Georges Canguilhem and the Concept of Autism

Pathology at the border of the vital and the social

Masterscriptie Wijsbegeerte van een Wetenschapsgebied Berend Verhoeff

Faculteit der Geesteswetenschappen Studentnummer 0611484
Universiteit van Amsterdam berendverhoeff@gmail.com
05.06.2009 Begeleider: prof. dr. G.H. de Vries
Contents

1 Introduction 3
2 Georges Canguilhem 6
2.1 History and Philosophy of Science 8
2.1.1 The History of Scientific Reason, a History of Concepts 11
2.1.2 Truthful Discourse Governed by Critical Correction 13
2.1.3 Vitalism and the Concept of Life 15
2.2 The Normal and the Pathological 20
2.2.1 The Identity of Pathology with Physiology 21
2.2.2 Normality, Normativity, and Pathology 25
2.2.3 Vital Norms and Social Norms 30
2.2.4 Error, a New Concept in Pathology 33
2.2.5 Concluding Remarks on Canguilhem and the Normal and the Pathological 35
3 Autism Spectrum Disorders 40
3.1 Introduction to Autism Spectrum Disorders 41
3.1.1 Autistic Disorder: three domains of impairment 43
3.1.2 Autism Light: Aspergers Disorder and PDD-NOS 46
3.1.3 DSM Classification and the Demarcation Problem 48
3.2 The Autism Epidemic 50
3.2.1 Two Lines of Explanation 52
3.2.2 Two Assumptions 57
3.3 Autism as a Brain Disorder 59
4 A Canguilhemian Perspective on the Concept of Autism 64
4.1 The Shifting Concept of Autism 64
4.2 Autism: a Paradigm Example 66
4.3 New Explanations for the Autism Epidemic 68
Bibliography 74
Appendix: DSM-VI Diagnostic Criteria for Autism Spectrum Disorders 79

2
Chapter 1
Introduction

It is notoriously difficult to determine what psychiatric disorder fundamentally is and how it

can be distinguished from mental health or merely socially defined problems (see Bolton
2008). How do we know that certain psychological or behavioral conditions are pathological
and in need of medical attention and intervention? What makes certain conditions
pathological and not just deviations from provisional social norms? These are the central
questions I would like to address in this masters thesis. The work of the French philosopher
of medicine Georges Canguilhem (1904-1995), especially his main work The Normal and the
Pathological (1989), will serve as a philosophical guidance in discussing the problem how to
demarcate between health and disease. His work will subsequently be related to two
relatively recent developments in psychiatry. To illustrate these two developments and to
point out how the problem of demarcating between health and disease becomes current in
the light of these two developments, I will elaborate the psychiatric conditions that are
covered by the concept of autism, and currently classified in diagnostic manuals as Autism
Spectrum Disorders (ASDs).

The first development concerns the apparent increase in the rates of several psychiatric
disorders in approximately the last two decades. Especially in child psychiatry the rates of
ASD and attention deficit hyperactivity disorder (ADHD) have increased significantly and
some psychiatrists and epidemiologists even speak of a true epidemic of these disorders.
Other researchers and clinicians (Fombonne 2005) are more careful with drawing
conclusions about whether there is a real epidemic of ASD and ADHD going on or whether
the increase in rates of these disorders is more likely to represent changing concepts,
definitions, service availability and awareness of ASD and ADHD. The second development
is the recent dominance of a biologically orientated psychiatry. Biological psychiatry tries to
explain psychiatric disorders in biological terms, and research, therapy and the way in which
psychiatric disorders are currently conceptualized are increasingly being done within a
biomedical model. I suggest that these two important developments are a good reason to
rethink the problem of what a psychiatric disorder is and what norms and criteria are

3
 1Introduction

important in demarcating between health and disease. The problem of how to determine
whether we are dealing with a real psychiatric disorder or not can be illustrated with the
following case:

Pete, an 8 year old boy, lives with his single mother and has no brothers or sisters. His
mother was advised by Petes school teacher to take her son to consult a child psychiatrist.
Pete was having some difficulties in class; he didnt have any friends and was not interested
in playing with other children. He always sat on the same seat in class and when someone
accidentally took his seat he reacted disproportionally with physical aggressive behavior. He
manifested this behavior regularly, especially when things were not going the way he wanted
them to go. This behavior resulted in many conflicts in class between Pete and his
classmates. Pete was usually calm when he was let alone and when he was able to read in his
favorite fantasy books. Furthermore, he was extremely interested in birds. He knew more
than 30 different bird species by heart and he was able to give detailed descriptions about the
colors and sizes of these birds, which was quite exceptional for a boy of his age. After a
clinical interview with the psychiatrist including information from Petes mother and teacher,
Pete was diagnosed with autism.

An important question that arises is how we should be able to know that Pete has a real
disease that might need medical attention and perhaps intervention? Couldnt Petes
behavior be just different or socially deviant instead of an expression of a disease process
and what could be the norms or criteria that could determine this? With a growing body of
knowledge about the biological underpinnings of psychological and behavioral conditions
like those of Pete and with a striking increase in children being diagnosed with an ASD, a
critical evaluation of the concept of psychiatric disorder becomes not only desirable but
necessary for psychiatrys self-understanding.

Before discussing (in the third chapter) these current developments in psychiatry, in
particular concerning the range of autism disorders, I will elaborate Canguilhems work in
the second chapter with an emphasis on his work on the distinction between health and
disease. His work is of particular interest, because it accurately shows that the distinction
between health and disease is neither a positivist scientific fact, nor just a constructivist
artifact at the mercy of social processes. The distinction turns out to be fundamentally value

4
 1Introduction

laden without being reduced to merely arbitrary human preferences. In the fourth and final
chapter, the second and third chapter will come together and conclusions will be drawn on
how biological knowledge of psychiatric disorders and the current increase in children
diagnosed with an ASD should be interpreted in the light of the general problem of the
distinction between health and disease. In the context of Canguilhems notion of pathology,
preliminary new explanations for the rising rates of children with ASD will be given.

5
Chapter 2
Georges Canguilhem

Georges Canguilhem was born in Castelnaudary in southwestern France in 1904 and he died
in 1995. He has, in France at least, the reputation of an in many ways very important
philosopher of science. He succeeded Gaston Bachelard as a Professor of Philosophy at the
Sorbonne and he was the Director of the Institute for the History of Science and
Technology at the University of Paris. He is considered to be one of the founders of current
French epistemological thought, and of the method of historical epistemology. Furthermore,
he is the author of some of the most insightful and important philosophical works on the
life sciences. Despite the translation of his important work The normal and the Pathological
(Canguilhem 1989) and the translation of a valuable collection of his writings (Canguilhem
1994), he still remains relatively unknown to Anglo-American philosophers and historians of
science.

According to his better known student Michel Foucault, who was supervised by Canguilhem
on his doctoral thesis on the history of madness, Canguilhem had a profound influence on a
group of French philosophers in the second half of the twentieth century. Even though
Canguilhems work was limited to a particular domain in the philosophy and history of
science, his thoughts and methods reached far beyond the borders of this limited domain;
medicine and biology. Take away Canguilhem and you will no longer understand much
about Althusser, Althusserism and a whole series of discussions which have taken place
among French Marxists; you will no longer grasp what is specific to sociologists such as
Bourdieu, Castel, Passerson and what marks them so strongly within sociology; you will miss
an entire aspect of the theoretical work done by psychoanalysts, particularly by the followers
of Lacan (Foucault in Canguilhem 1989: 8). Not in the least, Foucaults own work in the
history of science was strongly influenced by his teacher and supervisor. Especially
Canguilhems history of concepts (more on this history of concepts later in this chapter)
was a model for Foucaults archeology of knowledge. Although Foucaults primary focus
was the social sciences, his Birth of the Clinic and his The Order of Things can be read as a
Canguilhemian history of concepts. Moreover, Canguilhems work was concerned with the

6
 2Georges Canguilhem

status of norms and the normal in the life sciences, an issue equally important for Foucault.
In a broader perspective, Foucault (Canguilhem 1989) distinguishes two branches of
contemporary philosophy in France and he positions Canguilhem as an important
philosopher in the branch of philosophy of knowledge, of rationality and of the concept.
This branch of philosophy could be distinguished from a philosophy of experience, of sense
and of subject. The separation of these two networks of thought took place in the first half
of the twentieth century following the introduction of transcendental phenomenology by
Husserl in France. On the one hand, one network is that of Sartre and Merleau-Ponty; and
than another is that of Cavaills, Bachelard and Canguilhem. In other words, we are dealing
with two modalities according to which phenomenology was taken up in France (ibid.: 8).

The reason for me to use Canguilhem in this masters thesis is mainly because of his
profound work on the distinction between normal and pathological conditions of human
beings. Even though Canguilhem focused more on somatic disorders and scarcely talked
about psychiatric disorders, his work is very relevant in evaluating scientific and conceptual
developments in psychiatry today. In the introduction of A vital Rationalist (Canguilhem
1994), Paul Rabinow asks the question why Canguilhem should be read today? In answering
this question, he points to the more and more dominant position of the biosciences in the
scientific and the social arena. Canguilhem himself has spent his life tracing the liniments of
a history of the concepts of the sciences of life. Let us suggest that today it is the
biosciences, with a renewed elaboration of such concepts of norms and life, death and
information, that hold center stage in the scientific and social arena; hence the renewed
relevance of George Canguilhem (ibid.: 19). The biosciences are precisely the sciences,
especially molecular genetics, the neurosciences and pharmacology, that have entered the
complex and hybrid field of psychiatry the last decade. A biological approach of life with its
core concepts of life, norms, normality, abnormality and anomaly has entered the broad
context of psychiatry and the way psychiatric problems and disorders are explained, treated
and conceptualized. Apart from the questions whether these concepts can be simply
transferred from one disciplinary context to another, what kind of changes they go through
in the process, and what kind of relationships there are between scientific research and
explanation on the one hand, and of treatment and cure on the other, these developments

7
 2Georges Canguilhem

ask for a renewed valuation of the distinction between normal and pathological psychiatric
conditions.

Before continuing Canguilhems discussion on the distinction between the normal and the
pathological and specifically the distinction and relationship between vital (biological) norms
and social norms in psychiatry, I will take a closer look at his ideas on the history and
philosophy of science. His method of epistemological history, the history of concepts, the
concept of life and his position in the vitalism versus mechanism debate will be discussed.
This broader discussion of his work will put his more specific work on different themes in
the life sciences in an intelligible context.

2.1 History and Philosophy of Science

According to Canguilhem, the history of scientific thought is a fundamental and integral part
of philosophy of science, and his epistemological position, a historical epistemology,
provided a new approach to the study of scientific knowledge. This historical approach in
the philosophy of science was a continuation of the work of his predecessor as a Professor
of Philosophy at the Sorbonne, Gaston Bachelard. Bachelard, on his turn, continued an
approach in the philosophy of science that was initiated by Lon Brunschvicg. He is
considered the most important figure (Gutting 2005) for French philosophy of science and
in particular for French neo-Kantianism. Brunschvicg combined a broad neo-Kantian
philosophical point of view with a strong accent on the importance of the history of science.
He insisted on understanding science through its historical development. Until then, history
played only a minor role in philosophical reflection. Many philosophers had engaged with
the history of their own discipline, but this history was largely seen as a succession of ideas
and thinkers in their own time, without interruption by external events. Brunschvicg
combined in his approach of knowledge a positivist and an idealist position. He rejected a
nave empiricism that saw knowledge as the result of what the mind passively received and
he rejected that knowledge arises merely from the minds reflection on itself. Truth is
expressed in mixed judgments that combine what is given in experience with intellectual
frameworks developed, through scientific investigation, over the course of human history
(ibid.: 4). Knowledge of the world was seen by Brunschvicg as the result of the minds

8
 2Georges Canguilhem

historical reflection on scientists progressively more and more successful interpretations of

experience. With this interpretation of knowledge, he rejected Kants assumption that a set
of non-historical synthetic a priori could form the foundation of all following accounts of
the world.

Contrary to Brunschvicg, Bachelard (more than 30 years before Thomas Kuhn) read the
history of science (specifically physics) as a series of epistemological breaks or
discontinuities. Furthermore, he claimed that philosophy must develop new conceptions
related to each new scientific stage. The philosophy of an age of relativity and quantum
physics has to be essentially different from a philosophy of the Newtonian era, since
Newtonian concepts are now epistemological obstacles to an adequate understanding of
nature (ibid.: 4). Bachelards epistemological position is generally rationalist in that it puts
an emphasis on the active role of the mind in the construction of knowledge and philosophy
still has an irreducible role in reflecting on scientific results and their epistemological value.
He calls his point of view an applied rationalism. First, because the categories the mind
creates are relative to the historical circumstances. Second, because Bachelard sees the
minds construction of its objects as mediated through scientific instruments, which are
theories materialized (ibid.: 4). Contrary to the by that time widespread empiricist
conception, it is theoretical scientific ideas rather than common-sense experience that
constitute objects. Also contrary to Husserls phenomenological descriptions of the
constitution of objects, Bachelard needed a phenomeno-technics to describe how
instrumental technology constituted (scientific) objects. Bachelards main philosophical
project was to analyze the historical development of truth-producing practices. For
Bachelard, the philosophy of science became the study of regional epistemologies, the
historical reflection on the elaboration of theories and concepts by practicing scientists,
physicists, chemists, pathologists, anatomists and so on. The aim was not to attack science
[or dismantle the very idea of science] but to show it in action in its specificity and plurality
(Rabinow in Canguilhem 1994: 13). As we will see later, Canguilhem is deeply influenced by
Bachelards picture of science, his philosophy of science and the role he gives history in
analyzing science and reason. Two notions that play an essential role in Bachelards picture
of science also play a very fundamental role in Canguilhems philosophy, namely that of
norm and error. These notions will be discussed at length in evaluating Canguilhems work.

9
 2Georges Canguilhem

For Canguilhem, like for Bachelard, epistemology and the development of scientific thought
should be approached in a historical way. This point of view was related to the idea that
science produces its own norm of truth at each time of its history. His intention was not to
relativize science, but to show that there is no epistemology that is not historical and that the
history of science is basically an epistemological history. His view on epistemology and on
the history of science was critical towards the more popular way of doing history of science
that time. This popular history of science was a history of discoveries, individuals and
traditions. Contrary to this way of doing history of science, Canguilhem claimed that the
history of science should not identify the science with the scientist, the scientists with their
biographies, or the sciences with their results (Canguilhem 1994). Science had a history but
it was not marked out by the linear path of a chronology or the biographical path of life
stories, chance meetings, influences and conversations. (Rose 1998: 155). His objections
against this empirical historiography (in year X scientist Y made discovery Z et cetera) were
twofold (Hertogh 1986). His first objection against this tradition of history of science was
that the historians neglected the epistemology, while an implicit epistemology could be
discovered in the way that the historical facts were presented. The current knowledge
situation acted as an endpoint while science and knowledge developed in a linear and
chronological way towards this endpoint. This type of history of science required an implicit
positivist epistemology. His second objection concerned the poor way in which the object of
this type of history was defined. What was the history of science a history of? When it
concerned for example the history of medicine, was it a history of therapies, diseases,
medical institutions, doctors or medical scientists? Canguilhem concluded that this type
history of science was contradictory. This contradiction arises because in such a history there
is no logical development that becomes visible in the course of the history of a science. The
moments of explanation are luck, chance and genius, and this makes the development of
knowledge and science a contingent matter. At the same time, the final current scientific
point of view is in a necessary teleological way related to the historical development of the
science. This combination of chance, luck, contingency and necessity in the course of
scientific history leads to an obvious paradox (Hertogh 1986).

10
 2Georges Canguilhem

Canguilhem was clear about what the object of history of science should be. Since science is
for Canguilhem a discourse verified in a delimited sector of experience and an
exploration of the norm of rationality at work (Rabinow in Canguilhem 1994: 14), the
object of history of science is historical discourse. Unlike the task of the scientist, the task of
the (historical) epistemologist is to establish the order of conceptual progress that is visible
only after the fact and of which the present notion of scientific truth is the provisional point
of culmination. (ibid.: 14). Epistemology becomes not a final list of truths, but a thorough
description of the process by which truth is structured, and the task of historians and
philosophers of science is to analyze the historicity, provisionality and plurality of these
truths. In this process, the normativity of truth becomes apparent. With Canguilhem, the
object of history of science becomes historical discourse and the object of historical
discourse is the historicity of scientific discourse, in so much as that history effectuates a
project guided by its own internal norms but traversed by accidents interrupted by crises,
that is to say by moments of judgment and truth. (ibid.: 15). This historical discourse was
for Canguilhem first of all a history of concepts and relations between concepts.

2.1.1 The History of Scientific Reason, a History of Concepts

For Canguilhem, to do history of science was to describe the history of systems of concepts
and the complex conditions under which these concepts were formed. At the same time,
with studying the transformations and relations of concepts, the ways in which particular
scientific problems were able to be posed and resolved were studied (Rose 1998). For
Canguilhem, it was concepts more than theories that were important in studying historical
scientific discourse. Concepts were more important because the formation of concepts was
the first step in being able to ask a scientific question in the first place. Competing theories
could then provide answers to the scientific question asked using the same concept. A
certain theory could be introduced only after the concepts that function in that theory were
formed. It is important to emphasize the distinction between interpretation and theory in
placing concepts above theories as the main object of historical discourse. Particularly
analytic philosophers of science have emphasized that observation does not lead to pure
uninterpreted data. Scientific data are always given already interpreted. Canguilhem would
agree on this. It has been assumed (typically in Anglo-American philosophy of science) that
the interpretation of data is a matter of reading them in terms of certain theory. This way

11
 2Georges Canguilhem

interpretation derives from theoretical explanations of the phenomena you are interpreting.
On this view, concepts used to interpret data are obtained from theories that explain
phenomena. What follows is that the meaning of a concept is given by its function in a
theory. Canguilhem does not agree with this. For Canguilhem, it is essential to separate the
concepts that interpret data from the theories that explain them. A concept provides us with
the initial understanding of a phenomenon that allows us to formulate in a scientifically
useful way the question of how to explain it. (Gutting 1989: 34)

Theory and concept are relatively independent and the distinction between a concept, a
theory and the description of phenomena is crucial for Canguilhem. Galileo for example
introduced a new manner of conceiving the movement of a falling body. But to be able to
explain this movement, he, Descartes and Newton introduced different theories. The same
concept played a role in different theories. If a concept, a theory and a description of
phenomena are not clearly separated, a deformation of history can occur amongst other
things because of a double identification, on the one hand of the description of phenomena
with the concept and on the other hand of the concept with the theory in which the concept
functions. The formation of a concept is according to Canguilhem the first step in the
process from not-knowing to knowing, but this knowing doesnt have to take the form of a
theory; a concept can emerge independent of a theoretical context. Furthermore, like I said,
a concept can function in different theories and is therefore theoretically polyvalent. The
origin of a concept could be in a different theoretical context than the theoretical context in
which the concept functions at that moment. This brings us to one of the differences
between Bachelard and Canguilhem, namely that the former emphasizes discontinuities,
while the latter emphasizes both continuities (of concepts across incompatible theories) and
discontinuities (of both concepts and theories). Canguilhems view of discontinuities is
considered more partial, complex and fragmented than Bachelards (Chimisso 2003); A
theory is woven of many strands, some of which may be quite new while others are
borrowed from older fabrics. The Copernican and Galilean revolutions did not sweep away
tradition in one fell swoop. (Canguilhem 1994: 34)

Central to Canguilhems history of science is the premise that reason itself has a weight, a
density and a history of its own. Its concepts, explanatory forms, the objects it creates and

12
 2Georges Canguilhem

manipulates, the experiments and inventions it generates have a significance which is not just
at the level of ideas and ideologies. (Rose 1998: 158) A history of science is a history of
scientific reason and a history of scientific reason is essentially a history of concepts. This
history must be written in its own terms and it doesnt have a universal form. It is not a
linear or progressive history that leads to infallible truths; it is a non-linear, heterogeneous,
contingent, regional history in which the scientific concepts and objects that emerge and
transform are inserted with history themselves. This approach surpasses the usual opposition
in the history of science between externalist and internalist approaches of science and the
opposition in the philosophy of science between realist and relativist approaches of
knowledge. The study of formations, emergence, regularities, breaks, transpositions and
disappearance of concepts and explanatory structures in science, an epistemological history,
became a main concern for both historians and philosophers of science.

Canguilhem analyzed specific concepts in the life sciences such as the concept of the reflex,
the cell, regulation, adaptation and the milieu. These concepts were, for him, the
fundamental organizing features of scientific thought and discourse in the life sciences. They
organized an abstract field of relations, capable of manipulation of recombination, of
generating hypothesis and experimentationthus allow[ed] an intelligibility which goes
beyond experience. (Rose 1998: 159) This intelligibility effects in its turn the way
phenomena are interpreted and what is there to be experienced and because of the
autonomic character of concepts, they cannot be reduced to social interests or intentions of
scientists. From an epistemological point of view, according to Canguilhem, a history of
scientific reason can be written without referring to the experience of the subject. The
epistemological priority of scientific knowledge is not experience but it is found at the level
of concepts and this level is a break away from the obvious and intuitive. This break away
from the obvious to a conceptual level is a necessary condition for Canguilhem for reason
and thought to be scientific compared to non-scientific.

2.1.2 Truthful Discourse Governed by Critical Correction

I will not extensively discuss the distinction Canguilhem makes between scientific thought,
non-scientific thought and scientific ideology, but I will give it some attention because of its
relevance for the discussion on his concept of life and the distinction he makes between the

13
 2Georges Canguilhem

normal and the pathological condition. For Canguilhem, scientific reason has a very specific
character and his strong manner of distinguishing between scientific thought and non-
scientific thought has been criticized widely especially by sociologists of science (see Rose
1998). These critics are usually focused on the idea that within scientific thought or
discourse, there are many elements that are considered non-scientific, like certain
assumptions, presuppositions, dubious political or social influences et cetera, and the strong
division between scientific and non-scientific thought will not hold. For Canguilhem,
however, the history of scientific thought is a history of truthful discourses and, like for
Bachelard, scientific discourses are veridical discourses. To illustrate this Canguilhem quotes
from Bachelards last epistemological work, Le Materialisme rationnel: Contemporary science
is based on the search for true [vritable] facts and the synthesis of truthful [vridique] laws,
and Canguilhem continues: By truthful Bachelard does not mean that scientific laws
simply tell a truth permanently inscribed in objects or intellect. Truth is simply what science
speaks. The question that follows is how we can recognize whether a statement is scientific
or not? This can be recognized by the fact that scientific truth never springs fully blown
from the head of its creator. A science is a discourse governed by critical correction.
(Canguilhem 1994: 32)

Scientific discourse, unlike non-scientific discourse, is structured around a norm of truth and
it is constantly correcting and rectifying its own work around this axis of falsity and truth.
Canguilhem speaks of progress of scientific thought in a sense of progress of the discursive
project as measured against its own internal norm. But, This progress may, moreover, meet
with accidents, be delayed or diverted by obstacles, or be interrupted by crises, that is
moments of judgments and truth. (ibid.:26) This discontinuity of scientific thought, its
historicity, comes from the essentiality of error and its correction. In this process of
correcting and rectifying the norm of truth takes new shapes and not only the contents and
the methods of science change, but the very object of science is continuously adjusted.
Nature is not given to us as a set of discrete scientific objects and phenomena. Science
constitutes its objects by inventing a method of formulating, through propositions capable
of being combined integrally, a theory controlled by a concern of proving itself wrong.
(ibid.:26) The object of scientific discourse is not a static object, around which knowledge

14
 2Georges Canguilhem

turns. It is continuously changing due to new methods, explanations and hypothesis that are
linked with new ways of speaking the truth.

For Canguilhem, both scientific knowledge and its object are inescapably historical, revisable
and provisional, but this doesnt deny their objectivity or scientificity. It is rather the claims
of access to a final irrevocable truth that separate non-scientific discourse from scientific
veridical discourse. Central to scientific discourse is the constant dynamic of the
identification and rectification of error. Error is fundamental to science and with the
rectification of error new norms of truth and of speaking the truth are established. The norm
of truth is not a fixed norm but it varies depending on a variety of factors and it is in a
complex way related to the transformation and emergence of concepts, methods and
theories. Science itself is inherently a normative activity. This normativity of scientific
discourse is connected to Canguilhems concept of life and of human nature. Human beings
are normative in a sense that they establish norms and remove old norms in favor of new
norms. The human capacity to adapt to new circumstances by establishing new norms is
fundamental in Canguilhems concept of life. This concept of life and how it relates to the
normativity of science needs further clarification.

2.1.3 Vitalism and the Concept of Life

Canguilhems concept of life is worth discussing in more detail because his notions of
pathology and health are developed from his concept of life. As we have seen, with
Canguilhem, science is characterized by the episodic reinvention of its own norms. It is not
only science, but life itself that is conceived by Canguilhem as an irreducible normative
activity. For Canguilhem, biology is a science that should define life in a way that is not
reducible to the laws of physics and chemistry. Canguilhem holds a particular vitalist
position. In general, vitalism has been criticized widely because of being unscientific and
mysterious, but there are many forms of vitalism around and Canguilhems vitalism is of a
special kind. In the following his vitalist position will be discussed, not in the least because of
the importance for his concepts of health and pathology.

The term vitalism is usually contrasted with terms such as mechanism and reductionism. It
is often associated with obsolete 18th and 19th century debates between biologists with anti-

15
 2Georges Canguilhem

scientific attitudes, superstition and intellectual rigidity. Moreover, it is commonly assumed

that vitalism opposes mechanism, that is involves a certain degree of teleological thinking
and that it necessarily refers to metaphysical principles (Greco 2005). Although these aspects
do apply to some vitalists, Benton (1974) has illustrated that these aspects are a simplification
of the different meanings associated with the term vitalism. A broad distinction between
different forms of vitalism has been proposed by Wuketits (1989). He distinguishes animist
and naturalist varieties of vitalism. The animist version of vitalism is unambiguously
metaphysical and teleological in orientation, while the naturalist version posits organic laws
that transgress the range of physical explanations (Greco 2005: 17). Canguilhems vitalism is
different from both animist and naturalist vitalism. He clearly admits that what he called
classical vitalism was completely mistaken in the claim that organisms were exceptions to
the laws of physics and chemistry. Gutting (1989: 41) quotes Canguilhem to illustrate this:
One cannot defend the originality of biological phenomena [] by setting up, within the
domain of physics and chemistry, enclaves of indeterminism, zones of dissidence, centers of
heresy. Canguilhems vitalism can be divided in two not unrelated aspects.

A first aspect of Canguilhems vitalism refers to the critical use of vitalism during the history
of biology. Different forms of vitalism played an important role in the life sciences because
they kept biologists aware of the specific features of the organisms they tried to understand.
Vitalism always tried to defend the originality of biological phenomena compared to
reductionist mechanist theories of life. This positive role of vitalism was an implicit
correction of Bachelards idea of an epistemological obstacle. For Bachelard, an
epistemological obstacle always had a negative connotation, because it blocked a path of
inquiry and therefore the obstacle needed to be removed. Canguilhem, however, allowed
that obstacles could not only block a path of inquiry, but could also paradoxically be
important sources of scientific progress. A vitalism that tried to discard the use of physics
and chemistry in explaining organisms could be seen as an obstacle to scientific progress, but
in another sense, vitalism was and is a reminder that even when physics and chemistry are
fully relevant for organisms, organisms still have specific features that rule out any simplistic
reduction of organisms to inert systems (Gutting 1989).

16
 2Georges Canguilhem

Vitalism was simple the recognition of life as an original order of phenomena and hence a
recognition of the specificity of biological knowledge [The vitalists] vital principle, their
vis vitalis, vis insita, vis nervosa were so many names that they gave to their inability to accept
either pure mechanism or the action of the soul [animism] as explanations of the phenomena
of life. (Canguilhem quoted in Gutting 1989: 42)

According to Canguilhem, vitalism was, despite of its obvious scientific inadequate and
philosophically nave theories, a significant motor force in the history of biology and it
played an important role in the constitution of the specific domain of the life sciences
distinct from the domain of physics and chemistry. (Greco 2005) In the course of the history
of the life sciences, vitalism provided a form of recurring resistance to the recurring
possibility of reduction and it protected against the temptation of premature satisfaction of
mechanist and reductionist explanations of life. This first role of Canguilhems vitalism is a
critical one and he proposed that vitalism should be considered as an imperative rather than
a method and more of an ethical system, perhaps, than a theory (Canguilhem 1994: 288).
Monica Greco (2005: 18) accurately recapitulates Canguilhems vitalist position: it is not as
an account of life that vitalism appears viable; rather, it is as a symptom of the specificity of
life that its recurrence should be understood. But what is this specificity of life that
Canguilhem underlines? This is where the second aspect of Canguilhems vitalism, that
points beyond an epistemological question to an ontological one, comes into play.

Canguilhems vitalism does not only refer to a dialectic process internal to knowledge. His
vitalism also refers to a process that links knowledge of life with its condition of possibility,
life itself. When it comes to life itself, Canguilhem refused to accept a fundamental
distinction between the order of thought - of the concept and theoretical objects of the life
sciences - and the order of the referent - the extra-discursive vital order. (Rose 1998: 156)
Rose continues by quoting Canguilhem ([1952] 1965: 13): The thought of the living must
take from the living the very idea of the living. According to Canguilhem there is no
epistemological distinction between the order of reality and the order of thought. Rose
(1998: 163) recognizes in Canguilhems work the limits of both scientific realism and social
constructionism: neither is able to describe the ways in which scientific objects and concepts
connect with specific fields of reality because each operates, in a different way, with the view

17
 2Georges Canguilhem

that there is an irreducible epistemological divide between the order of thought and the
order of reality.

To be able to understand Canguilhems (second) type of vitalism compared to classical

vitalism means to engage in a search for the meaning of the relationship between life and
science in general (Canguilhem quoted in Greco 2005: 18). His vitalism is an affirmation of
the originality of life and science must be understood within the originality of life and the
activity of the living. The originality of life is logically prior to (scientific) knowledge.
According to Canguilhem, classical vitalism makes the philosophical mistake when it
approaches the specificity and originality of life to mean that life is an exception to the laws
of physics and chemistry. In this sense, classical vitalism implicitly accepts the logical priority
of the world explained by the sciences of physics and chemistry. The originality of life
cannot be claimed for a segment of reality, but only for reality as a whole. Biology must
affirm its own imperialism (ibid: 19). It is not science that is privileged to explain what life
is, it is life itself as a vital, original force that makes science possible, and at the same time the
concepts of life created by the life sciences must be seen as an integrated part of what life is.
The way in which the life sciences, including the therapeutic ones, have elaborated concepts
of life are one aspect of the originality and specificity of life. Canguilhem claims that human
specificity lies in the fact that it has created systematic knowledge and tools to help it cope
(Rabinow in Canguilhem 1994: 17). Human beings use concepts and tools to adapt to and to
structure their environment.

Life itself (not necessarily human life) has to cope with and adapt to an unpredictable and
variable environment and an organism is fundamentally not indifferent to its environment.
On the contrary, life is that which spontaneously valorizes aspects of the environment. At an
unconscious level life reacts adversely to that which threatens its existence, growth, stability
and reproduction, and it reacts positively to that which enhances these. Victoria Magree
(2003: 301) accurately summarizes Canguilhems second part of his vitalist position: life is
that which regulates its relationship to its environment through the adaptation of norms of
living, that is, patterns of behavior that express an evaluative relation to an environment, that
judge a phenomenon to be good or bad for the organisms survival. For Canguilhem, the
specificity of life is established by this unconscious and unteleological positing of value and

18
 2Georges Canguilhem

the possibility to create new vital norms when confronted with new situations. The organism
is not only capable of creating new norms of life; it is also capable of actively changing the
environment in favor of the organism. Human beings in particular, are able to create tools
and use (scientific) knowledge in order to increase the range of potential life forms. A human
being is condemned to adapt to an environment and to act using concepts and tools that
have no preestablished affinities with his surrounding world (Rabinow in Canguilhem 1994:
19). For Canguilhem, the capacity to create new vital norms and the capacity to create
concepts and tools, both in order to cope with a variable environment, are fundamental
activities of the vital force of life.

The spontaneous valorization by organisms in interaction with their environment makes

biology and the life sciences epistemologically distinct from chemistry and physics. In
chemistry and physics (unconscious) judgments of good and bad are not relevant, but in life,
things can and may go wrong. In life there is room for mistake, disease, deficit, monstrosity
and death. This possibility of error is intrinsic to life. As Foucault (1989: 22) states in his
introduction to Canguilhems The Normal and the Pathological: in the extreme, life is what is
capable of error. According to Canguilhem this notion of error is crucial for life itself and
for human beings in particular and this notion of error plays an important role in
Canguilhems concept of pathology. This will be discussed in more detail later. Apart from
being capable of error, life is capable of creating norms. The creation of norms through
spontaneous valorization makes life a normative activity, irreducible to the laws of physics
and chemistry. The normativity of life consists in the capacity to transcend established
norms of life as the environment changes or to impose ones own norms upon the
environment. It is from this understanding of normativity that Canguilhems concept of
health is formed. Health is a way of tackling existence as one feels that one is not only
possessor or bearer but also, if necessary, creator of value, establisher of vital norms
(Canguilhem 1989: 201).

19
 2Georges Canguilhem

2.2 The Normal and the Pathological

Canguilhems doctoral thesis for his degree in medicine, published in 1943, was an Essay on
Some Problems Concerning the Normal and the Pathological. It was supplemented with
later reflections on his thesis (1966) and an earlier mentioned introduction by Foucault and it
was translated for the first time in 1978 in The Normal and the Pathological. It is considered his
main work and it is the only one of his works that achieved some recognition in the English
speaking world. In this work he focuses on the very definition of the normal and the
pathological, as the founding concepts of modern research and clinical medicine. This work
indicated a major reverse in thinking about health. Previously, medical training in France
had privileged the normal; disease or malfunction was understood as the deviation from a
fixed norm, which was taken to be a constant. Medical practice was directed toward
establishing scientifically these norms and, practice following theory, toward returning the
patient to health, reestablishing the norm from which the patient had strayed. (Rabinow in
Canguilhem 1994: 15/16) Canguilhem started an attack on the theory and practice of
normalization that was fundamental to the procedures of a positivist science and medicine.
The fundamental questions posed by Canguilhem were: Is the pathological state merely a
quantitative modification of the normal state? and Do sciences of the normal and the
pathological exist? One of the aims of his work was to provide an attack on the possibility
of a pure scientifically based study of pathology. He criticizes the assumption that it is
possible to define health (where health is the normal state) in purely physicalist, physiological
terms and that pathology (abnormality) is merely a quantitative deviation from the normal
healthy state:

The ambition to make pathology, and consequently therapeutics, completely scientific by

making them derive from a previously established physiology would make sense only if, first,
the normal could be defined in a purely objective way, as a fact, and second, all the
differences between the normal state and the pathological state could be expressed in
quantitative terms, for only quantity can take into account both homogeneity and variation.
(Canguilhem 1989: 57)

20
 2Georges Canguilhem

2.2.1 The Identity of Pathology with Physiology

In the first part of The Normal and the Pathological Canguilhem carries out an historical analysis
and philosophical critique of a theory promoted by Broussais in the nineteenth century and
later taken up by August Comte and Claude Bernard. According to this theory, Broussaiss
Principle, pathological states (diseases) are only quantitative variations of normal (healthy)
states. Pathology is no more than a quantitative overload or deficiency of some factor
necessary for health. Canguilhem (1989: 71) cites Bernard to illustrate this position:

Health and disease are not two essentially different modes as the ancient physicians believed
and some practitioners still believe. They should not be made into distinct principles, entities
which fight over the living organism and make it the theater of their contest. These are
obsolete medical ideas. In reality, between these two modes of being, there are only
differences of degree: exaggeration, disproportion, discordance of normal phenomena
constitute the diseased state. There is no case where disease would have produced new
conditions, a complete change of scene, some new and special products.

According to Broussaiss Principle there is no fundamental distinction between pathology,

the study of diseases, and physiology, the study of normal bodily functions. Pathology is no
more than a branch, a result, a complement of physiology, or rather, physiology embraces
the study of vital actions at all stages of the existence of living things (Broussais cited in
Canguilhem 1989: 56). Diabetes, according to Broussaiss Principle, becomes no more than
an excess of sugar in the urine and blood, and anemia is a quantitative lack of red blood cells.
This conception of disease approached pathological processes not as qualitatively distinct
from normal processes, but as quantitative deviations from a set of constant states of affairs.
Canguilhem links this conception of health and disease with a broader world-view. From the
Renaissance onwards, the relation between man and nature had changed. Where in ancient
times human beings were part of nature, with Francis Bacon, the unity of man and nature
was broken down and the dominant view was that human beings should control and
intervene in nature in order to manipulate nature to improve the human condition. In the
case of medicine, the preferable human condition was health and the goal was to restore the
desirable healthy state. A quantitative conception of disease made it theoretically possible to
change the pathological state back into the normal healthy state through scientifically
informed human intervention.

21
 2Georges Canguilhem

Additionally, Canguilhem positions this project of normalization in the context of a more

general political effort to normalize and standardize. Between 1759, when the word
normal appeared, and 1834 when the word normalized appeared, a normative class had
won the power to identify a beautiful example of ideological illusion the function of
social norms, whose content it determined, with the use that that class made of them (1989:
246). Canguilhem argues that there is a fundamental distinction between social norms and
vital norms (this distinction will be discussed in more detail later) and these passages
allowed for his assimilation, together with Michel Foucault, within a sociological critique of
normalizing social control (Rose 1998: 157). Normalizing processes became an extensive
object of investigation in Foucaults later studies. For Rose however (1998: 158), and I agree
with him, the importance of Canguilhems work does not lie in its rather loose reflections
on normalizing social practices. Rather, it lies in the sustained argument concerning the
essential normativity of life and health.

In approaching pathology as a quantitative deviation from a constant state, the notion of the
pathological almost began to evaporate. The pathological state was no more than a statistical
state of affairs and health consisted in the preservation of the body within physiological
constants, determined statistically as average within the population. In opposition with these
positivist ideas of health and pathology, that tried to make the practice of medicine more
scientific and objective, Canguilhem emphasized the specificity and originality of the
pathological and he contrasted this quantitative concept of disease with previous concepts of
disease. The difference between pathology (disease) and health as a quantitative difference
succeeded earlier concepts of disease that were more qualitative in nature. Canguilhem
makes a broad distinction between an ontological and a dynamic conception of disease. The
ontological conception of disease conceives a disease as an entity that attacks the body from
the outside. The source of the disease is external and it is added to the body. Canguilhem
(1994: 322) claims that the germ theory of contagious disease has certainly owed much of
its success to the fact that it embodies an ontological representation of sickness. Greek
medicine, in the tradition of Hippocrates and Galen, offered a conception of disease which
was no longer ontological, but dynamic, no longer localizationist, but totalizing. Nature,
within as well as without man, is harmony and equilibrium. The disturbance of this

22
 2Georges Canguilhem

harmony, of this equilibrium, is called disease. In this case disease is not somewhere in man,
it is everywhere in him; it is the whole man (ibid.). Following Hippocratess tradition, mans
equilibrium consists of four humors and the disturbance of these humors causes disease. It is
not simple disequilibrium that causes disease; it is an effort on part of nature to effect a new
equilibrium in man. Disease is a generalized reaction designed to bring about a cure; the
organism develops a disease in order to get well (ibid.).

According to Canguilhem, medical thought has never stopped alternating these two
conceptions of disease. Deficiency diseases and all infectious or parasitic diseases favor the
ontological theory, while endocrine disturbances and all diseases beginning with dys- support
the dynamic or functional theory (ibid: 323). Nevertheless, Canguilhem claims that these
two conceptions of disease have something in common. In both conceptions, in disease, a
polemical situation arises; either there is a battle between the organism and a foreign
substance, or there is an internal struggle between opposing forces. Disease differs from a
state of health, the pathological from the normal, as one quality differs from another, either
by the presence or absence of a definite principle, or by an alteration of the total organism
(ibid). It turned out to be difficult to maintain a qualitative distinction between the normal
and the pathological that allowed man to be able to force nature and bend it to mans
normative desires. To manage disease meant to have knowledge of the normal state in order
to bring the pathological state back to the normal state. This is how Canguilhem understands
the theoretical need to establish a scientific pathology by linking it to physiology. Following
Broussaiss Principle, pathology became an extension of physiology and pathological
phenomena were nothing than quantitative variations to corresponding physiological
phenomena. The need to reestablish continuity in order to gain more knowledge for more
effective action is such that the concept of disease would finally vanish. The conviction that
one can scientifically restore the norm is such that, in the end, it annuls the pathological
(ibid: 325).

Canguilhem criticized Broussaiss Principle on both conceptual and empirical grounds

(Gutting 1989: 46). Empirically, he argues that there are many diseases that are not produced
by an overload or deficiency of some factor necessary for health. To illustrate this he gives a
detailed description of the nature and etiology of diabetes. Furthermore, he correctly states

23
 2Georges Canguilhem

that there are conditions we would call pathological, but do not deviate from a statistical
norm (in circumstances where diseases are spread widely in a population) and that there are
physiological conditions that do deviate from a statistical norm, but are not considered
pathological. There are many examples within medicine of statistically abnormal processes or
structures (Canguilhem calls them anomalies) which do not result in suffering or impairment
of an organism. Canguilhem mentions yogis who break physiological norms through mindful
control of functions that could by no means be called pathological. Rare variations within
species, as Darwins theory of natural selection shows, can even prove to be more
advantageous and adaptive to an organisms survival than the statistical norm from which
they diverge.

Conceptually, he claims that Broussaiss Principle involves a confusion of continuity and

homogeneity that is, a failure to realize that just because one state can be derived from
another by a continuous series of quantitative changes, it does not follow that the two states
do not differ qualitatively (ibid.). It is a philosophical error to conclude that quantitative
continuity between states implies qualitative identity:

One can deny that disease is a kind of violation of the organism and consider it as an event
which the organism creates through some trick of its permanent functions, without denying
that the trick is new. An organisms behavior can be in continuity with previous behaviors
and still be another behavior. The progressiveness of an event does not exclude the
originality of an event. The fact that a pathological symptom, considered by itself, expresses
the hyperactivity of a function whose product is exactly identical with the product of the
same function in so-called normal conditions, does not mean that an organic disturbance,
conceived as another aspect of the whole of functional totality and not as a summery of
symptoms, is not a new mode of behavior for the organism relative to its environment.
(Canguilhem 1989: 87)

Canguilhem concludes that quantitative deviation from statistical norms can not be
considered necessarily pathological. For Canguilhem, the pathological state is not
quantitative, but essentially qualitatively different from the healthy state.

24
 2Georges Canguilhem

2.2.2 Normality, Normativity, and Pathology

Apart from linking historical concepts of health, disease and normality to broader
philosophical systems and world-views, Canguilhem analyses past doctrines in order to reach
his own conclusions about the concepts of health, disease, the normal and the pathological.
The historical part of his work served mainly to show the narrowness and inadequacy of
the principle of pathology (Ibid: 277). For Canguilhem, the pathological state is qualitatively
different from the healthy state because it has an essentially different value for the organism.
This dimension of value, established by the nature of the organism, cannot be reduced to a
numerical quantity. A quantitative deviation from the statistical norm can only be considered
a pathological state when it has been previously evaluated to be undesirable. It is the implicit
evaluative and normative character of health and disease that cannot be accounted for in a
positivist quantitative concept of health and disease.

In the second part of his work (1989) Canguilhem elaborates his own concepts of health and
disease. Central in his view is his claim, as we have already seen in discussing his type of
vitalism, that biological (vital) norms are established by the organism itself and that life is
inherently a normative activity. Biological norms, according to Canguilhem, cannot be
equated with average in a statistical sense. In the quantitative conception of disease, the
normal is identical with the statistical average, which is considered the norm of health. The
abnormal is a deviation from the average and hence, a deviation from the norm. For
Canguilhem, however, the norm cannot be calculated from a set of data, because it stands
for a desired situation rather than a statistical average. Norms refer to values and therefore
cannot be reduced to a concept that is determinable scientifically. Certain states and manners
of functioning have a particular value from the organisms perspective, even though other
states and manners of functioning are just as likely and possibly more frequent. An anomaly
is not abnormality. Diversity does not signify sickness (Rabinow in Canguilhem 1994: 16).
Canguilhem thinks this is a crucial difference between on the one hand physics and
chemistry, and on the other hand biology. Life spontaneously searches for or stays away
from stimuli that bear different values for its existence and its goal of the preservation of life.
There is a polarity of motions; the organism is not indifferent with respect to directions and
variations in movement. There is a fundamental difference between biological states or
processes in an organism that improve the organisms functioning in the environment and

25
 2Georges Canguilhem

those that obstruct or reduce it. This fundamental distinction cannot be made for chemical
or physical processes.

Canguilhem claims that concepts of health and disease should be understood in terms of the
norms that are derived from the values established by the nature of the organism. Yet, he
does not suppose that health (the normal state) is life in harmony with these norms and that
disease is a violation of these norms. As Gutting (1989: 47) adequately formulates: To any
state of an organism there corresponds a certain mode of living, and this mode of living
defines norms appropriate to it. Hence, any state of an organism, even a pathological one, is
governed by norms; that is, there is a standard pattern of behavior appropriate for an
organism in the state. The question that now rises is how healthy norms can be
distinguished from pathological norms? For Canguilhem, the healthy state is characterized by
the fact that the organism is able to adjust to new situations by establishing new norms.

Being healthy means being not only normal in a given situation but also normative in this
and other eventual situations. What characterizes health is the possibility of transcending the
norm, which defines the momentary normal, the possibility of tolerating infractions of the
habitual norm and instituting new norms in new situations. (Canguilhem quoted in Gutting
1989: 47)

In this quote, Canguilhem emphasizes that it is being normative that is essential in the
distinction between being in a healthy state and being in a diseased state. Canguilhem argues
for the primacy of the normative over the normal in making the distinction between health
and disease. He explains the term normative as follows:

Normative, in philosophy, means every judgment which evaluates or qualifies a fact in

relation to a norm, but this mode of judgment is essentially subordinate to that which
establishes norms. Normative, in the fullest sense of the word, is that which establishes
norms. And it is in this sense that we plan to talk about biological normativity.
(Canguilhem 1989: 126-127)

there is no fact which is normal or pathological in itself. An anomaly or mutation is not in

itself pathological. These two express other possible norms of life. If these norms are inferior
to specific norms in terms of stability, fecundity, variability of life, they will be called

26
 2Georges Canguilhem

pathological. If these norms in the same environment should turn out to be equivalent, or in
another environment superior, they will be called normal. Their normality will come from
their normativity. The pathological is not the absence of a biological norm: it is another
norm which is, comparatively speaking, pushed aside by life. (ibid: 144)

Pathological norms, as opposed to healthy norms, are characterized by a decreased capacity

to tolerate change. Healthy norms are characterized by the capacity to response creatively to
the environment, which is particularly important because the environment is unstable and
requires specific reactions at different circumstances. In Canguilhems words: health is a
margin of tolerance for the inconstancies of the environment (ibid: 197). The life of a
patient with diabetes, for example, is regulated by norms that are generated by the new
relationship between the patient and the environment. The state the patient is in is called
pathological because the patient becomes dominated by these norms and looses the capacity
to adapt to new situations. Furthermore, the patient experiences that certain behaviors and
circumstances brings a decrease in the level of suffering. The patient, on both a conscious
and unconscious level, tends to limit her behavior to these norms. New or unexpected
situations in the patients environment are not experienced as new options or possibilities in
life, but they are experienced as a danger to a precarious control of the level of suffering and
attempts are made to keep the environment at a consistent level. The process of getting
better consists in generating new norms on life which are superior to the old ones, in the
sense that the patient is able to respond more creatively and positively to new situations in
the environment. What counts as normal in Canguilhems sense is not the statistical average,
but it depends on creative ways in which organisms have adapted to their environment.

A pathological state can still be normal in the sense that it regulates and controls ways of
being (pathological norms) according to a spontaneous valorization, but a pathological state
is not normal in Canguilhems sense of normative. This normativity refers to the ability to
revise and self-transcend existing norms. Pathological norms are conservative and intolerable
of change and health is characterized by variability and flexibility; normativity;

Each disease reduces the ability to face others, uses up the initial biological assurance
without which there would not even be life. Measles is nothing, but its bronchial pneumonia
that we dread. Syphilis is so feared only after it strikes the nervous system. Diabetes is not so

27
 2Georges Canguilhem

serious if it is just glycosuria. But coma? gangrene? what will happen if surgery is necessary?
Hemophilia is really nothing as long as a traumatism does not occur. But who isnt in the
shadow of a traumatism, barring a return to intrauterine existence? If even then! (ibid: 199)

To recapitulate earlier remarks of Canguilhems concept of life; the specificity of life is its
normative activity, its ability to react creatively and originally on new situations. It is with this
concept of life, as normative activity, that Canguilhem develops the concepts of pathology
and health. Health is normativity and not normality (normal as following a strict norm or as
the statistical average). Canguilhem concluded in the words of Rabinow that:

illness ultimately is defined by the very terms that had defined health, namely stable norms,
unchanging values. Life is not stasis, a fixed set of natural laws, set in advance and the same
for all, to which one must adhere in order to survive. Rather, life is action, mobility and
pathos, the constant but only partially successful effort to resist death, to use Bichats famous
definition: Life is the collection of functions that resist death. (Rabinow in Canguilhem
1994: 16-17)

What follows from Canguilhems clarification of the concept of biological (vital) norms is
that such norms cannot be reduced to an objective concept determinable by scientific
methods (Canguilhem quoted in Gutting 1989: 48). The labels normal and healthy do not
refer to some scientific physiological results, but they refer to the meaning of the states in
question for the individual itself. The labels pathology and disease are likewise based on
nonscientific judgments grounded in the experience of the individual. The criterion to
qualify someone or some state as pathological is not a deviation from a statistical norm, but
the decrease of the individuals capacity to interact with the environment, which is
experienced as suffering and limitation. For Canguilhem, this suffering, limitation and
incapacity to adapt to new circumstances and to interact with the environment are primary in
the definition of pathology. It is the patient who calls the doctor and the apparatus of
scientific medicine is oriented towards norms (health, the avoidance of disease) that are
defined independently of and prior to objective scientific analysis (Gutting 1989: 48). The
physician can obviously tell the patient he is sick even when the patient feels perfectly
healthy, but this is only possible because medical science has discovered causal connection
between patients present healthy state and one that the patient would judge unhealthy
(ibid: 49).

28
 2Georges Canguilhem

It is first and foremost a matter of value-orientation upon which the distinction between
health and disease is founded. This value-orientation cannot be determined by objective
measurement and what follows is that medicine is not a value-free science. The value-
orientation always concerns the specific meaning and value of a phenomenon for the general
functioning of the organism in a certain context. Biological characteristics can therefore
never be judged as pathological in isolation: There is no pathological disturbance in itself:
the abnormal can be evaluated only in terms of a relationship (Canguilhem 1989: 188). The
same biological state can be restrictive and maladaptive, or beneficial and creative depending
on the limitations and possibilities within a certain environmental context. Low blood
pressure, for example, can be considered pathological or healthy depending on the height
above sea level where the person is located. Since the value of a biological feature for an
organisms functioning changes depending on the context, it is impossible to establish
particular biological features as definitively pathological. Furthermore, Canguilhem states
that the fact that a living man reacts to a lesion, infection, functional anarchy by means
of a disease expresses the fundamental fact that life is not indifferent to the conditions in
which it is possible, that life is polarity and thereby even an unconscious position of value; in
short, life is in fact a normative activity (Canguilhem 1994: 339).

Even though Canguilhem maintains that value-orientation and the nature of biological
norms arise from the individual organism, he regards them not as arbitrary subjective
preferences, but as rooted in the essential nature of the kind of organism it is. He even talks
about an innate model for the behavior of an organism (ibid: 252). The ability to
discriminate between what is good and beneficial and what is not is subjective in the sense
that it must have its origin in the individual organism, but it is at the same time a real and
essential biological feature. Although biological norms are not objective in the sense of
conclusions from neutral scientific investigation, they are nonetheless firmly rooted in the
biological reality of the organisms they regulate. They are subjective only in the sense that
they derive from the organisms lived experience of this reality (Gutting 1989: 49).

In addition, for Canguilhem, even though the abnormal (as equivalent to pathological) is
grammatically and logically subsequent to the normal, it is existentially prior (ibid.). This is

29
 2Georges Canguilhem

because In anthropological experience a norm cannot be original. Rule begins to be rule

only in making rules and this function of correction arises from infraction itself and It is
not just the exception which proves the rule as rule, it is the infraction which provides it with
the occasion to be rule by making rules. In this sense the infraction is not the origin of the
rule but the origin of regulation. It is in the nature of the normative that its beginning lies in
its infraction (Canguilhem 1989: 241-242). The concept of health could only be formed as a
difference from the experience of disease. Health, in the widely quoted definition by Ren
Lriche, is life in the silence of the organs (ibid: 243). This silence does not require a
concept (of health) of itself.

2.2.3 Vital Norms and Social Norms

In one of the in 1966 added essays in The Normal and the Pathological, in particular the essay
From the social to the vital, Canguilhem delineates a clear distinction between vital (vital
and biological norms are used as synonyms) norms and social norms. The difference
emphasized by Canguilhem between vital and social norms is above all an ontological and
theoretical difference and not, like Rose (2007: 76) suggests, a fundamental ontological and
epistemological difference. Canguilhem maintains that vital norms may have their origin in
social norms and that social norms may have their origin in vital norms. I will get back to
this matter shortly, but first I will discuss the theoretical and ontological distinction
Canguilhem draws between vital and social norms.

In both vital and social norms, a norm is a term used to make a distinction between chaos
and order. Chaos is normlessness and where there is order there are norms. Like an
organism, a society is ordered by norms. The difference, however, is that for an organism the
norms are given with its life and the environment in which it lives (Mol 1998). For a society,
in contrast, norms are actively set. Social norms only mimic organic norms;

[] in a social organization, the rules for adjusting the parts into a collective which is more
or less clear as to its own final purpose be the parts individuals, groups or enterprises with
a limited objective are external to the adjusted multiple. Rules must be represented,
learned, remembered, applied, while in a living organism the rules for adjusting the parts
among themselves are immanent, presented without being represented, acting with neither
deliberation nor calculation. [] The social order is a set of rules with which the servants or

30
 2Georges Canguilhem

beneficiaries, in any case, the leaders, must be concerned. The order of life is made of a set
of rules lived without problems. (Canguilhem 1989: 250)

Vital norms, Canguilhem suggested, arose from and manifested the normativity of life itself,
of the organism as a living being, of its adaptability to its environment. Social norms, on the
other hand, manifested only adaptation to a particular artificial order of society and its
requirements for normativity, docility, productivity, harmony, and the like (Rose 2007: 76).
Despite this apparent strict distinction, Canguilhem acknowledges that it is practically
impossible to disassemble whether a human being is adapting to its environment with vital
norms or whether it is adapting to a particular order of society following certain social
norms. The human environment is a technologically constructed environment in which
social norms are partly originated in vital norms and vice versa. The environment and the
human body itself are to some extent the product of both vital and social norms. This
becomes clear in his discussion of biological facts in cultural and social terms;

[] from the human point of view, [where] social norms interfere with biological laws so
that the human individual is the product of a union subject to all kinds of customary and
matrimonial legislative prescriptions. [] Height, the characteristic studied by Quetelet,
would be a purely biological fact only if it were studied in a set of individuals constituting a
pure line, either animal or plant. [] But in the human species height is a phenomenon
inseparable biological and social. Even if height is a function of the environment, the
product of human activity must be seen, in a sense, in the geographical environment. Man is
a geographical agent and geography is thoroughly penetrated by history in the form of
collective technologies. (Canguilhem 1989: 159)

For Canguilhem, even an apparently biological feature such as height is inseparably social
and biological. The average height in a certain population reflects in part a certain social
appreciation of a certain height. Very small people or giant people could be socially
devaluated and isolated and as a result not be able to reproduce like people with heights that
are more accepted and appreciated. Height is also a reflection of certain cultural habits like
food habits or life style habits. In populations where for example smoking cigarettes is
accepted to a greater extend, people are averagely smaller. As a result, in the human species,
statistical frequency expresses not only biological but also social normativity. A human trait
would not be normal because frequent but frequent because normal, that is, normative in

31
 2Georges Canguilhem

one given kind of life (ibid: 160). What is meant is that if, in a certain form of life,
something is socially appreciated and accepted, it will occur more often and have its effect
on the statistical average of what seems to be a biological fact such as height. Average life
span, to give another example, depends amongst other things on social norms of hygiene,
food habits, health care systems, work conditions, et cetera. Canguilhem claims that the
techniques of collective hygiene which tend to prolong human life, or the habits of
negligence which result in shortening it, depending on the value attached to life in a given
society, are in the end a value judgment expressed in the abstract number which is the
average human life span (ibid: 161).

What seems to be a vital norm can have its origin in a social norm. Vice versa, a social norm
can have its origin in a vital norm. Canguilhem, for example, argues that there would be no
health care system if there would not be a vital norm that makes organisms care for
themselves. Despite the ontological distinction, Canguilhem does not try to precisely
disentangle vital and social norms. He claims that biological and social norms are
epistemologically inextricable. For Canguilhem, life, as an undivided whole including
biology and sociality, poses values on its own environment and on the organism itself. These
values are thus established by the human being and they depend not just on the biological
make up of the individual, but also on the cultural and social environment in which the
individual lives. This also means that the creation of new norms, on the basis of certain
values established by the human being, is not merely a biological process; it is a rather
complex process in which biological, social, cultural and political aspects of life contribute.

When a human being, for example, has to adapt to a new situation by establishing new
norms of behavior, it cannot strictly be said whether these norms are purely biological (vital)
or whether there are social norms involved. At the same time, when someone either
consciously or unconsciously valorizes a certain situation or state, this valorization is not
merely individual and biological, but amongst other things the consequence of someones
culture, social environment and history. Canguilhem acknowledges that experiences,
cognitions, behavior and emotions also depend on culture and history. The same situation or
state of affairs could lead to an experience of suffering in one culture and to a more neutral
experience in another culture, depending partly on the different values that are attached to

32
 2Georges Canguilhem

that situation or state of affairs in a given culture. The fact that vital and social norms are
inextricably intermingled is important for the discussion on how to make a distinction
between normal and pathological states. This distinction between normal and pathological
cannot be made by pointing at a strict distinction between vital and social norms and
claiming that health and disease are only concerned with vital and not social norms.
Furthermore, the critique that in medicine, especially in psychiatry (amongst others from the
anti-psychiatry movement), social norms are mistaken for vital norms does not hold when
vital and social norms cannot be distinguished properly. I will come back to this point in the
concluding remarks on Canguilhem.

2.2.4 Error, a New Concept in Pathology

Canguilhems essay A New Concept in Pathology: Error, also added in 1966 to The Normal
and the Pathological, treats the possible differences in understanding health and disease
introduced by the idea of a genetic or hereditary disease. Canguilhem reacted on the
scientific developments in medicine of that time, especially the developments in genetics and
the related identification of genetic factors that caused diseases like Down Syndrome, sickle-
cell anemia and glucose-6-phosphate-dehydrogenase. According to Canguilhem, the idea of a
genetic disease was an abstractive idealization from the whole relation between organism
and environment [] To speak of genetic disease is to locate disease within an aspect of an
individual organism and not in the relation between it and an environment (Trnka 2003:
440). Canguilhem believed it is a mistake to speak of an inborn error in the genetic
structure of an organism because it presumes that this genetic error has a negative valuation
in its own right, independent of any context. This is a wrong presumption since the
organism and the environment co-constitute each other and the valuation of both the
organism and the environment depend on each other. A genetic factor or any biological
feature can never be valuated negative or positive in its own right; the valuation always
depends on the organisms situation in a certain environment, and disease, as discussed
earlier, is located in the relation between an organism and its environment.

In order to value an isolated biological feature the environment must be approached as a

static, fixed condition and not as a changing, dynamic system. To eliminate genetic errors
through genetic manipulation assumes knowledge of how the conditions will be under which

33
 2Georges Canguilhem

organisms will live in the future. The idea that it is possible to (genetically) manipulate the
organism to perfection assumes a stability of the environment and it ignores possible
(unexpected) changes in nature and it does not take into account the creativity of science and
human beings to manipulate and structure the environment in order to create situations that
are advantageous for the organism. Like I mentioned earlier in discussing Canguilhems
concept of life, Canguilhem suggested that biology is different from chemistry and physics
because in biology there is room for error; biological processes and life itself can go wrong.
It becomes clear that with this notion of error, he doesnt mean just genetic or biological
errors. Other possible errors arise from maladaptions to the environment:

Mankind makes mistakes when it places itself in the wrong place, in the wrong relationship
with the environment, in the wrong place to receive the information needed to survive, to
act, to flourish. We must move, err, adapt to survive. This condition of erring or drifting is
not merely accidental or external to life but its fundamental form. Knowledge, following this
understanding of life, is an anxious quest for the right information. That information is only
partially to be found in the genes. Why and how the genetic code is activated and functions,
and what the results are, are questions that can be adequately posed or answered only in the
context of life (Rabinow in Canguilhem 1994: 20-21).

Canguilhem warns against the biased fixation on biological and genetic errors and the
scientific dream to correct and eliminate all possible biological errors and abnormalities in
the attempt to reach an ultimate healthy state:

At the beginning of this dream we have the generous ambition to spare innocent and
impotent living beings the atrocious burden of producing errors of life. At the end there are
the gene police, clad in the geneticists science. For all that it should not be concluded that
one is obliged to respect a genetic laisser-faire, laisser-passer, but only obliged to remind
medical consciousness that to dream of absolute remedies is often to dream of remedies
which are worse than the ill (Canguilhem 1989: 280-281).

Canguilhems main worry is not the possibility to eliminate or manipulate genetic errors,
but his worry is that in the attempt to reach an ultimate healthy state, the variability of the
organism will diminish and so reduce its possibilities to adapt to new unexpected situations,
hence reduce its health. He claims that normality protected in total stability and local,

34
 2Georges Canguilhem

momentary adaptation is the ancestor of total pathology. Canguilhem calls this nightmare a
paradoxical pathology of the normal man or the disease of the normal man (ibid: 285-
286), in which disease is thought to be completely eliminated:

By disease of the normal man we must understand the disturbance which arises in the
course of time from the permanence of the normal state, from the incorruptible uniformity
of the normal, the disease which arises from the deprivation of diseases, from an existence
almost incompatible with disease (ibid: 286).

This apparent healthy, normal and permanent state, that does not take the organisms
relation to its changing and unpredictable environment into account, could be reached with
the aid of continuing contemporary scientific and technological developments. In medicine,
and in particular in psychiatry, recent developments in molecular genetics, molecular
neurosciences and psychopharmacology opened up new ways that enable new interventions
that could be used to reach the mentioned ideal healthy state. With new developments in
the molecular biosciences, biological anomalies can be identified and normalized through
genetic engineering, chemical (for instance with neurotransmitters) and electric balancing of
the brain, and other possible new interventions at a molecular or biological level. On the one
hand, biological markers that are associated with pathological states can attribute to
successful new treatments and new ways to classify and specify diseases. On the other hand,
these recent developments in the biosciences force us to ask ourselves once again how
health can be distinguished from disease and for this reason affirm a renewed relevance of
the work of Georges Canguilhem.

2.2.5 Concluding Remarks on Canguilhem and the Normal and the Pathological
Canguilhem intelligently showed that disease cannot be understood as a quantitative
deviation from a statistical norm and that the study of pathology can never be reduced to a
pure scientific matter without referring to the patients value-orientation and experience.
Canguilhem illustrated that the pathological state is qualitatively different from the healthy
state because it has an essentially different value for the organism. This dimension of value,
established by the organism, cannot be reduced to a numerical quantity. A quantitative
deviation from the statistical norm can only be considered a pathological state when it has
been previously evaluated to be undesirable and it is this implicit evaluative and normative

35
 2Georges Canguilhem

character of health and disease that cannot be accounted for in a positivist quantitative
concept of health and disease.

Even though the concepts of health and disease cannot be understood in terms of deviations
from statistical norms, norms do play a key role in the demarcation Canguilhem makes
between the normal and the pathological state. For Canguilhem, the norm stands for a
desired situation (some states or modes of functioning have a positive value from the
organisms perspective) rather than a statistical average and the norm is established by the
organism itself. Nevertheless, he doesnt claim that health is life in harmony with these
norms and that the diseased state is a violation of these norms and therefore a normless
state. He claims that any state of an organism, either healthy or pathological, is governed by
norms. Even the pathological state of an organism corresponds with a mode of living that
defines norms that are appropriate for the organism. The difference between a healthy state
and a pathological state is that the healthy state is characterized by the fact that the organism
is able to adjust to new situations by establishing new norms. It is being normative (being
able to create or establish new norms) that is essential in the distinction between being
healthy and not healthy. The pathological state, on the other hand, is characterized by a
decreased capacity to tolerate change. Pathological norms are conservative and intolerable of
change; the patient looses the capacity to adapt to a new situation that might require a new
mode of functioning. The new situation will be negatively valued and experienced as
suffering and limitation, because the organism is not able to transcend the (pathological)
norm and interact in a new creative and positively valued way with its environment. This
suffering, limitation and incapacity to adapt to new circumstances (create new norms of life)
are primary in the definition of pathology. Healthy norms are characterized by the capacity
to respond creatively to new situations and changing circumstances (Canguilhems vitalist
position) in a way that the new condition that arises will be positively valued by the
organism. This new condition will not be experienced as suffering and limitation. It is this
value-orientation upon which the distinction between health and disease is founded. Being
healthy, for Canguilhem, is not a stable, fixed state; it is a flexible and creative activity.

The value-orientation of a human being, as concluded in the discussion on vital and social
norms, is not purely a biological capacity, but depends amongst other things on social and

36
 2Georges Canguilhem

cultural values that cannot be distinguished from biological functions in an absolute way.
Whether a certain state is valued positive or negative depends not only on biological features
of the individual, but also on the social and cultural context. New norms or modes of
functioning that are created to adapt to new situations are therefore not purely biological but
can have their origin in contingent and provisional social norms. Furthermore, whether
particular states or modes of functioning of an individual are positively or negatively
valuated, whether an individual is suffering and has feelings of limitation in certain situations,
depends on the social environment that is structured according to political or ideological
ends (Magree 2003). A child with Attention Deficit Hyperactivity Disorder (ADHD), for
instance, experiences feelings of suffering and limitation when he or she is not able to adapt
to the behavioral norms in the classroom (holding attention, sitting still, studying). If the
child is able to adapt to the norms in the classroom, the symptoms of ADHD have, by
definition of the disorder, disappeared. What follows is that there are cases in which
someone who tends to conform to a certain social, cultural and political environment will be
healthy (normative) in a sense that it is capable of adapting to a certain society. In the case of
a child with ADHD, social conformity to the rules in the class enhances the health of the
child with ADHD. At the same time, not adapting to particular social norms might result in
suffering and limitation (a negative valuation) because of social isolation, discrimination and
more direct because of the inability to adapt to the social environment. Social deviance turns
out to be less healthy and this makes the distinction between pathology and social deviance
problematic. Canguilhem tries to keep health normativization (creating norms with the goal
of health) and social normativization (adapting to already existing social norms) apart and
doesnt elaborate this problem between pathology and social deviance any further.

Magree (2003: 310) claims that the difference between pathology and social deviance can be
made clear by a strict use of the term normativity: An individual who is only able to act in
accordance with social norms is only apparently healthy because he has renounced that
capacity to institute other norms that is inscribed in full normativity as the openness to being
transcended [] the norm of following what is socially prescribed is itself not open to
change. But, although this distinction between pathology and social deviance can be
defended, Magree admits that there are zones in which this distinction is vague, for example
when adjustment to social norms provides health benefits, when not being able to adapt to

37
 2Georges Canguilhem

social norms is related to suffering and limitation, or when the pathology is intrinsically
concerned with behavioral problems (in the case of the child with ADHD). Magree suggests,
and I agree, that it is usually this zone in which psychiatric problems are situated, because
psychiatric problems are often problems in (social) behavior and in social interaction (for
instance ADHD and autism). The inability to adapt to changing environments and social
demands is one of the core symptoms of autism (more on this matter later).

We can conclude that there are, especially in psychiatry, zones in which pathology and social
deviance overlap. In patients where pathology and social deviance overlap, psychiatrists and
patients are required to make decisions about the relative increase of health of different types
of social adjustments. Therapeutic interventions become not only medical acts, but also
social and political acts, because they result in adjustment to social norms. In what cases and
to what amount are interventions that come to meet to particular social demands required?
This type of question gives at least certain psychiatric disorders an irreducible social (and
political) dimension. To take this problem a step further, the question that now arises
becomes whether the social norms and demands that are present in a given society are
reasonable to adapt to. It could turn out that it would be desirable to change not the
suffering patient in aiming at adapting to social norms or demands, but to change social
values and norms active in a society in a way that the patient stops being a patient because
social norms have changed in a direction beneficial to the patient. The intervention will not
be directed at the level of the ill individual, but at the level of the society. Changing social
structures and norms could result in a decrease or increase in the prevalence of psychiatric
disorders and in this sense, health, disease and society are intrinsically connected. I will get
back to this matter in discussing autism.

For Canguilhem, medicine is a practice that is inseparable from human evaluation and his
emphasis on values in his analysis of pathology and normality leads him to view medicine as
an art, according to which clinical practice can never become a science even when its means
are effective because of science (Trnka 2003: 434). Medicine is involved with treating
diseases and to decide whether someone is ill and should be treated is essentially linked with
the experience of the subject. In addition, whether a certain treatment is effective can only
be determined by the satisfaction of the patient. Science is ill-equipped for this normative

38
 2Georges Canguilhem

endeavor. Understanding the meaning of subjectivity is the contribution of arts to medicine,

just as interpreting dissatisfaction is the diagnostic art within medicine (ibid.). I would like
to add a by Canguilhem somewhat neglected social political dimension to medicine,
especially in psychiatry where pathology and social deviance are interwoven in a complex
manner. It should not exclusively be the individuals experiences, even with taking the
specific cultures and circumstances that influence these experiences into account, as the basis
upon which medicine operates; social norms and demands in a society should equally have
health practitioners attention. Medicine turns out to be a science, an art and a social practice
at the same time.

39
Chapter 3
Autism Spectrum Disorders:
Current developments of a concept in progress

The previous chapter on Canguilhem and the distinction between health and disease showed
us that disease or pathology can never be reduced to a pure scientific matter and that what
counts as disease depends on the subjects value-orientation, experience and normative
capacity. This normative capacity, the capacity to create new norms of functioning by
interacting in a more positively evaluated manner in the environment, has especially in the
case of human beings an inseparable social aspect. Whether a certain situation will be
evaluated positively by the individual depends partly on the social and cultural norms that are
present in the society in which the individual functions. It turned out that Canguilhems
strict distinction between health normativization (creating norms with the goal of health) and
social normativization (adapting to already existing social norms) is problematic, especially in
mental health where psychiatric problems are often problems in (social) behavior and in
social interaction. In psychiatry there are zones in which health normativization and social
normativization overlap, and in these zones the adaptation to social norms involves health
benefits and the inability to adapt to social norms involves suffering and limitation.

I suggest that Autism Spectrum Disorders (ASDs) can be found in such a zone. One of the
core symptoms of autism is the inability to adapt to changing environments and social
demands and in this chapter I will use ASDs to zoom in on the way in which pathology and
social deviance can overlap. ASDs are particularly relevant and interesting, first because they
are nowadays considered to be real brain disorders with apparent biological deficits and
second because the rates of children with an ASD have increased significantly the last two
decades and some psychiatrists and epidemiologists speak of a true epidemic of ASDs. The
distinction between health and disease, made clear in the previous chapter, will be used to
interpret these two recent developments. This distinction will help to understand and
interpret the increasing rates of children with an ASD. Furthermore, it will help to evaluate
the role of biological knowledge of ASD and the significance of biological deficits in calling
a certain condition a (brain) disorder. In the following I will first give a brief introduction to

40
 3Autism Spectrum Disorders

ASDs, after which I will discuss the possible epidemic of ASDs. Then the approach of
ASDs as a brain disorder will be discussed and related to Canguilhems ideas on the
distinction between health and disease. In the fourth and final chapter of this masters thesis
I will give a Canguilhemian perspective on the concept of autism. Consequences will be
drawn on the conclusion that the concept of autism is shifting and that the vital and the
social are inextricably connected in demarcating the normal from the pathological. These
conclusions open up new perspectives on the autism epidemic and I will finish with some
new possible explanations for the rising rates of ASD.

3.1 Introduction to Autism Spectrum Disorders

In respectively 1943 and 1944 Leo Kanner in Baltimore and Hans Asperger in Vienna
described, independently of each other, a group of children with remarkable behavioral
characteristics. Kanner called the cluster of remarkable symptoms early infantile autism and
Asperger talked about autistic psychopathy. These first descriptions of this remarkable
cluster of symptoms were considered a rare and fascinating psychiatric condition while at the
present time, autism has become a widely studied and rather widespread
neurodevelopmental disorder. Both Kanner and Asperger observed in these children a
combination of problems in social interaction, in communication and they observed that
these children exhibited striking stereotypical behavioral patterns. Wing and Could (1979)
showed that this triad of characteristics was present in different degrees in different children.
A person with autism could be completely non-verbal, cognitively severely impaired and
practically unable to make contact with even the closest family members. In contrast, a
higher functioning person with autism could have an average or above average intelligence,
very specific and idiosyncratic areas of interest and strong and stable relations with family or
friends. A broad range of autism-like symptoms was recognized in different children and
adults. This broad range of symptoms was labeled the autistic spectrum and this term later
changed into autism spectrum disorders.

It was around this time (1980) that autism became an official diagnosis in the DSM-III (third
version of the Diagnostic and Statistical Manual of Mental Disorders published by the
American Psychiatric Association) and it was classified within the broader category of

41
 3Autism Spectrum Disorders

pervasive developmental disorders (PDDs). Since then, the diagnostic criteria for ASDs
have been changing and new criteria were included for Aspergers disorder and for Pervasive
Developmental Disorder Not Otherwise Specified (PDD-NOS). There has been some
terminological confusion because PDDs and ASDs are frequently used as synonyms. Within
the DSM-IV (APA 1994), the category of PDDs covers five disorders of which three of
them are considered an ASD. I will discuss and elaborate on these three ASDs (autistic
disorder, Aspergers disorder and PDD-NOS) and not on the remaining two disorders in
this category (Rett syndrome and childhood disintegrative disorder). These two disorders are
somewhat different from the ASD because they have a clear monogenetic cause, they are
extremely rare and they dont have the three characteristic domains of impairment typical for
ASD.

The diagnosis of an ASD is entirely behaviorally based; there are no diagnostic laboratory
tests or well-defined biological markers that determinate or guide the diagnostic process.
Whether or not someone has an ASD is based on behavioral and psychological assessments.
Since Kanner described the first cases of autism, the standard definition of autism has
changed many times (Blaxill 2004). Together with the definition, the disease categories, the
diagnostic criteria and the nomenclature changed. This makes it particularly difficult to
compare epidemiological findings across and within different studies. In discussing the
autism epidemic I will get back to this methodological problem. At the present time, ASDs
are diagnosed one the basis of the diagnostic criteria developed in the DSM-IV (APA 1994)
and the in most respects identical ICD-10 (WHO 1992) and clinicians use standardized
diagnostic instruments like questionnaires and interviews (see Lord and Corsello 2005) that
are based one these diagnostic criteria.

The validity of the DSM-IV and ICD-10 classifications of mental disorders in general, and
more specifically of the ASD subtypes, has been widely criticized. It is highly questionable
whether the subtypes that are distinguished in the diagnostic manuals represent true
discrete disease entities. In the case of ASDs, there are clinicians and scientist who argue that
there are real discrete disease entities that correspond more or less with the ASDs
distinguished in the diagnostic manuals. Others claim that no such discrete disease entities
exist. They assume more a continuum of traits and characteristics, and having an ASD

42
 3Autism Spectrum Disorders

means being situated on the extreme ends of the behavioral and psychological continua of
social interaction abilities, communication skills and behavioral patterns, interests and
activities. This perspective complicates how to define the difference between normal
variance and an ASD, hence between health and disease. Despite its relevance, it is not my
intention to elaborate extensively on different ways of classifying ASDs or mental disorders
in general. Furthermore it is important to notice that to diagnose a disorder according to the
DSM or ICD is not identical with the more fundamental demarcation between health and
disease (this is also one of the validity problems of the diagnostic manuals). This issue will be
discussed in section 3.1.3 before discussing the growing amount of people being diagnosed
with an ASD.

3.1.1 Autistic Disorder: three domains of impairment

The core disorder of ASDs is known as classic autism, Kanners autism or nuclear autism,
and it is the form of autism in which the behavioral features are closest to those described by
Kanner (1943). In the currently used DSM-IV this core disorder is nowadays called autistic
disorder (AD). Aspergers disorder and PDD-NOS (also known as atypical autism) can be
considered variants from this core AD. Like I mentioned earlier, patients with an AD
experience problems in three domains. These three domains of impairment are generally
called the triad of autism characteristics and I will use the DSM-IV diagnostic criteria (for
exact DSM-IV criteria of all ASD; see appendix) to further explain these three domains of
impairment.

The first domain of impairment is a qualitative impairment in reciprocal social interaction. In

the DSM-IV this domain is divided in four criteria, of which at least two need to be present
for diagnosing an AD (together with several criteria of the remaining domains). First, there
may be impairment in the use of several non-verbal behaviors (for instance facial expression,
body postures and eye-to-eye gaze) that are needed to regulate social interaction and
communication. Second, there may be a failure to develop reciprocal relationships. Young
children may have little or no interest in playing with other children or in making friends. It
could be possible that there is an interest in making friends, but then there is a lack of
understanding of the conventions of social interaction. Furthermore, there may be a lack of
spontaneous seeking to share enjoyment, interests, or achievements with other people (e.g.,

43
 3Autism Spectrum Disorders

by a lack of showing, bringing, or pointing out objects of interest) (APA 1994: 70). The
fourth and last criterion of this domain states that there may be a lack of social or emotional
reciprocity. This includes preferring solitary activities or not actively participating in social
games or play. There is often a lack of awareness of others; a person with an autistic disorder
may have no or limited idea of the needs of others or may not be aware of another persons
distress.

The second domain of impairment is a qualitative impairment in communication. This

domain is again divided in four criteria, of which at least one needs to be present. The first
criterion is that there may be a delay in, or total lack of, the development of spoken
language. It may also be the case that there is an obvious impairment in the ability to initiate
or sustain a conversation with others or a stereotyped and repetitive use of language or
idiosyncratic language (second and third criteria). The fourth, somewhat different, criterion
states that there may be a lack of varied, spontaneous make-believe play or social imitative
play appropriate to developmental level (APA 1994). In addition, the speech of a person
with an autistic disorder often has an abnormal pitch, intonation, speed or rhythm. The tone
of voice, for example, is often monotonous. Furthermore, grammatical structures are
frequently immature and include repetitive and stereotyped use of language.

The third domain of impairment is behavioral. Individuals with AD have restricted,

repetitive, and stereotyped patterns of behavior, interests, and activities. This domain is again
divided in four criteria, of which at least one needs to be present. The first criterion is that
there may be an encompassing preoccupation with one or more stereotyped and restricted
patterns of interest that is abnormal either in intensity or focus. People with autistic
disorder often have very narrow interests that have something to do with predictable
systems or exact factual knowledge (e.g. soccer statistics or collecting facts about certain
animal species) and they are typically inflexible in changing their behavior appropriate to
changing contexts. Second and third, there may be an apparently inflexible adherence to
specific, nonfunctional routines or rituals or stereotyped and repetitive motor mannerisms.
They often insist on sameness and show resistance to trivial changes. A child with an AD
may have a disproportional and sometimes aggressive reaction to small changes in their
environment for instance in the way dinner is served or in the position of objects in the

44
 3Autism Spectrum Disorders

living room. Often there is a rigid pattern or inflexible ritual when doing everyday things like
taking the exact same route to school every day or always first tie the left shoe string and
then the right. Fourth, there may be a persistent preoccupation with parts of objects
(ibid.). There may be a strong attachment to or fascination for some inanimate objects e.g.
buttons, stones, pieces of plastic or fabric et cetera. It is important to emphasize that in
order to get the diagnosis AD, it is not mandatory to have all those symptoms just
mentioned. Two criteria in the social interaction domain and one in both the communication
and the behavioral domain are required and additionally a total of at least six criteria are
required. Furthermore, there must be a delay or abnormal functioning in social interaction,
language or symbolic or imaginative play, with onset prior to the age of three years.

Many people with AD, approximately 70%, are cognitively impaired and about 40% is
severely cognitively impaired (mentally retarded). A common distinction between high-
functioning and low-functioning AD is made based on intelligence or cognitive skills. Apart
from the diagnostic criteria, people with AD may have a variety of other behavioral
symptoms including impulsivity, aggressiveness, self-injurious behaviors, hyperactivity and
especially in young children, temper tantrums. In addition, there may be abnormal responses
to sensory stimuli e.g., a high threshold for pain, oversensitivity to sounds or being touched,
exaggerated reactions to light or colors. There may also be a lack of fear in response to
real dangers, and excessive fearfulness in response to harmless objects (ibid.). It has been
noticed that older individuals with AD may be very good in tasks involving long-term
memory, for instance in remembering historical dates, recalling exact words from a
conversation years ago or knowing public transport time-tables by heart. The tasks in which
they tend to perform well or above average are generally tasks that require systemizing skills
(analyzing variables in a lawful, finite and deterministic system and deriving the underlying
set of laws that govern the actions of a system). Baron-Cohen (2002) contrasts systemizing
with empathizing and he claims that men are generally better at systemizing and women are
generally better at empathizing. Since people with AD lack empathizing qualities and do
much better in systemizing tasks, he considers autism an extreme of normal male
characteristics, hence his extreme male brain theory of autism. This theory is compatible
with the fact that rates of AD are four to five times higher in males than in females. Rates of
Aspergers disorder are even ten times higher in males than in females.

45
 3Autism Spectrum Disorders

3.1.2 Autism Light: Aspergers Disorder and PPD-NOS

The DSM-III (APA 1980) introduced, in a broad and more formal sense, the concept of a
spectrum of autism-like conditions. They introduced the broader category of pervasive
developmental disorders (PDDs) and this category comprised two diagnostic categories;
infantile autism and atypical PDD. In 1987, when a revised version of the DSM-III was
published (DSM-III-R, APA 1987), the names of these two diagnostic categories changed
respectively into AD and PDD-NOS. It was in 1994 when Aspergers disorder entered the
category of PDDs in the fourth version of the DSM (APA 1994). PDD-NOS should be
diagnosed when there are impairments in some of the domains that characterize AD, but the
full criteria for AD are not met because of subthreshold symptomatology, late age at onset,
atypical symptoms, or a combination of these. According to the DSM-IV (ibid.), this
category should be used when there is a severe and pervasive impairment in the
development of reciprocal interaction associated with impairment in either verbal or
nonverbal communication skills or with the presence of stereotyped behavior, interest, and
activities, but the criteria are not met for a specific Pervasive Developmental Disorder.

Aspergers disorder is not just a light version of AD; it can be distinguished from AD by
some special exclusion criteria. Where AD requires impairment in three domains, Aspergers
disorder requires impairment in the social interaction domain and in the behavioral domain
in the same way these impairments are required for AD, but distinct form AD there should
not be a significant impairment in the communication domain. The DSM-IV explicitly states
that there shouldnt be a significant delay in language. Furthermore, there shouldnt be a
significant delay in cognitive development or in the development of age-appropriate self-
help skills, adaptive behavior (other than in social interaction), and curiosity about the
environment in childhood (APA 1994: 72). Additionally, it is important to notice that there
is no specific age of onset required and that the impairments in the two mentioned domains
need to cause impairment in social, occupational, or other important areas of functioning.
This last requirement is not required for AD.

Aspergers disorder and PDD-NOS are often regarded as the milder variants of AD and
these milder variants can currently account for almost three fourths of all ASD diagnoses

46
 3Autism Spectrum Disorders

(Chakrabarti and Fombonne 2001). Why was it necessary to create these two new diagnoses
as milder variants of AD? The exact reasons for this proliferation of diagnoses within the
autism spectrum are not yet clarified and objective to further exploration in this thesis, but
some factors can already be made clear. It was noticed that some people had serious
impairment in the autism domains, but they didnt fulfill all required criteria for infantile
autism (AD). Considering their amount of impairment and suffering, despite the fact that
they did not have a full-blown AD, it was desirable that this group got medical attention and
therefore an official status in the diagnostic manuals. PDD-NOS satisfied the diagnostic
need for children with pervasive and severe impairment in the development of the symptom
domains of autism, despite the fact that they did not meet the criteria for a specific PDD.
Aspergers disorder was introduced among other reasons after the recognition that autism-
like conditions exist among people with normal or above average intelligence and cognitive
capacities. It remains a matter of debate whether high-functioning AD and Aspergers
disorder should be considered two separate diagnoses or whether it is impossible to make a
clear distinction between the two different disorders.

A broadly acknowledged problem of the term PDD-NOS is that it includes subjects with
problems that most likely do not have phenotypical links with the other autism-like
conditions and etiological links with the other ASD are even more questionable. Examples
are children with non-verbal learning disorder or with severe specific language impairment
that cause persistent communication deficits. These children fulfill the criteria for PDD-
NOS but are generally not considered to have an ASD (Steyaert and De La Marche 2008).
The PDD-NOS category illustrates another already mentioned problem, namely that it
seems that the difference between AD and normal variance is not a clear (phenotypical)
distinction but rather a gradual distinction with PPD-NOS (and Aspergers disorder)
somewhere in between. Autism-like traits appear to be spread in different degrees in the
population and this complicates the demarcation between a disorder and normality. An
important question that rises is what determines whether we are speaking of real impairment
associated with those autism-like traits. At what point do these autism-like conditions, that
might as well be considered normal variance, become a disorder? Before discussing these
issues and the possible epidemic of ASD, I will pay some attention to the ways in which the
DSM classification system and the distinction between normal and pathological are related.

47
 3Autism Spectrum Disorders

3.1.3 DSM Classification and the Demarcation Problem

The DSM classification system is developed, among other reasons, to enhance the
communication between psychiatrists, psychologist and other mental health workers when
they exchange information on different harmful mental conditions (reliability). Furthermore,
it is developed to enhance systematic, reliable research on these different harmful conditions
of mental life. Although often used in this way, the DSM classification system is not
developed to draw the exact boundaries of the concept of mental disorder. Nevertheless,
conditions described in the manuals should be pathological conditions and it is of course
preferable that anyone who fulfills the criteria for a particular DSM diagnosis also has a
pathological condition. The DSM (APA 1994) does give a definition of the concept of
mental disorder and this definition is, like the one in the ICD (WHO 1992), primarily harm-
led. The first sentence of the DSM-IV definition of mental disorder is as follows:

In DSM-IV, each of the mental disorders is conceptualized as a clinically significant

behavioral or psychological syndrome or pattern that occurs in an individual and that is
associated with present distress (e.g., a painful symptom) or disability (i.e., impairment in one
or more important areas of functioning) or with a significant increased risk of suffering
death, pain, disability or an important loss of freedom (APA 1994: p. xxi-xxii)

Included are conditions that cause much distress or disability and this distress or disability
has to be pervasive across contexts. This is required in order to exclude frank social deviance
as a cause of the harm. It does not exclude distress and disability associated with pervasive,
less obvious (institutionalized) mechanisms of social exclusion, which makes the desired
distinction between mental disorder and social deviance somewhat inconsistent. But, as I
have mentioned earlier, it is impossible to fully demarcate between pathology and social
deviance and especially in psychiatry, an overlap is inevitable.

The tension between on the one hand the level of demarcating between normal and
pathological, and on the other hand the level of classifying different disorders, is often
referred to as one of the validity problems of the DSM classification system. It could be that
the DSM includes conditions that shouldnt be considered pathological or that they exclude
conditions that should be considered pathological. Whether this is the case is very hard to

48
 3Autism Spectrum Disorders

say, simply because the demarcation between normal and pathological is already very
complex and, according to Canguilhem, essentially normative and personal. These normative
and personal aspects of the process of demarcating between normal and pathological are in
controversy with the universalizing, collective and pragmatic aims of the process of
classifying. There seems to be an irreducible epistemological divide between the fundamental
level of demarcation between normal and pathological, and the more pragmatic level of
classification. I therefore suggest that it is unwise to try to fully integrate these to levels.
Rather, I would recommend a continuous process of evaluation, in which an individual
diagnosed according to the more universal DSM criteria should in addition be examined at
the more individual and fundamental level (in the diagnostic process between patient and
doctor) on whether the condition of the individual is pathological. This idea deserves more
attention than I can offer here, but it is important to emphasize that these two levels,
although strongly related, should conceptually and practically be kept apart.

There are critics (e.g., Horwitz & Wakefield 2007) who claim that in current times too many
people are diagnosed with a mental disorder according to the DSM criteria, while their
conditions shouldnt be considered pathological. They claim, for instance, that normal
sorrow and sadness is illegitimately pathologized and medicalized by labeling these
conditions as depressive disorder. It is not the border between mental disorder and social
deviance that is problematized here, but the border between mental disorder and normal
problems of everyday life. Whether it is true that many people are illegitimately pathologized
is hard to tell and could, in my opinion, only be judged in intensive interviews with those
that are supposed to be illegitimately pathologized, and even then, the boundary between the
normal and the pathological is not strict and factual, and constantly open to change. Bolton
(2008) has recently argued that many stakeholders become involved in where boundaries
are drawn in relation to diagnosis and the need to treat for conditions primarily involving
distress or impairment in the individual: the individuals concerned, but also families and
carers, advocacy groups, manufacturers of treatment technologies, and the funding bodies
(p. 238). Judgements of distress, disability and the need to treat are subject to many kinds of
social influences (pressure) from various stakeholders, some pressing for inclusion, some
for exclusion, which are best recognized and debated for what they are (p. 239). The
boundaries around harm are not fixed and subject to social, cultural and personal values and

49
 3Autism Spectrum Disorders

changing values result in changing boundaries of distress and disability. The extension of
these boundaries is not necessarily a medicalization of conditions that are essentially normal,
but could be considered a genuine extension of pathological conditions due to a changing
interaction between individuals and their environment. Whether this development (extension
of pathological conditions) is desirable and what the level of intervention (medical or social)
should be are different questions.

The rising rates of people diagnosed with ASD could be explained, following Horwitz and
Wakefield (2007), by an illegitimate extending of the boundaries of ASD. Part of what is
diagnosed as ASD, following this line of thought, should not be considered pathological.
This hypothesis, although not impossible, is hard to test, because again this can only be
judged in intensive interviews with those that are supposed to be illegitimately pathologized.
It would be very interesting to investigate whether this hypothesis is true, but in discussing
the possible epidemic of ASD I will assume that people diagnosed with an ASD (according
to DSM criteria) can be considered to be in a pathological condition and I will suggest that at
least part of the extension of the boundaries of autism is connected with an extension of
pathological conditions, and the extension of autism boundaries can partly be explained by
changing norms of social interaction and communication, while remaining a pathological
condition.

3.2 The Autism Epidemic

Since the 1960s, numerous autism researchers have published over 50 scientific articles on
the occurrence of autism in different populations. These articles show that there has been an
enormous increase in the prevalence of ASDs in the past 40 years (Fombonne 2005). In
1966, the rate of autism in Middlesex (UK) was 2.5 children out of 10.000 (Wing et al.,
1976). In 2001, the rate of AD (defined using DSM-IV) in the Midlands (UK) was 16.8 out
of 10.000 (Chakrabarti and Fombonne 2001). Some studies even found recent prevalence
rates of around 34 per 10.000 children for autism (Yeargin-Allsopp et al. 2003) and 60 per
10.000 for all ASDs (Chakrabarti and Fombonne 2005). A study by Baird et al., published in
the Lancet (2006), even found a prevalence of 1% (100/10.000) for ASDs in an urban
population. The scale of the observed increase of autism, and ASDs in general, has caused

50
 3Autism Spectrum Disorders

alarm about a possible epidemic of autism and it raised great public concern about the
possible (environmental) causes (Wazana 2007). This concern increased with popular media
talking about a mysterious upsurge of autism in children (New York Times, October 20,
2002, section 4: 10) and about new cases of autism exploding in number (Time, May 6,
2002: 48).

A particular study that increased the media and public concern of a possible autism epidemic
was an influential study by Wakefield et al. (1998) (not the Wakefield discussed above) who
explained the increasing incidence of autism with what is called the vaccine hypothesis. This
study hypothesized that new cases of autism were the consequence of brain damage caused
by the measles, mumps, rubella (MMR) vaccine or by thimerosal, an MMR vaccine
preservative that consists of 50% ethylmercury (Tan and Parkin 2000). The vaccine
hypothesis turned out to be very controversial and not based on sound scientific evidence
(Waterhouse 2008). In spite of this, until today, there are fierce debates about the different
causes of and reasons for the increased rates of children with ASD (Blaxill 2004,
Gernsbacher et al. 2005). In the literature on the increased rates of ASD, two broad lines of
explanation are given.

The first line of explanation is that there is a true secular rise of incidence of ASDs and this
line of explanation suggests a causal relation of some environmental factor (e.g. the MMR
vaccine) with the development of ASDs. This could include so called gene-environment
interactions, in which the disorder is expressed in individuals that have a genetic risk for
ASDs in combination with specific environmental factors that cause the disorder to become
apparent only in those with this genetic risk. The second line of explanation is that there is
no true secular rise of incidence, but the rise in rates of ASDs can be explained by changes
in diagnostic criteria, growing awareness and knowledge among doctors and parents, the
development of specialist services and diverse methods used in studies. A third possible
explanation, in line of Horwitz and Wakefield (2007), that an illegitimate extension of ASD
boundaries causes higher rates of people with ASD, is remarkably not discussed in the
epidemiological studies on ASD. Furthermore, what is meant with a true secular rise is
never made explicit in these epidemiological studies. I will suggest that there are some
implicit assumptions in both lines of explanation about disease and autism underlying the

51
 3Autism Spectrum Disorders

idea of a true secular rise of ASD. These assumptions narrow down possible perspectives,
explanations and treatments of ASD, and they restrain a new type of explanation for the rise
of ASDs; a type of explanation compatible with a Canguilhemian perspective on the
distinction between normal and the pathological, and health and disease. Before discussing
these assumptions and this different type of explanation for the rise of ASDs, the two lines
of explanation will be discussed in more detail.

3.2.1 Two Lines of Explanation

The first line of explanation is that there is a true increase of children with AD or other
ASDs. Many hypotheses have been suggested to explain a real increase in ASDs. All these
hypotheses assume that environmental factors, often in interaction or combination with
genetic factors, cause brain damage in some way or another and this brain damage correlates
with one of the ASDs depending on the specific type of brain damage. Many suggestions of
possible causes have been made, including neurotoxins such as mercury, antibiotics,
environmental pollutants, specific food and many more. The hypotheses for a real increase
in autism require a material modification at the level of the brain (that can be caused by
environmental factors) and ASD can ultimately and ideally be explained at this brain level
where each specific ASD relates to specific brain alterations. ASD are conceptualized as
concrete disease entities and to make a comparison with somatic disorders; different ASD
subtypes could be compared with different subtypes of e.g. influenza virus infections or
breast cancer.

One of the hypotheses I already mentioned is the vaccine hypothesis. This is the most
widely-circulated hypothesis of increasing rates of ASDs. This hypothesis was generated by a
study of Wakefield et al. (1998) which reported an association between autism and MMR
vaccination in combination with gastro-intestinal problems in 7 out of 12 children. The
results of this study were widely spread through a press and video news release and a
television press conference which caused an extreme concern about MMR vaccinations. This
resulted in increasing public concern about autism and it resulted in a decreasing number of
children that received the MMR vaccine (Deer 2008). The results of this study were never
replicated. Hornig et al. (2008) found no relation between the MMR vaccine and autism
onset. Honda et al. (2005) even found an increase in ASD in a part of Yokohama city in

52
 3Autism Spectrum Disorders

Japan, after they had stopped to give the MMR vaccine to children. Equally, Fombonne et al.
(2006) reported an increase in the rate of ASD and a decrease in the rate of MMR
vaccinations in a population during an 11-year study interval. Furthermore, apart from the
lack of additional scientific support for the suggested link between MMR vaccination and
ASD, Wakefields study turned out to be unethical and methodologically unsound.
Additional research uncovered that Wakefield didnt have the approval of the subjects for
the study and the recruitment of the subjects was biased because the parents of 10 out of 12
subjects were clients of R. Barr, a British lawyer who tried to sue the pharmaceutical
companies that made the MMR vaccine (Deer 2008). There are many more fraudulent
aspects associated with this study by Wakefield, but the essence is that despite its influence
and popularity, until today there is no evidence to support the hypothesis that the MMR
vaccine can induce ASD (Waterhouse 2008).

Among other theories that try to explain a real ASD increase there is a theory of Rogers
(2008) who proposes that the increased rates of ASD is the result of the increased use of
folate supplement by pregnant women. He suggests that children with a certain genetic
polymorphism (genetic variance) are unable to maintain normal folate levels which results in
abnormal neurodevelopment which results in ASD. Theoharides et al. (2008) hypothesized
that the increasing rates of ASD could be the result of the activation of mast cells that
release molecules that disrupt the gut-blood-brain barrier which, in turn, leads to neurotoxic
effects that generate ASD. Intestinal toxins would enter the brain and cause brain damage
leading to ASD. These theories are preliminary and not supported by strong scientific
evidence, but these three theories have in common that they explain the rise in ASD rates by
factors that damage or modify the brain at a biological level. ASDs are first and foremost
conceptualized as brain or neurodevelopmental disorders (see section 3.3) therefore a true
rise in the number of these brain or neurodevelopmental disorders requires an increase of
discrete molecular or biological deficits at this brain or neurodevelopmental level associated
with ASD. I want to regard this as the first assumption of a true secular rise of ASD.

The second line of explanation is that there is no true secular rise of incidence, but the rise
in rates of ASDs can be explained by factors such as changes in diagnostic criteria, growing
awareness and knowledge among doctors and parents, the development of specialist services

53
 3Autism Spectrum Disorders

and diverse methods used in studies. Fombonne (2005) and Wing and Potter (2002),
amongst others, concluded that there is indeed a rise in rates of ASD, but this is not a true
rise of incidence, because changes in diagnostic criteria, improved identification of children
with ASD and growing knowledge about ASD can explain the rise of ASD. As mentioned
earlier, there are no clear-cut diagnostic tests for ASDs. The diagnosis is made using detailed
information from parents about the development of the child, by observing the behavior of
the child in different situations and by using behavioral and psychological tests based on
diagnostic criteria in diagnostic manuals. This process is full of complexities of definition
and standardization and since Kanner first described his diagnostic criteria for early infantile
autism, the criteria, nomenclature and definition have changed substantially.

Kanner and Eisenberg (1956) emphasized two behavioral characteristics as necessary and
sufficient diagnostic criteria for early infantile autism. The first was aloofness and
indifference to others, the second was extreme resistance to change in the patients repetitive
routines. In addition, this behavior had to be present before 24 months. Rutter (1978)
slightly modified these criteria and he renamed early infantile autism into childhood autism.
His criteria were an impaired social development, insistence on sameness, impaired language
development and an onset before 30 months. The next change in the concept of autism
came soon after in 1980 in the DSM-III (APA 1980). The name changed again, now into
infantile autism and this diagnosis was classified as a subgroup of the broader category of
pervasive developmental disorders (PDD). The concept of autism shifted from a regular
psychiatric disorder to a developmental disorder and the criteria were for the most part
based on Rutters criteria. An additional change of the concept of autism was the
introduction of the concept of a spectrum of autistic disorders. Wing and Gould (1979)
pointed out that each of the three essential elements of autism (the triad of autism
characteristics) could occur in different degrees of severity and in diverse manifestations.
Social impairment could, for instance, be shown as passive behavior in social interaction or
as inappropriate but active behavior with repetitive inflexible approaches to others. Social
impairment became a much wider concept then just aloofness as in Kanners criteria for
early infantile autism.

54
 3Autism Spectrum Disorders

The idea of a spectrum of autism disorders began to be implemented in the revised version
of DSM-III, the DSM-III-R (APA 1987). This version kept the PDD category but the
subgroups were named autistic disorder (AD) and PDD-NOS and the diagnostic criteria for
these two diagnoses now covered a whole range of behavior of the triad of autism features.
Additionally, the criterion for the required age of onset of 30 months was removed.
Subsequently, the DSM-IV (APA 1994) introduced some new disorders in the PDD
category including Aspergers disorder. With the inclusion of Aspergers disorder, considered
a milder variant of AD, the idea of a spectrum of autism disorders got stronger and more
solid. It becomes apparent that the loosening in the diagnostic criteria and the inclusion of
milder variants of AD have contributed to the increased rate of people that include the
current spectrum concept of autism.

Wing and Potter (2002: 154-155) clearly analyzed that both DSM-III-R and DSM-IV/ICD-
10 allowed for a wide range of types of social and communication impairment and of
repetitive activities even within the subgroup of autistic disorder. Concerning the
communication domain, they concluded that unlike Kanners and Rutters criteria, DSM-
III-R and DSM-IV/ICD-10 did not insist on language delay or deviance as long as some
other type of communication impairment was present, such as poor intonation or
inappropriate use of speech in relation to social context (ibid.). In addition, comparing the
DSM-III with the DSM-IV criteria for AD, the differences in the way the criteria are
formulated shows the loosening of specific criteria. For example, concerning the social
interaction domain, in the DSM-III an individual needed to show signs of a pervasive lack
of responsiveness to other people (APA 1980: 89), whereas, according to the DSM-IV, an
individual only needs to have a lack of spontaneous seeking to share enjoyment, interest, or
achievements with other people and a failure to develop peer relationships appropriate to
developmental level (APA 1994: 70). Concerning the communication domain, the DSM-III
required gross deficits in language development and if speech was present, peculiar
speech patterns such as immediate and delayed echolalia, metaphorical language, and
pronominal reversal were required (APA 1980: 89). These criteria were replaced in the
DSM-IV by the looser requirements of impairment in the ability to initiate or sustain a
conversation with others or a lack of varied, spontaneous make-believe play or social
imitative play (APA 1994: 70).

55
 3Autism Spectrum Disorders

According to Fombonne (2005) and Wing and Potter (2002), together with an improved
identification of children with ASD and growing knowledge about ASD, the loosening of
the diagnostic criteria can most likely explain the enormous rise of people with ASD. In
addition, they support the idea that at least part of the rise in ASD prevalence is due to
diagnostic substitution. This is especially the case with severe mental retardation. Part of the
children who were diagnosed as severe mental retardation are currently diagnosed as ASD,
which resulted in an increase of children with ASD and a decrease of children with severe
mental retardation (Coo 2008). Consequently, they argue, there is no evidence for a true
secular rise of ASD. What a true secular rise of ASD eventually is is not made explicit and
they unfortunately dont try to explain why the diagnostic criteria for ASD have changed so
much the past 40 years. Following their line of argument we can infer that in order to have a
true secular rise of ASD it is needed that an increasing number of people include the
concept of autism over a specific range of time and that the criteria required to include this
concept of autism are stable over this specific range of time. From a behavioral point of
view, more people need to exhibit comparable behavior associated with autism or ASD over
this specific range of time. Since the criteria required to include this concept of autism are
not stable over time, it is hard to say whether there is a true increase in this sense.

Attempts to by-pass this problem are made by applying both Kanners criteria and the DSM-
IV criteria at the same group of people with ASD and this resulted in a couple of studies in
which proportions with Kanners syndrome ranged from 33 to 45% of all those diagnosed
as having DSM-IV/ICD-10 autistic disorder (Wing and Potter 2002: 154). Despite these
attempts to by-pass the problem of changing autism criteria, a true increase in autism,
according to the discussed authors, requires at least the idea of a stable concept of autism (if
necessary hypothetical). This concept of autism can in theory only be stable with unchanging
criteria over time either when it can be defined in a naturalist value-free way, independent of
a social context, or when the social context and the values that turn out to be bound to the
concept of autism, are perfectly stable. Both options are problematic. I want to regard this
idea of a stable concept of autism, with fixed criteria required for inclusion, as the second
assumption of the discussion on a true secular rise of ASD.

56
 3Autism Spectrum Disorders

3.2.2 Two Assumptions

It has become clear that there are at least two assumptions underlying the current debate on
whether there is a recent autism epidemic or not. These assumptions enforce a limited
perspective on the increasing rates of ASD and they restrain a social and critical evaluation
of the increasing rates of ASD in which ASD is essentially defined in a social context where
pathology and social norms are connected. The first assumption is linked with the idea of
ASD as brain or neurodevelopmental disorder and a true rise in the number of brain or
neurodevelopmental disorders requires an increase of discrete molecular or biological
deficits at this brain or neurodevelopmental level. I will get back to this in section 3.3. The
second assumption is the idea of a stable concept of autism, with fixed criteria required for
inclusion. This requires that the concept of autism is defined in a naturalist value-free way or
it requires a stability of the values associated with the concept when the concept of autism
cannot be defined in a value-free way. In the light of Canguilhems distinction between
normal and pathological, the idea of a stable disease concept, with fixed criteria for inclusion,
is problematic.

The concept of autism, before the process of classifying different subgroups on the basis of
specific characteristics, should at the first stage include people who are in a pathological or
diseased condition. As pointed out in the previous chapter, to be in a pathological condition
depends amongst other things on the subjects value-orientation. A pathological condition is
a negatively evaluated situation, associated with suffering and limitation, and the inability to
create new behavior in order to overcome the condition of suffering and limitation. This
condition of suffering and limitation is incorporated in most of the DSM diagnostic
categories. The AD diagnosis, for instance, requires qualitative impairment in social
interaction and impairments in communication, and Aspergers disorder requires
significant impairment in social, occupational, or other important areas of functioning
(emphasis added: APA 1994, p. 70). These conditions of impairment refer to conditions of
suffering and limitation and the impairment is linked with the criteria that describe the
behavior specific for the concerning disorder. The lack of social or emotional reciprocity and
the failure to develop peer relationships, for example, are associated with impairment in

57
 3Autism Spectrum Disorders

social interaction in the case of AD. This means that the specific behavioral or psychological
criteria of a disorder are essentially linked to the required impairment.

Whether certain behavior or lack of behavior, e.g. social or emotional reciprocity or

developing peer relationships, is linked with impairment depends partly on the social and
cultural norms that are present in the society in which the individual functions. Especially in
ASD, characterized by problems in social interaction, communication and behavioral
patterns, the suffering and limitation depend on norms of social interaction, communication
and behavior. Adaptation to norms (social normativization) of social interaction, for
instance, involves health benefits, and the inability to adapt to these norms involves suffering
and limitation and particularly this inability is regarded as the crucial problem of people with
ASD. This makes ASD situated in the zone where social deviance and pathology overlap and
moving boundaries of social norms in a certain society influence the boundaries of what is
considered pathological in that society.

Deviation from certain behavioral or psychological norms is never intrinsically pathological.

The for ASD characteristic lack of social or emotional reciprocity, the stereotyped and
repetitive use of language or the persistent preoccupation with parts of objects can be
considered pathological only when it is associated with or causes impairment. It is
conceivable that not developing peer relationships or the (pre)occupation with parts of
objects does not cause any impairment in a society that doesnt evaluate this specific
behavior in a negative way. This behavior will therefore not be considered abnormal or
deficit. With variable norms of social interaction and communication, the specific behavior
associated with suffering and limitation (impairment) will vary and the criteria for ASD need
to be adjusted over time. In a way, changing criteria of ASD reflect changing values and
norms of western societies concerning social interaction and communication. It is therefore,
that the concept of autism cannot refer to stable (behavioral, psychological or biological)
criteria independent of a social context. The idea of a true secular rise in ASD turns out to
be based upon this untenable assumption and a different perspective on the concept of
autism opens up different possible explanations for the current loosening of diagnostic
criteria and the rising number of people with ASD.

58
 3Autism Spectrum Disorders

3.3 Autism as a Brain Disorder

ASD is nowadays considered a biological disorder of the brain (Muhle et al. 2004). An
important reason for this biological approach is the scientific evidence that ASDs have a
very high heritability (Freitag 2007). Twin studies with monozygotic and dizygotic twins
showed that the variance between someone with an ASD and someone without an ASD was
for 70-90% explained by genetic factors (Le Couteur et al. 1996). The biological approach of
ASD is not exceptional in psychiatry; scientific research of practically all psychiatric disorders
is currently dominated by a biological approach of mental functioning. Biological psychiatry,
the subdivision of general psychiatry that tries to explain psychiatric disorders and mental
functioning in biological terms, is a fast growing research front and it is also the main form
of therapy in contemporary psychiatry in the form of medication.

Advances in the biomedical sciences such as the neurosciences and genomics created many
new possibilities to investigate the biological bases of psychiatric disorders. With these
(technological) advances in biomedicine the mind became visible in multiple new ways. Brain
imaging, for instance, shows different regions in the brain lighting up in the normal and the
autistic brain (Di Martino et al. 2009). Molecular neuroscience visualizes neurons, receptors,
neurotransmitters and cell membranes, and identifies specific patterns of functioning at this
molecular level to each normal and pathological mental state. Genetics look for gene
sequences and varieties in base pairs that correlate with specific diagnostic categories, like the
different ASD categories. Several genes that play a role in brain and neurodevelopmental
processes, like synapse formation, plasticity and neuron growth, have been linked with ASD
(Kim et al. 2008) and these findings support the idea of ASD as a brain disorder.

There are probably many social, economical, political, technological and historical ways to
understand and appraise these developments towards a more biologically orientated
psychiatry (see Rose 2007). For now I will focus on how a biological approach of ASD
relates to the distinction between normal and pathological as described in the previous
chapter. With increasing biological knowledge about ASD, a dominant reductionist
perspective on what ASD is and how it can be distinguished from normality is being shaped.
Biological psychiatry not only identifies biological factors associated with ASD, it shapes and

59
 3Autism Spectrum Disorders

delimits what there is to be explained. What there is to be explained becomes an internal

biological thing that is located within the brain (Rose 2007). Even though we may not exactly
know what this internal biological thing is, and even though ASD manifests itself in certain
behavior or in social functioning, ASD becomes something to be explained in biological
terms within the boundaries of the body. This style of reasoning underlies part of the idea of
a true secular rise in ASD that requires an increase of discrete molecular or biological
deficits at this brain or neurodevelopmental level (first assumption). In this style of
reasoning (Hacking 1992) the concept of autism is being shaped and I suggest that this
concept of autism as a brain disorder has narrowed down and simplified the discussion on
ASD and whether there is an epidemic of ASD going on or not. The concept of autism
exclusively as a brain disorder is in need of serious revision; a revision that incorporates
aspects of suffering and limitation in the concept of autism. These aspects are essential in
defining pathology and have been marginalized by biological reductionist approaches.

As Canguilhem (1989) showed in his essay A New Concept in Pathology: Error, that to
speak of a genetic or biological disorder, is to locate disease within the individual organism
and not in the relationship between the organism and the environment. It is a mistake, he
claims, to speak of a biological error or deficit, because it presumes that this biological deficit
has a negative valuation in its own right, independent of any context. A genetic factor, a
biological feature or even isolated behavior (see 3.2.2) can never be valued negative or
positive in its own right; the value always depends on the organisms situation in a certain
environment, and pathology, as discussed earlier, is located in the relation between an
organism and its environment. In order to value an isolated biological feature the
environment must be approached as a static, fixed condition and not as a changing, dynamic
system. In ASD, the scientifically determined abnormal biological features might cause
certain behavior that causes impairment in the three domains of autism, but it is not possible
to talk about impairment or suffering, by evaluating exclusively these biological features. The
impairment arises in the evaluation of the subjects situation or state in a particular
environment and the impairment is therefore irreducible to value-free biological features.
Biological features become errors or deficits only in a context. In isolation, these biological
features are nothing more than variations at a biochemical or biophysical level.

60
 3Autism Spectrum Disorders

What does it accordingly mean that 70-90% of the difference between someone with an
ASD and someone without an ASD is explained by genetic factors? Somewhat simplified,
this means that there are genes that account, for a substantial part, for the difference
between the behavioral and psychological phenomena required for an ASD diagnosis and
the absence of these behavioral and psychological phenomena. The fact that there are genes
involved in differences in certain phenomena, tells us nothing about whether these
phenomena are pathological or not. To make a comparison; difference in hair color is
explained almost completely by genetic factors, but this will not tell us anything about
whether a certain hair color, which is obviously not the case, is pathological or not. This
argument can be applied not only to genetic factors, but to all biological features. A
biological feature in itself cannot make the fundamental distinction between normal and
pathological. Cancer, a process of uncontrolled cell division, is without doubt a pathological
process, but we can only speak of it as a pathological process because of the very strong
causal link between the process of uncontrolled cell division and a negatively valued situation
of extreme pain, losing weight, loosing appetite, suffering, limitation and eventually death.
Only the consequences of the process of uncontrolled cell division make the process a
pathological process; it is not the specific biological feature of the process itself that make it
pathological.

There are some important conceptual and empirical differences in the biological features
significant in for instance cancer and in ASD. First, the causal link between the biological
features in cancer (uncontrolled cell division) and the resulting impairment and suffering is
very strong. Practically all people with this type of uncontrolled cell division will at some
point experience impairment and suffering because of this uncontrolled cell division. In
contrast, variations in genes associated with ASD (e.g. neurexin (see Kim 2008) and
neuroligin (see Jamain 2003)) only have a small statistical association with the ASD
phenotype. Having a detected genetic variation associated with ASD will increase the chance
of eventually getting ASD with less than a few percent. There are many people that have a
genetic susceptibility for ASD, but who will never get ASD. A genetic variation associated
with ASD will therefore not be considered pathological in a sense that a cellular cancer
process will be considered pathological.

61
 3Autism Spectrum Disorders

A second important difference is that in cancer, the caused impairment and suffering is quite
unambiguous. It often involves (if untreated) severe pain, severe loss of weight and appetite,
immobility, fatigue, painful complication in the organs close to the cancer process, and
eventually death. There is not much discussion on whether all these consequences really
have a negative value for the individual and this negative evaluation depends not so much, if
at all, on social and cultural differences and changes. The suffering and limitation associated
with ASD, on the other hand, is not that unambiguous. Not developing peer relationships,
not spontaneously seeking to share enjoyment with other people or having restricted
patterns of interest is not always evaluated negatively. The values in ASD seem to be more
vulnerable for controversy and for modification when norms of social interaction change.
The correlated biological features will consequently shift from the normal to the pathological
domain (or vice versa) when the values attached to the behavioral consequences of these
biological features change from positive to negative. This is clearly not the case with
biological features in cancer.

We can conclude that the distinction between normal and pathological cannot be made at
the biological level. Furthermore, the causes of impairment and suffering in ASD cannot be
found solely at the biological level, even when environmental factors, such as the MMR
vaccine, that might cause damage at the biological level, are included. Trying to explain the
rise of ASD only by an (hypothetical) increase in biological deficits or provable brain damage
will therefore at the most be a limited perspective on the possible explanations. Other
possible explanations that can be found in the social domain get little attention in the current
dominant biomedical concept of autism. The focus on biological features leaves little room
for critical reflection on social and cultural developments that might play a role in the rise of
not only ASD, but of ADHD, depression and addiction as well. Some preliminary social
explanations for the rise in ASD will be discussed in the final chapter.

Canguilhem (1989) already warned against the biased fixation on biological and genetic
errors and the scientific dream to correct and eliminate all possible biological errors and
abnormalities in the attempt to reach an ultimate healthy state that erroneously does not take
the organisms relation to its changing and unpredictable environment into account. His
worries and warnings are nowadays of great relevance; recent developments in molecular

62
 3Autism Spectrum Disorders

genetics, molecular neurosciences and psychopharmacology opened up many new ways that
enable new interventions at a molecular or biological level. Biological anomalies can be
identified and normalized through genetic engineering, chemical and electric balancing of the
brain. With a limited but dominant biomedical view on psychiatric disorders, other causes,
explanations and treatments that are to be found outside the biological domain, will be
unjustifiably ignored. Canguilhems fundamental work on the essence of normality and
pathology helps to resist the possibility of reduction and it protects against the temptation of
premature satisfaction of exclusively mechanist and reductionist explanations of psychiatric
disorder and ASD in this particular case.

63
Chapter 4
A Canguilhemian Perspective on the Concept of Autism

4.1 The Shifting Concept of Autism

With Kanner and Asperger first describing autism in 1943 and 1944, the concept of autism
emerged and Canguilhem would consider this (the formation of a concept) the first step in
the process from not-knowing to knowing and a break away from the obvious and intuitive.
Since its emergence, the concept of autism has functioned in different theories and it can be
considered theoretically polyvalent. In describing the history of the autism concept the
current dominant biomedical models, in which autism is explained, shouldnt be considered
an endpoint of a linear and chronological scientific development towards this endpoint. This
type of history of science would imply a positivist epistemology. Rather, Canguilhem
supports a historical epistemology in which the object of history of science is historical
discourse and this historical discourse is first of all a history of concepts. The task of the
historical epistemologist is to establish the order of conceptual progress that is visible only
after the fact and of which the present notion of scientific truth is the provisional point of
culmination (Rabinow in Canguilhem 1994: 14). A historical analysis of the concept of
autism should emphasize its historicity, provisionality and in this process the normativity of
truth about the concept of autism becomes apparent.

In the discussion on the current autism epidemic, the concept of autism is not approached in
this historical provisional way, but in a positivist way that assumes the idea of a stable
concept of autism and a definite materialist explanation of autism at a biological level.
Scientific development on the concept of autism will, contrary to a positivist conception of
autism, not lead to infallible truths; it is a non-linear, heterogeneous, contingent, regional
history in which the transforming concept of autism is inserted with history itself. The
normativity of scientific truth about autism parallels the essential (historical and provisional)
normativity of life itself. According to Canguilhem, there is no fundamental distinction
between the order of thought of the concept and theoretical objects of the life sciences
and the order of the referent the extra-discursive vital order (Rose 1998: 156). The
capacity to create new vital norms and the capacity to create scientific concepts and tools,

64
 4A Canguilhemian Perspective on the Concept of Autism

both in order to cope with a variable environment, are fundamental activities of the vital
force of life. Scientific concepts and specific fields of reality become connected and
Canguilhem hereby avoids the limits of both scientific realism and social constructionism in
which there is an irreducible epistemological divide between the order of reality and the
order of thought (ibid.: 163).

As discussed in the previous chapter, the relatively young concept of autism demonstrates
both continuities and discontinuities in its progress towards the present notion of ASD.
There is a clear continuity in the type of problem the concept refers to. Since the first
descriptions of the new concept by Kanner and Asperger, a remarkable indifference to
others and a resistance to change in the patients stereotypical repetitive routines have been
central to the concept of autism. Furthermore, the concept has always intended to describe a
pathological state that has been associated with a certain degree of impairment in the social
interaction and communication domains. Clearly discontinuous is the later introduced idea
of a spectrum of autism disorders. Furthermore, the specific content of the pathological
behavioral and psychological phenomena have partially changed in the progression of the
concept of autism. Again, this is not a progression in the direction of a higher truth; it is a
contingent, revisable and historical shift because of the inherent normativity of both science
and life. In finding an explanation for the rise of people with ASD, we have to realize that
the concept of autism in 1943 is not identical with the concept of autism used today.

The concept of autism is not a static concept, around which knowledge turns. It is
continuously shifting; first at the scientific level due to new methods, explanations and
hypothesis that are linked with new ways of speaking the truth, and second (as a
consequence of the inherent normativity of life itself) due to a varying relationship between
the subject and its environment and it is in this relationship that the pathological state is
defined. A true secular rise of ASD needs the pathological state called ASD to be defined
independent of its relationship with a varying environment and this turned out to be
incompatible with the essential demarcation between the normal and the pathological. The
notion of a true secular rise of ASD is an empty notion that could only (hypothetically)
have content when the behavioral or biological state (labeled autism) is cut loose from its
associated pathological condition. This is obviously undesirable and could lead to

65
 4A Canguilhemian Perspective on the Concept of Autism

normalizing variation and difference instead of treating what is essentially pathological. The
dominant biomedical notion of autism carries exactly this risk of focusing exclusively on
biological features and not taking into account the essential difference between the normal
and the pathological that cannot be made in the biological domain and that is necessarily
defined in the relation between the subject and the environment.

4.2 Autism: A Paradigm Example

Despite the fact that the concept of autism is necessarily changing, and despite the fact that
vital and social factors cannot be clearly separated in describing autism as a pathological
condition, I suggest that, at a behavioral and psychological level, autism could serve as a
paradigm example for Canguilhems concept of pathology defined in contrast with health.
For Canguilhem, the healthy state is characterized by the fact that the organism is able to
adjust to new situations by establishing new norms; new modes of functioning appropriate
for the organism in a new situation. Health is characterized by the possibility of transcending
the norm and by tolerating infractions of the habitual mode of functioning by reacting in a
creative way at different circumstances. Pathology, in contrast, is characterized by a
decreased capacity to tolerate change and a decreased capacity to adapt to new situations that
are negatively evaluated by the organism. This results in feelings of suffering and limitation
that are minimalized by an inflexible holding on to the earlier modes of functioning. New
and unexpected situations in the organisms environment are not experienced as new options
or possibilities, but as a danger to an unstable control of the level of suffering. With this
concept of pathology, the condition of people with autism can be considered pathological.

Since Kanners first descriptions of autism, the concept referred to a fierce resistance to
change in the patients stereotypical repetitive routines and until today, ASDs are
characterized by inflexible and repetitive modes of communication and social interaction, by
inflexible preoccupations with restricted patterns of interest and by inflexible and repetitive
adherence to specific routines or rituals. Infractions of e.g. these specific routines or rituals
result in an immediate increase in the level of stress and suffering experienced by the
individual and there is a lack of creative flexibility to cope with the new unexpected situation
that emerged. Pete, the 8 year old boy from the case described in the first chapter, reacted
with intense stress, suffering and eventually aggressive behavior when a classmate

66
 4A Canguilhemian Perspective on the Concept of Autism

accidentally took his seat. Such a seemingly small change in Petes expectations and situation,
that in a normal boy or girl would have resulted in picking another chair or simply asking
the teacher for a new chair, resulted in Petes case in intense stress and suffering. There are
numerous examples of the inflexibility and the incapacity to creatively cope with changing
situations of people with ASD. A striking example is one of my patients who has a strict
ritual when going to the supermarket. She makes an exact list of the things she needs and she
always takes the same route to the supermarket. When something happens on her way going
shopping, for instance a flat tire on her bike or when the road she always takes is blocked,
she completely stresses out and has to go back home to calm down. A little change in her
plan and expectations cannot be overcome in a creative way in order to finish the task she
had initiated. For someone with ASD, the suffering and limitation is minimalized when
circumstances are as stable or expected as possible. For Canguilhem, health is a margin of
tolerance for the inconstancies of the environment (Canguilhem 1989: 197) and in ASD
this margin of tolerance is minimal.

An important question that so far has remained unanswered is what factors are linked with
the progression of the concept of autism, especially concerning the changes in the specific
content of the pathological behavioral and psychological phenomena required for diagnosing
an ASD. What can explain the loosening of the diagnostic criteria and the emergence of the
idea of a whole spectrum of autistic disorders and at what point do autism-like conditions,
that might as well be considered normal variance in behavioral and psychological
phenomena become real disorders? With Canguilhem, pathology is not just a biological thing
or process inside the subject; it is, particularly in psychiatric disorders, essentially defined in
the relationship with a changing environment, and in human beings this environment is a
social environment in which social norms are inextricably linked with experiences of
suffering and limitation. This opens up new explanations for the rise of ASD that can be
found not exclusively inside the subject, but in the currently (by biomedical approaches of
ASD) neglected social domain. Furthermore, this opens up interventions aiming at
decreasing individual suffering and limitation, not just at the individual level, but at the
societal level as well. In the final paragraph, I will give some preliminary social explanations
for the rise of ASD that is connected with the loosening of the diagnostic criteria and a
shifting concept of autism.

67
 4A Canguilhemian Perspective on the Concept of Autism

4.3 New Explanations for the Autism Epidemic

Autism-like conditions, like inflexible and repetitive behavior and particular modes of
communicating and interacting, can become pathological conditions when valued negatively
by the individual and when the individual is not able to change this behavior or mode of
functioning. Whether these autism-like conditions are valued negative depends not in the
least on the norms of communication and social interaction. Furthermore, whether certain
behavior is labeled inflexible, depends on the demands of a society on the different roles
and types of behavior a citizen or a child is expected to exhibit. With this perspective on
pathology and autism, the rising rates of children who are limited and impaired because of
their inflexible behavior and their typical way of communicating and interacting with other
children, can be explained by changing social norms and changing demands on and roles of
children the last few decades. I will briefly discuss two social/contextual explanations for the
rising rates of children whose impairment due to certain behavior is related to social,
economical and political values in the current modern Western society.

The first explanation is partly based on the ideas of Sami Timimi, a critical child and
adolescent psychiatrist. He (Timimi 2008) recently emphasized the fact that the gender
distribution for psychiatric disorders in children is approximately four boys to every girl.
This crooked distribution is mainly caused by behavioral disorders such as ADHD and ASD.
He argues that the reason why we have become more troubled by boys than by girls
behavior has to do with shifting ideas of femininity and masculinity, the role and position of
men and woman in society, and the persistent progress of individualism in Western society.
He claims that the boundaries between what is considered normal and pathological
development (by educational and psychological authorities) of childhood is being shaped by
rapid social, economic and political changes in the last decades such as: the movement into
smaller family and social networks, decreasing amounts of time that parents spend with their
children, aggressive consumerism preying on childrens desire for stimulation, greater
involvement of professionals in childrearing activities (and advice on childrearing), and a
sense of panic about boys development (ibid.: 174).

68
 4A Canguilhemian Perspective on the Concept of Autism

Timimi argues that free market capitalist culture, a system mainly based on values of
masculinity like competitiveness, putting individual needs above social responsibility, and the
use of rational scientific analysis, results in contrast of these masculine values in a
feminization of certain educational and psychological aspects of childhood. Women are
more and more brought into the workplace (to increase workforce), which resulted in a
changing role of motherhood and a clear movement out of the family sphere and into the
public sphere. At the same time there has been a movement of childcare becoming a
professional and mainly female activity. There is a growing body of literature that supports
the idea that there is an increasing feminization of educational and psychological aspects of
childhood. Current educational methods used in most Western schools, like socially oriented
assignments and frequent assessment, are favored more by girls than by boys and girls are on
average achieving higher grades than boys at school (Burman 2005). In addition, the
diminishing size of families requires more intense emotional contact between members of
these smaller units, and less opportunity for learning to negotiate differing types of
relationships that contact with the wider range of people that extended families provide
(Timimi 2008: 175). The feminization of aspects of the capitalist culture not only has an
impact on childhood, but also on working settings. Ideas such as emotional intelligence in
management and professional relations have become popular since the 1990s (Gordo-Lopez
& Burman 2004) and this idea can be seen as part of a development creating new and better
ways to manipulate the consumer and to motivate the labor force. These are all signs of a
modern Western culture demanding more complicated forms of social interaction and
communication and children and adults who perform worse on social skills are clearly
disadvantaged.

As I mentioned earlier in chapter 3, part of what characterizes ASD is a lack of empathizing

skills (care about how others feel and predict and interpret a persons behavior) and people
with ASD generally perform good on systemizing skills. Baron-Cohen (2002) suggested that
the brain of someone with ASD is an extreme male brain, because men generally perform
better at systemizing and women generally perform better at empathizing. I suggest that the
current rise of people with ASD is connected with the increasing importance of the more
feminine social and empathizing skills, which leads to impairment in those who perform
worse on those skills. In earlier times or in different non-western areas in the world, these

69
 4A Canguilhemian Perspective on the Concept of Autism

skills might not be valued the way they are valued in a modern capitalist society and the lack
of empathizing skills might not result in impairment. This makes especially the lighter
disorders in the autism spectrum, like Aspergers disorder and PDD-NOS, essentially culture
specific. The hypothesis that the growing importance of empathizing skills in our society
influences the amount of children diagnosed with ASD, is supported by the fact that the
gender distribution in Aspergers disorder, the most recent ASD that became an official
DSM diagnosis in 1994, is nine boys for every girl. The hypothesis could further be studied
and supported by precisely relating changes in the criteria for ASD in different editions of
the DSM with for instance changes in educational methods in the same period. Furthermore,
the meaning of childhood, the norms of child development and the different roles children
have and are expected to exhibit in society need to be evaluated in a historical perspective.
Have the norms of psychological and behavioral child development changed the last forty
years and how did these changes effect educational methods? What kind of changes in the
meaning of childhood can be observed the last few decades? These are important questions
that need to be addressed to examine the hypothesis of a growing social importance of
empathizing skills that influences the amount of children with ASD. With this hypothesis,
Im not suggesting that Aspergers disorder is not a real pathological condition; Im just
trying to give an alternative (from a biomedical) explanation for the emergence of many
children with an ASD. Timimi is more resolute and he talks about a relentless medicalization
of children. He (2004: 226) claims that without any tangible evidence of organic pathology
and any biological tests to substantiate our hypothesis of a neurological dysfunction, the
boundaries of the disorder can expand endlessly and are dependent on the subjective
opinion of the person making the diagnosis, but as I have argued, the boundaries of the
disorder can never be made with organic pathology or biological tests and whether this
medicalization of children is legitimate or not depends more on the amount of suffering and
limitation of these children and not on some biological fact.

The systemizing skills (analyzing variables in a lawful, finite and deterministic system and
deriving the underlying set of laws that govern the actions of a system) of especially the
higher functioning people diagnosed with ASD are often above average. These skills are
beneficial in certain environments and the higher rates of people with ASD in areas where
high-technology industry is located, such as the city of Eindhoven in the Netherlands where

70
 4A Canguilhemian Perspective on the Concept of Autism

a technical university, many IT-companies and the Philips factory (producer of technical
equipment) are located, are explained by the advantage of having systemizing skills in these
high-technology working environments. In Eindhoven there are even special employment
agencies for people with ASD and there are employers who look for employees that have
these special systemizing capacities. An example of such a job is correcting computer
software for possible mistakes; a job that requires a precise, detailed, systematical and
focused approach, in which the context and social or communication skills are not relevant
and even distracting. The lack of empathy and narrow interests and activities will not lead to
impairment and suffering in an environment in which these skills or capacities are not
required. In a different environment, for example at school or at a job that requires social
contact with customers, the same behavioral and psychological characteristics would result in
impairment and would be considered pathological.

The earlier mentioned professor and well-known autism researcher Baron-Cohen (2000)
proposes that Aspergers disorder and high-functioning autistic disorder are not necessarily a
disability and should be conceived as differences rather than disabilities. However, I have
shown that the notion of disability is already incorporated in the notion of (autistic) disorder
and when there is autistic disorder without disability, there shouldnt be autistic disorder. I
suggest that this proposal of Baron-Cohen, that gives the impression of an emancipation of
people with ASD, results in an illegitimate extension of the boundaries of both autism and
pathology. Individuals who might have certain statistically deviant behavioral characteristics,
but who manage to find the appropriate environment to function without disability or
impairment, shouldnt be considered pathological nor have a psychiatric diagnosis.
Moreover, this contextual approach of pathology emphasizes the importance of actively
creating or finding an appropriate environment for people who experience autistic-like
behavioral or mental problems, instead of focusing exclusively on internal and biological
factors.

The second explanation for the rise of ASD has to do with the rise of the information
society, in which communication, interacting and working together in a globalizing world are
very important aspects. In this information society, it is not very detailed and restricted
expert knowledge, but communication, presentation and a flexible attitude that are important

71
 4A Canguilhemian Perspective on the Concept of Autism

and valuable. Educational methods have equally changed to satisfy changing economical
demands. At school there is a growing emphasis on working together in groups,
communication and presentation. People with autism-like traits, who can perform well in
one restricted area, but are poor in communication and working together, might become
impaired in such an environment. Furthermore, in current (neoliberal) times, in which
companies and institutions are constantly reorganizing (which makes a long-term strategy for
both the companies and the individual employees impossible), progression and change are
important values. In such a working environment, a modern employee should be willing to
change earlier plans and to let gained experience and knowledge go. People who prefer to
stick to certain routines and who are not willing or able to change their restricted activities
and plans are nowadays considered inflexible and rigid. They will experience more problems
and eventually impairment and limitation in these flexibility demanding environments. In
earlier times these people might have been craftsmen or skilled workers who turned to
account their ability to work in solitary for maybe years and years on the same project, while
a different environment with different social and economical demands turns their condition
into a pathological one (Haegens 2009). This idea needs further research, especially on how
working environments have changed the last forty to fifty years and on how this is related to
experiences of suffering and impairment. It appears that the space to creatively choose an
environment, in which the individuals mode of functioning is valued positively by the
individual (an essential part of Canguilhems concept of health), has diminished in current
times. This could to a certain extent explain the current rise of psychiatric pathology.

Something similar seems to be the case for children in current times. Not only employees
need to be flexible, but children as well are expected to do many different things in their so-
called carefree existence. Apart from going to school, they are more and more expected to
develop themselves in different domains such as sports, music, creativity and so forth. These
norms of development are hard to escape from and especially children (compared to adults)
have little freedom and opportunities to ignore the prevailing social norms and to cope in
their own preferable creative way with their environment and the corresponding social
demands. Children are more vulnerable and less capable than adults to resist social changes
that might not be in their benefit. A child is not yet able to create an environment that is
more suitable for his behavioral and psychological conditions and a child does not have the

72
 4A Canguilhemian Perspective on the Concept of Autism

freedom and opportunity to say: I quit school and start living in the country instead of living
in a busy urban area. An interesting paradox arises; the social system a child is living in is for
the child very inflexible and unchangeable, while this same system currently demands a child
to be very flexible and to be active on different domains of life. Because of this tension, I
suggest that it is not surprising that the rise of psychiatric disorders is predominantly seen in
children, with disorders such as ADHD, ASD and child depression. Especially children that
favor stability, routines, little social pressure and that have narrow interests, cannot keep up
and I suggest that this is reflected in the growing number of children with ASD.

These two briefly discussed explanations need further exploration and research, but I suggest
that they have played an important role in the loosening of the diagnostic criteria, in the
emergence of the idea of a whole spectrum of autistic disorders and in the rising rates of
children with ASD. Canguilhems fundamental work on the essence of normality and
pathology has helped and can in the future help to open up discussions on developments in
psychiatry, such as the discussed rise of children with psychiatric disorders and the dominant
biomedical approach of mental problems. Furthermore, Canguilhems work has helped to
open up new possible explanations for psychiatric disorders and as a consequence, new
possible levels of intervention. His main philosophical project has been a vitalist project; to
resist the possibility of reduction, to protect against the temptation of premature satisfaction
of exclusively mechanist and reductionist explanations, and to acknowledge the essential
normativity of both life and science. I hope this thesis made a small contribution to this
wide-ranging project.

73
Bibliography

APA (1980). Diagnostic and Statistical Manual of Mental Disorders, III edn, Washington:
American Psychiatric Association.
APA (1987). Diagnostic and Statistical Manual of Mental Disorders, III-TR edn, Washington:
American Psychiatric.
APA (1994). Diagnostic and Statistical Manual of Mental Disorders, IV edn, Washington:
American Psychiatric Association.
APA (2000). Diagnostic and Statistical Manual of Mental Disorders, IV-TR edn, Washington:
American Psychiatric Association.
Baird, G., Simonoff, E., Pickles, A., Chandler, S., Loucas, T., Meldrum, D., & Charman, T.
(2006). Prevalence of disorders of the autism spectrum in a population cohort of children in
South Thames: the Special Needs and Autism Project (SNAP). The Lancet, 368, 210215.
Baron-Cohen, S. (2000). Is Asperger syndrome/high-functioning autism necessarily a
disability? Development and Psychopathology, 12, 489-500.
Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in Cognitive
Science, 6:248-254.
Benton, E. (1974). Vitalism in Nineteenth-Century Scientific Thought: A Typology and
Reassessment. Studies in the History and Philosophy of Science 5, 17-48.
Blaxill, M. F. (2004). Whats going on? The question of time trends in autism. Public Health
Reports, 119, 536551.
Bolton, D. (2008). What is Mental Disorder? Oxford: University Press.
Burman, E. (2005). Childhood, neo-liberalism and the feminization of education. Gender and
Education, 17, 351-367.
Canguilhem, G. (1965). La Connaissance de la vie. Paris: Vrin.
Canguilhem, G. (1989). The normal and the Pathological (G.R. Fawcett, Trans.; R.S. Cohen, Ed.;
with an introduction by M. Foucault). New York: Zone Books.
Canguilhem, G. (1994). A Vital Rationalist. New York: Zone Books.
Chakrabarti, S., & Fombonne, E. (2001). Pervasive developmental disorders in preschool
children. Journal of the American Medical Association, 285, 30933099.
Chakrabarti S. & Fombonne E. (2005). Pervasive developmental disorders in preschool
children: high prevalence confirmed. American Journal of Psychiatry 162, 11331141.

74
 Bibliography

Chimisso, C. (2003). The tribunal of philosophy and its norms: history and philosophy in
Georges Canguilhems historical epistemology. Studies in History and Philosophy of Biological and
Biomedical Sciences 34, 297-327.
Coo, H., Ouellette-Kuntz, H., Lloyd, J.E.V., Kasmara, L., Holden, J.J.A., Lewis, M.E.S.
(2008). Trends in Autism Prevalence: Diagnostic Substitution Revisited. J Autism Dev Disord,
38:1036-1046.
Deer, B. (2008). The MMR-autism crisis-our story so far. The Sunday Times of London February
22 2004 updated August 10, 2008. Retrieved from http://briandeer.com/mmr/lancet-
summary.htm.
Di Martino A., Ross K., Uddin L.Q., Sklar A.B., Castellanos F.X., Milham M.P. (2009).
Functional brain correlates of social and nonsocial processes in autism spectrum disorders:
an activation likelihood estimation meta-analysis. Biol Psychiatry, 1;65(1):63-74.
Fombonne, E. (2005). The Changing Epidemiology of Autism. Journal of Applied Research in
Intellectual Disabilities 18, 281294.
Fombonne E, Zakarian R, Bennett A, Meng LY, McLean-Heywood D. (2006). Pervasive
developmental disorders in Montreal, Quebec, Canada: prevalence and links with
immunizations. Pediatrics 118:E139E150
Freitag CM (2007). The genetics of autistic disorders and its clinical relevance: a review of
the literature. Mol Psychiatry 12:222
Gernsbacher, M.A., Dawson, M., Goldsmith, H.H. (2005). Three Reasons Not to Believe in
an Autism Epidemic. Current Directions in Psychological Science, 14-2: 55-58.
Gordo-Lopez, A. & Burman, E. (2004). Emotional capital and information technologies in
the changing rhetoric around children and childhoods. New Directions in Child Development,
105, 63-80.
Greco, M. (2005). On the Vitality of Vitalism. Theory, Culture and Society 22 (1), 15-27.
Gutting, G. (1989). Michel Foucaults Archaeology of Scientific Reason. Cambridge: University
Press.
Gutting, G. (2005). Continental Philosophy of Science. Oxford: Blackwell Publishing.
Hacking, I. (1992). Style for historians and philosophers. Studies In History and Philosophy of
Science. 23-1.
Haegens, K. (2009). Toon me uw gekken. De Groene Amsterdammer, 133 (9), 22-25.

75
 Bibliography

Hertogh, C.M.P.M. (1986). Bachelard en Canguilhem: Epistemologische discontinuteit en het medisch

normbegrip. Amsterdam: VU Uitgeverij.
Honda, H., Shimizu, Y., & Rutter, M. (2005). No effect of MMR withdrawal on the
incidence of autism: a total population study. Journal of Child Psychology and Psychiatry, 46, 572
579.
Hornig, M., Briese, T., Buie, T., Bauman, M. L., Lauwers, G., Siemetski, U., et al. (2008).
Lack of association between measles virus vaccine and autism with enteropathy: A case-
control study. PLoS ONE 3(9): e3140 Retrieved from
http://www.plosone.org/article/info%3Adoi%2F10.1371% 2Fjournal.pone.0003140.
Horowitz, A.V., Wakefield, J.C. (2007). The Loss of Sadness: How Psychiatry transformed normal
sorrow into depressive disorder. New York: Oxford University Press.
Kanner L. (1943). Autistic disturbances of affective contact. Nervous Child 2:217250.
Kanner L, Eisenberg L. (1956). Early infantile autism 19431955. Am J Orthopsychiatry 26:
5565.
Kim HG, Kishikawa S, Higgins AW, Seong IS, Donovan DJ, Shen Y, Lally E, Weiss LA,
Najm J, Kutsche K, Descartes M, Holt L, Braddock S, Troxell R, Kaplan L, Volkmar FR,
Klin A, Tsatsanis K, Harris DJ, Noens I, Pauls DL, Daly MJ, MacDonald ME, Morton CC,
Quade BJ Gusella JF. (2008). Disruption of neurexin 1 associated with autism spectrum
disorder. Am J Hum Genet 82:199207
Le Couteur A, Bailey A, Goode S, Pickles A, Robertson S, Gottesman I, Rutter M. (1996). A
broader phenotype of autism: the clinical spectrum in twins. J Child Psychol Psychiatry 37:785
801.
Lord C, Corsello C (2005). Diagnostic instruments in autistic spectrum disorders. In:
Volkmar FR, Paul R, Klin A, Cohen DJ (eds) Handbook of autism and pervasive
developmental disorders, 3rd edn. Wiley, Hoboken, NJ
Magree, V. (2003). Normal and Abnormal: Georges Canguilhem and the Question of Mental
Pathology. Philosophy, Psychiatry and Psychology 2 (1), 1-13.
Mol, A. (1998). Lived reality and the multiplicity of norms: a critical tribute to George
Canguilhem. Economy and Society 27 (2-3), 274-284.
Muhle R, Trentacoste SV, Rapin I (2004). The genetics of autism. Pediatrics 113:e472e486
New York Times, (2002). October 20, section 4: 10

76
 Bibliography

Rogers, E. J. (2008). Has enhanced folate status during pregnancy altered natural selection
and possibly Autism prevalence? A closer look at a possible link. Medical Hypotheses, 71, 406
410.
Rose, N. (1998). Life, reason and history: reading Georges Canguilhem today. Economy and
Society 27 (2-3), 154-170.
Rose, N. (2007). The Politics of Life Itself. Princeton: University Press.
Rutter M. (1978). Diagnosis and definition. In Rutter M, Schopler E, editors. Autism: a
reappraisal of concepts and treatment. New York: Plenum Press. p 125.
Steyaert, J.G. & De La Marche, W. (2008). Whats new in autism? Eur J Pediatr, 167:1091-
1101.
Tan, M., & Parkin, J. E. (2000). Route of decomposition of thiomersal (Thimerosal).
International Journal of Pharmacology, 208, 2334.
Theoharides, T. C., Doyle, R., Francis, K., Conti, P., & Kalogeromitros, D. (2008). Novel
therapeutic targets for autism. Trends in Pharmacological Sciences, 29, 375382.
Time (2002). May 6: 48
Timimi, S. (2004). Diagnosis of autism: current epidemic has social context. British Medical
Journal, 328: 226.
Timimi, S. (2008). Childrens mental health and the global market: an ecological analysis. In
Liberatory Psychiatry, ed. C.I. Cohen & S.Timimi. Cambridge: University Press.
Trnka, P. (2003). Subjectivity and Values in Medicine: The Case of Canguilhem. Journal of
Medicine and Philosophy 28 (4), 427-446.
Wakefield AJ, Murch SH, Anthony A, et al. (1998). Ileal-lymphoid-nodular hyperplasia, non-
specific colitis and pervasive developmental disorder in children. Lancet 351: 637641
Waterhouse, L. (2008). Autism Overflows: Increasing Prevalence and Proliferating Theories.
Neuropsychol Rev, 18:273-286
Wazana, A., Bresnahan, M., Kline, J. (2007). The Autism Epidemic: Fact or Artifact? J. Am
Acad. Child Adolesc. Psychiatry, 46:6, 721-730.
WHO (World Health Organization) (1992). International classification of diseases (10th ed.).
Geneva, Switzerland: World Health Organization.
Wing L., Yeates S. R., Brierly L. M. & Gould J. (1976). The prevalence of early childhood
autism: comparison of administrative and epidemiological studies. Psychological Medicine 6,
89100.

77
 Bibliography

Wing L. & Gould J. (1979). Severe impairments of social interactions and associated
abnormalities in children: epidemiology and classification. Journal of Autism and Developmental
Disorders 9, 1129.
Wing, L., & Potter, D. (2002). The epidemiology of autistic spectrum disorders: Is the
prevalence rising? Mental Retardation and Developmental Disabilities Research Reviews, 8, 151162.
Wuketits, F.M. (1989). Organisms, Vital Forces, and Machines: Classical Controversies and
the Contemporary Discussion Holism vs. Reductionism, in P. Hoyningen-Huene and
F.M. Wuketits (eds) Reductionism and Systems Theory in the Life Sciences. Dordrecht: Kluwer
Academic Publishers.
Yeargin-Allsopp M, Rice C, Karapurkar T, Doernberg N, Boyle C, Murphy C (2003).
Prevalence of autism in a US metropolitan area. JAMA 289:4955

78
Appendix
DSM-IV Diagnostic Criteria for Autism Spectrum Disorders

Diagnostic Criteria for Autistic Disorder

A. A total of six (or more) items from (1), (2), and (3), with at least two from (1), and
one each from (2) and (3):
1. qualitative impairment in social interaction, as manifested by at least two of
the following:
a. marked impairment in the use of multiple nonverbal behaviors such
as eye-to-eye gaze, facial expression, body postures, and gestures
to regulate social interaction
b. failure to develop peer relationships appropriate to developmental
level
c. a lack of spontaneous seeking to share enjoyment, interests, or
achievements with other people (e.g., by a lack of showing,
bringing, or pointing out objects of interest)
d. lack of social or emotional reciprocity
2. qualitative impairments in communication as manifested by at least one of
the following:
a. delay in, or total lack of, the development of spoken language (not
accompanied by an attempt to compensate through alternative
modes of communication such as gesture or mime)
b. in individuals with adequate speech, marked impairment in the
ability to initiate or sustain a conversation with others
c. stereotyped and repetitive use of language or idiosyncratic language
d. lack of varied, spontaneous make-believe play or social imitative
play appropriate to developmental level
3. restricted repetitive and stereotyped patterns of behavior, interests, and
activities, as manifested by at least one of the following:
a. encompassing preoccupation with one or more stereotyped and
restricted patterns of interest that is abnormal either in intensity or
focus
b. apparently inflexible adherence to specific, nonfunctional routines
or rituals
c. stereotyped and repetitive motor manners (e.g., hand or finger
flapping or twisting, or complex whole-body movements)
d. persistent preoccupation with parts of objects
B. Delays or abnormal functioning in at least one of the following areas, with onset
prior to age 3 years: (1) social interaction, (2) language as used in social
communication, or (3) symbolic or imaginative play.
C. The disturbance is not better accounted for by Retts Disorder or Childhood
Disintegrative Disorder.

79
 DSM-IV Diagnostic Criteria for Autism Spectrum Disorders

Diagnostic Criteria for Asperger's Disorder

A. Qualitative impairment in social interaction, as manifested by at least two of the

following:
1. marked impairment in the use of multiple nonverbal behaviors such as eye-
to eye gaze, facial expression, body postures, and gestures to regulate social
interaction
2. failure to develop peer relationships appropriate to developmental level
3. a lack of spontaneous seeking to share enjoyment, interests, or achievements
with other people (e.g., by a lack of showing, bringing, or pointing out
objects of interest to other people)
4. lack of social or emotional reciprocity
B. Restricted repetitive and stereotyped patterns of behavior, interests and activities, as
manifested by at least one of the following:
1. encompassing preoccupation with one or more stereotyped and restricted
patterns of interest that is abnormal either in intensity of focus
2. apparently inflexible adherence to specific, nonfunctional routines or rituals
3. stereotyped and repetitive motor mannerisms (e.g., hand or finger flapping
or twisting, or complex whole-body movements)
4. persistent preoccupation with parts of objects
C. The disturbance causes clinically significant impairment in social, occupational, or
other important areas of functioning.
D. There is no clinically significant general delay in language (e.g., single words used by
age 2 years, communicative phrases used by age 3 years).
E. There is no clinically significant delay in cognitive development or in the
development of age-appropriate self-help skills, adaptive behavior (other than in
social interaction), and curiosity about the environment in childhood.
F. Criteria are not met for another specific Pervasive Developmental Disorder or
Schizophrenia.

Pervasive Developmental Disorder Not Otherwise Specified (Including Atypical

Autism)

This category should be used when there is a severe and pervasive impairment in the
development of reciprocal social interaction associated with impairment in either verbal or
nonverbal communication skills or with the presence of stereotyped behavior, interests, and
activities, but the criteria are not met for a specific Pervasive Developmental Disorder,
Schizophrenia, Schizotypal Personality Disorder, or Avoidant Personality Disorder. For
example, this category includes "atypical autism" - presentations that do not meet the criteria
for Autistic Disorder because of late age at onset, atypical symptomatology, or subthreshold
symptomatology, or all of these.

Source: APA (2000)

80