the agent is unconvincing.
This conflict of opinion is theoretically unsolvable unless the agent and
the external observer have together made some common decision about the
criteria for the existence of minimal "real" happiness. If they have agreed
on such criteria and been able to formulate them in intersubjectively
verifiable terms, then they have transformed the evaluative issue to a
theoretically decidable one.


(i) The dimensional aspect

By introducing the welfare notion of health we have not automatically

solved the problem of the dimensionality of health. Is health, as Porn
suggests, an absolute concept, and illness a dimensional concept admitting
degrees, or is health in itself a dimensional concept admitting a continuum
from optimal health to a minimal degree of health?
Let us first consider one aspect of letting happiness be the touchstone
of health. We have observed that happiness is in itself a dimensional
concept. We have also noted that in some of its dimensions it is open-end-
ed. And if health should be linked to happiness it would seem to be a direct
consequence that health is analogously dimensional.
There is, however, one important feature in our conception which prev-
ents our arriving at this consequence. Observe tl1at the notion of a vital goal
is tied to the notion of a minimal degree of happiness. There are a number
of goals - some of which are attainable through the agent's own action's
~ whose fulfillment would contribute to the agent's happiness, but which
do not belong to the vital goals of the agent. The reason why they are not
vital goals is that their fulfillment is not necessary for the agent's minimal
happiness in the long run: The fulfillment of these goals is not so essential
to the agent; it does not constitute the basis of his welfare, although it can
add to his welfare.
It is essential to our conception of health that there be a fairly sharp -
although in a way artificially sharp - border between minimal happiness
and additional happiness. If we did not make this demarcation we would
blur the distinction between health and ability (in its excellence sense). An
extremely able person can fulfill goals well beyond his vital ones, but this
ability, we would say, does not add to his health.
There is, then, a border - to be decided upon - which in the ideal case
defines a fixed set of vital goals (or, more precisely, a set which is fixed given
the circumstances). As a result health is at least not dimensional in the
open-ended sense. Thus, there is room for talking about health both as an
absolute state and as an optimal state.
The idea that health is an absolute state would in our terms amount to
the following: A is in health if, and only if, A has the ability, given the
circumstances, to fulfill all his vital goals. If this condition is not met A is

in some degree of illness, the degree depending upon the number and to
some extent the nature of the unfulfilled vital goals.
If health is interpreted as a dimensional concept the degree of optimal
health should be equated with the state described above simply as absolute
health. The idea of optimal health requires in turn some minimal degree of
health. This forces us again to make a decision. We must consider some
partial fulfillment of some goals to be minimally satisfactory, or we must
consider some goals as being less indispensable than others.
There is nothing in our analysis which forces us to decide in favour of
either of the two interpretations. A purely linguistic study would give the
answer that both the absolute and the dimensional discourses exist.
The English language, like many other languages, permits both the
locutions "complete health" and "optimal health". To speak of complete
health implies an absolute notion of health. Every state of incomplete
health is not really a state of health but a state of illness. Talking about
optimal health implies, as we have said, different stages of health which are
all stages of "real" health.
It seems as though the two discourses occur in different contexts. The
idea of health in an absolute sense (the idea of complete health) is favoured
in policy documents (including the famous document of the WHO).65 When
a policy document highlights the health of people as a goal, the goal
certainly is complete health (or, given the other conception, optimal
The idea of health as a dimensional concept is much more natural in the
pragmatic context of actual health care. The health-care system is forced
to take care of the most severe instances of illness. It must, therefore, make
pragmatic decisions in many cases as to when a patient is fit enough to
leave the system. The border thus drawn could be said to be that of a
minimal degree of health, which is indeed far from what everybody would
consider to be optimal health.

(ii) On contexts and the dimension of health

Observe now that there is a further important way to interpret the idea
that health is a dimensional concept. This is by dropping the relativization
of health to one particular environment. Instead of saying that A can fulfill
his vital goals merely in E, we can contend that A can fulfill them in a great
number of standard environments E 1 .•• En. In the latter case A has a higher

degree of health than in the former case. Optimal health, on this interpre-
tation, occurs when the subject is able to realize his vital goals in all
standard environments.
This observation calls for the introduction of a new notation. Health-
ascriptions which are tied to a particular environment should be explicitly
noted as "A is in health" or "A is in health EJ ••• En". The absolute statement,
"A is in health", should then - strictly speaking - be reserved for a person's
ability irrespective of standard environment.
The absolute variant of health ascriptions is not merely of academic
interest. We are sometimes interested in the degree of a person's adaptabili-
ty to different cultures and environments. Ifhe has a high degree of adapta-
bility he has a high degree of health in this sense.
Note that adaptability does not merely depend on the person's psycho-
physical strength. Adaptability is also a matter of realism: to what extent
is the subject prepared to change his ambitions, to reformulate them in the
light of a new environment. If the subject is adaptable in this sense some
lack of psychophysical strength can be compensated for and the equilib-
rium between abinty and vital goals can remain.
Observe the consequences of this for the idea of curing a person who is
ill. The process of curing normally implies the manipulation of some ability-
grounding factors, typically parts of human anatomy or physiology. But
given our present observation we can see how curing can be effected by the
manipulation of a person's ambitions (presupposing, of course, that the
ambitions are not reduced to a level under which no "real" happiness is

(iii) The distinction between mental and physical health

Health understood as a person's ability to fulfill his vital goals is neutral

as to the common distinction between physical and mental health. A person
who is healthy in our sense is both in physical and mental health as these
terms are traditionally understood.
Health always presupposes physical and mental order. This is immedi-
ately seen from our analysis of abilities. Practically all actions require for
their completion the execution of certain physical and mental mechanisms .
.Practically all actions, however physical they may seem, presuppose a
considerable set of mental conditions for their performance, in particular,
different items of knowledge and perception. (See our analysis in Chapter
three, section 2.) This means that all these actions, and as a consequence

the fulfillment of certain vital goals, can be prevented by mental factors,

as well as by purely physical ones. An attack of schizophrenia may, just as
well as a broken leg, prevent one and the same action and thereby the
fulfillment of one and the same vital goal.
This observation certainly does not prevent us from using the terms
"physical" and "mental health" for certain pr~ctical purposes. We can also
give an explication of what they should mean within the framework of our
theory. This is most easily done if we start with the cases of mental illness
and physical illness.
Mental illness is illness due to certain mental factors, for instance the
lack of rationality or a state of depression. Physical illness is illness due to
physical factors, for instance a broken leg or an infection. 66
What, then, is the sense of talking about mental and physical health
positively? That A is mentally healthy does not mean that he is generally
healthy; mental health certainly is compatible with physical illness. To say
that a person is mentally healthy can on our platform only mean that certain
compromisers of health are excluded. Ascribing mental health to A means
that, if A is ill, his health is not compromised by mental factors. The mental
background for his health is in order. Similarly, if A is physically healthy,
yet still ill, his health is not compromised by physical factors. We shall
return to the issue of compromisers of health in the next chapter.

(iv) Health and the notion of a defective will

Our concept of health is completely focused on the notion of ability. Have

we not then missed something of importance? What about a person who
has - as we say - a defective will, who sets goals for himself which are not
proper goals?
The concept of a defective will can have a number of interpretations and
our answers will be dependent on which interpretation is chosen.
(a) The sense ofa morally defective will. We have already responded to this
case. A healthy person need not be a moral person.
(b) The sense of an unrealistic will. We have discussed this sense at length
in connection with the theories of Porn and Whitbeck. The conclusions
from our point of view will be the following. The having of an unrealistic
will is not in itself a sign of illness. First, the will may be of the "idle wish"
or the "want" type, where the agent is aware of the impossibility of its
realization. Secondly, the unrealistic will may not, for other reasons, form
a vital goal of the agent. That is, the goal included in the will may not be

of the kind such that its fulfillment, ceteris paribus, is necessary for the
minimal happiness of the agent. This may be so, for example, because the
fulfillment of this will stands in conflict with the realization of some higher
order goal in the hierarchy of the agent's preferences.
But an unrealistic goal may be a vital goal. It may be the case that a
person sets an unrealistic goal of a high-priority kind. It need not be set
under compulsion and it need not stand in conflict with any higher priority
goal of the agent. Thus, the character of the agent may be such that the
realization of this goal is a necessary condition for his minimal happiness.
If now the lack of realism is genuine so that the agent does not even have
a second-order ability to reach this goal, then the conclusion must be that
he is ill to some degree. This conclusion can be derived directly from our
own conception of health. The agent is ill because of his lack of ability to
fulfill one of his vital goals. We need no extra supposition about defective
This conclusion does not mean that we should ignore unrealistic wills.
The agent's lack of ability certainly depends on his unrealistic will. The
unrealistic will is what compromises his health in this case. It is therefore
a candidate for being a mental disease.
(c) The sense of a "mad" will. Certain human beings want to do things
which are highly unusual; they may want to climb Mount Everest, for
example, or sail around the globe in a mini-boat. Sometimes the objects of
these wants will even have the status of vital goals for these persons. Would
we say that the existence of such unusual wills are ever sufficient conditions
for the occurrence of illness? Our theory gives a very definite answer. The
fact that a person's goals in general, or vital goals in particular, are unusual
is never in itself a sign of illness.
Sometimes, however, the pursuing of an unusual goal- as is certainly the
case in pursuing many usual goals - can create illness. This can be done by
means of two very different mechanisms. The first mechanism is the simple
causal one, which entails the pursuit of a certain goal damaging the agent's
mind or body so that his basic abilities are reduced. The other mechanism
involves the pursuit of the goal in a way which so absorbs the agent's
attention and energy that he is unable to fulfill some of his other vital goals.
Observe that this general reasoning also applies to such an extreme will
as the will to take one's own life. The having of this will does not in itself
entail illness. A difference from most other cases of mad wills is, however,
the following: the goal of taking one's own life can never become a vital goal.
Observe our definition of a vital goal. The satisfaction of a vital goal is a

necessary condition for the subject's minimal happiness. A necessary con-

dition for A's minimal happiness is that A remains alive. Thus, death can
never become a necessary condition for happiness.
(d) The sense of a weakfaculty of will. In one important sense of defective
will the whole faculty of will (and not any particular want or intention) is
not functioning well. According to one interpretation of this, the agent sets
too few goals for himself, or goals which are too low. This in its turn leads
to a very low degree of activity and vitality on the part of the agent. Can
such a case not in itself be sufficient to ascribe illness to the agent?
Our answer to this is that the sole fact that one sets very low goals for
oneself is not sufficient for illness. On the other hand, if we have reasons
to suspect that the agent cannot, for reasons internal to his body or mind,
set at least those goals which are identical with his vital goals, then there
is a case for his being ill.The argument for this contention is the following:
actions are in the standard case intentional; an action Ftherefore normally
presupposes for its existence the formation of an intention (a decision) to
perform F. If one is unable to decide to perform F, then one is unable to
perform F. The reaching of a vital goal V can, as we have explained, be
described as a complex action. Hence, if one is unable to decide to reach
V, then one is unable to reach V.
To this the following counter-argument could perhaps be raised: the
inability to decide to reach V is compatible with an ability to perform all
actions necessary and together sufficient for reaching V. And if one is able
to perform all such actions, then one must be able to reach V. The reply
to this is that an ability to reach V also requires an ability to perform all
actions necessary and together sufficient for V in a sequence and in the right
order. Unless the agent has decided to reach V and formed a plan for this
task, the execution of such a sequence in the right order will only be realized
accidentally. And if the goal is realized only accidentally, then we cannot
speak of an ability to reach the goal.
Our conclusion then is that if a person's faculty of will is defective in the
sense that he cannot decide to reach his vital goals, then he is ill. Observe,
however, that this does not imply that he must actually decide to reach his
vital goals in order to be healthy.

(v) Can there be a conflict between vital goals?

Can two vital goals come into conflict with each other? Or is a conflict
necessarily a sign that one of the two is not a vital goal? We have given a

partial answer above in the context of counterproductive goals. If it is

obvious to a minimally rational man that two intended goals are in conflict
with each other, then both goals cannot be vital goals of one and the same
Let us now qualify this standpoint a little. We must distinguish between
different kinds of intended goals and take the dimension of time into
consideration. First, two intended goals may be in conflict in the weak sense
that they cannot be realized exactly simultaneously. The fulfillment of one
must necessarily postpone the fulfillment of the other. This weak kind of
conflict is very common. In fact, few goals are such that they can be realized
exactly simultaneously. But, on the other hand, one rarely intends that
one's goals should be realized exactly simultaneously. Thus, this weak
sense of conflict rarely affects a goal's status as a vital goal.
A more important sense of conflict is the following: A intends to realize
goal G within the interval to - tlO; A intends to realize goal G t within the
interval to - t 10' The realization of G between to - t 10 blocks the realization
of Gt between to - tlO' Here G and G t are in a strong sense in conflict. Does
this fact prevent one of them from being a vital goal?
We need not draw this negative conclusion as long as the agent is
prepared to modify his goal-setting, perhaps in the light of some order of
preferences. Assume that the realization of G between to - t 10 has an
obvious priority, according to the subject, as compared to the realization
of G t between to - tlO.(As a consequence the subject modifies his intention
to realize G t by changing the time to tlO - t 20 . Thus the conflict is resolved.
GI may still be a vital goal of the subject.
GI will lose its status of being a vital goal only if the following condition
(i) G has a priority over G I ; the realization of G (as time-specified)
prevents the realization of G t (as time-specified) and th~ sub-
ject is not prepared to modify his intention concerning G I .
So far we have only discussed conflicts between intended goals. But the
set of vital goals also includes goals which need not be intended, viz. the
other goals necessary for minimal happiness. What about a conflict
between an intended goal and such a goal? If the non-intended goal really
is necessary for minimal happiness then it follows that an intended goal,
which is in conflict with the non-intended one in a strong sense, cannot be
a vital goal. In order to become a candidate for being a vital goal the
intended goal has to be modified in the way envisaged above.

The general conclusion thus is that two vital goals cannot be genuinely
in conflict with each other.

(vi) The welfare concept of health and the case of infants.

We noted earlier the following shortcoming of the subject-goal theory:

Since infants do not form intentions - and it is doubtful whether we can
ascribe any will to them whatsoever - the SG-theory of health seems
inapplicable to them. Can the welfare-theory overcome this difficulty?
The welfare theory stands a much better chance. First of all, it makes
sense to ascribe happiness to infants. Infants can feel pleasure and delight.
There are certain states of affairs which are necessary and jointly sufficient
conditions for their state of happiness. Some of these states - in particular
the necessities for survival- are universal for all babies. Others are individ-
ual. Some infants may, for instance, be harmonious merely in the company
of their mother; others feel security with several persons.
But how shall the relation between health and happiness be construed
in the case of infants? We cannot ascribe abilities to them. It is not true
that a healthy infant is able to realize its vital goals. In order for the vital
goals to be fulfilled, the infant needs continuous support from an adult. The
relation must be expressed in a weaker fashion. We suggest the following:
Infant I is in health if, and only if, the internal constitution and
development of I is such that, given standard adult support, the
necessary and jointly sufficient conditions for rs minimal hap-
piness are realized.
Similar reconstructions of the concept of health are net!ded for the case of
certain non-human beings. (See the discussion in Chapter five, section 3.)



So far we have suggested definitions of the general notions of health and

illness. We have stated that these are molar concepts pertaining to an
individual as a whole. A subject's health cannot be derived from an in-
spection of the details or the detailed functioning of his body. A person is
healthy if, and only if, he has the second-order ability (given standard
circumstances in his environment) to fulfill his vital goals. He is ill to
various degrees if this is not the case. But what is the relation between
health and the typical concepts of medicine and pathology? How are
"disease", "impairment", "injury" and "defect" related to health and illness?
Let us consider these concepts on a par here and collect them under the
common heading of maladies. 67 (For the choice of this term we are indebted
to C.M. Culver and B. Gert [27]). Maladies have one imp9rtant feature in
common apart from their occurrence in medical textbooks. They are simi-
lar in that they are entities internal to the subject's body or mind which tend
to compromise his health. A disease, etc., when occurring in S, tends to
cause some disability in S with respect to one or more of S's vital goals. 68
There are a number of features in this characterization which deserve
comment. First, the clause "tends to cause". The basic idea here of course
is that diseases do not always cause illness in their bearers. This can have
different explanations. It can be due to the fact that the progress of the
disease is aborted at a very early stage. The disease may be the kind that
causes disability in its bearer only in the late stages of its development.
A similar but not identical reason might be that a particular instance of
the disease is mild, that its effects on the subject's general ability are
negligible. Another reason would be that the particular bearer has internal
compensatory mechanisms, which are such that the disease-process does
not affect him.
This analysis, however, is not quite clear until the following ambiguity
is resolved: on the one hand, a disease can be understood as a type of
phenomenon; on the other hand, a disease can be understood as an instance
incorporated in a particular body. The normal labels of diseases, such as

"common cold", "cancer" and "tuberculosis", are general labels referring to

disease-types. Tuberculosis is, thus, a disease-type. This type can be in-
stanced in several individuals.
Now our dictum that a disease tends to cause illness can be understood
in two ways: first, disease-type D causes illness in most cases when it is
instanced. Second, disease-instance I of D will with high probability lead
to illness in the bearer of I. (We can conceive that D does not generally lead
to illness, but that I does so because of some properties peculiar to the
The first interpretation is to be adopted here. We shall say that some-
thing is a disease-type only if most instances of it actually compromise
health. Something is a disease-instance only if it belongs to a disease-type.
If we did not have this requirement it would not be possible to have a
general science of diseases. This does not exclude the possibility that other
kinds of processes can be of great medical interest. There may, for instance,
be a process-type which in a few instances compromises health. This
process-type may perhaps be called a medical risk factor.
(The case envisaged above, where there is a process-type of which there
are few instances contributing to illness, calls for a deeper analysis. What
are the properties of the subject who develops ill health from a process
which is normally harmless? Does he not have some basic impairment or
other kind of disease which allows this process?)
Considerations completely analogous to these apply also to other mal-
adies, such as impairment, injury and defect. A presupposition of a general
science of maladies is that it deals with types of entities, most instances of
which have illness as a consequence.
But the following crucial question could now be asked. How could such
a statistical fact be determined if health is an evaluative concept, and the
nature of vital goals can vary from individual to individual and from culture
to culture? The answer to this question can be given in two steps. On the
one hand, it must be conceded that there is some - but presumably not great
- variation as to the recognition of diseases between different cultures.
However, this is mainly due to different background situations in the
different cultures. But on the other hand, and most importantly, most
maladies strike their subjects in such a basic and general way that the
performance of most kinds of activities is impaired. For the task of identify-
ing these states as maladies then it does not matter very much what specific
goals one chooses as vital goals.

Consider first a physiological state which has the status of a malady in

a country like Sweden, but not in most North African countries. An
example of this is the state oflactase deficiency. Lactase deficiency causes,
in combination with ordinary consumption of milk, diarrhea and abdomi-
nal pain which both in their turn lead to illness. In the Scandinavian
cultures where practically everybody drinks milk, lactase deficiency in an
individual will cause illness in that individual. Therefore, in Scandinavia,
lactase deficiency ought to be included in a list of maladies.
In North Africa, however, people rarely drink milk. Therefore, lactase
deficiency seldom leads to illness. Consequently, in North Africa it would
be misleading to consider lactase deficiency as a malady. Lactase deficiency
there is an ordinary biochemical state of affairs, indeed statistically com-
mon, which has only very hypothetical relations to illness. 69 Note, however,
that this cultural variation in the recognition of diseases is not due to
differences in the choice of vital goals but to different background si-
As has already been indicated above it is difficult to see that different
choices of vital goals will very much influence the nomenclature of diseases.
this point can be sustained by some simple observations.
A great many maladies, perhaps a majority, cause pain, fatigue and
general unease. Such sensations and moods have a tendency to affect all
kinds of activity. They affect the general conditions for performing all basic
actions. And since every action requires the performance of some basic
action it does not matter very much exactly what the proper vital goals are.
Similar considerations could be raised concerning many diseases which
directly cause disability. The disability caused often concerns basic actions,
as for instance in general paralysis or a complete coma.
The cases where variation in goal-setting can affect the panorama of
diseases are special. The paradigm case would be the local disease which
does not give general pain or unease but only affects a particular ability to
reach a very particular goal, deemed vital in one context but not vital in
another. An example would be the following. A person is - for physiological
reasons, and not just due to lack of training - unable to move his ears.
Assume that in the culture where he lives the movement of one's ears is
an important element in the religious ritual. The person is then unable to
do something which is considered important in his society. He is disabled
and his disability is due to a disease.
In most cultures the ability to move one's ears is entirely irrelevant. This
particular talent is almost never practised. Therefore, a local defect pertain-

ing to the ear muscles would almost never be discovered and would not
even, we should argue, be deemed a disease or a defect. In the example
described, however, there is a case for speaking of a culture-bound disease
due to a culture-specific vital goal.
We shall now suggest the following characterization of the notion of
D is a disease-type in environment E if, and only if, D is a type
of physical or mental process which, when instanced in a person
Pin E,would with high probability cause illness in p.70

The above concerns the notion of disease, which has been chosen as a
paradigm notion among the maladies. However, almost everything in the
preceding analysis can be seen also to be valid for the other maladies:
impairments, injuries and defects. The essential difference between them
has to do with their ontological status?!
Impairments are mostly taken to be state-like entities, and are not, like
processes, continually variable. In the International Classification of Impair-
ments, Disabilities, and Handicaps (ICIDH) [144] an impairment is said,
normally, to refer to the endstate of diseases, to what remains after the
disease-process has terminated. According to the view of the ICIDH, an
impairment is also an "exteriorized" state of affairs, a state which has
become obvious and which is normally also a problem for its bearer. 72
Both Porn and Whitbeck suggest that an injury constitutes a change (cf.
our notion of event in the Appendix). 73 As we see it, "injury" can receive
either of two interpretations; according to the first, injury is a mental or
bodily state; a way of distinguishing an injury from an impairment - which
seems defensible from the point of view of ordinary linguistic usage - is to
reserve the term "injury" for states caused externally. According to the
second interpretation, an injury is a change; by using the term "injury" one
then focuses on the moment of incidence when the injury, as a state, is
induced in the mind or body.
By a defect is typically meant a mental or bodily state of affairs which
is congenital.
A possible reconstruction of this conceptual apparatus for technical
purposes, then, is the following:
A disease is a bodily or mental process which tends to com-
promise health.

An impairment is the endstate of a disease, which tends to

compromise health.
An injury is a bodily or mental change or state, caused external-
ly, which tends to compromise health.
A defect is a congenital bodily or mental state which tends to
compromise health. (For a more detailed discussion of the
concept of disease and the ontological categories mentioned
here, see the Appendix.)
Before considering some consequences of this apparatus for the concept
of health let us also consider briefly the notion of illness.
We have already adopted a general use of "illness", where illness is
equated with non-health. This was a use suggested in Porn's theory of
health (see Chapter three, section 7). To say that a person is ill is then
tantamount to saying that he is not healthy. Thus, illness is a matter of
degree. It can be mild or grave depending on how far the person's state is
from what is counted as the minimal or as the absolute state of health.
But "illness" is not only used generally, as in locutions such as "being ill".
It is also used to refer to specific entities, various particular illnesses.
What does "illness" in this sense designate? In ordinary discourse there
is here some vagueness and overlap with regard to the concept of disease.
In some contexts "illness" and "disease" are used interchangeably. This
holds particularly for the context of mental illness.
We shall propose here an explication according to which "illness" and
"disease" can be clearly distinguished. In this explication "disease" will be
characterized in the way already described; it refers to the psycho-physical
basis of a disability. "Illness", on the other hand, should be used to refer
to the various kinds of disabilities that there are. The general concept of
illness cannot be used for this purpose. It refers to the overall state of the
person. There is, however, also a need - not least for therapeutic purposes
- for being able to refer to a person's more specific disabilities. In what
respect is he disabled? What is the vital goal that he cannot realize?
The term "illness" can be used to combine a number of disabilities into
. a cluster (or a syndrome, which is the term often used in medical context).
Instead of enumerating each of the disabilities involved in a cluster one can
use a common term for the cluster, the illness-term, when this serves
scientific or therapeutic purposes. It is particularly convenient to use the
illness-term when a cluster is believed to have a common cause, in particu-

lar when all the disabilities involved in an illness arise from the same
The conceptual suggestion made here provides, in principle, a sharp
distinction between illness and disease. Illnesses are typically effocts of
diseases and not identical with them. This scheme has important conse-
quences for the field of mental pathology. In fact, most mental "diseases"
would turn out as illnesses in our terminology. The objects of psychiatric
classification are primarily clusters of disabilities, such as disabilities to
think coherently, to communicate, to socialize, etc. The real mental diseases
must be found among the causes of these disabilities.74
The conception of maladies presented here calls for a reconsideration of
the concept of health. In particular, a distinction could be made between
some different senses of ,health' which would do justice to some prevalent
medical intuitions concerning the notion.
In our main analysis we delimited a concept of health which was not
obviously a medical concept. Illness was defined as being due to internal
psycho-physical causes, but not necessarily to disease or to any other kind
of malady. The concept was deliberately kept open on this point.
To this general discourse on health we shall now add the specifically
medical one. According to it, the notion of health is typically tied to the
notion of malady in general, and the notion of disease in particular.
We are now in the position to clarify our discourse by introducing at least
two more narrow technical concepts of illness.

A is ill m if, and only if, A is, in standard circumstances, disabled

from realizing his vital goals due to the presence of a malady.

A is illd if, and only if, A is, in standard circumstances, disabled

from realizing his vital goals due to the presence of disease.

An interesting consequence of this conception is that A is healthy (when this

is qualified) cannot be characterized as A's actually being able to realize
his vital goals. It is still possible that something which is internal to him,
prevents him. Thus, the definition must be of the following kind:

A is healthyd if, and only if, A is, in standard circumstances, able

to realize his vital goals or if he is disabled from doing so, his
disability is due to something other than disease.

The characterization of "A is healthYm" would be analogous.

First a formal comment on these definitions. One might think that the
definitions are circular, since "disease" as well as "malady" are in their turn
defined in terms of illness. The difficulty here is, however, superficial. The
general concepts of health and disease have been defined in such a way that
no circularity appears. In the present context we are just making a partition
between those cases of illness which are due to diseases, i.e. psychophysical
processes which normally cause disability, and those which are due to other
A second comment is that we have presented merely two examples of
more technical, medical, concepts of health. The combinatorial possibilities
are great. Not all combinations are, however, equally plausible as interpre-
tations of standard discourses. Let us, in particular, notice the following
feature in ordinary discourse. There is a certain kind of mental defect which
is excluded from normal medical health care; this is the general intellectual
defect. Imbeciles and idiots are not treated in hospitals; in fact, they are
not treated at all, they are taken care of in special institutions subordinate
to social authorities other than the medical ones. The reason why they are
not treated is that there is no illusion of a "cure". General intellectual
disability is practically always believed to be incurable. (The typical ex-
ception to this is when this disability is caused by a temporary disease.)
Still, the intellectual defects would fulfill our requirements of being
maladies. The observation above then shows that the concept of malady
is still too general to capture the concept of illness as this is normally
referred to in the context of medical health care.
A refinement of the medical concept of illness could be the following:

A is ill (in a medical sense) if, and only if, A in standard

circumstances is disabled from realizing his vital goals, because
of at least one malady which is believed to be in principle

This definition also has interesting repercussions for our possibilities of

treating the phenomenon of old age within our general theory. This will be
discussed shortly.



Maladies are the most typical internal factors which compromise health.
But we have not excluded other factors. There may be other phenomena
internal to a subject which can compromise his health. The important
question is then: are there other typical internal compromisers of health
and, if there are, how are they to be distinguished from maladies? We shall
consider this question by analysing some phenomena which constitute
classical test cases in the philosophy of health: (1) old age or senility, (2)
pregnancy and (3) grief (together with some other negative emotions).
Consider first the degenerating process of old age. Prima facie it is a
candidate for being a disease: it is a bodily or mental process which tends
to reduce a person's abilities. It has some typical characteristics: it affects
the epidermic cells, which lose their plasticity; it accelerates the process of
atherosclerosis; it accelerates the process of necrosis in the cerebral
All these concrete degenerating changes could very well constitute or be
part of pathological changes. Still, old age has traditionally been dis-
tinguished from disease. Why? And how could this be expressed in the
theoretical framework of this essay? The major (pragmatic) reason for a
distinction between old age and disease seems to be that old age is the
inevitable fate of all human beings. In the long run, for reasons of principle,
senility cannot be cured. The "cannot" here is - at least so we believe -
irreversible. It is to be distinguished from what we today call incurable or
chronic diseases. In the latter cases we do not consider it impossible to be
able to cure them at some point in the future.
We do not know the mechanism of senile degeneration in all details.
(Although there have been notable advances recently, see for instance [22],
[34] and [52]).75 But we do in general believe that there is a special cause,
quite clearly distinguishable from most causes of diseases, which accounts
for this process. A plausible hypothesis is that this cause is to be located
in the genetic code as a specific program determining the life-span of man.
But how should senility be accounted for in our theory of health? There
seem to be two perhaps equally plausible strategies. The first one - suggest-

ed by our short discussion here - is to include senile degeneration among

maladies in general but exclude it from the specifically medical maladies.
If senile degeneration is believed to be incurable in principle then it is not
an object for treatment and not a medical affair in the restricted sense. If
we then wish to reserve the term "disease" for the kind of processes which
are in principle curable, we must acknowledge that there are other internal
processes apart from the diseases which compromise health.
The other important strategy takes its starting point in the variability of
human goal-setting, and thus in the variability of the requirements of
health. Even if it is true that the process of senility continually reduces
human ability, this is not equivalent to saying that it reduces health. The
explanation of this is that the ability required for health normally varies with
age. A perfectly healthy man of 75 cannot achieve as much as a perfectly
healthy man of 45.
The basic reason for this difference - at least as interpreted in our
theoretical framework - is that the old man adjusts his ambitions to the
new realities. As a consequence his vital goals do not require as much of
his basic abilities.
We can see that this observation does not hold exclusively for the final
stages oflife but equally as much for its initial stages. The abilities of a small
child must be measured according to a scale different from the one which
we apply to the adultlhe vital goals of the small child are different from
those of the adult.

Consider now the process of pregnancy. The nine months of pregnancy

involve for the normal woman some periods of disturbance. There is
initially some risk of nausea and towards the very end practically all women
have periods of severe pain. All these states cause in their turn disability;
the pregnant woman cannot normally go to work on the days just before
the birth of the baby; she may not be able to run her household. This
disability is due to an internal process in her body. But is it a disease?
In accounting for this we must see that the situation is complex. Most
women who go through the process of pregnancy want to have a child. The
goal of having a child is for them a high-priority goal, thus normally a vital
goal. The only way of realizing this goal is to pass through the process of
pregnancy. (In certain individual cases adoption is a way of "reproduction" ,
but for obvious reasons this is not a universal way out.) The process of
pregnancy is, therefore, in fact a necessary evil for the sake of a good end.

The important point in our characterization is that the process of preg-

nancy has (at least in most cases) been chosen voluntarily by individuals
for the sake of realizing an intended goal. An ordinary disease is not chosen
(apart from very untypical and extreme cases) for the realization of an end.
In a way the situation constitutes a conflict. It is not completely ideal.
A woman does not want the pain of pregnancy per se. But she considers
the goal of having a child to be so important that it is worth a sacrifice. It
is a goal which will give her happiness in the long run, well outweighing the
temporary unhappiness caused by the pain.
The disability of pregnancy is thus not an ordinary disability due to
illness; it is a disability which has been chosen in the activity of realizing
a highly important vital goal, the goal of reproduction. The case ofpregnan-
cy is a case where there is a conflict between two (or more) prima facie vital
goals; for instance, practising one's profession and having a baby. These
goals cannot be realized simultaneously. They could still both remain vital
goals if the woman is prepared to adjust her require~ents concerning
time-specification. This is indeed also what is normally done. Most preg-
nant women are content to postpone their professional activities.
But what then about the unwanted or unexpected pregnancy? This is not
an internal process which has been chosen by a subject in order to fulfill
the end of reproduction. Rather, it seems to be an accident, a highly
undesirable state of affairs.
Before jumping to the immediate conclusion that the unwanted pregnan-
cy is a malady, or even a disease, we should remember that according to
our conception the subject's choice does not completely determine his vital
goals. Having a child may be unwanted by A at a particular time, while it
may in the long run create happiness for A.
If, on the other hand, there is no prospect of the mother's becoming
happy with her child, then the situation could properly be called a state of
Observe, however, that from this it does not follow that pregnancy
becomes a disease. For something to be a disease we must require that it
belongs to a type which tends to create illness in its bearers. As we have
just stated, pregnancy does not typically lead to illness.

Consider now the states of grief, despair andfrustration. One of the criteria
of such negative emotions is that they tend to reduce the subject's general
ability. This reduction is accomplished mainly in two ways. One is that they
cause complete passivity, the kind of state we call depression. Another way

is that they cause outbursts which in their turn block systematic activity.
These emotions are thus mental states or processes which tend to com-
promise the subject's ability. But are they, ipso facto, maladies?
A cautious answer to this question will again tell us something about the
specification of human vital goals. First we make one distinction. There are
instances of negative moods, in particular of depression, which we believe
to be consequences of underlying impairments or diseases of the subject
himself. Such cases of depression belong to the pathological picture. They
qualify as illnesses.
Negative emotions, however, are responses to situations. Grief is a re-
sponse to a situation which involves a loss of something that the subject
holds dear. Frustration is a response to a situation which prevents the
subject from doing something which he considers to be important.
Emotions are, as we have previously said, directed towards objects and
these objects may be very limited. The loss of a particular physical object
may cause grief; the prevention of a very special course of action may cause
frustration. But although the objects (which are normally identical with the
causes) of these emotions may be very limited they are similar to happiness
in that they may colour the whole personality of the individual. This is why
an emotion may affect the general ability of the subject and thus have
consequences similar to the consequences of diseases.
But why then do we normally refuse to label the negative emotions
"diseases"? A tentative answer runs as follows. In contrast to other mental
categories, emotions are linked to the moral character of their bearers.
Emotions can be justified or unjustified in particular situations. It is highly
morally justified - or even morally commendable - to feel grief when a close
relative has died. Similarly, it is highly morally justified - or even morally
commendable - to feel happiness at the success of a loved one.
In general, emotional sensitivity, i.e. the disposition to have a rich em-
otional life including both positive and negative emotions, must be a goal
for every moral agent. He must prepare himself to become a sensible
person. This involves a disposition for acquiring negative emotions, such
as grief, should the course of events go against his desires. Thus, justified
grief is an indirect consequence of moral education. According to ordinary
intuitions, illnesses do not have such links to adequate moral training.
Although this may explain the common intuitions concerning grief and
some other negative emotions, it does not settle the status of these em-
otions from the viewpoint of the welfare theory of health. How can grief,

which may sometimes be highly disabling, be compatible with health as

conceived in the welfare theory?
The account to be proposed here has some affinities with the moral
analysis of emotions. It will be assumed that a person's welfare, i.e. his
happiness in the long run, presupposes some variety and richness in his
emotional life. Such richness and variety involves at times negative em-
otions; in fact, a disposition for positive emotions presupposes a corre-
sponding disposition for negative emotions. It is not possible to feel great
happiness at one's success without a disposition for disappointment or grief
at a corresponding failure.
Thus preparation for a life of overall happiness involves preparation and
training to become a sensitive person. Being a sensitive person involves a
disposition for acquiring negative emotions should the course of events go
against one's desires.
According to this line of thought a state of disablement caused by grief
is not illness, since it is an almost necessary concomitant of the emotional
sensitivity which is a precondition for a minimally happy life. For the same
reason grief is not a candidate for being a disease or illness.
To this should be added the following qualification. We sometimes say
that an emotion is exaggerated, either in intensity or duration; it does not
constitute an adequate response to the kind of situation which has evoked
it. If this is so the emotion is not an expression of the kind of emotional
sensitivity which is a vital goal of all humans. Rather, the emotion expresses
some weakness of character, and is a candidate for being a malady, possi-
bly a mental impairment.
Thus the solution proposed here parallels to some extent the solution for
the case of pregnancy. An important difference, however, is the following:
The state of pregnancy is directly chosen by most women who become
pregnant. Grief is never chosen for its own sake. What is chosen is that
emotional sensitivity which disposes one for a variety of emotions, but
which will also almost necessarily, some day or the other, lead to a state
of grief. 76

Let us now give a summary of our conclusions concerning the problemat-

ic compromisers of health:
(i) Senile degeneration allows different strategies within our theoretical
framework. First of all, it need not compromise health at all, since the vital
goals of the elderly may be fewer than those of the young. If it does
compromise health, and if it does so normally, then it is a candidate for

being a malady. There are, however, good reasons for putting this malady
into a special category. Contrary to the ordinary maladies we believe that
(at least many aspects of) senile degeneration is an inevitable process. Thus
it must have a very special status in medical treatment. Ifwe wish to reserve
the term "medical malady" for such states, events, and processes which are,
in principle, treatable, then senile degeneration is not a medical malady.
(ii) The standard case with pregnancy is that it does not compromise
health. The reason for this is that for most pregnant women the having of
a baby is a vital goal. Pregnancy is a necessary condition for this vital goal.
Thus, pregnancy cannot be a disease.
We did not, however, exclude the case where pregnancy is a cause of
illness. For some women bearing a child is not a vital goal. This does not
give pregnancy the status of a disease. To be a disease a process must
typically cause illness. This is not true of pregnancy.
(iii) Grief, and most other negative emotions, are not maladies. The main
reason for this is that the emotional sensitivity which disposes one for grief
is a precondition for a minimally happy life. Thus the disablement occasion-
ed by grief does not qualify as illness. We speak of illness in certain extreme
cases, when the emotions are exaggerated and prolonged. Then they may
be viewed as an expression of weakness of character, which in its turn is
a candidate for being a malady.




It is often claimed that the notion of health is related to a societal context;

indeed that every society determines its own notion of health. Analogous
claims have been put forward concerning disease and the other concepts
of malady.
On the basis of our analysis we are in the position to make a judgment
about the truth of this idea. We shall try to answer the question: in what
sense does a society determine its notion of health; in particular, to what
extent does this depend upon a decision on the part of the society?

(i) Society as a platform for a language game

According to a very influential theory of epistemology our capacity for

conceptualization, as well as the very process of conceptualization, is
dependent upon our language. We can form concepts, such as the concepts
of man, life and health, because we can produce symbols, for instance the
words "man", "life" and "health", representing these concepts. These sym-
bols are the building bricks of a language.
Ludwig Wittgenstein has a famous dictum according to which a lan-
guage, in order to exist, requires a context of communication between at
least two individuals. 77 There are no private languages, as he puts it. We
need not here argue for his position. Suffice it to say that all important
languages are as a matter of fact public; the natural languages are spoken
by very large groups of people living in one or more societies. The Sense
of the words in these languages, i.e. the concepts that they represent, are
determined by the way the words are actually used in communication. If
a lexicographer wants to define the concept of man in the English language,
he must study carefully how the word "man" is used in an English-speaking
society. As a result of this study he can characterize the concept of man.
In this basic sense, then, the concept of health is determined by the use
of the term "health", or its cognates in other languages, in the society which
employs it.

But observe that this dependence between the concept of health and a
language-using society is nothing peculiar to the concept of health. It
concerns all concepts as conceived in public languages. In the same sense
the concepts of stone, moon, electron, and animal are society-dependent.

(ii) Society as a background for action

A more interesting dependence between health and society is the one

analysed in our previous action-theoretic study (see Chapter three, section'
2). As was pointed out there, an ability is always an ability against a certain
background. A background may enable a person to perform a certain
action, but it may also make this action difficult, or even prevent him from
performing it. It is more difficult to build a house on top of the Himalayas
than on the prairies of Iowa. To claim that somebody is able to build a
house for himself then amounts to very different things in these two regions.
The same kind of ability (described generically) can in actual fact be
manifest as many different kinds of ability depending upon the environ-
In addition to the physical environment we have the societal environ-
ment. A society can enable one to pursue a particular course of action, or
render it impossible. What, then, are the consequences of this fact for the
concept of health?
Assume that G is a vital goal common to most people in both the
societies S \ and S2' and that G is much more easily fulfilled in S 1 than in
S2' The psycho-physical resources needed for realizing Gin S2 are much
greater than in 8\. To be able to achieve G in 8\ is then significantly
different from being able to achieve Gin 8 2 , To be healthy in 8\ (at least
as far as this variable is concerned) means then something different from
being healthy in 8 2 ,
The society, together with the natural environment, is one of the terms
in the relation constituting health. Thus, a man in good health in 8\ may
have properties quite different from a man in good health in 8 2 ,
Observe, however, that strictly speaking this does not mean that the two
societies have different concepts of health. In both 8 1 and 8 2 health is
defined in the following way: A is in health if, and only if, A is, given
standard physical and societal circumstances, able to fulfill his vital goals.
The concept is the same, but since what counts as standard societal cir-
cumstances differs in 8 1 and 8 2 the reference class of healthy people will
differ in nature for the two.

In conclusion, society determines some of the requirements for health by

constituting part of the background for action. This, however, does not
directly influence the notion of health.
Consider now the influence on the notions of disease, impairment, etc.,
in this respect. As we have said, if the requirements for health differ, this
also has, in principle, consequences for the application of the other notions.
It has probably quite little significance for the stratification of the notions,
i.e. for the variety signified by the taxonomy of diseases and impairments.
But it may have considerable significance for the drawing of the border
between disease and non-disease.
A bodily process does not turn into a disease until it has reached a
certain degree of severity. Severity must be defined in terms of probability
of causing illness. Assume now that the maintenance of health requires a
higher degree of psycho-physical integrity in society SI than in S2' This
means that a process of disturbance reaches the degree of severity much
more easily in an individual in S 1 than in S2' Hence, the borderline for being
a disease will differ between S 1 and S 2'
The same reasoning can be applied, mutatis mutandis, to impairments,
injuries and defects.

(iii) Society as an evaluator

Society is not only a passive background for action. It is also a source of

values; it contributes to our way of evaluating matters in life. What conse-
quences can this have for our way of conceiving of health ? We have already
noticed the evaluative component in the notion of health. To ascribe health
to a person A is to say that A can fulfill the necessary and jointly sufficient
conditions for his "real" happiness. We have said that what is to be counted
as conditions for minimal "real" happiness is partly a matter of evaluation.
But how does society perform this evaluation? Are not most ascriptions
of health made by individuals about other individuals? In what way is
society involved in this? First, a person's value system is influenced by
external sources. He is influenced by his educators, sometimes to the extent
that his value system exactly mirrors theirs. The educators in their turn
represent an ideology which may, but need not, be shared by a particular
society as a whole. It is important here to distinguish between particular
subsets of a society and the society as a whole. Societies are very rarely
homogeneous entities. In expressing an evaluation concerning the vital

goals of man, most individuals are highly influenced by the evaluations of

their fellow-citizens, which sometimes can express a common ideology.
The situation can, however, be different. In many cases when ascribing
health to an individual the ascriber does not really express an evaluation
of his own. Instead he presupposes the existence of the kind of evaluation
common to most individuals in the society. He presupposes some tacit
agreement that, for instance, the ability to run one's own household belongs
among the "vital" abilities, and hence must be a prerequisite for health. The
ascriber does not carry out the evaluation; he uses a given system of values
and makes a theoretical application in particular cases. If this is the case,
his dependence on external sources is greater than in the former case. Here
he completely takes over a system allegedly shaped by some subsection of
the society and uses it without adding an individual judgment of his own.
Consider now these cases from a conceptual point of view. How should
this influence be described? Shall we say that society, by expressing eval-
uations concerning human welfare, can in various ways change the concept
of health? But this cannot be a correct description of the situation.
By influencing a person's evaluations concerning what the vital goals in
life are, society does not ipso facto influence the sense of the expression
"vital goal" or of "health". Analogously, by telling someone what is good,
society does not give a new meaning to the word "goodness".
What can happen, on the other hand, is that the reference-class can be
changed. Depending upon one's choice of vital goals the class of people
supposed to be healthy will to some extent differ. Whether an advanced
athlete's extreme ambitions are or are not to be included among his vital
goals will definitely affect his "health-status". If they are, he will much more
easily enter a state of illness than if they are not.

(iv) Society as an explicit evaluator

We turn now to the interesting and difficult question of whether society in

fact makes basic evaluations concerning human welfare. This question
should be divided according to different specifications.
(a) The evaluations of political parties. Various subsections of society, in
particular political parties and other ideological organizations, often make
very explicit evaluations concerning human welfare. In their social policy
programs they specify certain goals which should be attainable for the
members of the society.

One thing to notice here, though, is that these goals are not always
attainable given the society here and now. The goals of such programs
normally presuppose the effecting of important societal changes. One can-
not therefore immediately interpret these goals as the vital goals characte-
rizing health in society as it is. They may perhaps characterize the vital
goals of tomorrow.
The programs can, however, give a clue to the general dimensions along
which the vital goals may be found. Sometimes the desired goals oftomor-
row can be seen as extensions and further specifications of the vital goals
of today. Consider, for instance, the general dimension of education. Ac-
cording to all political programs, a certain minimal degree of education
belongs to the vital goals of man. What distinguishes the programs and
what distinguishes the situation today from a desired situation tomorrow
is the exact position of a minimal degree of education. Still, the programs
tell us that education and learning as general dimensions belong to the vital
goals of man.
(b) The evaluations of society as a whole. A party's political program may
become the program of a parliament. Some of its aspects may even become
law. When this is the case we have the paradigm of an evaluation made and
codified by society as a whole. The use which we can make of such
programs and laws is analogous to what we said above about programs in
(c) Evaluations and decisions made by health authorities. It could be pre-
sumed that the most direct way to judge what the evaluations of the society
concerning welfare are is to observe the decisions and actions actually
taken by the health authorities. Now this procedure is both direct and
indirect. It is direct in the sense that the actions and decisions of health
authorities really concern what is to be evaluated as health today. It is
indirect, however, in the sense that the health authorities do not normally
formulate the vital goals of man in the way political parties try to. Instead
they make concrete decisions in particular cases. They decide from case to
case whether an individual fulfills the criteria for societal health care. From
these criteria one can only indirectly draw conclusions concerning the vital
goals of the individual.
It is important to note that when we talk about the health authorities we
almost invariably refer to medical health authorities, consisting mainly of
medically-trained personnel. Such authorities tend to use one of our more
narrow notions of health. That is, illness to them is a kind of disability that

is caused by any of the "medical" compromisers, the maladies: disease,

impairment, defect, etc.
This means that their attention is mainly drawn to the biological (or
psychological) states or processes themselves. The occurrence of these
states or processes tend to be the fundamental criterion of illness; and the
disability criterion, which we have singled out as the key criterion, is
seldom used.
However, no health authority works exclusively this way. Obvious cases
of disability are taken care of even in the absence of a known disease,
impairment or defect. A person who cannot move from his bed or a person
who cannot take care of himself in the most basic respects is considered
to be ill from a medical point of view, irrespective of observations of disease
or impairment.
Decisions on this latter ground would then give us better clues for
jUdging the evaluations of the health authorities. But this procedure also
has limitations. Health care is as a rule offered to people who seek help,
who have themselves recognized a disability and want the health-care
system to rehabilitate them.
This is in general a wise procedure from more than one point of view.
The most important aspect is the ethical one. Public authorities should not
intrude on an individual's integrity. The general rule must be that only those
people who wish to be helped should be helped. The procedure is also wise
inasmuch as it actually covers most of the existing illness in the population.
As we said in our discussion of vital goals, most vital goals are recognized
as such by the subject himself. If the subject is disabled so that he cannot
realize them, he will consequently suffer from this fact and seek competent
On the other hand, the fact that health care has this fundamentally
voluntary nature (with the exception of gravely mentally-handicapped and
diseased people) prevents it from being a sharp tool in judging prevalent
health evaluations. There is no presumption that the health-care system
takes care of everybody who is ill in a population (nor even all types of
illness). Moreover, the health-care system sometimes - perhaps for huma-
nitarian reasons - takes care of people who do not fulfill the conditions of
being ill. A paradigm case of this latter sort would be people who have
become disabled because of extreme environmental pressure (who live in
non-standard circumstances) and who indeed need help but not necessarily
medical help.

Let us, however, also consider those untypical cases where society
"offers" health care without the subject's having asked for it. The paradigm
here concerns people who are gravely ill or handicapped for mental rea-
sons. The majority are people whose disability is obvious to everyone
irrespective of ideological background; the disability concerns here such
basic vital goals as being able to take care of oneself, being able to orient
oneself and communicate with others.
There are, however, interesting exceptions to this rule. We have reasons
to suspect that the health-care system in some countries is used to take care
of people who are not handicapped in this obvious sense. The typical
characteristics of the subjects in question are that they do not conform to
certain norms laid down by the state. They may oppose the prevailing
ideology of the state, or they may simply be criminal or immoral.
When such procedures are customary in the health-care system they
become relevant to our conceptual issue. This depends, however, on how
the situation is interpreted by the state and the health-care system in
question. Assume that the situation is interpreted in the following way: the
deviant person under scrutiny is not merely a political opponent. He is
cognitively disabled; he does not understand the important vital goal of
contributing to the x-ist ideology. This cognitive disability must be cured
and he must be attended by medical personnel.
If this is the interpretation, the society in question is still using the
standard concept of health. What is peculiar here is the specification of a
human vital goal in conformity with a particular political ideology.
Consider now a different interpretation: the political opponent is not
considered disabled. He is viewed simply as a nonconformist who is dan-
gerous to the state. The most convenient - and perhaps the most efficient
- way for the state to take care of him is to let the health-care system do
it. This can be done either in a completely pragmatic way without involving
the notions of health and illness, or it can be done by actually extending
the notion of illness to cover cases like these. People who are dangerous
to the state can be defined as being ill.
If the latter procedure is the one which is adopted, then there has indeed
been a conceptual change effected by a society. We have a completely new
notion of health added to the standard notion accepted in all societies. It
can however be doubted whether this redefinition is ever made, since
standard pathological terms are often used to signify the mental states of
the subjects in question. The label "schizophrenia" is a common label
attached to such individuals. And a typical feature of schizophrenia, as this

term is normally used, is that it involves grave disabilities with regard to

thinking and communication. It would therefore be surprising if the
"political schizophrenics" should not be regarded in a way analogous to
ordinary schizophrenics. This would mean that the concept of health
employed in the case of dissidents is still the traditional concept. 78
But if this latter interpretation is correct, what is then the ground for
considering this a misuse of the concept of health? The answer is that
"misuse" is a proper description of the procedure, if the whole undertaking
is insincere. It is quite probable that the government in question wholly
understands that the political dissidents are not disabled in the ordinary
sense of being ill. It may also be the case that it does not seriously propose
a new vital goal in terms of conformity to a certain ideology. The govern-
ment or its health authorities may quite consciously decide to take care of
the dissidents under afalse label. The rationale of this procedure may be
efficiency; the ordinary awkward system of justice can be circumvented.
Ultimately it can perhaps be justified in terms of long-range political
reasons. But if the condition of insincerity is fulfilled no new concept of
health has been introduced. Instead, the old concept has been consciously
In summary: we have noticed that there are cases where society makes
explicit evaluations concerning the vital goals of man, but that these eval-
uations are rarely developed in a systematic way. The vital goals of man
are either formulated in a utopian or at least forward-looking way (as in
political programs), or they are merely expressed in an implicit and incom-
plete way in the actions and decisions of health authorities.
But what conclusions can we draw from this concerning our conceptual
issue? The conclusion is analogous to the one expressed in subsection (iii).
Explicit evaluations concerning the vital goals of man have obvious effects
on the reference class of healthy people. The basic concept of health is not
changed in any way.

(v) Society as a decision-maker

In the previous chapter we discussed the role of society as an explicit

evaluator of the vital goals of man. We concluded that there is probably
no section of society which has made a systematic attempt to spell out the
basic vital goals in detail. This of course does not prevent its being done
in the future. Indeed one can say that the upshot of our analysis is that in
order to be clear about health, we must make a number of decisions. If it is

true that health is not an absolute biological concept, but instead a three-
place predicate with argument-places for a standard environment and for
a set of vital goals, the necessity becomes great for further specification and
evaluation. Ifwe want discourse on health to become clear, we must know
what are to be counted as standard circumstances and we must have the
vital goals more exactly specified.
For certain technical health-concepts the request may already have been
satisfactorily met. Consider the specifications made by health insurance
authorities. Here the situation is simplified since there is only one vital goal
in focus. And this goal is quite clear: it is the fulfillment of the subject's
professional role. A person who is unable to fulfill his professional role is
ill from the point of view of health insurance. This is a simple and clear
criterion as well as an individually sensitive one. The professional role does
not mean one and the same thing but very different things to different
There are, however, more ambitious attempts to respond to the require-
ment expressed here. Consider, for instance, the previously mentioned
work issued by the WHO, which is a tentative manual for the Classification
of Impairments. Disabilities and HandiCaps (ICIDH) [144]. Admittedly, this
work does not express itself exactly in the terms suggested in this essay.
Nor is its explicit purpose to define health, or to list the vital goals involved
in health. The purpose, however, when properly analysed, is quite similar.
In an attempt to arrive at an exhaustive taxonomy of handicaps the
authors of the ICIDH suggest that every individual living in a modern
society must fulfill a number of roles in order to survive (or at least in order
to survive without direct assistance). These are called survival roles.
If the individual turns out to be more or less unable to fulfill these roles
then he is more or less handicapped. The manual even suggests procedures
for measuring the degrees of handicap along a scale from 0 - 9. We can
illustrate the technique of the ICIDH by describing the survival role of
economic self-sufficiency and the various degrees of handicaps which can
pertain to it. 79
Definition: Economic self-sufficiency is the individual's ability to sustain customary socioeco-
nomic activity and independence. Scale categories: [The text is here highly abbreviated].

o Wealthy
1 Comfortably well-off

2 Fully self-sufficient

3 Adjusted self-sufficiency

4 Precariously self-sufficient

Includes: individuals who remain self-sufficient only by virtue of appreciable support from
or dependence on financial or material aid from other individuals or the community ...

5 Economically deprived
Includes: individuals who economically are only partially self-sufficient because their income
or possession of financial or material aid from other individuals or the community meets only
part of their needs ...

6 Impoverished

Includes: individuals who economically are not self-sufficient by virtue of being totally
dependent for financial or material aid on the goodwill of other individuals or the communi-
ty ...

7 Destitute
Includes: individuals who economically are not self-sufficient and to whom support from
others is not available, so that their disability status is further aggravated.

8 Economically inactive

9 Unspecified.

The general procedure here is similar to what we suggest. To be handicap-

ped is - in our terminology - tantamount to being ill. Fulfilling survival roles
is (roughly) the same as fulfilling some basic vital goals. The manual thus
provides a list of vital goals, which are the ones to be realized by healthy
people. The manual goes even further in offering some methods for the
measurement of illness.
The ICIDH is thus a step in the right direction as regards improving our
discourse on health. This does not mean that the manual is altogether
satisfactory. As has been argued in Chapter three, section 4, the set of vital
goals cannot be identical with the set of survival roles. There are many
goals seriously aimed for by single individuals which are not connected with
mere survival. Moreover it can be doubted whether all the so-called
survival roles of the ICIDH are well-defined, or that they are on the same
level of abstraction. 80



General Consequences
The analysis of health and society presented here has already shown some
obvious consequences of our conception of health for health care and
A most crucial element in the analysis is that health is not a purely
biological concept. It is a three-place notion involving evaluations concern-
ing vital human goals; moreover it contains an essential reference to an
environment involving a society. Thus the clarification of the discourse on
health cannot be made merely by improving our biological or broadly
medical conceptual apparatus. As was noted in the previous chapter, there
must be a standing request to each health-care system to specify the basic
vital goals to be adopted in a particular society. These requests are
profound and entail a major demand on the social politics of the governing
bodies in society.
Without such a specification health care and clinical medicine cannot be
a clearly defined enterprise. As a result there may be very different and
unclear ambitions on the part of health-care personnel. Some of them may
interpret their task as consisting only of bringing a particular disease
process to an end; others may include in their task an attempt to rehabili-
tate a subject to a certain level offitness; yet others may include an attempt
to prevent new health risks from occurring, and so on.
But what then are the consequences for clinical medicine in a more
limited sense, and for the science of medicine? By clinical medicine in a
limited sense we mean the enterprise of diagnosing diseases and impair-
ments, treating disease and making prognoses. Is this enterprise affected
in a major way by our notion of health? Does it, for instance, mean that
the physician himself must take part in the procedure of formulating the
vital goals of health?
Our answer is no. The physician could and should remain a technician.
The formulation of vital goals should be in the hands of the individuals
themselves and - concerning certain basic vital goals - in the hands of
politicians and policy-forming health authorities. The clinician and the
medical scientist should instead work in the light of such a well-defined
concept ofhealth. Given a well-defined set of vital goals the clinician can use
his expertise in theoretical judgment: what bodily and mental states are
risk-factors for a subject's ability to realize his given set of vital goals? What

states have such a high probability of disablement that we should call them
Observe here the clear connection between the bodily and mental states
and disablement. This involves a sharpening of the request to the medical
practitioners. Not any abnormal bodily state is a disease. There is, accord-
ing to the philosophy presented here, no reason to make a bodily state into
a disease just because it functions in some sense abnormally or subnormal-
ly. We should do so only if experience tells us that there is a high probability
of the state's causing disablement in the subject.
The choice of vital goals can, in principle, also affect the set of diseases
and impairments. There may be some cultural variations in deciding what
should be counted as diseases. But, as we have already noticed, it is not
likely that this variation will be very great. The main reason for this is that
most acknowledged diseases and impairments strike their subjects in a
basic and general way. Consider, in particular, all those diseases and
impairments which cause pain and fatigue. Pain (even "local" pain) and
fatigue strike the subject as a whole. These sensations make all kinds of
activities (except perhaps certain kinds of omissions) difficult for him. The
result then is general passivity. Then the exact nature of the vital goals that
the subject should achieve will be of little importance. He will be ill in
whatever culture he finds himself.
When it comes to the stratification of diseases and impairments, when
we deal with particularities of taxonomy, the influence of our philosophy
will be even less. The grounds for the division of diseases and impairments
are not connected with particular vital goals. They concern essentially
anatomical localization or etiology. Out of the 17 main categories in the
International Classification of Diseases (ICD) [143] about 10 are formed
according to the principle of anatomical localization, whereas 2 or possibly
3 are formed according to an etiological principle. These facts are certainly
purely biological (or, in the case of mental diseases, psychological) and will
remain so.
lt is therefore important to stress that the vital goal notion of health is
not a notion which stands in opposition to biological medical research or
to a biological understanding of diseases. The biological order is a necessa-
ry condition for health. Our conception, however, means that health is not
identical with the biological order. We need a set of goals for action in order
to understand what health is. And before we have the goals we cannot be
quite clear about the notion of biological order. As we understand the
notion, biological order is that set of biological functions which, given

standard circumstances, makes it possible for the subject to satisfy his vital
Let us then summarize our general conclusions concerning the conse-
quences of our philosophy for the medical profession:
(i) The determination of what is to be counted as health is not an
exclusively biomedical affair. It is an evaluation of a social and
political character, which should ultimately be decided upon by
central political organs.
(ii) Biomedical work should be pursued against the background of
such political decisions. The medical categories, diseases,
impairments, injuries, etc., should be identified in the light of a
given specification of health.
(iii) A clear definition of health will have a particularly salient effect
on treatment and rehabilitation. If the vital goals of health be-
come explicit, medical personnel- including physicians, nurses,
physiotherapists and occupational therapists - will also know
the goal of their work. The goal is no longer simply the termi-
nation of a particular process of disease; it is the realization of
a particular level of ability on the part of the subject.
A Test Case for Diseasehood: Homosexuality
It might now be of some interest to test the welfare theory of health on some
matter of controversy. In the general political and medical debate the status
of some bodily and mental states has been unclear. The typical cases are
various kinds of deviances, states which express themselves in deviant
behaviour, i.e. criminal, immoral or otherwise undesirable behaviour. Par-
adigm cases are alcoholism, psychopathy and homosexuality. In different
countries and at different times these states have moved in and out of the
sphere of pathology. Alcoholism, at least in its severe forms, has perhaps
come to stay; psychopathy, although hardly ever clearly defined, has per-
haps never been completely out, but its importance as a state of pathology
has varied. Homosexuality is particularly interesting; it has entered and left
many lists of diseases and impairments; it remains in some of them. It has
also been a more or less permanent guest in the criminal codes of different
In the discussion to follow we shall use homosexuality as our own test
case. We shall try to answer the following questions: On what has one
based, and on what can one base, the idea that homosexuality is a disease?

What is the internal coherence of this reasoning? How should homosexuali-

ty be characterized from the viewpoint of the welfare theory of health? 81

In most societies homosexuality has been considered a highly undesir-

able inclination. From an orthodox Christian point of view performing
homosexual actions involves a breach of the laws of God; the homosexual
act is the most abominable sin. In Leviticus 20:13 it says:
If a man also lie with mankind, as he lies with a woman, both of them have committed an
abomination: they shall surely be put to death; their blood shall be upon them.

The religious standpoint was for a long time supported in the Western
criminal codes; and as religious authority began to wane secular law took
over the persecution. The death penalty for homosexual acts remained in
most countries until the 19th century, when it was replaced by milder forms
of penalty. Now, in the year 1985, there are a great many places where
homosexuality is no longer a criminal offense. On the other hand public
discrimination of homosexuals often continues, for instance by labelling
homosexuality a pathological condition. 82
The first significant scientific works on homosexuality were those of
Freud. In fact, the psychoanalytic school of thought is still the dominant
school among those who consider homosexuality to be a pathological
Freud and his followers held it as almost self-evident that a heterosexual
disposition was the normal and natural outcome of a person's psychosexual
growth. Thus, homosexuality was an inversion, a trait due to some stop-
page in the person's development to being a normal adult. 83
According to Freud, however, homosexuality is not a completely unna-
tural condition. In fact, he says, all children experience homosexual phases,
which must be gone through before the completion of their development.
In emphasizing this feature Freud distinguished himself from most of his
predecessors and followers. He opposed the view that homosexuals repre-
sented an extreme form of degeneracy or that they were in general deviant
or disabled. In fact, Freud noticed how many homosexuals were dis-
tinguished members of the society displaying high intellectual and moral
abilities. As a result he rejected the idea that homosexuals should be barred
from membership of psychoanalytic societies. 84
What, then, is the nature of that psycho-biological process which termi-
nates in a homosexual inclination? Consider here Freud's basic ideas on
male homosexuality.85

In the beginning there is a young man who has strongly fixated on his
mother. After puberty he changes his attitude; he identifies with his mother
and looks for love-objects whom he can love as his mother loved him. The
male object should be of the same age as himself, when this new phase of
his development starts.
The main causes of this identification with the mother are, according to
Freud, the following: fixation on the mother, which makes it difficult to
transfer affection to other women; fixation on the male organ - a tendency
not to tolerate its absence in a love-object; respect for the father and a fear
of him; thus all rivalry with the father - and with other men - is avoided.
Many of the later psychoanalysts have followed Freud in tracing the
origin of homosexuality to psychosexual development. There are, however,
also important differences. The most influential psychoanalysts, like San-
dor Rado, Irving Bieber and Charles Socarides, consider homosexuality to
be a grave disorder. According to them homosexuality is not simply an
abnormal variant. It is a highly disabling state of affairs connected with
massive fears. Socarides writes:
Homosexuality is a masquerade of life in which certain psychic energies are neutralized and
held in a marginally quiescent state. However, the unconscious manifestations of hate,
destructiveness, incest, and fear always threaten to break through. Instead of union, cooper-
ation, solace, stimulation, emotional enrichment, and a maximum opportunity for creative
interpersonal maturation and realistic fulfillment, there are multiple underlying factors which
constantly threaten any ~going homosexual relationship: destruction, mutual defeat, exploi-
tation of the partner anti the self, oral sadistic incorporation, aggressive onslaughts, and
attempts to alleviate anxiety - all comprising a pseudo-solution to the aggressive and libidinal
conflicts that dominate and torment the individuals involved ([121], p. 119).

The elements of fear and hatred are important also in Socarides' causal
explanation of male homosexuality. The very young boy's fear of his mother
and aggression towards his father (already during the preoedipal phase)
prevents him from separating from his mother and establishing an identity
of his own. As a result he cannot identify his true gender, and retains in
a sense a feminine personality.86
For Socarides there can be no doubt about the pathological status of
homosexuality. Moreover, he claims that most homosexuals, because of
their homosexuality, have other mental diseases as well, such as schizoph-
renia and paranoia. 87
The psychoanalytic theories, although dominant in particular in the
American debate, have no monopoly in explaining homosexuality. There
are also a number of purely biological hypotheses. One of the most favoured

is the endocrinological hypothesis, which is well developed by the German

endocrinologist G. Dorner. From a number of experiments with rats he has
arrived at the following hypothesis: it is absolutely crucial for the proper
sexual development of a person that there is a particular amount and ratio
of sex-hormones surrounding (and in) the fetus during a period of its
growth. In particular, "androgen deficiency during a critical hypothalamic
differentiation period [in a male rat] may lead to a predominantly female
differentiation of the brain".88 Such a brain in its turn creates a disposition
for female-like sexual behaviour. This is so even if the level of androgen
later becomes normal or even supernormal. In a female-differentiated brain
androgens stimulate sexual activities in general but not their specific di-
rection. 89
It is significant that Dorner performs his studies with the explicit aim of
revealing the aetiopathogenesis of homosexuality. The pathological status of
the condition is already presupposed in his research. 90 In this presuppo-
sition he is a representative of many scientists in the field including, as we
have seen, the psychoanalysts.
Against this background it is a striking fact that homosexuality was
abolished from the official American list of mental disorders, DSM-III
(Diagnostic Statistical Manual of Mental Disorders) [30], as a result of a
formal decision by the American Psychiatric Association in 1973. Mainly
due to strong influence from the Gay Liberation Front, many psychiatrists
were persuaded to reconsider their views on non-standard sexual
behaviour. The presuppositions and contentions of the later psychoana-
lysts were questioned. Moreover, there was a substantial discussion con-
cerning the general concept of a mental disorder and the defining criteria
of such a disorder. 91
The psychiatrist R.L. Spitzer, who was the leading theoretician behind
the new outlook, summarized his reasoning in the following way:
In 1973 I reviewed the characteristics of the various mental disorders and concluded that, with
the exception of homosexuality and perhaps some of the other "sexual deviations", they all
regularly caused subjective distress or were associated with generalized impairment in social
effectiveness or functioning. It was argued that the consequences of a condition, and not its
etiology, determined whether the condition should be considered as a disorder and therefore
that it is irrelevant whether a condition is the result of childhood conflicts and intrapsychic
anxieties, since many desirable conditions that noone would suggest are disorders, such as
ambition and self-discipline, may also result from conflict «(124], p. 211).

Spitzer's dictum is a suitable starting-point for our discussion of homose-

xuality as a pathological condition. His general conception of illness has

obvious affinities with ours in that he focuses on the phenomena of distress

and impairment in functioning, i.e. disability. Like us, he contends that the
decisive criterion of a pathological condition is whether it has distressing
or disabling consequences; the genesis of the state has no bearing on the
It seems to be very easy to commit the genetic fallacy pointed out by
Spitzer. If it can be shown that extreme fear and aggression playa major
causal role in the development of homosexuality, as the psychoanalysts
claim to have shown, then it is natural to assume that homosexuality itself
is a disturbance. Likewise, if it can be shown that some statistically abnor-
mal proportion of sex-hormones is a causal condition of homosexuality, as
Dorner claims to have shown, then it is tempting to conclude that homose-
xuality itself is an abnormal phenomenon in the sense of being an illness
or a disease. But, as Spitzer notices, the most dreadful and abnormal
causes can have the most advantageous effects. A causal analysis can
therefore not by itself solve the issue of whether homosexuality is pathologi-
calor not.
In the following we shall pursue the analysis on the basis of the theory
of health presented in this essay. 92
(a) Homosexuals. their general ability. and their general state of happiness.
There are a number of important empirical issues to resolve. Some of the
debaters have conflicting opinions on pure matters of fact. This is most
strikingly the case as regards the homosexual's general abilities and his
degree of satisfaction with life. Contrast Socarides' statement about "ag-
gressive and libidinal conflicts which dominate and torment the individuals
involved" with the findings of Spitzer and others that homosexuals function
socially equally well as heterosexuals and are on the whole quite satisfied
with their situation.
How can the conclusions differ to this extent? One explanation seems
to be that the psychoanalysts on the whole base their impression on
findings from homosexuals who have actively consulted them. In other
words, they have met a very special minority of homosexuals, viz. those
who experience problems in connection with their sexual inclination. 93
It is not reasonable to make generalizations from samples such as these.
On the other hand it is quite difficult to obtain completely unbiased selec-
tions from the homosexual population. This is well discussed by Saghir et
al.[112]. The interviewer here approached three homophile organizations
and recruited a number of volunteers. Only one fourth of the people

approached agreed to take part in the investigation. The sample thus

obtained may be distorted but for reasons very different from the case of
the psychoanalytic patients. 94
The findings of Saghir et al. do not support the pathological hypothesis.
They showed that distress and problems were slightly more prevalent in the
homosexual group than in the heterosexual control-group. On the other
hand, "the homosexual subjects were able to achieve an educational,
occupational, and economic status similar to that of the controls and
significantly higher than that of single employed males" ([112], p. 1086).
One of the most comprehensive studies so far focusing on the emotional
status of homosexuals is Bell and Weinberg [10]. The authors gave exten-
sive questionnaires to almost 1000 homosexuals as well as to just under 500
heterosexuals. Two of the basic questions given to the homosexuals were
the following: Do you accept your homosexuality? Taking things altogeth-
er, how would you say you are feeling these days? About 80 % claimed that
the amount of regret they felt about their inclination was none or very little.
About 80 % answered "pretty happy" or "very happy" to the latter
question. There were subgroups among the homosexuals, in particular the
ones who lived monogamously with a partner, who reported greater happi-
ness than the heterosexual group.95
On the other hand, the study showed that among those who have
problems there are many who have serious problems. The proportion of
people who have attempted suicide or considered suicide is much greater
in the homosexual group than in the heterosexual group. The reason for
considering suicide was reported to be connected with the homosexual
inclination itself in roughly 50 % of the cases. 96
Limited investigations such as these must be treated with caution. On
the other hand, the followin& statement seems to be warranted: the case
for claiming that a homosexual person is in general a disabled and unhappy
person is weak. It is true that some homosexuals display considerable
distress. There is, however, a forceful external explanation of this fact
which rivals the explanation that homosexuality is an illness or a disease.
The external pressure on homosexuals is still considerable in most
societies. Few homosexuals can openly display their inclination. The risks
of negative reactions are great. Therefore, much of the distress of homose-
xuals can be explained in terms of severe circumstances rather than in
terms of inner constitution.

(b ) Homosexuality as an obstacle to fulfilling specific vital goals :familyform-

ing and reproduction. Thus if there is little ground for arguing that homose-
xuality involves general unhappiness or general disability one might consid-
er the idea that homosexuality is associated with some more specific
disabilities, viz. disabilities in relation to some specific vital goals. Let us
here consider the related goals of familyforming and reproduction.
As a preliminary to this discussion let us look more closely at the
ontology of homosexuality. Exactly what kind of entity is referred to by the
term "homosexuality"? It is of importance to investigate this for the follow-
ing reason: if homosexuality is a pathological condition, what kind of
pathological condition is it? Is it, for instance, an illness, a disease or an
111 reviewing some of the literature in the field we have referred to
homosexuality as a sexual attitude or a sexual inclination. Thus it is
understood as a kind of mental state. We have also mentioned some
possible physical or mental causes of this inclination. These causes, wheth-
er they be processes or states, could also possibly be referred to (or partially
referred to) by the term "homosexuality". On both of these interpretations
homosexuality could - from an ontological point of view - qualify as a
disease or an impairment.
Some authors, however, call homosexuality an illness. How would this
be interpreted in our conceptual framework? We have said that "illness"
is a label for a particular disability or a cluster of disabilities. Ifhomosexual-
ity is to qualify as an illness in our sense, the term must then refer to a set
of disabilities with respect to some vital goals.
Assume now, for instance, that reproduction is a vital goal for all adult
people. Assume also that all homosexuals are unable to h~lVe children.
Then all homosexuals must experience some degree of illness. The inability
to have children would be included in the illness of homosexuality. More-
over, assume that this inability is always, or almost always, due to the
homosexual inclination; then the homosexual inclination is a disease or an
impairment. Similarly, the bodily or mental causes of the homosexual
inclination itself would qualify as being diseases or impairments.
Let us now consider the assumptions made above. Is reproduction a
universal vital goal? Is every adult's minimal happiness dependent on his
or her having children? How can we find out?
The Bell and Weinberg investigation [10] can at least give us some
indication. Most homosexuals are unmarried and do not have children. On
the other hand, most homosexuals are quite content with their situation

and show no regret. It must then be unreasonable to believe that repro-

duction is a necessary prerequisite of their minimal happiness. On the other
hand, a considerable minority of homosexuals are deeply depressed about
their relative inability to form a family and have children. This latter group
considers family-forming and reproduction to be vital goals for them. They
are obviously ill and their illness is probably due to their homosexual
From this fact, however, we cannot draw the conclusion that the homos-
exual inclination is a disease or an impairment. As we recall, a necessary
condition for something to be a malady (for instance, a disease or an
impairment) is that it cause some disability in the majority of its bearers.
Ifwe are to believe the Bell-Weinberg investigation this is not the case with
the homosexual inclination.
Assume for the sake of argument, however, that the happiness of homos-
exuals is not considered to be "real" happiness. Assume that we share an
evaluation according to which family-forming and reproduction must be
vital goals, i.e. that they are basic vital goals. This brings us to considering
the second premise in the argument that the homosexual inclination is a
disease or an impairment. This is the premise that homosexuals are in
general unable to form families and have children.
This premise is obviously false. Many homosexuals have married and
have had children. Homosexuality does not entail sterility; it only entails
the subject's not obtaining adequate sexual satisfaction in heterosexual
intercourse. It is certainly true that homosexuals form families and have
children to a much lesser extent than heterosexuals. They normally choose
not to realize these vital goals, if they are vital goals. But according to our
theory, this is not a sufficient condition for their being ill. Illness presup-
poses an inability to realize a vital goal, not just a tendency not to realize
In summary we can say that the case for maintaining that homosexuality
is a pathological condition of some kind is very weak if we accept the
welfare theory of health. There are few indications that homosexuality
involves (or is associated with) general disability and misery. And in the
cases where it is associated with unhappiness there are good rival explan-
ations of this fact. Many homosexuals live in severe circumstances; the
societal rejection of the homosexual inclination is sometimes considerable.
A more plausible hypothesis is that homosexuality involves some specific
disability, for instance with respect to the vital goals of forming a family
and having children. Two questions could then be put: are these goals

universal vital goals; and, are homosexuals in general strictly unable to form
families and have children?
It could seriously be doubted - given some recent empirical investi-
gations - that family-forming and reproduction are universal vital goals.
Moreover, it is simply not true that homosexuals are, in general, unable to
form a family and have children. Thus, there is no support for claiming that
the homosexual inclination (or its causes) is a disease or an impairment.
There are good reasons for believing that family-forming and repro-
duction are vital goals for some homosexuals. There are also good reasons
for believing that many of these - mainly for psychological reasons - have
great difficulties or are even unable to realize these vital goals. These
homosexuals qualify as being ill in our system. But even if the homosexual
inclination is a cause of this illness, we cannot conclude that it is a disease
or an impairment. The reason is that the homosexual inclination does not
cause illness in the majority of its subjects.
On the Notion of Health Outside the Realm of Human Beings
We have noticed at several stages in our discussion that the concept of
health as well as all the malady-concepts have applications outside the
context of human affairs; in fact, they seem to apply to all living beings.
Animals as well as plants can be healthy, have diseases and be impaired.
It is also obvious that the way these concepts are applied to non-human
living beings is closely related to the way they are applied to humans. This
relation is particularly obvious when we compare the diseases and impair-
ments of higher animals with the diseases and impairments of humans. The
nomenclature of pathology is partly identical and otherwise very similar.
Dogs and horses can break legs and have skull fractures; they can be
infected, acquire influenza and common colds; they can have renal and
cardiac troubles and their tissues can be transformed into neoplasms.
Much knowledge about human diseases is in fact acquired through the
study of animal diseases. Current medical ethics still allows some experi-
mentation with animals. Certain knowledge is therefore more easily acces-
sible through the study of animals. But this study of animal disease would
be pointless unless we believed that there was a close conceptual and
empirical connection between animal pathology and human pathology.
But the relation is certainly not restricted to the malady-concepts. There
is an obvious relation also between animal and human health. We say that
a dog or a horse is healthy when it is active and alert, when it expresses

joy, when it does the things that we expect and want it to do. All this sounds
rather close to a rough description of human health.
Now, what about plants? When is a flower healthy, and what do its
diseases and impairments amount to? A flower is healthy when it
"flourishes", when it displays vitality, when it grows and develops well. Its
diseases and impairments are all those processes and states within it which
tend to diminish its vitality or even threaten its life altogether.
These elementary observations converge on one conclusion: the health-
concepts of man are not completely distinct from those of animals and
plants. Our observations in fact unambiguously suggest that they are highly
similar and must belong to one and the same family. This raises a criterion
of adequacy for the theorist. As we have said, a good theory of health
should be able to account for the whole family of health-concepts. 97
We observed that the biostatistical conception of health had the advan-
tage of offering a universal notion also applicable to animals and plants.
The subject-goal theory, however, was limited in this respect. A notion of
health based upon the fact that the subject sets goals for himself cannot
be applied to the whole realm of animals, and obviously not to plants. But
what are the qualities of the welfare theory in this respect? How should we
describe the welfare of animals and plants?
Let us first note the very salient analogies which can be drawn between
humans and the higher animals. It is proper to say, for instance, that
monkeys, foxes, dogs and cats have wants and intentions. We can see how
they decide to reach certain goals, such as acquiring food or sheltering their
offspring, and successfully carry out these intentions. We can also see how
some of these goals display some stability. They have the character of
standing goals and particularly important goals. We can also observe what
happens when the animal in question does not succeed in realizing or
maintaining these goals. It expresses aggression or anxiety; it nervously
seeks for alternative routes to realizing the goal; if the preventative factor
is an intruder, it attacks the intruder or flees from it; ifit is some other kind
of physical circumstance it would try to manipulate it or again flee from it.
The analogies with the human case are very salient. Higher animals have
goals similar in certain respects to those of humans. Their behaviour is
systematically directed to these goals. As a result of reaching the goals the
animals express satisfaction in a way partly similar to human expressions
of happiness. And when the goals are compromised, they display emotions,
for instance, anxiety and aggression, in a way similar to humans. The
important difference between humans and animals lies in complexity,

variety and richness. Humans can deliberate in a way that animals cannot;
they can decide on the relative importance of goals and consciously form
intentions. Moreover, humans have a much greater variety of goals, far
beyond those connected with survival and reproduction. A far greater
number of things are important to humans than to animals. As a result the
happiness of humans is much richer and far more complex than that of
Our general conclusion is that the whole welfare-conception of health
(even keeping the identification between welfare and minimal happiness)
can in all its essentials be applied to the higher animals. We can talk in
terms of their ability to reach their vital goals.
But how must we modify our analysis to take in the health of lower
animals and plants? There seem to be three features to consider, two of
which have already been treated in the discussion of the health of infants.
(i) Lower animals and plants do not form intentions,
(ii) Lower animals and plants do not have abilities in the sense
analysed in this essay,
(iii) Lower animals and plants cannot experience happiness.

(i) is a fundamental problem for the SG-theory but it does not disturb the
welfare theory, as was pointed out in the case of infants. The states
necessary and jointly sufficient for a living being's welfare need not be the
objects of his intentions.
(ii) can be met in a way analogous to the solution for infants:
A lower animal or plant LP is in health if, and only if,
the inner constitution and development of LP is such that, given
standard circumstances, the necessary and jointly sufficient
conditions for LP's welfare are fulfilled.
The crucial problem then concerns (iii). How do we determine the welfare
of a lower animal or a plant when there is no criterion of happiness? Two
ideas for the solution of this problem will be suggested here. First, the
welfare of lower animals and plants may be modelled on that of humans
and higher animals. Second, their welfare may be evaluated on the basis
of their efficiency in contributing to human welfare.
The first suggestion amounts to the following: From the happiness
concept of health, in the case of human beings, we can derive a secondary
concept in terms of a certain biological order. There are certain well-known

biological developments in human beings and animals which lay the

foundation for their general abilities, in particular their abilities to perform
a great variety of basic actions. Thus, they are important prerequisities for
the achievement of most conceivable vital goals. Some of these internal
processes or states are such that they are also direct sources of happiness,
for human beings both in the emotion and the mood sense, for animals
normally in the mood sense. Consider such processes as steady growth, the
acts and processes of reproduction and in general the development of
internal potentialities.'
Biological realities such as steady growth, reproduction and develop-
ment of potentialities are thus in general associated with health. Therefore,
when they occur in lower animals and plants we tend to make an inference
by analogy. They become criteria for the welfare of these living entities.
The second suggestion is the following: there is an important sense in
which humans let their own plans and purposes determine what is to be
counted as the welfare of non-human nature. This observation gives a new
dimension to health discourse concerning non-human life. It also shows the
paternalistic and egocentric attitude of humans concerning such life.
Many species of animals and plants are bred or cultivated by humans,
and are used for very specific purposes. Cows are bred to give milk and
meat; pigs to give meat, dogs to help us hunt, guard our homes or simply
to keep us company. Wheat is cultivated to give us bread, some plants are
cultivated to give us medicine, certain flowers are cultivated simply to give
us pleasure. The purposes thus induced in nature will become identical with
the vital goals of the animals and plants in question.
Human wants and purposes can also be important for those sectors of
life which are not consciously cultivated for the sake of some human
ambition. We can observe all the good things that various wild animals and
plants do for us. We observe how the hawks by their hunting activities keep
rats and mice in check. We see what many wild plants such as berry bushes
and mushrooms can provide in terms of food.
We humans definitely have a view of nature as nature for us. Our eval-
uation of this nature from a welfare point of view is then easily and quite
obviously coloured by our judgment of nature's capacity to fulfill our
purposes. If a cow cannot, for reasons internal to it, give us the amount of
milk that we expect, then there is a presumption of illness. If the wheat
fields do not produce wheat according to plan, then there is a presumption
of illness with respect to the wheat plants. In general, if animals and plants
which are in the service of humans do not fulfill the goals for which they

have been cultivated and trained, and their failure is not due to extraordina-
ry circumstances, then these species of life are considered to be ill.
Our reasoning can be summarized. The welfare theory of health is also
a plausible theory from the point of non-human life. First, such notions as
intentions and wants have application to many higher animals; a happiness
criterion can therefore be used also in these contexts. Second, we evaluate
the welfare of animals and plants also on the basis of other criteria. These
criteria may either be completely modelled on the human case or they may
be attributed to nature on the ground of nature's capacity to be, in an
expected way, useful to mankind.
This line of reasoning has revealed the following basic tenet in our
philosophy of health. The concept of health has its fundamental place in
the specifically human discourse. The term "health" is obviously also used
outside the universe of humans. According to our suggestion, however, it
is used there in a parasitic sense. The idea of health can be extended to
cover all animals and plants only by using certain incomplete analogies.
This conclusion constitutes another basic difference between the welfare
theory of health and the biostatistical theory as proposed by Christopher





In Chapter one, section 2, we listed six basic questions which every good
theory of health should be able to answer. Let us now summarize our own
investigation by letting the welfare-theory answer these questions.

1. What are the logical relations between the health-concepts?

Health is the basic concept in the welfare theory. It is defined as a person's
ability, in standard circumstances, to reach his vital goals. His vital goals
are those states of affairs the realization of which are necessary and jointly
sufficient for his minimal welfare. In the human case welfare is equated with
Illness is the complement of health. A person is ill to some degree if, and
only if, there is at least one vital goal which, given standard circumstances,
he cannot reach.
"Illness", as when one speaks of different illnesses, refers to clusters of
disabilities, i.e. syndromes on the ability-level.
Disease, impairment, injury, and defect make up the set of maladies. It
holds for all maladies that they are such states of affairs as tend to com-
promise health. The existence of maladies in a person A is, however,
compatible with the health of A. On the other hand, health may be com-
promised by factors other than maladies. Unwanted pregnancy is not a
malady. On the other hand, it may cause illness.
The malady-concepts are secondary to the concept of health. The wel-
fare-theory is therefore a holistic theory of health.

2. What are the logical relations between the concept of health and some other
central humanistic concepts?
Health is defined as an ability-concept. It is, however, easily distinguished
from the concepts of excellence such as those of intelligence, strength, or
talent. Health is the ability to reach a certain basic level, i.e. the level of


the realization of one's vital goals. The excellence-concepts have, however,

no upper limit.
An exception to this dictum is the case where the further ability involves
adaptability to a greater number of standard environments. When health
is not relativized to a particular standard environment, it has no upper
Health is easily distinguishable from activity in accordance with norms.
First, health is not a kind of activity but a kind of ability. Second, the vital
goals of a person need not be identical with any goals prescribed by society.
Health is easily distinguishable from happiness. Health is, by definition,
an important contributor to happiness, but is not identical with it. A person
in health may fail to be happy mainly for the following two reasons: (a)
although he is able to reach his vital goals, he does not try to reach (all of)
them; (b) the external circumstances are so severe that they do not provide
him with the opportunity to reach all his vital goals.

3. What is the relation between human health and the health of other living
The theory states in its most abstract form that a person's vital goals are
necessary and jointly sufficient for his minimal welfare. The concept of
welfare is applicable to all living creatures, including plants.
For humans and higher animals we have interpreted welfare as happi-
ness. For lower animals and plants we must use some other interpretation.
Our suggestion is that there is an ideal of welfare imposed on lower animals
and plants in analogy with the one used with respect to higher animals and
ourselves. Steady growth, development of potentials and reproduction are
important conditions of our own minimal happiness. We therefore consider
such features as essential criteria for health in, for instance, plants.
We have also suggested that with some animals and plants, in particular
those actually raised and cultivated by humans, we tend to use a notion of
welfare which is parasitic on our own welfare. According to this notion, a
plant or a lower animal is healthy if it fulfills the purposes of the breeder
or cultivator, i.e. some of the goals necessary for his welfare.

4. What is the relation between mental and somatic health?

The welfare theory provides a very close connection between somatic
health and mental health. In fact, the theory is a unified theory of health
covering both aspects.

Still, ordinary intuitions about mental illness and mental health can be
expressed within the theory: a person is mentally ill if his health has been
compromised by states or processes in his mind. A person is mentally
healthy if there is no mental compromiser of his health (i.e. if he is either
in complete health or ifhis health has been compromised merely by somatic

5. What is the relation between health and the environment?

The concept of health, or, strictly speaking, its application in a particular
context, is affected by environmental factors or judgments in at least the
following respects:
(i) What is to be regarded as standard circumstances for ability must
ultimately be decided upon by the members of a particular society. Such
a decision is in its turn influenced by the social and physical environment
as well as by prevalent values.
(ii) The boundaries for minimal happiness are analogously to be decided
upon by members of the society. This decision is partly based on eval-
uations about what a minimally human life must require in terms of happi-
ness. But it must also be restricted by realities: what can reasonably be
required, given the circumstances; how large a part of the popUlation can
be the object of health care?
The malady-concepts are also affected by these considerations but to a
much lesser extent. The major reason for this is that most maladies strike
in such a basic and general way that health is normally affected by them,
regardless of how we determine standard circumstances or the level of
minimal happiness. And that health is normally compromised by x is
enough for x to be a malady where x is a process or state internal to a
person's body or mind.

6. What is the place of the health-concepts in science?

Health is a partly evaluative concept to the extent and for the reasons
developed in this essay. Since the malady-concepts are secondary to health,
they too are partly evaluative. These facts do not preclude their use in the
medical sciences. As soon as the matters which are open for evaluation are
decided upon, a standard is established which can be used for purely
empirical investigation. That part of medicine which deals exclusively with
the study of maladies does not normally even need this standard. This is
so for the reason given in our answer to question 5. The existence of a

standard is, however, essential for the science of rehabilitation and absol-
utely necessary for the general enterprise of health care.


(i) Basic action

A basic action is the kind of action which is not performed by performing
some other action. In the standard case the basic action involves merely
the movement of a part of the body.

(ii) Accomplishment
An accomplishment is an action generated by at least one further action.
This generation can be causal or conventional.

(iii) Activity
An activity is a set of actions on the part of an agent A, constituting a
spatio-temporal sequence.

(iv) Practical possibility

The practical possibility of performing an action consists of a set of internal
conditions, viz. ability (or first-order ability), as well as a set of external
conditions, viz. opportunity.

(v) Second-order ability

A has the second-order ability to perform an action F if, and only if, A has
a first-order ability to acquire a first-order ability to perform F.

(vi) Need (abstract characterization)

A has a need of y in situation S if, and only if, A has a goal G and y is a
necessary condition for A in S to reach G.

(vii) Health (abstract characterization)

A is healthy if, and only if, A has the second-order ability, given standard
circumstances, to realize all his vital goals.

(viii) Health (the welfare concept of health applied to human beings)

A is healthy if, and only if, A has the second-order ability, given standard
circumstances, to realize all the goals necessary for his minimal happiness.

(ix) Malady
M is a type of malady in environment E if, and only if, M is an episode-type
which, when instanced in a person A in E, causes with high probability
illness in A.

(x) Types of maladies

The following types of maladies are distinguished with regard to their
ontological status:
disease: a bodily or mental process,
impairment: a bodily or mental endstate of a disease,
injury: a bodily or mental event or state caused externally,
defect: a congenital bodily or mental state.

(xi) Three medical concepts of health

(a) The malady-concept

A is healthyM if, and only if, A's ability to realize his vital goals is not
compromised by his having a malady.

(b) The disease-concept

A is healthyD if, and only if, A's ability to realize his vital goals is not
compromised by his having a disease.

(c) The curable malady-concept

A is healthy eM if, and only if, A's ability to realize his vital goals is not
compromised by a malady of his which is in principle curable.



In this essay we have so far only spoken of the general concepts of health
and illness and the general malady concepts. We have, for instance, sug-
gested generally that diseases are internal processes such as tend to com-
promise health. However, the medical discourse does not just contain a
general notion of disease; there are also a great number of specific disease
notions, i.e. notions standing for all the various kinds of diseases that there
are, for instance, cancer, tuberculosis, diabetes mellitus and coronary
The ontology of these diseases or disease-kinds has been the subject of
a long and interesting debate which has played an important role in medical
thinking. 98 It is customary to characterize this medico-ontological dis-
cussion as a debate between two parties, normally called the ontologists and
the physiologists. Another term for the latter party is "functionalists". Both
parties have a long tradition, but perhaps the physiological school is the
older of the two. The tradition of this school dates back to Hippocrates and
his ideas about the physiology of the healthy and the ill person. The healthy
person was characterized by physiological balance, a balance sustained by
the right mixture of bodily elements, including the bodily humours. A
person who was ill was characterized by an imbalance in these respects.
The various disease labels, which certainly also existed in ancient times,
were then taken to refer to the different ways in which an imbalance could
manifest itself. The disease or illness of melancholia could, for instance, be
identified as the imbalance consisting of too much of the humour black bile
and too little of other humours. 99 (Similarly, according to the modern
physiological way of thinking, diseases could be viewed as different kinds
of disturbances in the ordinary physiological functioning of the body.) 100
In the history of medical ideas "physiologism" often stands for a particu-
lar attitude towards the task of medical classification too. The physiologists
- given their way of viewing diseases - have difficulties in ordering disease
phenomena into clear categories. They are bewildered by the enormous
individual differences; they are struck by the fact that no one case is exactly

like another. As a consequence they find categorial labels which distort

reality more than they illuminate it.
The ontological school, which in fact includes a wide variety of thinkers,
opposes the physiologists by emphasizing that diseases are entities of some
kind in their own right. To speak of a disease is in some sense to speak
about a thing, not a manifestation of imbalance or a mode of disturbance
in ordinary physiological functioning.
Now, what sort of thing is a disease according to the ontologists? Here,
the answers differ greatly. We shall here briefly discern two major variants
of ontologism; first the view that diseases are clusters of signs and symp-
toms and, second, the view that diseases are physical substances.

(i) The idea ofdisease-species as clusters ofsigns and symptoms. It is important

to stress that the idea that diseases were things did not necessarily mean
that diseases were concrete things. Indeed, classical ontologists, like Tho-
mas Sydenham, did not at all conceive diseases as physical substances
comparable to animals or plants. 101
At the time of Sydenham a physician could in examining a patient
observe only signs and symptoms and clusters of such signs and symptoms.
Pathological anatomy had not been invented. Diseases could, in general,
not be localized to any particular part of the human body. Thus diseases
could not be identified with parts of the body, nor were they identified with
any other physical bodies.
Still, Sydenham as well as the famous nosologists of the 18th century
claimed that diseases were entities that could in some sense be identified.
Often the signs and symptoms of different individuals displayed regular-
ities. The symptoms appeared in typical clusters which were similar from
individual to individual. In a famous passage Sydenham writes:
Let a person seriously and accurately consider the phenomena which accompany such a fever
as a quartan ague. It begins almost always in autumn; it keeps to a regular course of
succession; it preserves a definite type; its periodical revolutions, occurring on the fourth day,
if undisturbed by external influences, are as regular as those of a watch or any other piece
of machinery; it sets in with shivers and a notable feeling of cold, which are succeeded by an
equally decided sensation of heat, and it is terminated by a most profuse perspiration. ([126],
p. 19).

To Sydenham these observations provide reasons for believing that this

disease is a species. These reasons, he claimed, are comparable to the ones
we have for believing that a plant is a species. Thus, according to him, not
only substances can be given general names and be classified; certain

regular successions of events, for instance signs and symptoms in a human

body, are proper objects for the same theoretical treatment. As a result
there is an ontology of diseases as a kind of abstract objects. 102
Sydenham's followers, the nosologists of the 18th century, in general
sustained this view of diseases as abstract objects. Moreover, they normal-
ly viewed them in an Aristotelian way as eternal and unchangeable; the
disease-species were considered to be preformed and given once and for
all. Diseases were to be discovered and not invented.

(ii) The idea of diseases as substances. Other theoreticians were actually

searching for physical substances. They pursued their quests along two
different lines. According to the first line, diseases were to be identified with
agents invading the human body. The disease was a separate individual,
a parasite. This idea was particularly nourished during the heyday of
bacteriology, when a disease could be traced back to certain specific
infectious microbes.
It is important to see how this view differs from the first ontological idea.
Diseases were no longer identified with the clinical phenomena themselves,
the clusters of signs and symptoms. Instead they were identified with their
causes; and these causes were believed to be made of concrete material,
whether they be parasites, microbes or inanimate agents.
This view of diseases(Nas criticized by proponents of the second line of
thought. Rudolf Virchow, the founder of cell-pathology, maintained that a
disease should not be confused with its causes: a disease is not an agent
which is distinguishable from its bearer. A disease is rather a part of the
bearer's body, although an altered part of the body. Virchow [138] says the
... here suffice it to say that, in my view, the disease entity is an altered body-part, or, expressed
in first principles, an altered cell or aggregate of cells, whether tissue or organ. In this sense
I am a thoroughgoing ontologist, and I have always regarded it as a merit to have brought
the old and essentially justifiable requirement that disease should be a living entity, and lead
a parasitic existence, into harmony with genuine scientific knowledge (p. 192).

Those are some classical ontological conceptions of disease. It is obvious

that they all point to characteristics which every reasonable philosophy of
disease must give an account of: practically all diseases display signs and
symptoms; invading objects - be they of a biological, mechanical or chemi-
cal kind - play an important role in the genesis of disease. It is certainly
also true that many diseases involve as a basic ingredient changes on a

cellular level. Still, none of these conceptions will do as general characteri-

zations of the ontology of diseases. How many diseases, as recognized by
modern medicine, could be defined exclusively in terms of signs and symp-
toms? And which are the invadingphysical agents and what cellular change
is involved in the case of many mental diseases? A more abstract and
comprehensive machinery seems to be required for a general understand-
ing of the nature of diseases.


Before pursuing this discussion let us consider a further problem in the

theory of disease-concepts. This is the historical problem of conceptual
change. It is an important and interesting fact about disease-concepts that
they have undergone an enormous change during the history of medical
evolution. Such a change is not unique to medical science. All scientific
evolutions are characterized by conceptual change. A zoologist of today
would define the species of cat or elephant very differently from a zoologist
of 500 years ago. He knows much more about cats and elephants and he
has found it useful to use new types of defining characteristics.
However, there is an important difference between zoology and medi-
cine. The cat-species and the elephant-species have remained almost com-
pletely the same over this period, with respect to the class of reference, i.e.
the individuals referred to by the terms "cat" and "elephant". The zoologi-
cal species remain unchanged although the formal characterization of them
has changed over time.
This is, unfortunately not, true of diseases. It is not simply that the
diabetes mellitus of antiquity was defined differently than today; to a great
extent this disease-term also referred to different things than it does today.
There is not - as in the case of cats - a common reference-class to point
to. In order to support this contention let us sketch the history of the
concept of diabetes mellitus. 103
The Greek physician Aretaeus of Cappado cia, during the second century
A.D., is the first known author of a comprehensive description of the
diabetic syndrome.
Diabetes is a wonderful affection, not very frequent among men, being a melting down of the
flesh and limbs into urine ... the patients never stop making water, but the flow is incessant,
as if the opening of aqueducts ... life is disgusting and painful; thirst unquenchable; excessive
drinking, which, however, is disproportionate to the large quantity of urine, for more urine
is passed; and one cannot stop them either from drinking nor making water. Or if for a time

they abstain from drinking, their mouth becomes parched and their body dry; the viscera seem
as if scorched up; they are affected with nausea, restlessness and a burning thirst; and at no
distant term they expire ([2], p 338).

The syndrome described, then, is characterized mainly by excessive uri-

nation, excessive thirst and drinking, a tendency for dryness, nausea and
restlessness. To this picture was eventually added the important symptom
of the sweetness of the urine. One of the first to emphasize this was the
famous Arab philosopher and physician Avicenna (960-1037).
Although the clinical picture, and thereby the disease concept, of diabetes
mellitus became much richer during the following centuries, it remained a
syndrome concept until the 19th century. Then, however, there was a rapid
evolution, which can be summarized in the following steps.

Claude Bernard, ca 18S0: Diabetes is a disturbance in normal nutrition

in which an excessive amount of sugar appears in the urine.

Oscar Minkowski, ca 1890: Diabetes is a disturbance in normal nu-

trition, where an excessive amount of grape sugar appears in the urine, and
is caused by malfunction in the pancreas.

Frederick Banting and Charles Best, 1921: Diabetes is a disturbance in

normal nutrition, in which an excessive amount of grape sugar appears in
the urine, and is caused by the lack of secretion of insulin from the islets
of Langerhans in the pancreas.

George F. Cahill, 1976: Diabetes is a disturbance in glucose homeosta-

sis, which is secondary to a deficiency in the betacells of the endocrine

One tempting way of viewing this development is to say that it is the

"same thing" which has only been better understood during the centuries,
and particularly during the last century: we know now much better the
"true" nature of diabetes, which has remained the same from Aretaeus to
Banting and Best.
This is, however, a misleading view of this evolution. The clinical syn-
drome of Aretaeus was the disease of diabetes mellitus in the second century.
Excessive urination, thirst and drinking was diabetes irrespective of the
internal" causes of this syndrome. And we know now that these symptoms

can appear without the pancreas being involved. Therefore, some of the old
"true" cases of diabetes would be labelled "false" diabetes today.
Nor is the specification of diabetes as a defect in the glucose homeostasis
sufficient to identify it with the modern concept. A disturbance in glucose
homeostasis can be effected by other means. The metabolism of sacchar-
ides depends on many factors outside the pancreas, for instance the endo-
crine glands and the liver. Therefore, not all those disturbances, which look
very much like diabetes clinically, need be "true" instances of the modern
concept of diabetes mellitus.
Nor again are all those cases of deficient glucose homeostasis which are
due to lack of insulin, true instances of diabetes mellitus. The insulin pro-
duction can be blocked by external factors and need not be due to specific
internal failures, which the modern concept - strictly speaking - re-
quires.105 ,106
It appears, then, that the conceptual change which is due to medical
discoveries does not just result in better descriptions of the "same" thing.
This evolution results in an, at least partial, change of the reference-class.
Let us consider this in some detail via a more abstract analysis of the
A disease-species was first identified as a cluster of signs and symptoms.
In the days of pathoanatomy and pathophysiology some causes of these
signs and symptoms were sought in the structure or function of certain
organs. If such changes were found the disease was identified with these
causes. The disease had taken its seat in the organs. Sometimes cellular
biology traced the etiology deeper down in the microbiology of the orga-
nism. If such cellular changes were found, as were likely to have caused
both the dysfunction of the organs and the external signs and symptoms,
the real locus of the disease moved to the cells. But the process can go
further; we may try to look for the external (or internal) cause of the cellular
change. If such a cause is found, this normally has consequences for the
disease-concept. The disease tends. to be etiologically defined, although it
is not normally literally moved to the locus of the external object.
This evolution has a number of conceptual consequences. They can be
described slightly formally in the following way: A particular cause (on the
physiological level) of a cluster of signs and symptoms is found. A former
symptom-disease D is redefined in terms of this cause. It eventually ap-
pears, however, that this cause can account for only a portion of the
manifestations of the sign-symptom cluster. The old disease-species D is
now divided into the new disease-species D and something else which lacks

a particular cause. For the new D one may find in its turn a cellular cause
or an· external cause, which may give rise to yet another conception of D.
Again these causes can most probably account for only a certain percent-
age of the new D. The new D should, then, in its turn be divided into a
further D and a residue disease which lacks the microscopic or external
causes in question. If such a conceptual split is made over a long period
and in several steps - and if, furthermore, an ancient term is carried over
to connote a modern concept - then we have rich sources for conceptual
confusion. In this case, the reference class formerly denoted by the term
has a rr.uch greater extension than the same term today.
The conceptual development which we have outlined here is the most
typical one. Another important case, though, is the one in which several
kinds of symptoms have been found to have a common cause. This is the
case with the disease of Tuberculosis, where many different symptoms -
earlier viewed as different diseases - were shown to be caused either by
Mycobacterium tuberculus or Mycobacterium bovis. Here instead of a split, we
get a conceptual unification.
What does this story tell us about disease species and the possibility of
characterizing the ontology of diseases? It gives us numerous reasons for
First, the facts regarding conceptual evolution constitute a good argu-
ment against an Aristotelian view of disease-types. Given the great histori-
cal change described above, it is highly implausible that the terms
"diabetes" or "tuberculosis" represent essences which are given once and
for all. Second, the historical outline implies that a disease as identified at
one time may have an ontology distinct from the disease as identified at
another time.


Returning now to the main issue, what strategy for ontological inquiry
should be chosen?
Let us make two preparations for this discussion, first a pragmatic one,
then a philosophical one. The pragmatic consideration will concern the
already existing terminology used to refer to diseases. The philosophical
consideration will introduce some concepts which might be useful for the
analysis of disease-concepts.

There has always been and must always be a need for medical classifi-
cation and medical definition. This is a sine qua non for medical communi-
cation which, in its turn, is a sine qua non for scientific development.
Medical history has given us a vocabulary of diseases, together with a
variety of classifications of these diseases. Some of these classifications
have for a number of pragmatic reasons become more influential than
others, and have been codified and accepted by medical congresses. A
relatively recent important codification of terminology took place in Paris
in 1948, when different disease-classifications were fused with the Inter-
national Classification of Causes of Death into the International Classification
of Diseases, Injuries and Causes of Death (ICD) [143].
According to this classification there are 17 classes of disease and -
depending on the level of specification - several thousand disease-species
to be subsumed under these classes.
The given classification and its associated vocabulary is not considered
sacrosanct. Modern taxonomists are quite aware of the historical evolution
of the system and have anticipated a revision of this taxonomy once every
decade. There have been three such revisions since 1948.107
What do these classifications tell us? The principal message is that every
term - on the species-level - refers to some internationally accepted dis-
ease-phenomenon. These terms can be used in international medical com-
munication; they can be used in international medical statistics, such as the
statistics on causes of death.
If this international communication and these international statistics are
to be of any value, there must be some common characterization of the
concepts referred to by these disease-labels. And most importantly, these
characterizations or concepts must remain stable as long as no explicit
notice concerning change has been given.
If we admit this much, we admit more than the extreme physiologist
does. We mean that it is essential to be able to speak about disease-species
and tokens or examples of these species. This in its turn presupposes that
there are disease-concepts to be defined. It becomes sensible to ask: what
kinds of entities are the disease-tokens thus characterized? What is their
The above historical outline indicated that the classical versions of
ontologism were not good enough. What could we offer instead? Let us here
insert some philosophical considerations by way of preparing the ground.
As we observed, the basic trouble for medical ontology and classification
is that, whatever diseases are, they are not physical objects. The naive

ontologism, entailing that diseases are parasites invading the body breaks
down for many reasons. First, we know of parasites or invading agents only
in the case of a few types of diseases. Second, and most importantly, it is
absurd to say that the parasite is the disease. Whose disease is the parasite
when it is outside the human body? And what about the state of affairs
remaining when the parasite has left? To say that disease-terms refer to
altered parts of a human body is also to simplify matters. Disease-terms
have a much broader reference than transformed cells or tissues.
The fact that diseases are not physical objects immediately differentiates
medicine from the other biological disciplines. The zoologist and the bota-
nist classify subjects which have definite positions in space and time and
fairly clear-cut boundaries. Another feature which has simplified - although
definitely not solved - definition and classification in these disciplines is the
facts of reproduction. The rule of thumb that only members of the same
species can reproduce has formed the basis for something which could
justifiably be called natural classification.
Such helpful features are absent in medicine. Diseases have no clear
location in space and time; they definitely have no salient boundaries; there
is no analogue to animal reproduction in the case of diseases. (The only
conceivable analogue would be ordinary causation. But although causation
plays an important part in the formation of disease-concepts, it does not
do so in a way analogous to animal reproduction.)
An adequate characterization of diseases thus requires ontological cate-
gories other than that of physical objects. We shall suggest that the fundam-
ental category needed for this purpose is the category of episodes. By an
episode we mean, roughly, something that happens to an object, something
which comes to be, remains, or ceases to be about this object. This ontolog-
ical category thus concerns properties of objects, not the objects them-
selves; and it concerns these properties in a way in which their temporal
aspects are important.
We shall distinguish between three kinds of episodes: states, events and
processes. 108 The basic concept to be introduced is the concept of a state.
It can be viewed as a primitive notion within the theory. A state can,
however, be informally characterized in the following way:

an object x is in a state s at time t, s(x)t if, and only if, x at t has

a contingent property or stands in a contingent relation of some

By a contingent property (or relation) we mean a property (relation) which

does not belong to its bearer as a logical consequence of the simple fact
that the bearer is an object of a particular kind. For example, "x is red" or
"x is hot" denote that x is in a particular state. "x is identical with himself',
however, does not indicate a state.
This is the idea of a momentary state. Usually, however, when we talk
about states, we mean that an object is in a state (or has a property) for
some time. Thus, the states of ordinary discourse are normally enduring.
In contrast to states, processes and events represent the dynamic aspect
of the world. Consider first the concept of a process. When an object is
(involved) in a process it undergoes a continuous change along some proper-
ty- or relation-dimension. The dimension can be distance in space (for the
process of movement), or it can be a quality spectrum, like colour, magni-
tude or shape (for the various processes of quality change).
The continuity of change is essential. When an object moves, it contin-
uously changes its location in space. Typically the locations passed are all
different from each other, i.e. when the object covers some distance without
at any point returning to a previous position. (This need not be the case.
We say that a ball is rolling even if it is moving in a circle and thereby
returns to positions that it has previously passed.) A necessary prerequisite
for x's being in a process of movement is, however, that there are no
consecutive times t - 1 and t such that both s(x)t _ I and s(x)t> where s refers
to a specific location in space. If x is at s at two consecutive times we will
have to say that x remains at s and is no longer in a process of movement.
Although processes are in an obvious way dynamic and involve change
they also have similarities to states. Like states they can extend over
periods of time. Like states processes can be kept and maintained. A
plumber can keep the water running and an umpire can keep a game going.
Moreover, processes can, like states, be initiated and come to an end.
By contrast, events form the category of real change, viz. a change from
one state of the world to another. Simply put, one could say that an event
is the coming about of a state. Coming into existence, coming to a location,
becoming a member, receiving a quality and entering into a relation are
certainly all events.
Conversely, an event may also be the cessation of a state. Leaving a
location, relinquishing a membership and losing a quality are all events.
However, most of the cessations ofa state are at the same time the coming
about of a new state. Leaving a location is coming to some other location,
losing a quality is acquiring some other contrary quality, etc. Moreover, a

transition from a state to a process qualifies as an event: starting to walk,

starting to smoke and starting to work are events which initiate the pro-
cesses of walking, smoking and working. Conversely, of course, the termi-
nations of processes are events. Such terminations are typically the corning
about of states. When John stops walking he enters into a state of rest and
when he stops laughing he enters into a state of silence.
In general, an event is taken here to signify the transition between states
and processes. It may be a transition between two different states, between
a state and a process or between a process and a state.
At this stage the following question naturally arises: How do we dis-
tinguish a process from a succession of events? A first answer which follows
from our definitions is: not any series of events can qualify as a process.
The processes, as we have so far defined them, are sequences of events
along a single dimension. For instance, the series: a door is opened, a bird
sings, it starts raining, does not qualify as a process. There is no single
dimension which covers these transitions. Therefore, it is just a series of
events. We could, however, pursue the question of the cases where the
processes are clear-cut. What prevents us from saying that a ball's rolling
is a succession of events involving the movements of the ball from one
position to another? Nothing really, but the reason for retaining the notion
of a process is precisely that it indicates an interesting, and for many
purposes useful, distinction between those series of state transitions which
are in some way structured and clearly restricted to one dimension and
those which are not.
This is then the basic machinery of episode-concepts. Note that these
were in fact the concepts employed in our endeavour to distinguish between
different kinds of maladies. Impairments and defects were described as
enduring states, injuries as states or events and diseases as processes (see
Chapter four, section 1).
But in order to justify the idea that diseases are processes, a substantial
qualification must be made. The processes characterized so far constitute
continuous change along one single dimension. Few diseases have such a
simple structure. Most of them involve changes along several dimensions
as well as transitions between such dimensions.
There is still a case for retaining the label "process" for diseases. Diseases
have the same static feature as do simple processes. Like the processes of
movement and quality-change, diseases last for periods of time; similarly
diseases start and corne to an end. As with the simple processes, there must

therefore be some unifying principle(s) holding together that sequence of

events which constitutes a disease.
But ifthere is no single dimension covering a disease-process, what could
its unifying principle( s) be? We shall return to this question in discussing
the properties of some specific disease-concepts. Let us here mention only
two important unifying principles, which seem to be essential for complex
processes in general and for diseases in particular.
The first principle concerns the causal origin of the episode. A sequence
of events is often said to be a process if it has a common initiating cause.
This can be understood in two different senses. In one sense the sequence
simply constitutes a causal chain, one event causing the next (or partly
causing it). In another sense, the constituting events each follow from an
initial common cause without, however, any of them causing any of the
others. The unifying principle in both these cases is the common initial
cause. 109
It is, however, doubtful whether the sole fact that events have a common
cause is ever in practice sufficient for a process-label. According to a
prevalent metaphysics, causal sequences do not have a definite end; hence,
all causal processes would be infinite. We would end up with an infinite
number of infinite processes.
A unifying principle which may partly solve this problem is the idea of
the bearer of a process. A process is normally located in one particular body,
or one particular region. We say that an object undergoes (or is involved
in) a process. This sets important limits on a causal chain of events. When
a causal chain of events leaves a body or a place, it is no longer a part of
a process belonging to this body or this place. Therefore, if a disease is a
process belonging to a body, then no part of that causal chain of which the
disease is a part, which lies outside the human body - be it before the onset
of disease or after - belongs to the disease-process.Here, then, the bearer
of the process is a unifying principle in addition to the cause of the process.
Let us now consider some specific disease-concepts.


The examples which follow are taken from the well-known Textbook of
Medicine by Paul B. Beeson and Walsh McDermott [9]. In the very careful
characterizations which are given in this book, the attempt is often made
to give concise definitions. Mostly, these are actually called "definitions".
Sometimes other terms are used, for instance, "general considerations".

These different terms do not, however, seem to mark any theoretical

First, some examples of infectious diseases:
I. Pneumococcal pneumonia

Definition: Pneumococcal pneumonia is an acute bacterial infection of the lungs caused by

Streptococcus pneumoniae (the pneumococcus) and characterized clinically by an abrupt onset,
rigor, fever, chest pain, cough, and bloody sputum (p. 277).

2. The Common Cold (Acute Coryza)

Definition: The common cold is a symptom complex caused by viral infection of the upper
respiratory passages. Most precisely, the term applies to a febrile, acute coryza of viral origin.
In the broadest sense, the common cold refers to any undifferentiated upper respiratory
infection. The terms rhinitis, pharyngitis,laryngitis, and "chest cold" are sometimes used to
designate the principal anatomic site of infection. The main difference between the common
cold and other viral or bacterial respiratory infections is the absence offever and the relatively
mild constitutional symptoms and signs (p. 184).

3. Tuberculosis
General considerations

Tuberculosis is a chronic infection, potentially of lifelong duration, caused by two species of

mycobacteria, M. tuberculosis, and, rarely, M. bovis. It is almost always initiated by inhalation
of infectious material, rarely by ingestion, ... Early in infection a silent bloodstream spread seeds
the lymphatic system and other organs throughout the body, leaving foci which may cause
clinical illness after long periods oflatency. Tuberculosis must be differentiated on bacteriolog-
ic grounds from chronic infections caused by other species of mycobacteria (p. 391).

Second, a disorder of the nervous system:

4. ParalysiS agitans (Parkinson's disease)

The disease begins insidiously with any of its three cardinal manifestations either alone or in
combination. Tremor usually in one or sometimes in both hands, involving the fingers in a
pill-rolling motion, is the most common initial symptom. This is often followed by stiffness in
the limbs, general slowing of movements and inability to carry out normal and routine daily
functions with ease. As the disease progresses the face becomes "masklike", with a loss of
eyeblinking and a failure to express emotional feeling; ...
Although paralysis agitans is invariably progressive, the rate at which symptoms develop and
disability ensues is extremely variable ...The major neurologic findings are the following:
l. Lack offacial expression ...

2. Tremor of the distal segments of the limbs at rest...

3. Muscular rigidity ...

4. Akinesia, the tendency to slowness in the initiation of movement...

5. Postural abnormalities; ... the patient's body has a tendency to fall forward or backward ... (p.

Third, a cardiovascular disease:

5. Acute myocardial infarction

Definition: Acute myocardial infarction is a clinical syndrome resulting from deficient coronary
arterial flow to an area of myocardium with eventual cellular death and necrosis. It is
characterized by severe and prolonged precordial pain similar to, but more intensive than, that
of angina pectoris, and signs of myocardial damage,including acute electrocardiographic
changes and a rise in level of certain serum enzymes. Atherosclerosis oqhe coronary arteries
is the common denominator in the overwhelming majority of patients with acute myocardial
infarction (pp. 1006-1007).

Fourth, a metabolic disease:

6. Diabetes mellitus

General considerations

Diabetes mellitus is a diagnostic term applied to a constellation of anatomic and biochemical

abnormalities which share in common, as part of a syndrome, a disturbance in glucose
homeostasis, which is secondary to a deficiency in the beta cells of the endocrine pancreas
(p. 1599).

Certain general facts about these disease-characterizations are striking.

They are not formulated in an exact language, as very precise definitions.
Some of the characteristics mentioned are not universal characteristics
(applicable to all instances of a disease). This is particularly so with paraly-
sis agitans, the description of which concerned a typical sequence of symp-
toms and signs.
Moreover, there is at times an explicit uncertainty with respect to ontol-
ogical commitment. This is most salient in the case of the description of
the common cold. Initially, the common cold is called a symptom complex
caused by a viral infection. Later on, the author expresses a difference

between the common cold and other infections. We can ask the question:
is the common cold an infection, or is it caused by an infection?
Before entering more deeply into the ontological considerations let us
first consider the disease-notions emerging from these definitions, mainly
from the point of view of similarities and differences. This procedure can
be made easier by some formal simplifications:

( 1) Pneumococcal pneumonia
An acute bacterial infection of the lungs caused by Streptococcus

(2) The Common Cold

A symptom complex caused by viral infection of the upper
respiratory passages (alternatively: a viral infection of the upper
respiratory passages). Differentiated from other viral respirato-
ry infections by absence of fever and mild constitutional symp-

(3) Tuberculosis
A chronic infection caused by either Mycobacterium tuberculosis
or Mycobacterium bovis.

(4) Paralysis agitans

A symptom complex of neurological origin where most of the
following symptoms are present: a) lack of facial expression; b)
tremor of the distal segments of the limbs; c) muscular rigidity;
d) akinesia; e) postal abnormalities.

(5) Acute myocardial infarction

A symptom complex = clinical syndrome, resulting from def-
icient coronary arterial flow to an area of the myocardium.

(6) Diabetes mellitus

A constellation of anatomic and biochemical abnormalities
which have in common a disturbance in glucose homeostasis,
caused by a deficiency in the beta cells of the endocrine pan-

This short list underlines the pattern found in the historical development
of the modern concept of diabetes mellitus. There is a very strong tendency
to use etiological means in the specification of disease-concepts. A disease
is favourably defined as an X caused by a Y. Later on the disease may be
identified with the Yor preferrably as a Y caused by a Z.
What are then the ontological categories found in our examples? Three
examples refer to infections caused by certain viral or bacterial agents.
Three examples refer to symptom complexes (clinical syndromes) of various
(partly unknown) origin. One example mentions biochemical abnormalities
(alternatively: a disturbance of homeostasis). How should these categories
be viewed? Is there any common feature to be found which could give a
clue to the basic ontology of diseases? These questions will be answered
in this and the following section.
Let us first make a preliminary analysis of the categories mentioned.
What is an infection? By definition, an infection is a series of events
initiated by some microscopic living organism which has entered the hu-
man body and spread toxic substances with the effect that the body re-
sponds with a series of typical measures; for instance, a rise in temperature,
excessive blood flow to the affected part of the body, and production of
antibodies against the toxic substances.
What is a symptom complex or a clinical syndrome? 110 Normally, symp-
toms are defined as features of a disease directly accessible to the bearer.
These features can be of an almost exclusively subjective kind, such as
when they are mental phenomena like sensations and emotions. But they
can also be intersubjective properties. In that case they should be easily
accessible to an external observer without the assistance of any specific
medical observational apparatus; Signs or features which require advanced
apparatus for their detection would not be called symptoms. Paradigmatic
symptoms are changes on the surface of the body, as to figure, colour,
dryness and warmth. Other symptoms are changes in the bearer's ability
to move or perceive. Symptoms, then, are phenomena such as sensations,
emotions, changes on the surface of the human body or changes in human
abilities. A symptom complex is the existence of at least two such symp-
toms, occurring either contemporaneously or in sequence.
Consider, finally, the concepts of 'biochemical abnormality' and 'distur-
bance of homeostasis'. These are difficult since they seem to be, in different
ways, ambiguous. In one sense a biochemical abnormality can amount to
a static property (of an organ, tissue or liquid) which consists in the
existence of abnormally much or abnormally little of certain substances. In

another sense, the term can refer to a dynamic process of abnormal pro-
duction and evolution of substances on the biochemical level.
The term "disturbance in homeostasis" is also, although differently,
ambiguous. ori the one hand, a disturbance can be a particular momentary
event. Such an event presupposes the existence of a normal, undisturbed,
state or process. In this state or process a change occurs, normally caused
from the outside. This change is the momentary event of disturbance. (Such
a change can be repetitive; there may be a continuous sequence of disturb-
ing events.)
On the other hand, a disturbance could be the episode resulting from
such an initiating (sequence of) disturbing event(s).


What do these analyses of the key concepts tell us? How could we use our
episode concepts to describe them? Consider first the easiest case, in-
fection~. The rough definition which we have given would seem to be
sufficient. An infection is a bodily process with certain typical characteris-
tics, causally triggered by a microscopic living creature. These characteris-
tic changes certainly do not fall under a single dimension. They involve
biochemical changes, changes of temperature, as well as visible external
changes of the body.
There are, however, unifying principles. The principle of a common cause
is obvious: the microbe. The general principle of a single bearer is also
obvious. The infection belongs to a person and terminates outside the
person. Is there a further unifying principle?
There certainly is and this is the principle which gives the process its
status as a disease, in contradistinction to other possible processes
contained in a human body. A further unifying principle is that the elements
of the infection, the individual state-transitions tend to limit the bearer's
abilities, either directly or indirectly by causing pain or other kinds of
Let us now proceed to an analysis of symptoms. Do they have a place
in the ontology of episodes? Consider first mental symptoms. In ordinary
parlance we acknowledge such locutions as being in a state of pain or in
a state of depression. We can also experience a stroke of pain or a sudden
pang of anguish. If we consider the situation more deeply this seems to be
quite a proper manner of speaking. To have a mental property is, if the
property remains unchanged, to be in a state. If it changes continuously or

if it comes and goes, the situation is a kind of process .. A sudden stroke or

a pang of feeling would be a typical event. This is uncontroversial and does
not commit us to any view about the nature of the mental "substance".
Similarly, the intersubjective symptoms, such as the rise in temperature,
can take the shape of all three episodic forms. They can be of the event kind
- the temperature rises suddenly; they can be static - the temperature
remains at a high level; or they can be of a process kind - the temperature
varies continuously.
So far we have only considered a single symptom and its possible
ontology. What about a constellation of symptoms, or a syndrome, a basic
concept in the characterization of diseases? A syndrome can be regarded
as having two dimensions, first a spatial dimension, and second a temporal
one. From the point of view of the first dimension we consider the spatial
distribution of the symptoms. A person can have a number of contempora-
neous symptoms, partly distinguished by their location: headache, nausea,
pain in the stomach, sweating, etc. In the second dimension we mean by
a constellation the temporal distribution of symptoms: a headache at 11 is
followed by nausea at 12 , which in its turn is followed by a pain in the
stomach at 13 ,
In the extreme case where there is no change over time in the picture of
symptoms - when all symptoms remain static - then the whole constel-
lation is a (complex) state. In all other cases (i.e. the most probable
situation) the syndrome should be viewed as a series of events, possibly as
a process.
What then would we require for it to be a process? What is the unifying
principle for a syndrome? There is the identity of the bearer who has the
syndrome. It may also be the case that all or most elements in the syndrome
have the typical "disease"-character of either being intrinsically painful or
disabling or being believed to have such effects.
But could these features be sufficient to identify one disease-process?
What prevents us from saying that the patient has a number of diseases
simultaneously? This question takes us to the root of our conception of a
We talk about a cluster of symptoms as defining a disease species or
constituting a disease-process only if we believe that the symptoms includ-
ed are in some sense connected with each other. We have acquired such
a belief by noticing from a number of instances that these symptoms tend
to go together. We have, for instance, noticed that there is a typical

symptom complex constituted by lack of facial expression, tremor of the

distal segments of the limbs and muscular rigidity.
What kind of connection do we believe in? There is only one plausible
answer: we believe that the various symptoms have a common cause. And
we believe that this common cause, as yet unknown, is the factor which
holds this disease together and identifies it as one disease-process. There-
fore, although we do not know the cause of the symptoms constituting
paralysis agitans (Parkinson's disease) we believe that there is such a cause.
And, most importantly, had we not believed in the existence of such a
cause, we would never have developed a uniform conception of paralysis
agitans. The various symptoms would then have no status beyond their own
This finding is interesting since it entails the concept of causation being
present also in those disease-concepts where the specific etiological
background is unknown. Observe here that we certainly do not claim that
there really is a common cause for all the symptoms constituting a case of
paralysis agitans. The formation of the concept of paralYSis agitans may be
the result Of ignorance or of systematic mistakes. We may, in fact, eventual-
ly find good reasons for splitting up this disease into a great number of
diseases to which we can assign different causes. The bearing of our
observation is rather the following: as long as we believe that a particular
constellation of symptoms, instanced in an individual, are the symptoms
of one disease, as distinct from a number of diseases, then we presuppose
that these symptoms have a common cause. Also in this case, then, we have
good reasons for characterizing the disease as a process identified by a
cause, a bearer and the particular disease-distinguishing characteristics.
What then about a disease constituted by a biochemical abnormality?
There are two major possibilities of interpretation. One is to view the
biochemical abnormality as the ultimate cause (at least from all pragmatic
points of view) of a particular disease-process. (The biochemical abnormal-
ity could function as a cause irrespective of whether it is itself a process
or a state.) This move is particularly natural if the abnormality is a genetic
defect and there is no obvious environmental cause to be found. In such
a case the abnormality in itselffunctions as the idea which unifies a disease-
process, a process partly characterized by other biochemical changes, and
partly by sequences of symptoms.
According to another interpretation the biochemical abnormality is
viewed as secondary to some further internal or external cause. This cause
may be known or unknown. In the case where it is known, a characteri-

zation of a biochemical disease can be performed along the same lines as

infections; in the case where it is unknown, the characterization will be
more similar to the syndrome case. A biochemical abnormality will, howev-
er, always be better off than the syndrome, since it can in itself constitute
a unifying basis for a disease-process. The biochemical abnormality is, if
not the ultimate cause, a secondary cause of an ensuing sequence of
internal bodily changes and symptoms.
A story similar to this can be told for the case where the disease is
characterized as a disturbance in some homeostatic function. A difference
from the biochemical case is that the disturbance is perhaps less likely to
be viewed as a primary cause of the disease-process; a disturbance in
homeostasis, however, could well function as a disease-unifying secondary
What picture of disease ontology has now emerged as a result of this
analysis? The particular disease-concepts which we have chosen seem, in
general, to indicate that diseases are kinds of processes, in our extended
sense of this term. The unifying features of these processes are: a common
cause, a common bearer and the disease-distinguishing features of tending
to cause disability in the bearer.
The disease-concepts chosen here differ, however, in the way in which
they refer to these processes:

(1) Pneumococcal pneumonia is an acute infectious process caused

by a particular bacterium.
(2) Likewise, Tuberculosis is an infectious process caused by one of
two possible bacteria.
(3) The Common cold, somewhat differently, is a symptom complex
caused by an infectious process which in its turn is caused by
one of a great number of possible types of virus.
(4) Paralysis agitans is a symptom complex believed to have a
common cause of a neurological kind.
(5) Acute myocardial infarction is a symptom complex caused by a
physiologically abnormal process.
(6) Diabetes mellitus is a biochemical process (or state) (at least
secondarily) causing a constellation of symptoms.

Our short list indicates first an interesting difference in disease-concepts.

Partly for historical reasons, partly for reasons of our general knowledge

or ignorance, they refer primarily to different stages in pathological pro-

cesses. In some cases, the syndrome cases, they refer primarily to certain
final stages, viz. stages where the pathological process has become exter-
nalized and is apparent to the patient. In other cases, for instance regarding
infections and biochemical abnormalities, the disease-terms refer primarily
to much earlier stages in such a pathological process. In the instance of a
genetic biochemical disease, we can even say that the disease term refers
to the very first part of the disease-process.
What consequence does this finding have for our view of the ontology
of diseases? We shall suggest that it motivates a distinction between a
kernel sense of a disease-concept and an extended sense of the same con-
cept. Consider the following applications. In the kernel sense of'Jmeumon-
ia" and "tuberculosis" these terms refer to infectious processes; in the kernel
sense of "the common cole!', "paralysis agitans" and "myocardial infection"
these terms refer to symptom complexes; finally, in the kernel sense of
"diabetes mellitus" this term refers to a biochemical process.
But all these terms have an extended sense and this sense can be
extracted from our conceptual analysis. In this sense they can be said to
refer to more complex, but similar, episode-structures. Abstractly, and
expressed in a simplified way, these episode-structures are of the following

(1) An external object enters the human body; alternatively, there

is a basic genetic defect.

The external object or the genetic defect causes a succession of events on

a physiological or biochemical level; viz.

(2) There is an abnormal biochemical or physiological process.

The abnormal process causes a succession of events on the level of symp-

toms; viz.

(3) A constellation of symptoms (some of which have the disease-

distinguishing character) appears.

The disease-terms in the extended sense refer to the whole complex

process starting with an initial cause and terminating in a constellation of
symptoms. But how is this possible with those diseases where the etiology
is unknown? Our answer is that in such cases the disease-term does not
refer to any specified cause. But it refers in the general way suggested
above. Paralysis agitans, for instance, is, in the extended sense, that process
which has a unitary cause and which terminates in a well-known constel-
lation of symptoms of a neurological kind. III
There is a point in saying that "Paralysis agitans" refers to this whole
process. Once we know more about the underlying process and can specify
what it is, perhaps even in an ostensive way, we shall be immediately able
to use the term, "Paralysis agitans", for it. Another point, essential to the
purpose of this essay, is that we can give a general answer to our initial
question about the ontology of diseases. We shall now be able to say the
following: In an extended sense, all disease-terms "D" refer to a process of
the form 1 - 3 above. Not all elements in this process need to be known
or specified. In their kernel sense most disease terms refer to subsets of
elements of the structure 1 - 3.


We began this analysis by reviewing the classical debate between ontolog-

ists and physiologists concerning the ontology of diseases. We noticed that
the views of these two parties differed in two fundamental respects. First,
they had different opinions about the very nature of diseases, the ontolog-
ists claiming that a disease is a kind of object, the physiologists holding that
a disease is a disturbance in normal bodily functioning.
Second, they had different opinions about the possibility of defining and
categorizing disease-species. The ontologists firmly believed that it was
possible to do so; the physiologists would either deny this possibility, or
stress the theoretical difficulties in attempting to define diseases.
Our analysis has found a kernel of truth in both positions. In our view
the physiologists are closer to the truth in denying that diseases are sub-
stances or physical objects. They seem to be essentially right in holding that
diseases are some kind of disturbance in human functioning, i.e. that
diseases are kinds of episodes. On the other hand they seem to be wrong
if they conclude from this that diseases cannot be defined. If disease-con-
cepts are to have any function at all- which they obviously have - they must

be susceptible of definition. The ontologists are right in insisting on this

We drew attention to the fact that disease-concepts undergo historical
change. This fact has sometimes been used in arguments against the possi-
bility of defining disease-concepts. We have tried to settle the validity of
such arguments. An argument from historical change shows that there is
no single definition of the concept of, say, diabetes mellitus extending over
all history. The reason for this is, simply, that there is no single concept of
diabetes mellitus from antiquity to the 20th century. But in noticing this fact
we have removed the "historical" obstacle against definition. The diabetes
mellitus of Aretaeus, of Minkowski or of Banting and Best are certainly
concepts which can be defined. To attempt a definition of an ahistoric
concept of diabetes mellitus would, however, be futile.
As we have noted, some ontologists hold that disease-species are eternal
"essences". History gives us very little reason to believe that this is the case.
It is much more reasonable to believe that disease-species are convention-
ally formed by medical cultures on the basis of their knowledge, prejudice
and ignorance. This is, in any case, the view advocated in this essay.
In our constructive enterprise we have suggested that diseases belong to
the general ontological category of episodes. Episodes were subdivided into
states, events, and processes. Among processes we could distinguish
between simple processes and more complex ones. A simple process is
determined by a dimension along which a sequence of changes takes place.
A complex process requires a set of unifying principles holding it together
as a single episode.
For disease-processes we have suggested that there are three kinds of
unifying principles: 1. a common cause; 2. a bearer, the patient; 3. a set of
disabling effects in the bearer.
The various disease-concepts differ, however, in one important respect;
this holds more for contemporary theory than for pre-19th century theory.
They refer primarily to different stages in pathological development. Some
concepts refer to initial phases in such a development, others to intermedia-
ry phases, yet others to the final phases. These differences are mainly due
to the state of our present medical knowledge.
Our major suggestion is that most disease-concepts, in an extended
sense, also refer to more complex episodes of a process kind which have
a uniform structure. This structure has the following elements: 1 An exter-
nal cause and/or basic genetic defect; 2. An abnormal biochemical or
physiological process; 3. A constellation of signs and symptoms.

I For useful introductions to the history of the philosophy of health and disease. see [6S). [69).

[8S), [lOS), (108), (128), [129) and [130); concerning antiquity and the middle ages see [9S).
2 For a modern discussion ofthe legal and ethical aspects of involuntary psychiatric hospitali-
zation, see [80). The Swedish regulations relating to the use of force in psychiatry are to be
found in Lagen om beredande av sluten psyldatrisk vdrd i vissa fall (The Act on Mandatory
Institutional Psychiatric Care in Certain Cases.) Swedish Government Official Reports, SOU
3 For a thorough analysis of the procedures and decisions of Swedish insurance authorities
concerning sickness benefits, see [137).
A classical analysis of prewar German illness insurance policies is [SI).
4 The Swedish Public Health Act (Swedish Government Official Reports, SOU 1982:763) is
presented and commented on in (113).
5 The expression "medical concepts" may seem more natural here. For reasons elaborated
in Chapter four, section I, we wish, however, to reserve this expression for a subset of the
6 Aristotle's views on definition can be found for instance in his Topies, I 4-6, VI, VII, and

his Posterior Analyties, I 2, 10, II 3-13.

(106) is a detailed discussion of this and other kinds of definitions. For a short introduction
to the philosophy of definition, see [S4).
7 A proponent of what we call the strong version of nominalism is the Danish philosopher

Uffe Juul Jensen. Jensen (60) makes a distinction between two kinds of concepts, viz. concepts
of praxis and ideal concepts.
By an ideal concept he means a concept determined by an explicit definition. If we define
the concept of man as a rational animal, then we have created an ideal concept of man. But,
Jensen contends, few concepts are defined in such a precise way. However, we maintain that
these concepts exist. They exist as concepts of praxis; they exist in the sense that people use
them. This is tantamount to saying that people use certain linguistic expressions according
to certain, often implicit, rules. The concept of 'table' exists in the sense that people use the
word "table" according to well-known linguistic rules.
Now an important idea in Jensen is that the praxis concepts must precede the ideal concepts.
First comes language use; then we can make definitions and formulate ideal concepts.
This distinction lies at the basis of Jensen's program for conceptual analysis. His program
implies that we ought to study praxis-concepts and not ideal concepts. If we are interested
in the concept of health we should pay attention to how we, in our praxis, use the concept.
It is, he contends, much less interesting to study a particular explicit definition, viz. an ideal
concept, and it can even lead us in the wrong direction.
A program of this kind has important consequences, according to Jensen. Since the
conceptual praxis is partly changed over time, and since there may exist several contemporary
forms of praxis, the goal of the program is not to offer one single definition of health, i.e. to


create an ideal concept of health. The goal must instead be to mirror the conceptual
multiplicity and the conceptual development without formulating explicit (iefinitions.
Jensen complains about the fact that much ofthe discussion of the theory of health has been
misguided due to attention being focused on one particular concept of health, what he calls
the "machine-concept" of health. (As a paradigm for this concept he uses some formulations
of the Danish philosopher Alf Ross [107]. Ross' concept is in many ways similar to Boorse's
concept of health, which is discussed at length in Chapter two.)
According to this conception man can be viewed as a machine, a man is healthy if his
"machine" functions like the majority of other "machines" of his kind. He is diseased if the
function deviates from this norm, i.e. is abnormal. (See [60], pp. 142-156.)
This ideal concept, however, has very little to do with praxis, says Jensen. At most it
illuminates parts of praxis, i.e. the praxis of medical doctors oriented towards pure biological
Let us here comment on Jensen's program. We agree with Jensen in the following respects:
(i) The primary object of study are the praxis-concepts of health and disease.
The starting-point for a conceptual analysis is normally a given conceptual praxis (except
for the cases of pure stipulation). Already formulated ideal concepts are of secondary interest.
They would constitute the primary focus of study if we were to write the history of health
theory or a survey of modern health theories.
(ii) The praxis changes; the praxis-concepts of health and disease are changeable.
From these two common platforms Jensen draws, however, a number of conclusions which
differ from ours. Contrary to Jensen we find ideal concepts to have an important place in
conceptual analysis. We endorse the following thesis:
(iii) In some cases ideal concepts are the results of sharp reflection over praxis
concepts; if our goal is to become clear about praxis concepts, it is often
expedient to scrutinize some ideal concepts and compare them with praxis.
Jensen is certainly right in criticizing the naive acceptance of some ideal concept of health
(such as the "machine-concept"). Nevertheless, such concepts may be helpful in conceptual
research. A careful and precise ideal concept is an excellent platform for the further analysis
of existing praxis concepts.
We would like to sharpen this thesis. It is in practice necessary to think against a background
of ideal concepts, be they proposed in the literature or by the analyst himself. Every process
of conceptual analysis must contain tentative characterizations of the concept in focus.
These tentative characterizations are, at least, a kind of partial definition. Thus, they fulfill
the main criteria for being ideal concepts. The process of analysis then proceeds by way of
a criticism of these tentative characterizations. We may show that a certain conceptual praxis
does not agree with the proposed characterization. The latter must therefore be remodelled
in order to incorporate the counterinstance.
(iv) The result of a conceptual analysis should, if possible, be summarized in a
definition. In other words, the result of a successful conceptual analysis is one
(or a number of) ideal concept(s).

Our program deviates most significantly from Jensen's on this issue. Jensen seems to reject
definitions completely. His main argument against defining seems to be that definitions, in an
illegitimate way, "freeze" reality; they do not do justice to the mUltiplicity and the evolving
nature of concepts. .
We do not accept this argument. The extent to which concepts evolve can be mirrored by
a series of definitions which can capture every step in the change. In a similar way a number
of definitions can mirror the contemporary multiplicity of concepts.
We shall furthermore endorse the following thesis:

(v) The purpose of a conceptual analysis is not only to "mirror" given concepts of
praxis as sensitively as possible. The purpose is also to find fruitful simplifi-
cations, which can sharpen the concepts in their future scientific and technical

We consider it legitimate to make some marginal stipulations in the characterization of a

concept, if this involves some significant theoretical simplification. It is particularly legitimate
to make such a stipulation in those cases where the praxis is unclear, or when our own
linguistic intuitions do not give unequivocal results.
For a summary of Jensen's arguments, see also (61).
Since the first publication of the present volume Jensen [60] has been translated into English
[158]. Here Jensen uses the terms "abstract concepts" and "prototype concepts" for ideal
concepts. Similarly, "concrete concepts" is the term used to denote the praxis concepts.
8 For some classical definitions of health, see (130). Cpo (66), in particular pp. 131-145. For
contemporary characterizations stressing psychological and anthropological aspects, see (41),
[43) and [67).
9 Wittgenstein's theory of family-resemblance is presented in [1481, section 66.

10 An excellent example of such a socio-linguistic study is [146].

II There are a great number of authors who have proposed holistic views and holistic theories

of health and disease. Among philosophical works the following could be mentioned: [1), [20],
[33], [36], [38], [48], [99], [140] and [1411. Other important texts expressing holistic views
are [41] and [43].
The theories mentioned are often called non-neutralist since they combine the holistic view
with the idea that health is an evaluative (non-neutral) concept.
12 Christopher Boorse is the clearest and most widely cited representative of an analytic,

neutralist theory of health. His most important works are [13], [141 and [15). Penetrating
critical analyses of his work are (1], [48], (70) and [76].
Well before Boorse a British physician, J.G. Scadding [115), (116) and (117), presented an
analytic definition of disease. Scadding's general definition runs as follows:
A disease is the sum of the abnormal phenomena displayed by a group ofliving
organisms in association with a specified common characteristic or set of
characteristics by which they differ from the norm for their species in such a
way as to place them at a biological disadvantage ([116], p. 877).
This definition is very similar to Boorse's. It even suggests the notion of a species-typical
goal (in the phrase "biological disadvantage").
In Scandinavia the philosopher oflaw, AlfRoss [107] has presented a theory quite similar
to Boorse's. For a critical discussion of Ross, see [60], pp. 159-172.

13 Lucid expositions of the Hippocratic-Galenic conceptions of health are [95], pp. 127-135,

and [130].
14 The most celebrated revival of the idea of balance is Walter Cannon's masterpiece The
Wisdom of the Body, [21]. This balance is today mostly referred to as bodily homeostasis or
15 The widely used textbook on pathology by Hopps [57] displays both ideas:

In our study of pathology, we Will be concerned with deviations from normal

structure, chemical composition and function. To recognize these differences,
and to measure them, one must know the normal; hence the required prerequi-
sites of anatomy, biochemistry, and physiology. There is only one normal, within
relatively narrow limits; unfortunately there are many, many abnormals - an
infinite number of complex variations (p.2).
Perhaps the best concept of disease is that of imbalance ... In disease the patient
is out of balance, and his cells and tissues and organs are altered because of
Health then is a state of balance, and, since neither we nor our environment
is static, this is a constantly shifting balance - a state of dynamic equilibrium
16 Notice. for instance, the evaluative term "biological disadvantage" in the definition propos-

ed in [116]. p. 877:
A disease is the sum of abnormal phenomena displayed by a group of living
organisms ... by which they differ from the norm for their species in such a way
as to place them at a biological disadvantage.
Evaluative notions are particularly common in psychiatric discourse. The well-known
textbook by Redlich and Freedman [103] does not even mention statistical normality in its
characterization of mental diseases or behavior disorders:
What constitutes manifest behavior disorders depends on the culture or value
system. The culture also sanctions those interventions that are referred to as
psychiatric treatment (p. 2).
17 For a more thorough discussion of the notion of biological goals, see [59], pp. 101-124.
18 A similar distinction, expressed in terms of systems of actuality and systems of ideality,
is made in [99]. pp. 3-5.
18 This discussion is found in [15]. pp. 566-567.

20 In a note added to the reprint of [13] in [23] Boorse withdraws some of his statements in
the article:
... the view that illness is disease laden with values ... now seems a mistaken
concession to normativism. Illness is better analyzed simply as systemically
incapacitating disease, hence as no more normative than disease itself(p. 560).
21 It is now common to make some distinction between disease and illness roughly in the way
indicated by Boorse. From our philosophical point of view the differences in interpretation

can. however, be significant. Contrast Boorse's distinction with POrn's. as described in

Chapter three, section 7.
An early explicit distinction between disease and illness was made by Feinstein [44].
For each patient who ~ndergoes treatment. a clinician observes at least three
different types of data. The first type of data describes a disease in morphologic,
chemical. microbiologic. physiologic. or other impersonal terms. The second
type of data describes the hOSI in whom the disease occurs ... The third type of
data describes the illness that occurs in the interaction between the disease,!lIld
its environmental host. The illness consists of clinical phenomena: the host's
subjective sensations, which are caIled"symptoms". and certain findings. called
"signs", which are discerned objectively during the physical examination of the
diseased host.
22 For the conception of "Ianthanic disease". see [44]. p. 145.
23 We are indebted to [70], pp. 218-221. for taking the dynamic aspect of normality into
24 Agich [I] advances some further arguments against Boorse's philosophy of health. Among
other things, Agich discusses possible conflicts between the two biological goals acknowledg-
ed by Boorse, viz. individual survival and species survival. Similar observations are made in
[109], p. 712.
25 The quotation is taken from [1281. p.637.
26 A classical holistic characterization of health is the following by Herman Boerhaave:

The person who can perform the several actions proper to the human body with
ease, pleasure, and a certain constancy. is said to be well; and that condition
of the be dy is termed health. But if he either cannot perform those actions; or
ifhe performs them but with difficulty, pain. and sudden weariness; he is then
said to be ill: and that state of the body is called a disease ([121. I. 2-3).
21 The lOCUS classicus for this kind of position is Ludwig Wittgenstein's paragraph 580 in [1481:
An 'inner process' stands in need of outward criteria.
About this sentence Armstrong [51. p. 55, says the following:
When Wittgenstein speaks of'outward criteria' he means bodily behaviour. The
phrase 'inner process' refers to mental happenings of the sort that. prima/ode,
seem quite different from bodily behaviour: such things as thoughts and sen-
sations. In saying that 'inner processes' stand in need of outward criteria
Wittgenstein seems to be saying that there is a logically necessary connection
between the former and the latter.
Armstrong goes on to conclude that Wittgenstein must identify the 'inner process' with some
behaviour. As a result he is an analytical behaviourist.
Our position, however, is not behaviourist. A sensation like pain is not identical with
pain-behaviour. There is still a conceptual !ink between the concept of pain and the concept
of behaviour. Pain is a mental state which lends to result in certain sorts of behaviour. If we

identify a mental state as pain we commit ourselves, for conceptual reasons, to predicting a
certain kind of behaviour as well as a certain kind of inability to perform intentional actions.
28 For the term "stratification" see Shwayder's title [118], The Stratification of Behavior.
29 The concept of basic action was introduced by Arthur Danto (28]. The philosophy of basic

action and action-generation essentially adopted here is developed by Goldman [47]. For
similar treatments, see [98], pp. 43-63 and [132]. pp. 8-35.
30 See [47], pp. 20-48.

31 For more elaborate discussions of insufficient causes, see [72], pp. 29-58 and (88]. pp. 24-

32 The terms "accomplishment", "activity", "action-chain", and "action-sequence" have other

references in the case of other authors. The present use of ihese terms was introduced by me
in [90], pp.25-33.
33 An analysis of conventional generation requires a further concept, viz. the concept of
QUChorilY. An authority can be viewed as a "conventional power" attributed to a person or a
corporation. This power may last for a short while, as that of an umpire in a game, but it may
also last for a lifetime, as is the case with most monarchs. For some conventional accomplish-
ments authority is crucial. Only a judge can sentence a criminal; only a doctor can sign a
certificate of illness, and only a priest can administer the Holy Communion.
34 This list of distinctions has been inspired by [64], p. 131.
35 Ayers [6] argues strongly, in particular on pp. 125-144, against a conditional analysis of

power-concepts in general. A central argument is the following:

No kind of possibility-statement can conceivably be analysed into hypotheticals
like those suggested, simply because, as we have previously remarked,p entails
that p is possible whereas p does not entail that if q then p (p. 128).
Ayers is particularly suspicious of analysing human ability in terms of a conditional where
trying is the antecedent. Trying, he says, is no event or action antecedent to the "real" action.
Trying is identical with the action tried. It is only that some instances of trying are abortive
actions (i.e. when trying results in failure).
Kenny [64] disputes the general validity of the inference: A does F entails A can do F. Thus
he disputes Ayers' main argument against conditional analyses of the concept of ability.
A hopeless darts player may, once in a lifetime, hit the bull, but be unable to
repeat the performance because he does not have the ability to hit the bull.
... the distinction between luck and skill is not a marginal matter in this context:
it is precisely what we are interested in when our concern is ability, as opposed
to logical possibility or opportunity (p. 136).
36 Observe that this test is not always available. As is shown in the paragraphs to follow there

are instances when one cannOl even try. .

37 The essential ideas in the analysis presented in Chapter three, section 2, were first

published in [90].
38 In later writings I have modified my analysis of the background set of circumstances in the
following way. As the general concept I now propose the concept of accepted circumstances.
By "accepted" I mean that the circumstances are accepted by the person who ascribes ability or
disability to another person. These circumstances may be standard in the sense of being
commonly accepted in a particular culture but they need not be. For a more complete
introduction of this concept, see the Postscript of this volume, pp. 212-213.
39 The distinction between internal and external outcomes of an action has been well develop-
ed in [15 I), pp. 86-88. The author there introduces the term "result" for the outcome which
is included in the action itself. and "consequence" for an outcome which is excluded from the
40 Cf. the points made by Spicker [123] on the idea of a family-concept of health.

41 For a good philological account of the term "need" and its cognates in other languages.

see [131], pp. 149-155.

42 Important contributions to the general analysis of the concept of need are [17], [79], [131)
and [142].
43 Springborg [125] is a comprehensive historical treatment of theories of need. A standard

introduction to Marx's theory of needs is [53]. [7] makes a useful comparison between Marx's
and Maslow's theories of needs.
The main source for the discussion of Maslow's theory is [78]. See, in particular, pp. 35-58.
43 In [77), p. 21 the connection between the concepts of need and health is elaborated even
The concept "basic need" can be defined in terms of the questions which it
answers and the operations which uncovered it. My original question was about
psychopathogenesis. "What makes people neurotic?" My answer ... was, in brief,
that neurosis seemed at its core to be a deficiency disease; that it was born out
of being deprived of certain satisfactions which I called needs in the same sense
that water and amino acids and calcium are needs, namely that their absence
produces illness.
45 Since the present volume was first published Per-Erik Liss [159] has provided a thorough
examination of the notion of health-care need and related notions. Liss [160], Anton Aggernaes
[152] and Torbjorn Mourn [161] trace from various points of view the relations between the
notions of need and quality of life.
46 This conceptual circle was the object of my study in [92]. Some of the arguments advanced
there have been carried over to this chapter.
47 The treatment of Porn's and Whitbeck's theories is wholly based on the texts referred to.
viz. [99], [100), [140] and [141]. [93] contains a short critical discussion of these theories. [19]
is a recently published study relating Whitbeck's theory to Boorse's.
48 In a recent article [166] Porn has developed his theory of health considerably. Some ideas in
his present theory of health-as-adaptedness are presented and commented on in the Postscript of
this volume, pp. 206-208.
49 A more precise analysis of the concepts of intention and want has been given in [86],

pp. 61-93. For intentions the analysis runs as follows:

A intends to bring about P if, and only if,
A is in a.state which is such that
(i) A believes that he can bring about P
(ii) For all actions h, if A believes that his doing h is necessary for the bringing about
of P, and if A can do h, then A does h.

In contrast. a want is given the following characterization:

A wants to bring about P if, and only if,
A is in a state which is such that if A believes that he can bring about P, and
if A can decide to bring about P, then A decides to bring about P (i.e. makes
it the case that he intends to bring about P.)
In short, an intention is a peculiar kind of disposition to act in the direction of a goal. A
mere want is a disposition to make a decision (i.e. form an intention). Intentions are then in
a sense closer to actions than are wants.
This analysis is partly in the spirit of the one suggested by von Wright [151], pp. 95-96.
50 There is a subtle problem here which is difficult to reconcile within the framework of the
SG-theory. A person may drop a particular intention as a result of acquiring a disease. Thus,
the disease need not any longer compromise his health. A consequence of this is that the
person, who is completely adaptable - as far as his profile of goals is concerned - to negative
changes in his body, will continuously remain in health. This problem is partly identical with
the one called "the case of the person with very low ambitions" which is dealt with in Chapter
three, section 7.
It is the contention of this essay that the welfare-theory can handle this problem in an
intuitively more plausible way than the SG-theory.
51 This thesis needs the following qualification: according to some evaluations the vital goals
must fulfill some minimal moral requirements. The satisfaction of a profoundly evil man
cannot, according to this idea, qualify as happiness.
52 McGill [81] and Tatarkiewicz [127] are outstanding and comprehensive philosophical
treatments of the concept of happiness with large sections on the history of the concept.
Veenhoven [134] provides an excellent survey of modern empirical studies on happiness.
Other important philosophical treatments of happiness are [29] and [150], in particular
pp. 63-85 of the latter.
53 My analysis of happiness has been developed considerably in [164]. See also [165] for a
concise presentation. I there propose a version of a desire-satisfaction theory. The basic definition
is given in the following schema:
A is completely happy at time t, if and only if
(i) A wants at t that (xl ... xn) shall be the case at t,
(ii) (xl ...xn) constitutes the totality of A's wants at t,
(iii) A is convinced at t that (xl ...xn) is the case.
An intuitive way of expressing the contents of this schema is: A is completely happy, if and
only if A wants the conditions in the world to be just as he or she is convinced that they are.
[164] deals with a number of aspects not treated in the present volume, for instance the
dimensionality of happiness, the relation between happiness and pleasure, and ways of assessing
happiness. The latter analysis is placed in the context of the present-day discussion on
assessment and measurement of quality of life.
My basic view concerning the relation between the notions of health and happiness has
however remained intact. (For a summary see [164], pp. 96-101.)
54 For summaries of the Platonic and Aristotelian views on happiness, see [81], pp. 9-35. and
[127]. pp. 29-30.
ss Important modern introductions to the theory of emotions are [5]. [49]. [63], and [147].
56 The concept of formal objects of emotions was introduced in [63], pp. 187-202.
57 von Wright [150] pays explicit attention to the status of sensations:

We have pain in a tooth or pain in the stomach. But we do not commonly say
that we have pleasure in the mouth, when eating an apple. Pain, as has often
been observed, is much more sensation-like than pleasure. The word "pain" has
analogical uses, which resemble the use of "pleasure" in that they make the word
a value-attribute. But it seems to me right to say that, in its primary sense, "a
pain" refers to a kind of sensation and that "pain" names a sense-quality of
which, however, there are many shades. In this respect "pain" is on a different
logical level, both as compared with the substantive "pleasure" and as compared
with the adjectives "pleasant", "unpleasant", and "painful" (pp. 69-70).
58 Tatarkiewicz [127] writes:
Next, even complete satisfaction does not necessarily add up to happiness.
Satisfaction with particular things, however important - health or an untrou-
bled conscience, success or position - falls short of happiness if it is not
accompanied by other satisfactions. It is then only partial satisfaction; happi-
ness requires total satisfaction, that is satisfaction with life as a whole (p. 8).
59 Observe that Veenhoven [134] takes quite a different standpoint. On her view, happiness

cannot be simply a mood; instead it presupposes an intellectual judgment.

I use the word "happiness" where someone made an overall judgement about
the quality of his life. This implies an intellectual activity. Making an overall
judgement implies assessing past experiences and estimating future ones. Both
require marshalling facts into a convenient number of cognitive categories. It
also requires awarding relative values and setting priorities. Thus happiness is
not a simple sum of pleasures, but rather a cognitive construction which the
individual puts together from his various experiences (p.22).
60 The dimension of frequency suggests that it may be proper to ascribe happiness to a person
throughout a period to, ... ,1 10 without it's being true that this person has a continuous feeling
of happiness during this period. We shall accept this but at the same time claim the following:
a necessary condition for a person's being happy throughout a period is that he has a feeling
of happiness at least once during the period and that he has a disposition for such a feeling
during the whole period.
61 Tatarkiewicz [127] also assumes a dimension of depth.

Let us suppose that someone is satisfied with life, desires no change, has enough
money, good health, a loved wife, fine children, interesting work, and that this
is all he wants ... Though the majority will call him "happy", this view could be
questioned if we mean by happiness only that contentment which reaches the
depths of a man's consciousness and touches its innermost fibre (p. 21).

62 Rescher [104] is an important work dealing with many aspects of human welfare. The

conceptual structure developed there is, however, in some respects different from ours. In
other respects there are similar and parallel observations. One such observation concerns the
relation between health and the minimally good life.

Welfare is only thefoundation of such a life [the good life), not the structure itself.
Physical health, adequacy of resources, and mental and emotional well-being
are enormous - perhaps even indispensable - aids toward a meaningful and
satisfying life, but they are not in themselves sufficient for this purpose. This is
the reason why the components of the good life must extend far beyond the
province of welfare (p. 8).
Observe that the term "welfare" in Rescher's theory denotes certain basic causes of hap pi-
ness. In our theory happiness is the most important variant of welfare.
63 Caroline Whitbeck also attempts a distinction between health and happiness. We doubt,
however. that she succeeds in making it clear. Consider the following passage:
I argue, that to be happy, a person needs to be able to act in ways that serve
many goals, aspirations and projects simultaneously. The opportunity to do so
is a function of at least four things; the range and relative importance of a
person's goals ... ; the person's health ... ; the person's creativity; and the person's
access to resources of all types (social, economic and so forth) ([141), p. 620).
There are certainly important things here with which we agree. We agree that a person's
happiness depends on a number of factors of which health is only one. We also agree that all
the four factors mentioned are of importance for the creation of happiness.
The analysis as it stands gives, however, a very incomplete picture of the nature of happiness.
Moreover. it gives the impression that happiness is a concept very closely linked to health.
Whitbeck's analysis stresses the fact that a happy man needs to be able to act in ways
supportive of his goals. Happiness is presented as an ability-concept; the emotional aspect of
happiness is not mentioned. But ifhappiness is a kind of ability (or highly dependent on a kind
of ability) what would then distinguish happiness from health as understood by Whitbeck (i.e.
the subject-goal view of health)?
The answer seems not to be that the nature of the goals should be different. In both cases
Whitbeck speaks ofthe subject's goals, aspirations and projects. The fundamental difference
is that in order for a person to be happy he must be able to support many goals simultaneously;
this is not required in the case of health.
We shall thus envisage a man who is healthy but who is, according to Whitbeck's criterion,
unhappy. Consider a person with a great number of goals: he intends to perform well at his
job, he wants to take care of his children properly, he wants to go to the theatre once in a
while, he wants to travel to New York very soon, etc. Assume now that this person is
psychophysically capable of realizing each ufthese individual goals. Therefore, he is in health.
He does not, however, manage to complete the whole program within what he himself
considers to be a reasonable time interval. Therefore, he is unhappy.
There are at least two problems with this characterization. First, most goals have a (possibly
implicit) time clause built into them. If one has set oneself a goal one has set it to be realized,
roughly, at some time or within some time interval. When a person intends to pass an
examination. he normally intends to do so within the standard period of time. If a person
intends to get married, he normally intends to do so reasonably soon and not at any odd time
in life.
These considerations are important when we try to assess the state of health of the person

mentioned above. If he intends to realize all his goals within the coming week and fails to do
so for psychophysical reasons, then we are not entitled to say that he is in complete health
(unless the circumstances were extraordinary). In setting his goals, the person must, of course,
take into account the other goals he has set. In the case of his not grasping his whole profile
of goals and not understanding how unrealistic it is, he then qualifies as a man who is ill given
the subject-goal theory. The contrast between health and happiness is thus blurred.
The contrast between health and happiness, on Whitbeck~s platform, is blurred for yet
another reason: what is to be selected as one goal in contradistinction to many goals is a
completely arbitrary matter. Whatever state we designate as a goal can be subdivided into
many goals. This can be done in at least two ways. First, as we have shown above, every goal
has a (perhaps infinite) number of sub goals. In order to complete an examination one normally
passes a number of tests during the years of study. The passing of every such test can be
counted as a subgoal. Secondly, every ends tate can be divided into its parts. If an examination
consists of three salient final parts, then the passing of every such final part constitutes a part
of the endstate of completing the examination.
In a perfectly clear way, then, the serving of one goal involves the serving of several goals.
What is in one context considered to be one goal is in other contexts many goals. This certainly
also holds conversely. What is in one context considered to be many goals can in another
context be considered to be one. The person who intends to realize many goals during a week,
can be said to have one program to realize. To realize this program is his (single) goal.
We conclude therefore that Whitbeck's account of happiness cannot be adequate. Happiness
and health are distinct categories in a more profound way than her account suggests.
64 Let us parenthetically note that a theory which conceptually connects health with happi-
ness provides a further tool for the analysis of health. It provides a way of dealing with the
typical experiential concomitants of illness, such as suffering and pain, in a direct way.
In order to see this let us consider the relation between happiness and certain negative moods
such as anguish, and negative sensations such as pain. Is happiness compatible with these
mental phenomena?
There are two principal ways in which negative moods and sensations can disturb or even
annihilate happiness. One is direct and the other is more indirect. In the direct way the
negative feeling is so strong or intense that it "totalizes" the consciousness of a subject so that
there is no room for a feeling of happiness. To be able to feel happy it must be possible to
bring one's attention to some fact and contemplate it fairly undisturbed. In states of great pain
and depression this is impossible.
The other, indirect, way of preventing happiness is associated with the formal object of
happiness. Great pain and depression prevent many possibilities of goal-satisfaction, both via
one's own actions and otherwise. Therefore, one's own pain and depression are normally the
object of the emotion of unhappiness. If the depression and pain are great enough, then many
or most of one's goals are impossible to satisfy. Hence, there is a case for general unhappiness.
And general unhappiness must exclude general happiness.
Consider now how this observation could be used for an additional characterization of health
and illness: A is healthy if, and only if, A has the ability, given standard circumstances, to fulfill
his vital goals and A's happiness is not reduced under the level of minimal happiness by a
negative mood or a negative sensation.
Observe that we cannot include negative emotions in this characterization. A negative

emotion, such as disappointment or grief, can certainly create great unhappiness. But the
unhappiness about the outbreak of a war or about the loss of a close relative is not a criterion
of illness. This problem is discussed in Chapter four, section 2.
65 According to the Constitution o/the World Health Organization health should be defined in
the following way: Health is a state of complete physical, mental and social well-being and
not merely the absence of disease or infirmity.
This constitution was adopted by the International Health Conference which was held in
New York in 1946 and signed by the representatives of sixty-one countries. The Constitution
is reprinted in [23], pp. 83-84.
66 For a brilliant paper on the concept of mental health which panly seems to follow the same
general path as we do, see Moore [84]. He says:
The central idea behind illness is that of impairment: to be ill is to be impaired
from functioning in some of the wide varieties of ways we think to be normal
A main difference lies, however, in the importance he attaches to the concept of normality.
61The term "malady" has been introduced in the important work by Culver and Gen [27].
The precise definition of this concept is made in the following way:
A person has a malady if and only if he has a condition, other than his rational
beliefs and desires, such that he is suffering or at increased risk of suffering, an
evil (death, pain, disability, loss of freedom or opportunity, or loss of pleasure)
in the absence of a distinct sustaining cause ([27], p. 81).

The internal condition responsible for the evil can be of different kinds. It can be a wound,
a disorder, a defect, an affection and a lesion or a disease ([27], p. 65). Thus a malady is a
concept covering all typical causes of illness in the way we interpret illness.
Culver's and Gert's theory (the CG-theory) is in important respects similar to ours, but there
are also some notable differences. Let us try to collect some of these features.
(i) The CG-theory is holistic. The basic concept in the conceptual apparatus is the concept
of evil pertaining to the person as a whole.
(ii) Maladies are defined as internal states which cause or tend to cause some evil.
(iii) The CG-theory is wholly preoccupied with the negative aspect of health, viz. illness and
its causes. One can, however, say that there is an implicit theory of health. This implicit theory
is different from ours, since it seems completely tied to the notion of a malady. Where there
is no malady there is health, according to the CG·theory. We say instead, that this concept
of health is a special instance of a more general case which is not necessarily tied to the absence
of maladies.
(iv) The CG-theory seems to be more general than ours in another respect. The concept
of evil is more inclusive than the concept of disability. As Culver and Gert explicitly say, evil
covers such things as death, pain, disability, loss offreedom or opportunity or loss of pleasure.
We hope, however, that it has become clear from our discussion that the concept of ability
(or disability) has a much more central place than Culver and Gert suggest. Our analysis has

shown that loss of freedom, or loss of opportunity (when these phenomena are dependent on
internal conditions) cannot be distinct from disability. A condition which causes death or
threatens to cause death must, afortiori, also cause disability. We have previously discussed
the relation between pain (as well as other negative sensations and moods) and disability.
There is a case for maintaining an intimate relationship between these concepts, in the sense
that severe pain must necessarily lead to some disability. Thus. a notion of health based solely
on the concept of disability would be able to cover (almost) equally as much as Culver's and
Gerfs concept of evil with its superficially different special cases.
(v) The CG-theory stresses the fact that the malady is an internal condition which does
not have a sustaining (normally external) cause. By saying this Culver and Gert are partly
making the same qualification as we are when we speak ofa person's ability, given standard
circumstances. One of their examples concerns a man who is put in jail. He has some loss
of freedom which is due to an external cause. Thus, this loss of freedom does not constitute
a malady ([27). p. 79).
However. there is something odd about this example.lt does not describe any internal
condition of the agent whatsoever. The prevention is completely external, and there would
therefore be no question at all of maladies.
An interesting case would instead be the following: a finger is pricked by a needle which
is kept there. The subject feels pain as long as the needle remains.But here there is not just
an external condition. i.e. the needle's position on the body, but also an internal reaction to
this situation. The needle has actually cone some damage to the tissue and it has irritated the
pain receptors. The damage and the irritation are, however. of the kind that when the needle
is taken away. practically all reactions disappear. Some slight damage to the tissue remains
but it does not affect the bearer in a significant way.
Here we have an example of an internal condition existing as long as an external cause
actually sustains it. Culver and Gert have committed themselves to the view that this relation
entails the internal condition's not being a malady. In our analysis we have not committed
ourselves to this view; nor do we agree with it in general. To us damage is damage even if
it is constantly sustained by an external cause.
68 Our characterization of maladies is essentially in accordance with the one suggested by
Porn and Whitbeck. We shall here summarize their treatments of maladies.
Porn: I therefore think it is correct to characterize impairments, injuries, and diseases
as. respectively, states. changes. and processes of an anatomical, physiological,
or psychological kind which are evaluated as abnormal (poor, weak, etc.)
because of their causal tendency to restrict repertoires and thereby compromise
health ... However, it does not follow ... that a person is ill if he sustains an injury,
is affected by a disease, or is the bearer of an impairment. Owing to the
relational character of illness and the nature of relata, he is ill if and only ifthe
injury, disease, or impairment makes his repertoire inadequate relative to his
profiJe of goals ([99), pp. 6-7).

Whitbeck has a more complex characterization:

A disease is any type of psychophysiological process such that:

( I) People wish to be able to prevent or terminate the process because it interferes

with the bearer's psychophysiological capacity to do those things that people

commonly wish and expect to be able to do;
(2) either the process is statistically abnormal in those at risk or people have some
other basis for a reasonable hope of finding means to prevent or effectively treat
the prucess; and
(3) The pr:lcess is not also necessary for doing anything that people commonly want
and expect to be able to do «(140], p. 211).

Common to these ~haracterizations is then that diseases are psychophysiological pro-

cesses which tend to compromise people's psychophysiological ability to realize their goals.
Only Whitbeck, however. has worked out a theory of disease in any detail. The following
comments will exclusively concern her characterization of disease.
In the first clause disease is characterized as a process which interferes with the bearer's
psychophysiological capacity to do those things that people commonly wish and expect to do.
One element here seems to be too strong. Whitbeck actually claims that it interferes with some
capacity. But consider all early stages of diseases and all trivial diseases which need not
interfere with any capacity whatsoever. It would be more cautious to say that diseases tend
to interfere (that there is a high probability of interference) with certain abilities of their
Because of her strong requirements, Whitbeck is forced to relate the preventative force of
a disease not to the particular bearer's intentions but to "what people commonly wish and
expect to be able to do". Otherwise. on her view, a general science of diseases would be
impossible. But given the weaker formulation that a disease "tends to compromise the health
of the bearer" (in our interpretation) a clause referring to what people in general wish to do
is unnecessary.
Clause 2 is meant to exclude such human processes which are necessary parts of the human
lot. One could say that some features of old age are such necessary parts. They do not then
qualify as diseases. However. as Whitbeck explicitly mentions herself. other features of old
age are now being treated and at least partly "cured". These could then qualitY as diseases.
The third clause is intended to exclude the kind of disabling processes which are for various
reasons desirable. Pregnancy is perhaps the most obvious example. See our treatment ofthese
phenomena in Chapter four. section 2.
Whitbeck's theory of disease tries to find the right kind of dependence between disease and
health, an approach with which we sympathise. The dependence: should not be too tight. nor
too loose. We summarize some of her theses in the following way:
(I) A high degree of health should be compatible with the existence of diseases (as
well as with the existence of injuries and impairments; the latter concepts are.
mutatis mutandis. defined in the same fasnion as disease).
(2) Although the concrete conditions for health may vary from individual to individ-
ual. what counts as disease should not vary from individual to individual.
(3) We should have an explanation of the fact that most diseases tend to affect
health negatively in most individuals.

An important further feature of this theory, which may not seem equally plausible, is the
following: A person may be in very poor health without being affected by diseases, impair-
ments, or injuries. There may be a kind of process (or other condition) which negatively affects
a particular subject. It does not. however, qualify as a disease (impairment or injury) since
it does not affect what people commonly want to do. For instance. a unique and negative
experience may handicap a particular subject with regard to a particular ability. This ability
is of no importance for most people; it is. however. something that the subject wants to have.
Therefore, he is in bad health. although not affected by disease or any other malady. This
possiblity is also left open in our treatment of health and disease.
69 The example of lactase deficiency is taken from [561. pp. 189-190.

70 My reasoning here is simplified insofar as it presupposes a concept of pathological disease,

where diseases can be more or less completely identified with certain bodily processes.
Sundstrom [169] has shown that clinically interpreted diseases are much more complex entities,
partly entailing symptoms, signs and various kinds of disability. Given such a notion of disease
it would be improper to say that diseases cause illness. It is then more proper to say that
particular diseases partly entail certain forms of illness. I think that we can grant this very
important observation and still maintain that there is a distinction between the concepts of
disease and illness. To say that one has a disease is to say something different and more specific
than to say that one is ill.
71 In his very interesting account of the notion of disease, Reznek [167] presents an analysis
partly similar to mine. He does not, however, found his characterisation on a notion of health
but on the general notion of harm.
12 The ICIDH characterizes impairments in the following way:

Impairment is characterized by losses or abnormalities that may be temporary

or permanent, and that include the existence or occurrence of an anomaly,
defect, or loss in a limb, organ, tissue, or other structure of the body, induding
the system of mental function. Impairment represents exteriorization of a
pathological state, and in principle it reflects disturbances at the level of the
organ ([1441, p.47).

From this characterization it is not clear that an impairment is normally a consequence of

disease. This becomes, however, obvious in a chapter entitled The Consequences of Disease,
[144], pp. 23-43.
73 For an account of Porn's and Whitbeck's treatments of the concept of disease and similar
concepts, see footnote 68.
74 Observe the great overlap between mental diseases (or disorders) as classified by the
International Classification of Diseases. Injuries and Causes of Death, [1431, pp. 177-213 and
mental impairments and disabilities as classified by the ICIDH, [1441, pp. 48-50 and pp. 144-
147. For a discussion of this overlap, see [901, pp. 14-20.
7S The problem of old age and disease is well discussed in [221 and [341.
Engelhardt [34] notices, for instance, the relations between aging and the cardinal aspects
of the sick role (as described by Talcott Parsons). The latter can. Engelhardt says, be equally
well applied to aging as to normal disease.
( I) One is exempted from usual social responsibilities in a way not unlike the
excuses based on sickness; (2) the aged person is not considered immediately
responsible for his or her state ... (3) insofar as research focuses on postponing
aging or reversing its processes, aging, like sickness, is something to treat or

avoid; (4) the aged person, as the sick individual, becomes the one who would
seek the help of the health care establishment ([34], p. 190.)
76 For different treatments of grief in relation to health, see [27] and [32]. Engel [32] holds

that grieffulfills the conditions of being a disease, whereas Culver and Gert [27] argue against
this position. The reasons given by Culver and Gert are, however, different from ours ([27],
77 The thesis is most clearly summarized in [148], §243. In the paragraphs which follow there

is a long and very difficult discussion defending it.

78 The label "schizophrenia" seems, for instance, to be widely used in the Soviet Union. A

penetrating discussion of Soviet psychiatry is contained in [II]. pp. 322-341.

79 See [144], pp.202-205. For an attempt at measuring and specifying disabilities in the
context of neurological diseases, see [119].
80 For a discussion of the conceptual framework of the ICIDH, see [90], pp. 3-24. In [163] I
analyse this framework in the light of the contemporary discussion on disabilities and handicaps.
81 This discussion of homosexuality as a pathological condition owes much, both in terms of
arguments and empirical examples, to [109]. Other central sources for our presentation are
[8] and [50).
82 For a short summary of the history of homosexuality, see e.g. Bayer [8], pp. 15-21.
83 Clear accounts of Freud's views on homosexuality are given in his papers [45] and (46).
84 The difference between Freud and his followers is discussed in [8], pp. 21-40 and [69],
pp. 135-143.
85 Freud has summarized his views on the genesis of male homosexuality in this way in [46],

pp. 230-231.
86 See [121], pp. 118-119 and [122], pp.417-418.
87 See [120], p. 90.
88 [31], p. 199.
89 [31),p.200.

90 See [31), p.198: "For many centuries mankind has been trying to explain the

aetiopathogenesis of genuine homosexuality"(our italics).

91 The history of this diagnostic change has been written by Bayer [8]. See in particular
pp. 101-154.
92 An analysis similar to ours has been performed by Ruse [109]. pp. 693-722. Ruse discusses
whether homosexuality is a disease or illness from the point of view of two models of health
and disease. One is Christopher Boorse's naturalist model. The other is a normativist model.
which he mainly ascribes to Joseph Margolis and H. Tristram Engelhardt.
Ruse can show that homosexuality comes out very differently on the two platforms. He
can also reveal an internal tension in the naturalist model. The criterion of survival (in the
naturalist or the biostatistical model) need not support the same conclusion as the criterion
of reproduction. Male homosexuals naturally reproduce to a lesser extent than male heterose-
xuals. But, according to some endocrinological results, (presented, for instance. in the work
of Dorner) male homosexuals who have "fel'lale type" hypothalami tend to live longer than
male heterosexuals ([109], p. 712).
Ruse also discusses an interesting sociobiological explanation of the genesis of homosexuali-
ty. We do not, however, agree with the conclusions drawn by Ruse in this example. The
background assumptions for this type of explanation are the following.

The genes appear in pairs in an organism. If genes which are paired are identical, then the
organism is called "homozygous", otherwise "heterozygous". Ifin a population of organisms
there are just two kinds of genes which can occupy the two positions, then there are three
possible kinds of combinations, two homozygotes and one heterozygote. Let us illustrate the
combinations in the following way: .1\.110 .1\.12' A~2'
Assume now that A \.1 \ represents a low-reproducing kind of heterosexual. A \.12 is a repro-
ductively very fit heterosexual. And A~2 is a homosexual. Since a person of the .1\.12 type
is highly reproductive, he passes on a certain amount of .12 genes in the new generation, which
necessarily consists of a certain amount of A~2 combinations, viz. homosexuals. Given the
genetic structure this is bound to go on indefinitely. This means that the species design (of
the majority) has in itself secured and necessitated a certain number of homosexuals in each
The conclusion could then be drawn that the existence of a certain number of homosexuals
is in accordance with the species design. Thus, Ruse concludes, given a Boorsian naturalist
model of health, homosexuality would not qualify as a disease or illness ([109], p. 715).
This reasoning is, however, not obviously correct. Ruse introduces a notion of species design
which is different from Boorse's. Ruse's notion admits that there is more than one ideal type
for the individuals in the species. All the genetically determined types would be ideal or healthy
types. But this is not what Boorse says. The procedure he suggests for determining the species
design (i.e. the model for a healthy human being) is descriptive statistics. The design of the
majority of the population (with respect to each function) defines the criteria of health. Thus.
the properties of the individuals representing the most frequent genetic combination (which
is not the A~1 combination) would constitute the species design as we understand Boorse's
93 This was apparently one of the strongest arguments put forward by the Gay Liberation

Front in the American debate. See [8], pp. 67-100. Contrast, however, the contention of
Socarides [121). p. 122:
It is not true that homosexuals who seek treatment represent any special group
or skewed sample. The notion that the only homosexuals who enter therapy are
the "sick" homosexuals is erroneous in that often they are far less masochistic
and self-destructive than their partners or associates, who will not even attempt
any realistic effort to relieve their anguish.
94 See [112]. pp. 1079-1080.
See [10], pp. 337.432.
96 (10), pp. 450-457.

97 This general requirement distinguishes us from. for instance, Whitbeck. In [140], p.212,

she explicitly claims that her treatment of disease concerns human disease only.
98 [33]. [36], [58]. pp. 229-247, [66], pp. 165-183. [102] and [128] include good accounts of
the ontological and physiological views on the nature of disease.
99 For a discussion of the Hippocratic views on diseases, see e.g. [128], pp. 634-641.
\00 For a thorough analysis of the physiological thinking among the French medical theorists

of the 19th century. see [20], pp. 11-60.

10\ Bowman's [16] very valuable study on William Cullen emphasizes that the opposition
between ontologism and physiologism is often exaggerated. This holds in particular for the

time before the nineteenth century. For instance, Sydenham, Sauvages and Cullen all agreed
that diseases were processes. They saw no tension between the reification necessary for
classification and the view that diseases were variable processes [16], pp. 155-162.
102 For an illuminating discussion of Sydenham's views, see [102], pp. 353-369.
103 A principal source for this historical sketch is (96).
104 This summary is based on [96]; about Bernard, pp. 19-20, about Minkowski, pp. 26-27, and
about Banting and Best, pp. 47-61. The modem definition of diabetes mellitus is presented in
[9], p. 1599.
105 A comprehensive description of diabetes mellitus is presented in [9], pp. 1599-1619. The

development of the concept of diabetes mellitus is also discussed in [20], pp. 37-40.
106 This discussion presupposes Cahill's definition above. It appears, however, that the

modern definition may also be a matter of controversy. Papaspyros [96], p. 62, writes as
Insulin seemed at first to have solved the problem of the treatment of Diabetes
Mellitus. Soon, however, it became clear, that diabetes is not always the result
of insulin-lack. Other factors also, which block or destroy insulin may be
involved. For quite some years we know that the pancreas is not the only cause
of diabetes ...
107 For discussions of the history of modem classifications of diseases, see [89], pp. 13-22 and

108 The conceptual apparatus summarized here was introduced by me in [87], pp.43-50.

109 For the sake of simplicity we here speak of the causal origin in terms of a single cause.

The reasoning does not, however, depend on this assumption. The etiology may be multifac-
torial. Our thesis, however, presupposes that there is a set of causes determining a process
and thereby partially defining its limits.
110 Cf. Feinstein (44), p. 131, who makes a very sharp distinction between symptoms and signs.
He says:
A symptom is the name given to a subjective sensation or other observation that
a patient reports about his body or its products ... A sign is the name given to
an entity objectively observed by the clinician during physical examination of
the patient.
III Cf. ~lalmgren's discussion about unitary but unidentified etiology in [74], pp. 84-86.

[1] Agich, G. J.: 1983: 'Disease and Value: A Rejection of the Value-Neutrality Thesis',
Theoretical Medicine, 4, 27-41.
[2] Aretaeus: 1856, Aretaeus, The Cappadocian, The extant works, edited and translated
by F. Adams, The Sydenham Society, London.
[3] Aristotle: 1934, The Nicomachean Ethics, The Loeb Classical Library, Harvard
University Press, Cambridge, Massachusetts.
[4] Aristotle: 1960, II. Posterior Analytics, Topica, The Loeb Classical Library, Harvard
University Press, Cambridge, Massachusetts.
[5] Armstrong, D. M.: 1968,A Materialist Theory of the Mind, Routledge & Kegan Paul,
[6] Ayers, R. R: 1968, The Refutation of Determinism, Methuen & Co Ltd., London.
[7] Bay, c.: 1980, 'Human Needs, Wants, and Politics: Abraham Maslow, meet Karl
Marx', Social Praxis, 7, 233-252.
[8] Bayer, R: 1981, Homosexuality and American Psychiatry. The Politics 0/ Diagnosis,
Basic Books, Inc., New York.
[9] Beeson, P. B. and McDermott, W. (eds.): 1975, Textbook o/Medicine, 14th ed., W. B.
Saunders Company, Philadelphia.
rIO] Bell, A. P. and Weinberg, M. S.: 1978, Homosexualities: A Study 0/ Diversity Among
Men and Women, Mitchell Beazley Publishers Limited, London.
[11] Bloch, S. and Chodoff, P. (~s.): 1981, Psychiatric Ethics, Oxford University Press,
[12] Boerhaave, H.: 1751-1757, Academical Lectures on the Theory 0/ PhYSic, 6 vols.,
Rivington et aI., London.
[13] Boorse, c.: 1975, 'On the Distinction Between Disease and Illness', Philosophy and
Public Affairs, 5, 49-68.
[14] Boorse, c.: 1976, 'What a Theory of Mental Health Should Be', Journal/or the Theory
0/ Social Behavior, 6,61-84.
[15] Boorse, c.: 1977, 'Health as a Theoretical Concept', Philosophy 0/ Science, 44,
[16] Bowman, I. A.: 1975, William Cullen (1710-90) and the Primacy o/the Nervous System,
PhD Dissertation at Indiana University, Indiana.
[17] Braybrooke, D.: 1968, 'Let Needs Diminish That Preferences May Prosper', Ameri-
can Philosophical Quarterly, Monograph Series, 86-107.
[18] Bronaugh, R N.: 1968, 'The Logic of Ability Judgments', The Philosophical Quarterly,
18, 122-130.
[19] Brown, W. M.: 1985, 'On Defining 'Disease", The Journal o/Medicine and Philosophy,
[20] Canguilhem, G.: 1978, On the Nonnal and the Pathological, D. Reidel Publishing
Company, Dordrecht.


[21] Cannon, W. B: 1932, The Wisdom of the Body, Norton, New York.
[22] Caplan, A. L.: 1981, 'The "Unnaturalness" of Aging - A Sickness Unto Death?' in
Caplan A. L., Engelhardt, H. T jr., and McCartney, J. J. (eds.): Concepts of Health
and Disease; Interdisciplinary Perspectives, Addison-Wesley Publishing Company,
Reading, Massachusetts, pp. 725-737.
[23] Caplan A. L., Engelhardt H. T., Jr. and McCartney J. J. (eds.): 1981, Concepts of
Health and Disease; Interdisciplinary Perspectives, Addison-Wesley Publishing
Company, Reading, Massachusetts.
[24] Carr, D.: 1980, 'The Logic of Knowing How and Ability', Mind, 89, 394-409.
[25] Clouser, K. D., Culver, C. M. and Gert, E.: 1981, 'Malady: A New Treatment of
Disease', The Hastings Center Report, 11,29-37.
[26] Cullen, W.: 1769, Synopsis nosologiae methodicae, William Creech, Edinburgh, Scot-
[27] Culver, C. M. and Gert, E.: 1982, Philosophy in Medicine: Conceptual and Ethical
Issues in Medicine and Psychiatry, Oxford University Press, Oxford.
[28] Danto, A. c.: 1968, 'Basic Actions' in White, A. R. (ed.), The Philosophy of Action,
Oxford University Press, Oxford, pp. 43-58
[29] Davis, W.: 1981, 'A Theory of Happiness', American Philosophical Quarterly, 18,
[30] DSM-III: 1980, Diagnostic and Statistical Manual ofMenial Disorders (Third Edition),
The American Psychiatric Association, Washington.
[31] Dorner, G.: 1976, Hormones and Brain Differentiation, Elsevier Scientific Publishing
Company, Amsterdam.
[32] Engel, G. L.: 1960, 'A Unified Concept of Health and Disease',Perspectives in Biology
and Medicine, 3, 459-485.
[33] Engelhardt, H. T, Jr.: 1975, 'The Concepts of Health and Disease', in Engelhardt
H. T Jr., and Spieker, S. F. (eds.), Evaluation and Explanation in the Biomedical
Sciences, D. Reidel Publishing Company, Dordrecht, pp. 125-141.
[34] Engelhardt, H. T., Jr.: 1979, 'Is Aging a Disease?' in Veatch, R. M., (ed.), 1979, Life
Span, Values and Life-Extending Technologies, Harper & Row Publishers, New York,
pp. 184-194.
[35] Engelhardt, H. T, Jr.: 1981, 'Clinical Judgment', Metamedicine, 2,301-317.
[36] Engelhardt, H. T., Jr.: 1982, 'The Subordination of the Clinic', in Gruzalski, B. and
Nelson, C. (eds.), Value Conflicts in Health Care-Delivery, Ballugh Publishing Com-
pany, Cambridge, Massachusetts, pp. 41-56.
[37] Engelhardt, H. T, Jr.: 1982, 'Illnesses, Diseases, and Sicknesses', in Kostenbaum,
V. (ed.) The Humanity of the Ill, University of Tennessee Press, Knoxville, pp.
[38] Engelhardt, H. T., Jr.: 1984, 'Clinical Problems and the Concept of Disease', in [94],
[39] Erde, E. L.: 1979, 'Philosophical Considerations Regarding Defining "Health", "Dis-
ease", etc., and Their Bearing on Medical Practice', Ethics in Science and Medicine,
[40] Faber, K.: 1923, Nosography in Modem Internal Medicine, Paul F. Holber, New York.
[41] Fabrega, H., Jr.: 1972, 'Concepts of Disease: Logical Features and Social Impli-
cations', Perspectives in Biology and Medicine, 15,538-617.

[42] Fabrega, H., Jr.: 1975, 'The Position of Psychiatry in the Understanding of Human
Disease', Archives of General Psychiatry, 32,1500-1512.
[43] Fabrega, H., Jr.: 1979, 'The Scientific Usefulness of the Idea of Illness', Perspectives
in Biology and Medicine, 22, 545-558.
[44] Feinstein, A. R.: 1967, Clinical Judgment, Robert E. Krieger Publishing Company,
Huntington, New York.
[45] Freud, S.: 1920, 'The Psychogenesis of a Case of Homosexuality in a Woman', in
Strachey, J. (ed.): 1955, The Standard Edition of the Complete Psychological Works of
Sigmund Freud, vol. XVIII, The Hogarth Press, London.
[46] Freud, S.: 1922, 'Some Neurotic Mechanisms in Jealousy, Paranoia and Homose-
xuality', in Strachey, J. (ed.): 1955, The Standard Edition ofthe Complete Psychological
Works of Sigmund Freud, vol. XVIII, The Hogarth Press, London.
[47] Goldman, A.I.: 1970, A Theory of Human Action, Prentice-Hall Inc., Englewood
Cliffs, New Jersey.
[48] Goosens, W. K.: 1980, 'Values, Health, and Medicine', Philosophy of Science. 47,
[49] Gordon. R. M.: 1974, The Aboutness of Emotions', American Philosophical Quarter-
ly, 11, 27-36
[SO] Green, R.: 1972, 'Homosexuality as Mental Illness', The International Journal of
Psychiatry, 10, 77-98.
[51] Graff, S.: 1949, Medizinische und pathologisch-anatomische Forschung und Lehre:
Betrachtungsweisen und ihre Auswirkung in Arztlum, Rechtspftege und Versicherungs-
wesen, Stromverlag, Hamburg, Bergedorf.
[52] Hayflick, L.: 1976, The Cell Biology of Human Aging', The New England Journal of
Medicine, 295, 1302-1308.
[53] Heller, A.: 1974, The Theory of Need in Marx, Allison and Busbey, London.
[54] Hempel, c.: 1952, Fundamentals of Concept Formation in Empirical Science, Inter-
national Encyclopedia of Unified Science, The University of Chicago Press, Chica-
[55] Hesslow, G.: 1983, 'Explaining Differences and Weighing Causes', Theoria, 49,
[56] Hesslow, G.: 1984, 'What is A Genetic Disease? On the Relative Importance of
Causes', in [94], pp. 183-193.
[57] Hopps, H. c.: 1964, Principles of Pathology, Appleton-Century-Crofts, Inc., New
[58] Hudson, R.: 1983, Disease and Its Control: The Shaping of Modem Thought,
Greenwood Press, London.
[59] Hull, D. L.: 1974, Philosophy of Biological Science, Prentice-Hall Inc., Englewood
Cliffs, New Jersey.
[60] Jensen, U. J.: 1983, Sygdomsbegreber i praksis, Munksgaard, Copenhagen.
[61] Jensen, U. J.: 1984, 'A Critique of Essentialism in Medicine', in [94], pp. 63-73.
[62] Kass, L. R.: 1975, 'Regarding the End of Medicine and the Pursuit of Health', Public
Interest, 40, 11-42.
[63] Kenny, A.: 1963, Action, Emotion and Will, Routledge & Kegan Paul, London.
[64] Kenny, A.: 1975, Will, Freedom and Power, Basil Blackwell, Oxford.
[65] King, L. S.: 1954, 'What is a Disease?', Philosophy of Science. 21, 193-203.

[66] King, L. S.: 1982, Medical Thinking: A Historical Preface, Princeton University Press,
[67] Kraupl Taylor, F.: 1976, 'The Medical Model of the Disease Concept', British Journal
of Psychiatry. 128,588-594.
[68] Kraupl Taylor, R: 1979, The Concepts of Illness. Disease and Morbus, Cambridge
University Press, Cambridge.
[69] Lewis, H. B.: 1981, Freud and Modem Psychology, vol. 1: The Emotional Basis of
Mental Illness, Plenum Press, New York.
[70] Lie, R K.: 1982, Acceptance and Rejection ofHypotheses in Medical Science, Filosofisk
Institutt, University of Bergen, Bergen, Norway.
[71] Linnaeus, c.: 1763, Genera morborum in auditorum usum, Steinert, Upsaliae.
[72] Mackie, J. L.: 1974, The Cement of the Universe, Oxford University Press, Oxford.
[73] Macklin, R: 1972, 'Mental Health and Mental Illness: Some Problems of Definition
and Concept Formation', Philosophy of Science. 39,341-365.
[74] Malmgren, H.: 1984, 'Psychiatric Classification: The Status of So-called Diagnostic
Criteria', in [94], pp. 77-87.
[75] Margolis, J.: 1976, The Concept of Disease' , The Journal ofMedicine and Philosophy.
1, 239-255.
[76] Margolis, J.: 1980, The Concept of Mental Illness: A Philosophical Examination',
in Brody, B. A., and Engelhardt, H. T. Jr. (eds.), MentalIllness: Law and Public Policy,
D. Reidel Publishing Company, Dordrecht, pp. 3-23.
[77] Maslow, A. H.: 1968, Toward a Psychology of Being, 2nd ed., D. van Nostrand
Company, New York.
[78] Maslow, A. H.: 1970,Motivation and Personality, 2nd ed., Harper & Row Publishers,
New York.
[79] McCloskey, H. J.: 1976, 'Human Needs, Rights and Political Values', American
Philosophical Quarterly. 13, 1-11.
[80] McGarry, L. and Chodoff, P.: 1981, The Ethics ofInvoluntary Hospitalization', in
Bloch, S. and Chodoff, P. (eds.) Psychiatric Ethics, Oxford University Press, Oxford,
pp. 203-219.
[81] McGill, V. J.: 1967, The Idea of Happiness, Concepts in Western Thought Series,
Fredrick A. Praeger Publishers, New York.
[82] Mercer, J. R: 1972, 'Who is Normal? Two Perspectives of Mild Mental Retar-
dation', in Jaco, E. G. (ed.), Patients. Physicians. and Illness, The Free Press, New
York, pp. 66-85.
[83] Moore, M. S.: 1976, 'Some Myths about 'Mental Illness", Inquiry, 18,233-265.
[84] Moore, M. S.: 1980, 'Legal Conception,s of Mental Illness', in Brody, B. A., and
Engelhardt, H. T., Jr. (eds.), Mental Illness : Law and Public Policy, D. Reidel Publish-
ing Company, Dordrecht, pp. 25-69.
[85] Neale, S. A.: 1976, The Concept of Health in Medicine, Ph.D. Dissertation at George-
town University, Washington D.C.
[86] Nordenfelt, L.: 1974, Explanation ofHuman Actions, Philosophical Studies Published
by the Department of Philosophy and the Philosophical Society, 20, University of
Uppsala, Uppsala.
[87] Nordenfelt, L.: 1977, Events. Actio1)s, and Ordinary Language, Doxa Studies in the
Philosophy of Language, Lund.

[88] Nordenfelt, L.: 1981, Causation, Akademilitteratur, Stockholm.

[89] Nordenfelt, L.: 1983, Causes of Death, Swedish Council for Planning and Coordi-
nation of Research, Report 83:2, Stockholm.
[90] Nordenfelt, L.: 1983, On Disabilities and Their Classification, Studies on Health and
Society, SHS 1, University of LinkOping, LinkOping.
[91] Nordenfelt, L.: 1984, Five Studies in Action Theory, Studies in Communication, SIC
8, University' of LinkOping, LinkOping.
[92] Nordenfelt, L.: 1984, 'On the Circle of Health', in [94], pp. 15-23.
[93] Nordenfelt, L.: 1984, 'Comments on POrn's 'An Equilibrium Model of Health" in
[94], pp. 11-13.
[94] Nordenfelt, L., and Lindahl, B.1. B. (eds.): 1984, Health. Disease. and Causal Explan-
ations in Medicine, D. Reidel Publishing Company, Dordrecht.
[95] Ottosson, P-G.: 1984, Scholastic Medicine and Philosophy. A Study of Commentaries
on Galen's Tegni, Bibliopolis, Naples.
[96] Papaspyros, N. S.: 1964, The History of Diabetes Mellitus, Georg Thieme Verlag,
[97] Parsons, T.: 1972, 'Definitions of HeaIth and Illness in the Light of American Values
and Social Structure', in Jaco, E. G. (ed.), Patients. Physicians. and Illness, The Free
Press, New York, pp. 107-127.
[98] POrn, I.: 1977, Action Theory and Social Science, D. Reidel Publishing Company,
[99] POrn, I.: 1984a, 'An Equilibrium Model of Health', in [94], pp. 3-9.
[100] POT!), I.: 1984b, 'Vardllira och vardvetenskap', Socialmedicinsk Tidskrift. 61,302-306.
[101] Rather, L. J.: 1958, 'Harvey, Virchow, Bernard, and the Methodology of Science',
in Rather, L. J. (ed.) Disease. Life. and Man, Setected Essays by Rudolf Virchow,
Stanford University Press, Stanford.
[102] Rather, L. J.: 1959, 'Towards a Philosophical Study of the Idea of Disease' in
Brooks, C. M. and Cranefield, P. F: (eds.), The Historical Development ofPhysiological
Thought, Hafner, New York, pp. 351-373.
[103] Redlich, F. C. and Freedman, D. X.: 1966, The Theory and Practice of Psychiatry,
Basic Books Inc., New York.
[104] Rescher, N.: 1972, Welfare: The Social Issues in Philosophical Perspective, University
of Pittsburgh Press, Pittsburgh.
[105] Riese, W.: 1953, The Conception of Disease; Its History. Its Versions and Its Nature,
Philosophical Library, New York.
[106] Robinson, R.: 1954, Definition, The Clarendon Press, Oxford.
[107] Ross, A.: 1979, 'Sygdomsbegrebet', Bibliotekfor Laeger. 171/2, 111-129.
[108] Rotschuh, K. E. (ed.): 1975, Was ist krankheit? Erscheinung. Erklarung. Sinngebung,
Vol. CCCLXIII of 'Wege der Forschung', Wissenschaftliche Buchgesellschaft,
[109] Ruse, M.: 1981, 'Are Homosexuals Sick?', in Caplan, A. L., Engelhardt, H. T. Jr.,
and McCartney, J. J. (eds.), Concepts ofHealth and Disease; Interdisciplinary Perspec-
tives, Addison-Wesley Publishing Company, Reading, Massachusetts, pp. 693-723.
[110] Ryle, J. A.: 1961, 'The Meaning of NormaI', in Lush, B. (ed.), Concepts ofMedicine:
A Collection ofEssays on Aspects of Medicine, Pergamon Press, Oxford, pp. 137-149.

[l11] Sadegh-zadeh, K.: 1977, 'Krankheitsbegriffe und nosologische Systeme', Metamed,

[l12] Saghir, M. T., Robins, E., Walbran, B., and Gentry, K. A.: 1970, 'Homosexuality:
III. Psychiatric Disorders and Disability in the Male Homosexual', American Journal
of Psychiatry, 126, 1079-1086.
[113] Sahlin, J. (ed.): 1982, Htilso- och sjukvdrdslagen med kommentarer, Statens nlimnd for
utgivande av f6rvaltningsrattsliga publikationer, Stockholm.
[114] Sauvages de la Croix, Fran'rois Boissier de: 1763, Nosologia methodica sistens morbo-
rum classes juxte Sydenhami mentem et botanicorum ordinem, 5 vols., Fratrum de
Tournes, Amsterdam.
[115] Scadding, J. G.: 1963, 'Meaning of Diagnostic Terms in Bronchopulmonary Dis-
ease', British Medical Journal, 1425-1430.
[116] Scadding, J. G.: 1967, 'Diagnosis: The Clinician and the Computer', The Lancet, 21,
[117] Scadding, J. G.: 1972, 'Viewpoint: The Semantics of Medical Diagnosis' ,Internation-
al Journal of Biomedical Computing, 3, 83-90.
[118] Shwayder, D. S.: 1965, The Stratification of Behaviour, Routledge & Kegan Paul,
[119] Slater, R. J. and Raun, N. E. (eds.): 1984, Symposium on a Minimal Record of
Disability for Multiple Sclerosis, Acta Neurologica Scandinavica, Supplementum No.
101, vol. 70.
[120] Socarides, C. W.: 1968, The Overt Homosexual, Grune and Stratton, New York.
[121] Socarides, C. W.: 1972, 'Homosexuality - Basic Concepts and Psychodynamics',
The International Journal of Psychiatry, 10, 118-125.
[122] Socarides, C. W.: 1978, 'The Sexual Deviations and the Diagnostic Manual',Ameri-
can Journal of Psychotherapy, 32, 414-426.
[123] Spicker, S.: 1984, 'Comments on Nordenfe1t's 'On the Circle of Health", in [94],
[124] Spitzer, R. L.: 1981, 'The Diagnostic Status of Homosexuality in DSM-III: A
Reformulation of the Issues', American Journal of Psychiatry, 138, 210-215.
[125] Springborg, P.: 1981, The Problem of Human Needs and the Critique of Civilization,
Allen & Unwin, London.
[126] Sydenham, T.: 1848, 'Medical Observations Concerning the History and the Cure
of Acute Diseases', The Works of Thomas Sydenham, translated by R. G. Latham,
vol. I, The Sydenham Society, London.
[127] Tatarkiewicz, W.: 1976, Analysis of Happiness, Martinus Nijhoff, The Hague.
[128] Temkin, 0.: 1963, 'The Scientific Approach to Disease: Specific Entity and Individ-
ual Sickness', in Crombie, A. C. (ed.), Scientific Change: Historical Studies in the
Intellectual, Social and Technical Conditions for Scientific Discovery and Technical
Invention from Antiquity to the Present, Basic Books, Inc., New York, pp. 629-647.
[129] Temkin, 0.: 1973, Galenism: Rise and Decline ofMedical Philosophy, Cornell Univer-
sity Press, Ithaca, New York.
[130] Temkin, 0.: 1973, 'Health and Disease', Dictionary of the History of Ideas, 2, Scrib-
ner's, New York, pp. 395-407.
[131] Traney, K. E.: 1975, "Ought' implies 'Can': A Bridge From Fact to Norm', Part II,
Ratio, XVII, 147-175.

[132] Tuomela, R.: 1977, Human Action and Its Explanation, D. Reidel Publishing Compa-
ny, Dordrecht.
[133] Veatch, M. (ed.): 1979, Life Span. Values and Life-Extending Technologies, Harper
& Row Publishers, New York.
[134] Veenhoven, R.: 1984, Conditions of Happiness, D. Reidel Publishing Company,
[135] Walton, R.: 1969, 'Need: A Central Concept', Social Service Quarterly, summer,
[136) Warner, R.: 1976, 'The Relationship between Language and Disease Concepts',
International Journal of Psychiatry in Medicine. 7, 57-68.
[137) Westerhiill-Gisselsson, L.: 1983, Sjukdom och arbetsoformdga, P. A. Norstedts &
SOners fdrlag, Stockholm.
[138) Virchow, R.: 1958, Disease. Life. and Man, Selected Essays by Rudolf Virchow,
translated and with an introduction by L. I. Rather, Stanford University Press,
[139] Whitbeck, c.: 1977, 'Causation in Medicine: The Disease Entity ModeJ', Philosophy
of Science. 44, 619-637.
[140] Whitbeck, c.: 1978, 'Four Basic Concepts of Medical Science', PSA. 1,210-222.
[141] Whitbeck, c.: 1981, 'A Theory of Health', in Caplan,A. L., Engelhardt, H. T. Ir., and
McCartney, I. I., (eds.), Concepts ofHealth and Disease .. Interdisciplinary Perspectives,
Addison-Wesley Publishing Company, Reading, Massachusetts, pp. 611-626.
[142) White, A. R.: 1974, 'Needs and Wants', Philosophy of Education Society Proceedings.
8, Part 2, 159-180.
[143) WHO (World Health Organization): 1977, Manual of the International Statistical
Classification ofDiseases. Injuries. and Causes ofDeath, 2 vols., 9th Revision, Geneva.
[144) WHO (World Health Organization): 1980, International Classification of Impair-
ments. Disabilities. and Handicaps (ICIDH), WHO, Geneva.
[145) Williams, A.: 1974, "Need' as a Demand Concept (with special reference to health)"
in Culyer, A. I. (ed.), Economic Policies and Social Goals, Robertson, London, pp.
[146] Williams, R.: 1983, 'Concepts of Health: An Analysis of Lay Logic', Sociology. 17,
[147] Wilson, I. R. S.: 1972, Emotion and Object, Cambridge University Press, Cambridge.
[148] Wittgenstein, L.: 1953, Philosophical Investigations, Basil Blackwell, Oxford.
[149] Wood, P. H. N.: 1982, Advances in the Classification of Disease, in A. Smith (ed.),
Recent Advances in Community Medicine, Number Two, pp. 169-183.
[150] von Wright, G. H.: 1963, The Varieties of Goodness, Routledge & Kegan Paul,
[151] von Wright, G. H.: 1971, Explanation and Understanding, Routledge & Kegan Paul,

[152] Aggernaes, A: 1994, On General and Need-Related Quality of Life: A Psychological

Theory for Use in Medical Rehabilitation and Psychiatry, in Nordenfelt L. (ed.):
Concepts and Measurement of Quality of Life in Health Care, Kluwer Academic
Publishers, Dordrecht, pp.241-256.
[153] Braybrooke, D.: 1987, Meeting Needs: Studies in Moral, Political and Legal
Philosophy, Princeton University Press, Princeton.
[154] Green, O.H.: 1992, The Emotions: A Philosophical Theory, Kluwer Academic
Publishers, Dordrecht.
[155] Fulford, K.W.M.: 1989, Moral Theory and Medical Practice, Cambridge University
Press, Cambridge.
[156] Fulford, K.W.M.: 1993, Praxis Makes Perfect: Illness as a Bridge Between Biological
Concepts of Disease and Sociological Conceptions of Health, Theoretical Medicine,
14, No 4, 305-320.
[157] Griffin, J.: 1986, Wellbeing: Its Meaning, Measurement and Moral Importance,
Clarendon Press, Oxford.
[158] Jensen, v.J.: 1987, Practice and Progress: A Theory for the Modern Health Care
System, Blackwell Scientific Publications, Oxford.
[159] Liss, P-E.: 1993, Health Care Need, Avebury, Aldershot.
[160] Liss, P-E.: 1994, On Need and Quality of Life, inNordenfelt L. (ed.): Concepts and
Measurement of Quality of Life in Health Care, Kluwer Academic Publishers, pp.63-
[161] Moum, T.: 1994, Needs, Rights and Resources in Quality of Life Research, in
Nordenfelt L. (ed.): Concepts and Measurement of Quality of Life in Health Care,
Kluwer Academic Publishers, Dordrecht.
[162] Nordenfelt, L.: 1990, Towards a Theory ofHealth Promotion, Health Service Studies,
5, Linkoping Collaborating Centre for WHO programmes, Linkoping.
[163] Nordenfelt, L.: 1993, On the Notions of Disability and Handicap, Scandinavian
Journal ofSocial Welfare, 2, 17-24.
[164] Nordenfelt, L.: 1993, Quality ofLife, Health and Happiness, Avebury, Aldershot.
[165] Nordenfelt, L.: 1994, Towards a Theory of Happiness: A Subjectivist Notion of Quality
of Life, in Nordenfelt L.(ed.): Concepts and Measurement ofQuality ofLifo in Health
Care, Kluwer Academic Publishers, Dordrecht.
[166] Porn, 1.: 1993, Health and Adaptedness, TheoreticalMedicine, 14, No 4,295-303.
[167] Reznek, L.: 1987, The Nature ofDisease, Routledge & Kegan Paul, London and New
[168] Seedhouse, D,: 1986, Health: Foundations of Achievement, John Wiley & Sons,
[169] Sundstrom, P.: 1987, Icons of Disease, Linkoping Studies in Arts and Science,
Linkoping University, Linkoping.
[170] Telfer, E.: 1980, Happiness, MacMillan Press, London.


ability 35, 42-57, 205-206 Bowman,I.A. 191n

first order 49-53, 148 Braybrooke, D. 62
second order 49-53, 148
accomplishment 40, 44, 148 Cahill, G.F. 155, 192n
action 37-46 Canguilhem, G. XI,35
basic 37, 43, 148 Cannon, W. 178n
generated 38-40 capacity 35
opportunity for 41,42 causal generation 38
stratification of 37-41 causal relations 39, 54
theory 38 complexio 15
action-chain 40 clinical medicine 2,3, 129-131
action-sequence 41 common cold 163, 165, 170
activity 41, 44-45, 148 concepts
adaptability 68-69, 99 basic 13-14
Aggernaes, A. 181n descriptive 6
Agich, G. 179n evaluative 6, 21
ambitions 71, 74-75 family 9,55
American Psychiatric Association 134 humanistic 4,23, 145-146
analytic perspective (of health) XIII, ideal 175n-176n
12-14 praxis 175n-177n
Aretaeus of Cappadocia 154-155 technical 8, 55
Aristotelian philosophy 6 Constitution of the World Health
Aristotle 1,6, 15, 81, 175n Organization 186n
Armstrong, D. 179n conventional generation 38-39
Avicenna 155 coordination 45
Ayers, M. 180n counterproductivity
accidental 93
balance 15 internal 94
Banting, F. 155, 192n Cullen, W. 191n-192n
basic competence 51 Culver, C.M. XVI, 105, 186n-187n,
Bayer, R. 190n 190n
Beeson, P.B. 162
Bell, A.P. 136, 137-138 Danto, A. 180n
Bernard, C. 8, 155, 192n defect 4, 108-109
Bieber, I. 133 definitions
biochemical abnormality 166-167 lexical 8
biological set-up 43 multiplicity of 8
Boerhaave, H. 8, 179n stipulative 8, 9
Boorse, C. XIII, 16, 19-23,25, 31, 69, Descartes, R. 1
143, 177n, 178n, 179n, 181n, deviant behaviour 56
190n-192n, 205


diabetes mellitus 154-156, 164, 165, goals

166, 170, 173, 192n factual 17
Diagnostic Statistical Manual of Mental ideal 17
Disorders (DSM-III) 134 intended 72
disability 4, 32, 35, 46-47 (see also set by the agent 17, 55, 65-67,
ability) 70-73
disease 4,19-21,105-111,151-173 trivial 95
as a process 22, 161-162 ultimate 25, 30
the ontological conception of vital 53-57, 90-96
151-154 Goldman, A. 38, 180n
the physiological conception of grief 114-116
151-154 gross functions 27-28
disease-instance 105-106
disease-type 105-106 handicap 4, 127-128, 189n, 190n
drives (physiological and psychological) happiness 78-90, 182n
60-61 as an emotion 82-86
Dorner, G. 134, 135, 190n as a mood 82, 85-86
dimensions of 86-88
emotions 82-83 minimal 79, 89-90
moral justification of 115-116 molar notion of 85
empirical theories vs definitions 9 real 79, 95-96, 121
endocrinological hypothesis of health
homosexuality 134 absolute state of 67, 97-98
Engel, G.L. 190n analytic theories of 13-14
Engelhardt, H.T. Jr. XI, 189n, 190n biostatistical theory of XIII, 15-33
excellence 1,4, 59, 97, 145-146 circle of 63-65
external relations 54 dimensional aspect of 67-69, 97-99
extraordinary opportunities 48 discourses on 33, 55
event 159-161 and environment 5, 29-33, 49, 147
equilibrium theory 65-67
Feinstein, A. 21, 178n, 192n Galenic conception of 15-16
forensic psychiatry 3 holistic theories of 13-14
Freedman, DX 178n mental 5,20,99-100, 146-147
Freud, S. 132-133, 190n need theory of 63-65
Fulford, K.W.M. 205, 210-211, 214n optimal 98
function 23-29 and society 119-128
subnormal 24 somatic 5, 99-100, 146-147
supernormal 24 subject goal theory of 65-77
functional technical concept of 8, 55
ability 24-25 welfare theory of 76-80, 148
performance 26 WHO-definition of 98, 186n
with animals and plants 139-143
Galen 8, 9, 15, 35 with infants 104
Gay Liberation Front 134, 191n health care 2, 3
Gert, B. XVI, 105, 186n-187n, 190n health-care system 123-125
goals 17, 18, 58-59, 65-67 health-concepts 4, 14, 22-23, 140,
counterproductive 71, 75-76, 93-94 145-149

health evaluations 79-80, 121-126 McGill, V.I. 81, 182n

Hippocrates 8 medicine
holistic perspective (of health) XIII, art of 12
12-14 science of 12, 130-131, 147
homeostasis 178n Mill, I.S. 57
disturbance of 166-167 Minkowski, O. 155, 192n
Hopps, H.C. 178n Moore, M.S. 186n
homosexuality 131-139 morality 4, 56, 71
Mourn, T. 181n
illegal behaviour 2, 57 myocardial infarction 164, 165, 170
illness 4,21,31-32,67,69,80,109-111
long-term 52 natural state (of the body) 15
medical concept of 110-11 I needs 57-65, 148
illnesses as clusters of disabilities basic 57
109-110 human 55, 57, 59-63
immoral behaviour 2, 57 physiological 60-63
impairment 4, 108-109 vital 59-63
infection nominalism
the paradigm case of 30 strong version of 7
injury 4, 108-109 weak version of 7
intentional behaviour 37 nominalist approach to definitions 6-7
internal relations 54 normal state 15
International Classification of Diseases, functioning 19
Injuries and Causes of Death norms 4
(lCD) 130, 158, 189n biological 16
International Classification of
Impairments, Disabilities and old age 112-113
Handicaps (lCIDH) 108,
127-128, 189n, 190n pain 32,36,130, 179n
Papaspyros, N.S. 192n
Iensen, V.I. 175n-177n paralysis agitans (Parkinson's disease)
Kant, I. I Parsons, T. 35, 189n
Kenny, A. 81, 180n philosophical anthropology 1
King, L. 16 philosophy of science 2
pneumococcal pneumonia 163, 165, 170
lactase deficiency 107 practical possibility 42, 44-45, 148
language game 119 pregnancy 113-114
Liss, P.-E. 181n involuntary 114
Locke,1. 1 voluntary 114
process 159-162
malady 105-112, 149 Public Health Act of Sweden 3, 175n
Malmgren, H. 192n Porn, I. XI, 65-67, 69-70, 72, 80, 108,
Margolis, 1. 190n 178n, 181n, 187n, 189n, 205,
Marx, K. 57, 60, 181n 207,211,212
Maslow, A. 60-62, 181n
McDermott, W. 162

Rado, S. 133 symptom 152-153, 156, 166-168, 192n

realism syndrome 109, 166, 168
as a psychological notion 99
as a semantic notion 6 Tatarkiewicz, W. 81, 182n, 183n
realist approach to definitions 6 Temkin, O. 18
Redlich, F.e. 178n training 50
rehabilitation 131 Tranj1Sy, K.E. 62-63
reproduction 18,33, 113, 137-139, 190n trying 42, 180n
Rescher, N. 183n-184n tuberculosis 157, 163, 170
Reznek, L. 189n
Ross, A. 176n, 177n unifying principle 162
Rousseau, J.J. 57, 60 unrealistic person 73-74
Ruse, M. 190n-191n
values 72
Saghir, M.T. 135-136 Veenhoven, R. 183n
Sauvages de la Croix, F.B. de 192n Virchow, R. 153
Scadding, J.G. 177n vital goals 53-57, 90-96
schizophrenia 125-126, 190n basic 91
screening 3 conflicts between 102-104
Seedhouse, D. 205,211-212, 214n vitality 4, 75
Shwayder, D. 179n
sickness benefit 3 wants 72, 182n
sign (of disease) 152-153, 156, 192n Weinberg, M.S. 136, 137-138
Spicker, S.F. 181n well-being 1, 35
Spitzer, R.L. 134-135 Whitbeck, e. XI, 65-70, 72, 108, 181n,
Springborg, P. 181n 184n-185n, 187n-188n, 189n, 191n
standard circumstances 48, 51, 64, 213 will
strength 4 mad 101-102
state (as an episode concept) 159-160 morally defective 100
subgoal 17-18,25,54 unrealistic 100-10 I
suffering 35-36 weak 102
Sundstrom, P. 189n Wittgenstein, L. 9, 119, 177n, 179n
survival 18-19,32-33,62, 190n Wright, G.H. von 81, 181n, 182n
Sydenham, T. 152-153, 192n


The philosophical literature on the concept of health is now rapidly

growing. It is significant that the main bulk of this literature has taken a
stance against the traditional medical model of health, or certain
explications of it such as Christopher Boorse's [13]. It will be my main
concern in this postscript to present and assess some of the arguments and
theories proposed, both from a pragmatic and a logico-semantic point of
view. I shall also try to compare the theories with the one proposed in the
present book.
The core element in the theories of health to be discussed is the idea that
health consists, either wholly or partly, in the person's having the potential
or ability to perform a certain set of actions, or reach or maintain a certain
set of goals. If the person can perform all the actions included in this set (or
realize all the goals included), then, according to these theories, he or she is
completely healthy. If there is some action in the set that the person cannot
perform (or a goal that the person cannot realize) then he or she is to some
degree unhealthy.
Such a specification can be made in various ways and with greater or
lesser precision. This is indeed the case in the theories to be discussed in the
following, viz. the ones presented by Ingmar Porn, David Seedhouse, and
William Fulford. The works which I shall mainly draw upon are 'Health
and Adaptedness' by Porn [166], Health: The Foundations of Achievement
[168] by Seedhouse, and Moral Theory and Medical Practice [155] by


As a background strategy let me rehearse some fundamental elements in the

logic of ability. Ability is, as I see it, not an absolute concept. It is
uninformative and therefore pointless to say of a person that he or she is in
general able, or conversely, that he or she is in general disabled. The ability
has to be specified. First, one has to identify a particular agent A. Second,
one has to specify a project or goal of A's, i.e., something thatA wants to be

able to attain or maintain; in many cases this goal can be identified with a
simple action, e.g., A IS goal is just to be able to perform the action of
walking to the supermarket. And third, one has to identify the
circumstances in which A is assumed to be able to attain or maintain this
goal or perform this action.

Thus ability is a three-place relation with the following terms:

(1) the agent involved

(2) the goal ofthis agent

(3) the circumstances in which the agent acts.

Let me now illustrate this. Brown is able to do his own cooking when he is
on his own and is undisturbed. Here in this concrete case of an ability
relation, Brown is of course the person, the goal is the accomplishment of
cooking, and the circumstances include Brown's being on his own and
being undisturbed. This is clearly not a complete description of the required
circumstances, but these are the ones that are made explicit in this case.
It is now very easy to see how a case of disability in general shall be
construed by analogy with this: Brown is unable to do his own cooking
when he is disturbed. We can here detect the same terms in the relation: the
person, the goal, and the circumstances. The difference between ability and
disability can be expressed by the following analysis: When A is able to
reach G in C, then he or she reaches G in C if he or she tries; when A is
unable to reach G in C, then A does not reach G in C if he or she tries.
We can now see how we can use this schema for analyzing all kinds of
ability and disability relations, including the cases where disability
constitutes non-health or illness. We can, for instance, see how we can
express various cases of ability and disability. Smith, for instance, may be
unable to do his cooking in all kinds of circumstances. He then has an
extreme disability with respect to this action. Or he may be unable to do his
cooking in a specified set of circumstances. He is then disabled with respect
to this action in relation to this specified set of circumstances.


From the above it is clear that the success of an action is dependent on three
types of things: the agent with his or her biology and psychology, the nature
of the goal to be attained or maintained, and the nature of the circumstances

surrounding the action. A person may be prevented from success by the

manipulation of all three kinds of factors, and he or she may be helped to
success by the manipulation of all these factors.
Consider now the idea of simply reducing the philosophy of health to
being a subcategory of the philosophy of action. What would be the
consequences from the point of view of the logic of ability? Is health simply
a generalized three-place relation of the following kind?
A is completely healthy, if and only if A is able to realize all his or her
goals, given all kinds of circumstances.
In its actual content this specification is of course completely absurd and
has not been proposed in the literature. What I have here suggested is
perhaps a definition of the health of God but it is not the health of human
However, by relaxing just one of the terms we get a proposition which
has indeed been suggested:
A is completely healthy, if and only if A is able to realize all his or her
goals given the circumstances in which he or she happens to be placed.
This is the concept of health where health is identified with
adaptedness. This concept has been elaborated by Porn [166] in detail and
it has, if I understand him rightly, been seriously proposed by Seedhouse in
his health-theoretic work. (See [168], p. 61.)
According to this idea then, being healthy is being adapted to the
prevailing situation. The person is healthy now if it is practically possible
for him or her here and now to attain or maintain the goals that he or she
has. This possibility may be permanent but it may also be changed in three
types of ways. One way is a change in the biology or psychology of the
person - for instance, the person may contract a disease which makes him
or her unable to do what he or she was previously able to do. Another way
is a change in the goals of the person: the person may develop new goals
which are not realizable given the prevailing conditions. A third way is that
the circumstances may change - for instance may become much harsher so
that the agent is no longer able to realize all his or her goals.
Thus, if health is identified with the three-place relation of adaptedness,
then an important consequence is that health can be affected in these three
ways. It can be manipulated by traditional health care, i.e., by the treatment
of diseases and injuries or by other manipulations of a person's body or
mind. But it can also be changed positively or negatively by the change of
goals and the change of circumstances. Porn very clearly acknowledges this
when he coins the terms "goal care" and "environment care" as subspecies
of health care ([166], p.302). This idea is not made entirely explicit in

Seedhouse's book, but I think he would be prepared to agree with this



Now is there anything wrong with the idea of health as adaptedness? Let me
first focus on the circumstance term in the relation of being adapted. What
is wrong with including the actual set of circumstances in the notion of
health itself?
Is not, for instance, "environment care" very often considered to be an
important element in health work, in what is normally called health
promotion? Does not health promotion involve the provision of
opportunities for various kinds of healthy actions? Does it not involve
measures to purify the air, earth, and water surrounding us? Why should it
then be wrong to include the given set of circumstances surrounding a
person in the characterization of the health of that person?
I shall not dispute the importance of a certain kind of environment care.
I have argued at length elsewhere for the necessity of a diversified effort for
health [162]. It is obvious that environment care has a central place in the
work for health.
The standard kind of environment care, however, is logically different
from the care proposed in the definition of health noticed above. In order to see
this, we must make a distinction between the environment as a cause of
human health or illness, and the environment as a platform for action.
Let me explain. The environment can of course cause diseases in the
human body or mind. That is what toxins of various kinds often do and
what polluted air or water often does. The environment causes a change in
the body of a particular person and this, in turn, reduces the ability of the
person to perform a certain set of actions. We then say that the environment
affects health by causing changes within the person, for instance, by
causing diseases or injuries. Conversely, a person can be cured of a disease
and thereby promoted to a healthier state by purification of the air and
water surrounding him or her.
But this is not what is meant in the definition noticed above. Here the
environment does not play the role of a cause. It plays rather the role of a
platform for action. Let me illustrate. Assume that a man tries to enter his
work-place. He is brusquely prevented from doing so, however, by a number
of pickets, who claim that he is not permitted to enter. He then has no
platform for doing his work. He cannot do his work for external reasons.
This is obviously very different from the case where he cannot do his work

because he has contracted a disease, which mayor may not have external
The advocate of the notion of health-as-adaptedness-to-a-situation is
forced to hold that the strike case is also a case of non-health. The man who
has lost his platform for work because of a strike is also unhealthy
according to this theory. It is so, at least given the precondition that going
to work is a goal for the man in question. (We can of course imagine cases
where this is not included among a person's goals.)
Is this consequence counterintuitive? Is this consequence a sign that the
notion of health-as-adapted ness is a defective notion of health?
I think that it is obvious that it is very far from the notion of health as it
is used in ordinary language. I think that most of us would say that being
physically prevented is not tantamount to being unhealthy. In observing this
I do not wish to ally myself with the idea that philosophy should terminate
with the subtle analysis of the concepts of ordinary language. I think,
however, that every substantial deviation from ordinary usage should be
founded on sound arguments of some scientific or pragmatic kind.
I believe that there are good scientific and pragmatic arguments for
rejecting this notion. Perhaps the most important argument is that we need
some conceptual variety. It is important to be able to describe the general
idea of adaptedness which is a notion of freedom. When one is adapted one
is free to do the things which are required by one's wants. But I think that it
is also important to single out such an aspect of this freedom as stems from
the individual him or herself and distingush it from the external platform.
Health is by tradition semantically viewed as particularly linked to the
human body or mind. It can indeed be practical to keep this connection. If
we do, it will still be possible to say: this person is completely healthy but
he or she is at this particular moment prevented from doing certain things.
Let us also notice the converse kind of consequence. By providing
opportunities for a person ordinarily called "unhealthy" or "disabled", like
giving a wheelchair to a lame person, one would, according to the notion of
health-as-adaptedness, automatically be improving his or her health. This
cannot be a happy consequence. By calling a lame person who is helped by
a wheelchair a healthy person, we blur the fact that this person has some
basic characteristics which we ought to pay attention to. (From this
semantic point one should not draw the conclusion that it is not up to the
institutions of health care to provide means of assistance. On the contrary
there may be many pragmatic reasons for these institutions to help people
compensate for losses in health. The notion of compensation is not,
however, identical with the notion of restitution or rehabilitation.)

As I have argued here (as well as earlier in the present book), I therefore
suggest a restriction on the circumstance term when I try to describe the
ability that constitutes health and the disability that constitutes non-health
or illness. When one is ill one is not disabled given any set of
circumstances; one is disabled given a rather specific set of circumstances. I
shall shortly return to this issue.
In taking this stance I think that I have some substantial support from
William Fulford, who in his Moral Theory and Medical Practice [155]
provides a very detailed analysis of health and related concepts. One of his
most central statements is that "illness is the failure of action in the absence
of preventive factors or opposition" (p.109). Disablement or failure of
action, given external prevention or opposition, is not then counted as
illness by Fulford.


SO much for circumstances. What about goals? Which are the actions that a
healthy person should be able to perform given a specified set of
Let me here start by offering some comments on Fulford's theory. I wish
to do this because he expresses himself in other terms than the three other
scholars. Fulford does not explicitly talk in terms of goals. He, instead,
consistently uses action-language (or doing-language). Fulford coins the
term "ordinary" doing for the set of doings which is relevant for the
analysis of health.
The concept of "ordinary" doing includes certain intentional actions but
ranges over a broader spectrum from simple functioning to fully conscious
intentional actions. The basic idea seems to be that 'brdinary" doing should
fall somewhere in the middle of this range.
Fulford, however, abstains from providing a systematic analysis of this
notion of ordinary doing. Thereby he also leaves us without a set of explicit
criteria for delineating the set of doings and actions that a healthy person
should be able to perform. Some explication of what he means by ordinary
doing is given in the following:

...the patients who are ill are unable to do everyday things that people ordinarily just get on and
do, moving their arms and legs, remembering... things, finding their ways about familiar places and
so on ([155], p. 149).

What Fulford provides here is of course just an indication. He presents a

rudimentary list of examples, which could be expanded. But one wonders in

what direction it should be expanded. What could be included in the set of

"ordimiry" doings? Ordinary - in relation to what? Is "ordinary" to be
taken in relation to a specific cultural judgment, i.e., what happens to be
called "ordinary" in a particular cultural context? Or is the comparison to
be made with some quantitative measurement, so that ordinary is the same
as frequent? Or, should the judgment be related to the subject's own habits,
i.e., with what he or she ordinarily tends to dO?l.
There are at least two ways of interpreting Fulford's reluctance to
elaborate on the notion of ordinary doing. One of course is that he leaves
this out as being a further project to work on later. But another and
theoretically more interesting line of interpretation is the following. Fulford
may believe that a semantic analysis of the concept of illness does not give
us anything more precise than inability to perform 'brdinary" doing. The
rest is open to stipulation.
However, even if this interpretation were true there really is a need for
further philosophical explication. Some of this may not be purely
philosophical work (in the analytic sense of philosophy). It may better be
called ideological work. I think, however, that it is important for
philosophers to take part in such ideological work.
Porn, Seedhouse and I (in particular the latter two) are preoccupied with
the goal term of the health relation. Let me, as a starting point, consider
Seedhouse's definition of health:

A person's optimum state of health is equivalent to the state of the set of conditions which fulfil
or enable a person to work to fulfil his or her realistic chosen and biological potentials. Some of
these conditions are of the highest importance for all people. Others are variables dependent upon
individual abilities and circumstances ([168], p. 61.).

There are of course a number of details to discuss. But let me for my

particular purposes simplify matters and focus on the goal-term. Let me
therefore also modify the terminology and say the following:
A person is healthy, if and only if all conditions are fulfilled so that he
or she can fulfil his or her realistic chosen goals. 2 .
The defining criterion then is "realistic chosen goals". This is an
important specification of the general locution of something being
someone's goal. By simply referring to someone's goals one has not yet
defined who has set the goals or anything about the relation between goals
and the subject's own wants. Seedhouse's specification makes the issue
much clearer in this respect.
Let me now comment on this, first by praising one aspect of Seedhouse's
characterization, then by criticizing another aspect of it.

The addition of the word "realistic" is essential for the plausibility of the
notion. The individual may have very unrealistic goals - wanting to
become a Nobel laureate or to win an Olympic gold medal in the high jump.
Assume that he or she is very far from having the resources necessary for
attaining such goals. According to an unqualified theory of health this
individual would then be in a state of non-health. I think this is an absurd
conclusion. So, obviously, does Seedhouse. He qualifies by adding the
clause of realistic chosen goals. (I think that there are also other
qualifications to be made, having to do with distinctions between different
species of will [see this volume, pp. 72-73]. But I shall leave them aside
But there is a reverse problem which is not explicitly taken care of in
Seedhouse's theory. Nor has it been dealt with by Porn. This is the problem
of the person who does not choose anything at all, the lazybones or the
intellectually very weak person or the person with a very defective will.
If the subject has not chosen any goals at all, or really does not want to
have anything, then he or she can come out as healthy without having any
resources whatsoever. He or she can even have grave diseases or be highly
disabled, at least according to conventional ideas about disablement. But
still the subject can fulfil all his or her chosen goals, since he or she does
not have any such goals or at least only limited ones. This is indeed a
counterintuitive consequence.
If we decide on a subject-oriented notion of health, like Seedhouse's,
Porn's, and my own, we must find reasonable ways to avoid both kinds of
difficulties. We must define a notion of a person's goals which is not simply
identical with, but still related to, what a person wants to do or chooses to
do. My own notion of a vital goal of human beings involves an attempt to
play this theoretical role.



Let me now briefly rehearse my definition of health in the light of the above
A is completely healthy, if and only if A is in a bodily and mental state
which is such that A is able to realize all his or her vital goals, given
accepted circumstances.
I shall briefly elaborate on this definition. Consider first the
circumstance clause.
I criticized the characterisation of health as adaptedness to any situation.
The reason was that we cannot in such a case talk about a person's being

healthy in the presence of obstruction or opposition. My definition delimits

the class of circumstances. We do not judge a person unhealthy or disabled
unless the situation is a reasonable situation, in particular a situation
without obstruction.
What should be counted as a piece of obstruction or opposition, or in
general what should be counted as elements of an acceptable situation, is
not a neutral or objective affair. This involves an ideological judgment,
either made normatively by the person who ascribes a state of health or ill-
health to someone, or made descriptively with reference to a particular set
of societal norms. As I see it, there is a primary use of the locution "A is
disabled" and a secondary one. According to the primary use, it is indeed
true that the speaker him or herself defines the set of accepted
circumstances. The speaker claims that, given what he or she finds to be a
reasonable set of circumstances, A is disabled. Thus the expression "A is
disabled" has, in this case, a clearly evaluative element. Moreover, it is a
use that may fail to be communicative. Unless the speaker informs the
listener of the accepted set of circumstances, the listener may fail to
understand the message.
According to the secondary use, the speaker implicitly refers to some
commonly accepted background - what might be called standard
circumstances. The speaker does not here take a stand on the normative
issue concerning what is a reasonable background situation. He or she,
instead, refers to what other people, the politicians, the experts or most
other laymen, have accepted as a reasonable background situation for the
ascription of ability and disability. (Some implications of the notion of
standard circumstances have been analyzed in this volume, pp. 47-49.)
The introduction of the notion of accepted circumstances has the
following important consequence. In order to make ascriptions of health,
ill-health, ability, and disability unambiguous in the context of health care
as well as in the context of medical and caring sciences, we shall have to
make explicit (perhaps in a stipulative way) the boundaries of the accepted
situation or the accepted circumstances.
What is a vital goal? As has been argued at length in this volume (pp.
76-80), a person's vital goal is a state of affairs that is necessary for the
realization of this person's state of minimal long-term happiness. As a
consequence of this interpretation, many of the things that human beings
hope to realize or maintain belong to their vital goals. More precisely, all
states which have a high priority along a person's scale of preferences
belong to his or her vital goals. However, certain things that people happen
to want are not vital goals. First, we have trivial wants. People may casually
want something, but if they don't get it, it does not matter much. Second,

people sometimes have counterproductive wants. They may want to get

drunk, but getting drunk is not a vital goal. Instead of contributing to long-
tenn happiness, being drunk contributes in the long run to suffering and
thereby unhappiness. Third, we may have irrational wants, i.e., wants
which are in conflict with other more important wants. As soon as we
realize this conflict we also realize that the only candidates for vital goals
are the more important wants.
On the other hand, some things that we do not want may be contained in
our set of vital goals. The completely apathetic or lazy person who does not
have any conscious goals whatsoever for him- or herself will soon realize
that this fact creates suffering for him or her. This is particularly salient, for
instance, if the person does not even seek to get some food or some shelter
against a harsh climate. It must certainly belong to this person's long-term,
minimal happiness to have these basic matters organized. Therefore these
basic goals are among every person's vital goals.
With elaborations along these lines (some of them presented earlier in
this volume (pp.90-104» we may be able to establish a reasonable theory of
health based on the idea of the ability to realize one's vital goals.

1 Fulford at some places refers to the individual's preferences. But in the quotation above he
explicitly talks about what people ordinarily just get on and do.
2 As is indicated in the quotation, Seedbouse also mentions biological goals, but he has informed
me in conversation (April 1992) that in the case of conflict between biological and chosen goals the
chosen goals override the biological ones.
Philosophy and Medicine

1. H. Tristram Engelhardt, Jr. and S.P. Spicker (eds.): Evaluation and Explanation
in the Biomedical Sciences. 1975 ISBN 90-277-0553-4
2. S.F. Spicker and H. Tristram Engelhardt, Jr. (eds.): Philosophical Dimensions
o/the Neuro-Medical Sciences. 1976 ISBN 90-277-0672-7
3. S.P. Spicker and H. Tristram Engelhardt, Jr. (eds.): Philosophical Medical
Ethics: Its Nature and Significance. 1977 ISBN 90-277-0772-3
4. H. Tristram Engelhardt, Jr. and S.F. Spicker (eds.): Mental Health: Philosophi-
cal Perspectives. 1978 ISBN 90-277-0828-2
5. B.A. Brody and H. Tristram Engelhardt, Jr. (eds.): Mental Illness. Law and
Public Policy. 1980 ISBN 90-277-1057-0
6. H. Tristram Engelhardt, Jr., S.P. Spicker and B. Towers (eds.): Clinical
Judgment: A Critical Appraisal. 1979 ISBN 90-277-0952-1
7. S.F. Spicker (ed.): Organism, Medicine. and Metaphysics. Essays in Honor of
Hans Jonas on His 75th Birthday. 1978 ISBN 90-277-0823-1
8. E.E. Shelp (ed.): Justice and Health Care. 1981
ISBN 90-277-1207-7; Pb 90-277-1251-4
9. S.P. Spicker, J.M. Healey, Jr. and H. Tristram Engelhardt, Jr. (eds.): The Law-
Medicine Relation: A Philosophical Exploration. 1981 ISBN 90-277-1217-4
to. W.B. Bondeson, H. Tristram Engelhardt, Jr., S.F. Spicker and J.M. White, Jr.
(eds.): New Knowledge in the Biomedical Sciences. Some Moral Implications
ofIts Acquisition, Possession, and Use. 1982 ISBN 90-277-1319-7
11. E.E. Shelp (ed.): Beneficence and Health Care. 1982 ISBN 90-277-1377-4
12. G.J. Agich (ed.): Responsibility in Health Care. 1982 ISBN 90-277-1417-7
13. W.B. Bondeson, H. Tristram Engelhardt, Jr., S.F. Spicker and D.H. Winship:
Abortion and the Status o/the Fetus. 2nd printing, 1984 ISBN 90-277-1493-2
14. E.E. Shelp (ed.): The Clinical Encounter. The Moral Fabric of the Patient-
Physician Relationship. 1983 ISBN 90-277 -1593-9
15. L. Kopelman and J.C. Moskop (eds.): Ethics and Mental Retardation. 1984
ISBN 90-277-1630-7
16. L. Nordenfelt and B.I.B. Lindahl (eds.): Health, Disease, and Causal Explana-
tions in Medicine. 1984 ISBN 90-277-1660-9
17. E.E. Shelp (ed.): Virtue and Medicine. Explorations in the Character of
Medicine. 1985 ISBN 90-277-1808-3
18. P. Carrick: Medical Ethics in Antiquity. Philosophical Perspectives on Abortion
and Euthanasia. 1985 ISBN 90-277-1825-3; Pb 90-277-1915-2
19. J.C. Moskop and L. Kopelman (eds.): Ethics and Critical Care Medicine. 1985
ISBN 90-277-1820-2
20. E.E. Shelp (ed.): Theology and Bioethics. Exploring the Foundations and
Frontiers. 1985 ISBN 90-277-1857-1
21. G.J. Agich and C.E. Begley (eds.): The Price o/Health. 1986
ISBN 90-277-2285-4
22. E.E. Shelp (ed.): Sexuality and Medicine.
Vol. I: Conceptual Roots. 1987 ISBN 90-277-2290-0; Pb 90-277-2386-9
Philosophy and Medicine

23. E.E. Shelp (ed.): Sexuality and Medicine.

Vol. II: Ethieal Viewpoints in Transition. 1987
ISBN 1-55608-013-1; Pb 1-55608-016-6
24. R.C. McMillan, H. Tristram Engelhardt, Jr., and S.P. Spicker (eds.):
Euthanasia and the Newborn. Confliets Regarding Saving Lives. 1987
ISBN 90-277-2299-4; Pb 1-55608-039-5
25. S.F. Spieker, S.R. Ingman and I.R. Lawson (eds.): Ethical Dimensions of
Geriatric Care. Value Conflicts for the 21th Century. 1987
ISBN 1-55608-027-1
26. L. Nordenfelt: On the Nature of Health. An Action-Theoretic Approach. 2nd,
rev. ed. 1995 ISBN Hb 0-7923-3369-1; Pb 0-7923-3470-1
27. S.F. Spicker, W.B. Bondeson and H. Tristram Engelhardt, Jr. (eds.): The
Contraceptive Ethos. Reproductive Rights and Responsibilities. 1987
ISBN 1-55608-035-2
28. S.P. Spieker, I. Alon, A. de Vries and H. Tristram Engelhardt, Jr. (eds.): The
Use of Human Beings in Research. With Special Reference to Clinieal Trials.
1988 ISBN 1-55608-043-3
29. N.M.P. King, L.R. Churchill and A.W. Cross (eds.): The Physician as Captain
of the Ship. A Critical Reappraisal. 1988 ISBN 1-55608-044-1
30. H.-M. Sass and R.U. Massey (eds.): Health Care Systems. Moral Conflicts in
European and American Public Policy. 1988 ISBN 1-55608-045-X
31. R.M. Zaner (ed.): Death: Beyond Whole-Brain Criteria. 1988
ISBN 1-55608-053-0
32. B.A. Brody (ed.): Moral Theory and Moral Judgments in Medical Ethics. 1988
ISBN 1-55608-060-3
33. L.M. Kopelman and J.C. Moskop (eds.): Children and Health Care. Moral and
Social Issues. 1989 ISBN 1-55608-078-6
34. E.D. Pellegrino, J.P. Langan and J. Collins Harvey (eds.): Catholic Perspec-
tives on Medical Morals. Foundational Issues. 1989 ISBN 1-55608-083-2
35. B.A. Brody (ed.): Suicide and Euthanasia. Historical and Contemporary
Themes. 1989 ISBN 0-7923-0106-4
36. H.A.M.J. ten Have, G.K. Kimsma and S.F. Spieker (eds.): The Growth of
Medical Knowledge. 1990 ISBN 0-7923-0736-4
37. I. L6wy (ed.): The Polish School of Philosophy of Medicine. From Tytus
Chalubmski (1820--1889) to Ludwik Fleck (1896-1961).1990
ISBN 0-7923-0958-8
38. TJ. Bole III and W.B. Bondeson: Rights to Health Care. 1991
ISBN 0-7923-1137-X
39. M.A.G. Cutter and E.E. Shelp (eds.): Competency. A Study of Informal
Competency Determinations in Primary Care. 1991 ISBN 0-7923-1304-6
40. J.L. Peset and D. Gracia (eds.): The Ethics of Diagnosis. 1992
ISBN 0-7923-1544-8
Philosophy and Medicine

41. K.W. Wildes, SJ., F. Abel, SJ. and J.C. Harvey (eds.): Birth, Suffering, and
Death. Catholic Perspectives at the Edges of Life. 1992
ISBN 0-7923-1547-2; Pb 0-7923-2545-1
42. S.K. Toombs: The Meaning of Illness. A Phenomenological Account of the
Different Perspectives of Physician and Patient. 1992
ISBN 0-7923-1570-7; Pb 0-7923-2443-9
43. D. Leder (ed.): The Body in Medical Thought and Practice. 1992
ISBN 0-7923-1657-6
44. C. Delkeskamp-Hayes and M.A.G. Cutter (eds.): Science, Technology, and the
Art of Medicine. European-American Dialogues. 1993 ISBN 0-7923-1869-2
45. R. Baker, D. Porter and R. Porter (eds.): The Codification of Medical Morality.
Historical and Philosophical Studies of the Formalization of Western Medical
Morality in the Eighteenth and Nineteenth Centuries, Volume One: Medical
Ethics and Etiquette in the Eighteenth Century. 1993 ISBN 0-7923-1921-4
46. K. Bayertz (ed.): The Concept of Moral Consensus. The Case of Technological
Interventions in Human Reproduction. 1994 ISBN 0-7923-2615-6
47. L. Nordenfelt (ed.): Concepts and Measurement of Quality of Life in Health
Care. 1994 ISBN 0-7923-2824-8
48. R. Baker and M.A. Strosberg (eds.) with the assistance of J. Bynum:
Legislating Medical Ethics. A Study of the New York State Do-Not-Resus-
citate Law. 1995 ISBN 0-7923-2995-3