Vous êtes sur la page 1sur 187

Clarity and Confusion in Social Theory

Philosophy and Method in the
Social Sciences
Series Editor: Phil Hutchinson, Manchester Metropolitan University, UK
Engaging with the recent resurgence of interest in methodological and
philosophical issues in the human and social sciences, this series provides an
outlet for work that demonstrates both the intellectual import of philosophical
and methodological debates within the social sciences and their direct relevance
to questions of politics, ethics or policy. Philosophy and Method in the Social
Sciences welcomes work from sociologists, geographers, philosophers,
anthropologists, criminologists and political scientists with broad interest across
academic disciplines, that scrutinises contemporary perspectives within the
human and social sciences and explores their import for today’s social questions.
Other titles in the series:
Wittgenstein among the Sciences
Wittgensteinian Investigations into the ‘Scientific Method’
Rupert Read, Edited by Simon Summers

UK .Clarity and Confusion in Social Theory Taking Concepts Seriously Leonidas Tsilipakos University of Edinburgh.

2.ISBN 9781-4724-3242-1 (epub) 1.(Philosophy and method in the social sciences) Includes bibliographical references and index. HM585. DT1 1HD . The Library of Congress has cataloged the printed edition as follows: Tsilipakos. electronic. photocopying.T76 2015 301. to be identified as the author of this work. Sociology--Philosophy. 1988. Social sciences--Philosophy. Title. Dorchester. Leonidas Tsilipakos has asserted his right under the Copyright. No part of this publication may be reproduced. stored in a retrieval system or transmitted in any form or by any means. Designs and Patents Act. mechanical.com British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library.ashgate. recording or otherwise without the prior permission of the publisher.ISBN 978-1-4724-3241-4 (ebook) -.© Leonidas Tsilipakos 2015 All rights reserved. GU9 7PT USA England www. at the Dorset Press.01--dc23 2014029208 ISBN: 978-1-4724-3240-7 (hbk) ISBN: 978-1-4724-3241-4 (ebk – PDF) ISBN: 978-1-4724-3242-1 (ebk – ePUB) Printed in the United Kingdom by Henry Ling Limited. -. pages cm. Published by Ashgate Publishing Limited Ashgate Publishing Company Wey Court East 110 Cherry Street Union Road Suite 3-1 Farnham Burlington. I. Clarity and confusion in social theory : taking concepts seriously / by Leonidas Tsilipakos. VT 05401-3818 Surrey. Leonidas. ISBN 978-1-4724-3240-7 (hardback) -.

for Alexandra and Wes .

This page has been left blank intentionally .

Contents Preface   List of Abbreviations   ix xi Introduction   1 1 Sociology and Ordinary Language Philosophy: A Dialogue   9 2 What Achilles said to Hempel: Unrealistic Restrictions and the Idea of Explanation as a Matter of Form   35 3 Ordinary Language Philosophy as Logic   57 4 Ontological Confusion in Social Theory   69 5 Continuing Ontological Confusion: The Constitution and Application of an Ontological Scheme   99 6 Some Problems with ‘The Problem of Structure and Agency’   117 Conclusion   149 Postscript   References   Index   151 159 171 .

This page has been left blank intentionally .

If that were the case one could hope that studies might eventually reveal that this. But so is to offer a theoretical programme that will allegedly transform them into genuine sciences or into forms of enquiry that really matter. it is ultimately addressed to social scientists and to philosophers in the belief that they will all recognise the issues discussed herein as fundamental to the study of human beings living together and thus be able to relate them to their particular discipline and concerns with relative ease. This book is an attempt to demonstrate the force of that ‘appropriately’. I have incurred numerous intellectual as well as other kinds of debts. Accordingly. I will try to provide partial – though detailed – evidence for a simple proposition. The situation. that or even most doctrines do not accord with the facts and need to be replaced by ones that do. is not that it contains a number of empirically false doctrines. attempts at reconstruction end up reproducing the same problems as they are based on the same procedures and attitudes that are responsible for the logical incoherence.Preface To say that the social sciences are in a state of crisis is a worn-out cliché. especially as the latter pertains to sociology. In doing so it will make contrastive use of a revolutionary philosophical tradition which places concepts and the logic of language at centre stage and which can help us cultivate our logical sensibilities in a way that will open up alternative ways of proceeding. problems having to do with its relation to the concepts that constitute its subject matter. Michael Mair. I am exceptionally lucky to have found Wes who is a man of the rarest qualities. both academic and personal: his curiosity and understanding. My effort and attention will focus on social theory. The greatest debt of all I owe to Wes Sharrock who has guided me through my postgraduate studies over a period of five years and who continues to provide support and inspiration. The problem with social theory. Social theory is in logical disarray: it is most centrally plagued by problems of reasoning and sense. Sivamohan Valluvan and Christian Greiffenhagen have been immensely valuable intellectual companions and friends. his humility and kindness are all unmatched. In this book. namely that the social sciences need to start taking concepts seriously. Moreover. is largely different. . not least because it began life as a doctoral thesis. what one may hope for is that problems of logic will be recognised for what they are and given serious attention and that flawed procedures will be abandoned and the concepts involved will be handled appropriately. as I see it. instead of embracing either cliché. I would like to suggest. Despite the fact that the present book deals with social theory. In the course of writing this book. They have provided.

Luke Yates has been an exceptional conversationalist. Chris Elsey. Philosophy of the Social Sciences. I owe a great deal to Phil Hutchinson. the editor of this series. doi: 10. who has shown extraordinary trust in my abilities and has repeatedly given me the opportunity to take part in stimulating seminars and conferences.1177/0048393112461055. Patrick Watson. Steve Kemp has shown exceptional patience as well as open-mindedness in dealing with my thoughts and concerns. while also featuring a small fraction of ‘Realist social theory and its losing battle with concepts’.1177/0048393114525859. I should like to say. Edinburgh June 2014 . Martin Cleary. I hope they will forgive me. Responsibility for the book’s errors. of which Chapter 6 is an expanded version. Chapter 5 is reworked from ‘Theoretical procedures and Elder-Vass’s critical realist ontology’. My partner. especially Irmak KarademirHazir. I am deeply grateful to my parents Marianthi and Christos and to my brother Odysseas for their unwavering support throughout the years. is in an altogether different logical category. Steve Kemp. Paul Smith. Alexandra. Nikolia Kartalou. I have been fortunate enough to receive encouragement and support from various other colleagues at Manchester and Edinburgh. Bethan Harries and Heide Weishaar. however. doi: 10. Journal for the Theory of Social Behaviour. Michael Mair and Marianthi Makri-Tsilipakou have read parts of the manuscript and offered comments and suggestions for improvements. For the latter I would also like to thank Rupert Read as for his willingness to supply reference letters at short notice. Cristina Delgado and Rowann Bowcutt. as well as from my friends Alexandros Veloudis. Philosophy of the Social Sciences. I am grateful to them both for their generosity. I am deeply grateful to them both for their support and confidence in me. Thank you also to James Nazroo. Néstor Saiz. For his virtuous ways and for his benign overall guidance I am most grateful. Stephen Motrioni and to all those I may have neglected to mention. Isabelle Darmon and Carlos Frade’s reading group and long dinners have been sources of extreme pleasure and sometimes of frustration. judicious advice as well as numerous incisive observations on a plethora of subjects.T. Last but not least. 44(2): 201–19. I have benefited greatly from our not only disciplined but also caring discussions on numerous sociological questions and texts. Jeff Coulter. L. This book would not have been possible without them. I would like to thank publishers and journals for permission to use the following previously published material: Chapter 4 is based on ‘The poverty of ontological reasoning’.x Clarity and Confusion in Social Theory each in their own particular yet equally generous way. lies exclusively with me. Wes Sharrock.

List of Abbreviations CCR – Corporate Criminal Responsibility CR – Critical Realism CSR – Corporate Social Responsibility MC – Main Contention OLP – Ordinary Language Philosophy SC – Social Constructionism Works by Wittgenstein OC – On Certainty PI – Philosophical Investigations Z – Zettel BT – The Big Typescript: TS 213 .

This page has been left blank intentionally .

I will not produce a concise panorama of the occurrence of this pattern across different kinds of social theory. My strategy in order to accomplish this task will be to gradually build a detailed picture of the confused pattern of relating to concepts through the examination – at the necessary depth – of recent theoretical work and debate. grief (or ‘the emotions’). Similarly. time or the difference between our everyday concept of time and the one physics makes use of. on the other. this confusion is a form of logical incoherence. In all these reflective exercises. what do I mean by ‘theoretical attitude’? Those familiar with the invocation of various ‘attitudes’ within phenomenology may have already thought of the phrase as possibly contrasting to ‘the natural attitude’. filling in a tax form. which to be sure can be geared to 1 I am paraphrasing Sören Stenlund’s statement about philosophy: ‘the belief in the traditional ideals of metaphysical philosophy is basically an attitude towards a philosophical vocabulary. in the first instance.Introduction Doing social theory is primarily assuming a theoretical attitude towards language and only derivatively vindicating any proposition about social phenomena. Husserl came to develop his doctrines and methods by taking his lead from René Descartes and his radical method of doubt. it is worth noting that doubt-in-thenatural attitude encompasses both lay and technical activities – including science – as both employ a localised and occasioned manner of doubting thus contrasting sharply to the generalised doubting under the philosophical attitude (Sharrock and Anderson 1991). asking someone what time it is or calculating the time it takes for light to travel from the sun to the earth and. in short. Although this dimension is not in question in this book. I will claim. hence doubt has served as a dimension on which to distinguish the natural attitude from Descartes’ philosophical one. between. to the very idea of social theory. which Edmund Husserl used for the way we relate to the world in our everyday experience. not empirical falsity. that both the theoretical attitude and the resulting confused pattern are tied not to particular theoretical schools but to central social theoretical problems and procedures. reflecting on the concepts of the state. on the one hand.1 In this book I will try to demonstrate that adopting a theoretical attitude leads theorists into a confused relationship to the concepts that interest them and thus to their subject matter. by speaking of a theoretical attitude towards language I do not mean to draw a contrast between a natural attitude towards language and a conceivably technical one but rather. between both and second-order activities of reflecting on our concepts: for example. describing a friend as in grief. however. Then. . and only secondarily the vindication of some philosophical doctrine expressed in this vocabulary’ (1996: 197).

‘thought is deliberation with oneself in silence’ we press ourselves to recall how we speak in various situations we will see that ‘thinking about’ is connected to ‘deliberating’ and ‘pondering’. which forms the topic of this book. involves.2 Clarity and Confusion in Social Theory many a purpose across a number of disciplines. Ludwig Wittgenstein described the situation elegantly when he remarked: ‘One learns the word “think”. i. that would render responding to the resulting reflectiveabstractive attitude especially tricky. we may come to relate to them in ways that generate puzzlement precisely because we have severed the concepts from the practical circumstances in which they have their home. thinking of the good times we had and thinking that the film we watched last night was rubbish. ‘thinking about’ and ‘thinking that’. further. the various concepts we might say that they are used to express. as I will try to show. while we learn to use related concepts and groups thereof by doing ‘we do not learn ambulando how to coordinate team with team’ (1954: 124). a number of notable features: it involves the restriction of our conceptual resources to a small number of terms. This attitude contrasts sharply with the practical attitude of situated language use. If. however. yet once we pose the reflective question of. instead of year of publication and page number. we speak in our everyday life of thinking about what we should do next. its use. To take the example of ‘thinking’. and ‘thinking that’ to ‘believing’ and ‘being mistaken’. under certain circumstances. yet even if we identify the question as being about the concept of thought or thinking. memory or belief we find ourselves in the midst of puzzlement. The example of ‘thinking’ has been meant as an illustration of the kind of relation to our concepts we tend to assume under the reflective-abstractive or. Nor for that matter can we readily resort to an overview of the various uses of these expressions. however. how thought relates to deliberation. One obvious source of our difficulties is that it is far from clear what the question is asking. in short. which. The adoption of the theoretical attitude in social theory. to a theoretical vocabulary (or conceptual framework) which is adhered to and treated 2 References to Wittgenstein’s works throughout this book will display the abbreviated title and paragraph number. . Moreover.e. we may come to lose contact with our concepts as we abstract from the contextual features that are part of their use. or of the various relations between concepts.2 Not that he thought this presented any sort of practical problem but only that it would easily turn reflection into an abstracted exercise and. one does not learn to describe’ (Z: §114). our main response is to look for an abstracted theoretical scheme. the concepts expressed are connected in tortuous ways to notions such as the mind. as is conventional in the literature. In other words. for example. the brain. for example. human beings and animals. ‘thinking of’ to ‘remembering’ and ‘imagining’. it does not come natural to us to recount the contextual features that go together with forms of words such as ‘thinking of’. theoretical attitude. As Gilbert Ryle observed in a fashion similar to Wittgenstein. instead of reaching for an abstract formula such as.

They are the appeals to ‘the need to conceptualise’ and ‘the need to theorise’ which furnish self-evident grounds for introducing a ‘technical’ vocabulary which. ‘may lead to context-sensitive. both the commitment to a restricted vocabulary and the exclusion of other resources are tied to a number of dubious conceptual procedures. in other words. With regard to the restricted vocabulary. thus furnishing grounds for their dismissal while. To appreciate how the theoretical attitude is embedded in social theoretical practice it might be useful to briefly consider some of Nicos Mouzelis’s thoughts in the context of his diagnosis of what went wrong with sociological theory3 (1995). it is the former that should be at the centre of our preoccupation since the potential problems applying to generalities III. however. a distinction he marks with Althusser’s terms generalities II and III. it is hoped. any clear specification of the sense in which these terms can be said to apply to various cases. crucially. This restriction has a necessary corollary which justifies and is justified by it. 3 In this book I am concerned with social theory as it pertains to sociological concerns. I will argue that interpretative procedures under the theoretical attitude are responsible for the systematic misunderstanding of our concepts. I will show that certain conceptual procedures are made use of in order to fix the meaning of the chosen set of terms so as to guarantee their uniform application across circumstances – without. and that they can be taken to be so without any careful examination whatsoever. exacerbating conceptual confusion. Mouzelis is not alone of course in placing the construction of conceptual tools at the beginning of sociological investigation. that is. The adoption of the theoretical attitude and accompanying procedures in social theory leads to serious problems in reasoning and sense or. do not apply to generalities II. however. namely the presumption (which is false precisely because it is such) that available conceptual resources and the understanding they might afford are insufficient. their ‘aim being to provide us with conceptual tools for asking interesting questions and preparing the ground for the empirical investigation of the social world’ (1995: 42) which. in contrast to generalities III. According to Mouzelis. that as universal generalisations they are either trivial or wrong. There are two pieces of related reasoning one finds in abundance in sociology which point in the same direction and which are usually employed as argumentative bedrock. historicallyoriented comparative investigations’ (1995: 2). As regards the exclusion of other resources. I am thus using ‘sociological theory’ and ‘social theory’ interchangeably.Introduction 3 as indispensable for the formulation of problems and their solution as well as for the adequate explanation of social phenomena. Both ‘needs’. . to logical incoherence. namely the simultaneous exclusion (and sometimes complete dismissal) of other available conceptual resources. As I propose to demonstrate in detail in this book. if not irrelevant. are based on unwarranted assumption. Mouzelis thinks it is imperative that we distinguish between theory in the sense of conceptual framework and theory in the sense of substantive propositions. will provide understanding of what it is aimed at.

which is understanding our existing concepts when obstructed by the theoretical attitude. to be sure. Because of its recognisably central role in philosophical puzzlement. is already to smuggle in the assumption that the only understanding that matters is the outcome of sociological enquiry and therefore to merely reiterate rather than provide support for the conviction that in the absence of any of the latter. Ordinary . it has come under sustained and forceful attack by the school of thought of Ordinary Language Philosophy which. first. Moreover. insisting on the introduction of technical concepts is misleading as it obscures the real problem. Although the theoretical attitude towards language has been a central part of the social sciences since their development in the nineteenth century. by removing the exclusion and.L. Others will point to the sociological commonplace that ordinary concepts are unsuited for precise sociological uses. This. But such commonplace sweeping condemnations of ordinary concepts are themselves no more secure than what they are invoked to secure. Some will be quick to point to the paucity of sociological literature on whatever subject matter is in question as what really establishes both ‘needs’. Indeed. any of the former is also to be absent.4 Clarity and Confusion in Social Theory There are various ways. While it may be true that puzzlement establishes a need to do something about it. Yet others will cite the fact of their puzzlement and protest that something needs to be done about it. however. far from taking us some way in that direction and far from being enabling. will be understood quite broadly as consisting of Ludwig Wittgenstein. Austin and their followers. it originates in philosophy. historically. The point I wish to make by arguing in this manner is neither that our ordinary concepts are always adequate nor that our understanding might never need supplementing but rather. I will try to show that the special vocabulary social theorists are keen to introduce. The setting up of the theoretical machinery for description and explanation will be found to be the simultaneous setting up of machinery for puzzlement and confusion. For it will also be seen throughout this book that in many cases puzzlement is in fact generated only after our ordinary understanding and the resources feeding into it have been excluded and that. as do the social sciences themselves. as importantly. that before concepts can be dismissed or improved upon they need to be considered and properly understood. tends to alienate us from our concepts and thus ends up being profoundly disabling. that the question of adequacy is not to be debated once and for all. being either vague or ideologically tainted. instead. as will be seen throughout this book. in which such a bastion of sociological reasoning is upheld. Equally. Gilbert Ryle. we are actually able to dissolve puzzlement. for our present purposes. not only is the need for technical concepts manufactured and unsubstantiated but. J. being tied instead to particular concepts and particular goals and purposes and. it is to restate rather than to justify the idea that available conceptual resources are inadequate. it is also true that correctly responding to puzzlement presupposes a correct diagnosis of its sources. they are extremely debatable. second. As will also emerge by the end of this book. and that any alleged adequacy of available conceptual resources is belied by the very fact of puzzlement. cultivating our sensibility in the use of such resources.

however. It is because so much of social science gets the latter wrong that Winch’s arguments are as pertinent as ever. Lynch 2000. OLP. is deeply mistaken (cf. methods and animating worries. rather than creating any tools of its own it aims to liberate and put to judicious use the conceptual tools which are part of our linguistic competence. If they do they will find that many of the arguments that I make herein are identical with or constitute explications or elaborations of arguments that have been made therein. By virtue of this fact. Read 2012: 100. namely his insights on understanding as expressed in the concepts a society employs. it also constitutes an explication of OLP and an attempt to show the pertinence of its methods to social science. that Winch’s arguments are still pertinent is not to say that they are necessarily so with regard to everything that goes on within collections of a diverse set of concerns. For it will be said. The idea then that because Winch attacked an obsolete picture of science in some chapters of his book. on the relationship between philosophy and social science. I hope to defend Winch indirectly. as are the social sciences and even ‘single disciplines’ such as sociology. however. in the sense that I will attempt to show how much purchase his central insights still have. however. Winch’s arguments do not depend on attacking the precise picture of science social theorists resort to but rather on getting right what it takes not to misunderstand our concepts. To say. I believe that none of the above can withstand any scrutiny though I will not here undertake anything resembling a proper defence of Winch. on the latter’s conflation of empirical and conceptual matters and on the embeddedness of logic in social activities. (2013) both of which readers may wish to consult in conjunction with the present book. For that I defer to Hutchinson et al. current social theory has developed immunity to Winch either by not being scientistic or by adhering to an updated picture of science. was it not Peter Winch who argued that the philosophy of Wittgenstein has important implications for the social sciences and who even called a part of sociology ‘misbegotten epistemology’ (2008[1958]: 41)? Precisely because Winch has been debated ad nauseam it is easy to assume that what he had to say is well understood and that relevant bits have been either incorporated in standard social theoretical approaches or rightly rejected as false. It is only to .Introduction 5 Language Philosophy (OLP) constitutes a powerful antidote to the theoretical attitude which it counters with detailed descriptions of language use. Gunnell 2011: 105–28) and will be extensively demonstrated to be so throughout this book. Speaking of OLP’s methods should not lead us to expect that it can furnish us with a new set of special tools. purposes. The name of Wittgenstein especially may have evoked a feeling of déjà vu. I imagine that talk of OLP may have already spurred standard reactions among some readers. By clarifying OLP as a form of elucidation of the logic of language and applying it to social theory. the present book has two aspects to it: apart from being a radical critique of the theoretical attitude as the latter displays itself in the conceptual procedures employed in social theory. is frequently misunderstood by both philosophers and social scientists. (2008) and the unpublished Sharrock et al.

to prepare the ground for the application of OLP to social theory. address possible misunderstandings and provide necessary background arguments. this book. namely the problem of scientific explanation as it originates in Carl Hempel’s work. intelligibility and reality vary by the context of use (2008[1958]: 17ff. by extension. Summary The first three chapters of the book are designed to explicate my chosen way of proceeding. The stage is set for Chapters 4 to 6 in which conceptual elucidation is employed in order to shed light on various sources of confusion at the core of social theory. explicating its relation to sociological forms of enquiry and defending it against standard objections. I discuss in detail a prominent case of such a problem. The fact that all academic activities are nowadays subjected to the same managerial-speak has imposed a common language across disciplines. I have chosen to present the argument in the form of a dialogue between two sociologists. that the concepts of research. a language which is much better suited to the natural sciences than to the social sciences or the humanities. Given. following Winch’s point regarding how the concepts of understanding. data and theory too are to be understood very differently in connection with the social sciences. Before I proceed to give a brief summary of each chapter I should like to point to a different kind of thing that stands in the way of understanding Winch and.). The discussion of Hempel’s problem will also facilitate the transition from the transcendental understanding of logic he evinces to the understanding of logic as embedded in social life and to the ways in which logical features of the concepts constitutive of social life can be explicated. These ways are illustrated in Chapter 3 which also aims to clarify the senses in which OLP can be thought of as logic. In Chapter 2 I turn to locating the theoretical attitude and the confused relationship to our concepts in. the premium placed on theory by the idea that whatever else research needs to be it needs to be ‘theoretically informed’ the implications can be seen to also reach the empirical recesses of social science. in short. falsification of what the social sciences are and can be like has already taken place. In order to ease the reader into the book’s problematic. I should like to stress then. . knowledge.6 Clarity and Confusion in Social Theory say that since his arguments address the fundamental issue of the relation of social scientists to the concepts that constitute their subject matter. I begin in Chapter 1 by introducing OLP. ‘unrealistically restricted’ problems. they can have wide ranging and profound implications. In describing what all academics do as research which uses theory on empirical data to produce knowledge and which should then make some kind of impact under various ‘knowledge exchange’ activities. what I will argue are. This book will only concern itself with the implications they have for social theory. nevertheless. namely institutionally induced misunderstanding of the distinctive features of the social sciences. which has proved extremely influential regarding misconceptions of explanation in the social sciences and philosophy.

the real targets of these chapters are generic conceptual procedures that transcend any specific theoretical paradigm. The chapter buttresses the contention that social ontological projects constitutionally rely on incoherent methods. Moreover. who asserts that social structures may possess exclusive causal powers. Chapter 4 deals with a currently fashionable source of confusion. are examined and shown to be seriously flawed. such as the use of distortive models. including explanation. The focus is on how Elder-Vass’s ontological scheme (comprising entities.Introduction 7 Instances of theoretical reasoning and debate are discussed in considerable detail. It is contended that the ontological scheme is not and could not be grounded in science. Chapter 5 continues the attack on ontology and its constitutive procedures by scrutinising one of the latest versions of Critical Realism. the conflation of empirical and conceptual issues. and the questionable use of examples. A number of conceptual procedures that are employed by ontological projects. the reality of social structures and their relation to individual action. causality. this is necessary in order to combat the frequently used summary formulations of complex problems that end up obscuring and perpetuating conceptual muddles. will enable us to understand the various sources of confusion and. ontology loses its intelligibility. once this picture is exposed as incoherent. stems from the selfsame procedures. . Chapter 6 provides a culmination of the argument of this book by addressing ‘the problem of structure and agency’. namely ontology. The chapter concludes by offering ways of moving beyond ‘the problem of structure and agency’ and the conceptual confusion it gives rise to. This is argued via the detailed examination of a solution to the problem recently proposed by Elder-Vass. Despite detailed discussion and critique of specific theorists. however. it is argued that ontological projects are mainly fuelled by a misconceived picture of language and that. it is argued. relations between their parts and emergent properties) is constituted and on the methods through which it is applied to society. This claim is dissected and ensuing discussion of its components shows how they are tied to logically deficient procedures. Ontological questions have proliferated in social thought in the past decades mainly as a way of recasting traditional sociological questions. Far from constituting an advance in our understanding. by finding our way about. which are. however. The claim put forth is that both ‘the problem’ and the procedures typically brought into play in order to solve it are themselves deeply problematic. observation. and that its use distorts rather than illuminates what it is applied to. this form of reasoning has frequently brought confusion. Attention is also paid to the debate Elder-Vass’s work has sparked and its irresolvable character which. it is hoped. Finally. In more detail. the one propounded by Dave Elder-Vass who promises to place the social sciences on a sound ontological footing. approached from different starting points in criss-cross fashion. This way of approaching the issues. This is demonstrated with detailed reference to a recent debate centred on Rom Harré’s attempt to tackle the issue of whether social structures are real. More generally. to navigate an exit into clarity. including Critical Realism. Chapters 4 to 6 can be said to deal with the same compacted set of central social theoretical issues.

In this way. I can imagine that some social scientists might find it convenient to dub the book ‘a critique of Critical Realism’ and to say that the social theory they rely on or practice themselves is not susceptible to conceptual confusion. not in the sense of having been given a special apparatus to use but in the sense of having had our conceptual sensibilities cultivated and our linguistic abilities pointed in a certain direction. let me make a final cautionary remark. Given that social theory of the critical realist kind features prominently in this book. To anyone prone to readily respond in this way I should like to say: de te fabula narratur.8 Clarity and Confusion in Social Theory By the time we have reached the end of this book we will be in a position to understand the nature of social theoretical puzzlement and the logical dangers associated with the adoption of the theoretical attitude towards language. Before we begin our journey. we will be able to restore the connection to our subject matter and our capacity for understanding to its full potential. . such confusion being confined to Critical Realism or to social theories with explicit scientific ambitions. We will also be equipped with the means of countering the confusion the latter generates.

Moreover. are more susceptible to them since there is not much by way of application of OLP within the social sciences. Dramatis Personae: A and B. On a different level. . (b) social scientists. The thematic structure of the dialogue is as follows: Part I: Sociology and philosophy – Ordinary Language Philosophy – conceptual and empirical problems. Such preconceptions do not usually depend on careful scrutiny of OLP but rather on influential misunderstandings. a critical theory one) have preconceptions about. proposing at the same time a stance some sociological agendas (for example. Part II: Objections to OLP: language and reality – empirical evidence and handling disagreement – conservatism. what kind of work have you been doing today? Been reading journal articles? Maybe reading up on theory? Looking at your data? A: Quite … what’s the matter? 1 So as not to clutter the dialogue. Nevertheless. who I am mainly addressing in this book. objections based on such misunderstandings need to be considered for the following three reasons: (a) they are in fact widespread. Dealing with objections early on will hopefully help readers to move beyond them in the remainder of the book. How are you? B: That’s a pity … so.Chapter 1 Sociology and Ordinary Language Philosophy: A Dialogue The following dialogue1 provides an account of Ordinary Language Philosophy (OLP) and its relation to sociological theory and empirical research. but also in order to be able to substantiate and attribute to sources some of the things that are being said. sociologists. and (c) it is my purpose to display and thus problematise the typical level of (rather superficial) engagement with OLP. citations appear in footnotes. This is not in perfect unison with the referencing conventions that are used in the rest of the book but I take it that it is warranted by the peculiarities of the genre of writing employed in this chapter. it attempts to pre-empt possible misconstruals by dealing with a number of standard objections that have been raised against OLP. the dialogue is intended to register some of the difficulties in the encounter between someone who is committed to theoretical sociological practices and one who wants to call these practices into question. I need to get back to the office soon. A: I don’t have much time.

2 A: Do you mean to imply that we do not have a way of distinguishing between the two? B: Not at all. I need you to promise me you will listen carefully. Ordinary Language Philosophy. I won’t promise anything but I’ll meet you tomorrow when I have time to spare. (2008). A: Listen.e. not from the armchair. sociology is philosophy. In short. What is this about? B: Let’s just say for now it is about philosophy. But I would like to bring to your attention the fact that some philosophers have recently started using what they claim are empirical – even experimental – methods. a bit more sceptical about their direct relevance to empirical sociological research. I need to talk to you. and its importance for what we sociologists do … I A: Ok. in a sense. such as Marx or perhaps Foucault or even Bourdieu and by also saying that sociology is informed by or draws on philosophy – say Maurice Merleau-Ponty or Judith Butler. We may buttress this not very illuminating observation by pointing to the great sociologists who were also philosophers. And others would be inclined to say for a variety of reasons that. A: So we shall discuss Ordinary Language Philosophy. The only reason that I have pressed the question of their relation is because it enters into the proper understanding of Ordinary Language Philosophy (for convenience let us say OLP from now on). I was hoping you could spare some time for this.10 Clarity and Confusion in Social Theory B: I’m sorry. Louch (1966). or should I say. you are going to explain to me what it is and why you think it is important. B: The relationship between sociology and philosophy is a complex one. B: Indeed. Hutchinson et al. By saying this much I am already claiming that OLP has something to say about the nature of philosophy and that it is unlike the 2 Winch (2008[1958]). . I think as long as we remain at the level of disciplinary categories. we have as much time as we need. I am sure you will agree. Since it is a philosophy I take it that it has something to do with epistemological and ontological questions. Now is the time to be serious. Sociology. we are bound to say something very general.. One can understand OLP and its relevance to sociology only through grasping OLP’s critique of (the rest of) philosophy and how the latter forms a part of sociology. i. ‘sociology’ and ‘philosophy’. however. I think most sociologists are open to considering such questions in relation to the methodology of the social sciences. I am. such as that they are related but ultimately distinct. is all about studying society from the field. it is not philosophy.

explaining how what we say makes sense. I will come back to this. A: Well. but rather one of method. the employment of language during philosophising or theorising versus its employment in conducting our practical affairs. etc. One may say it is a systematic exercise of those abilities and a refocusing of them towards the elucidation of the concepts that are relevant to philosophy’s problems. As a first approximation it may also be said that its contribution is not substantive. in data collection and analysis. So it is not about ontological and epistemological questions in the way that the rest of philosophy is: OLP does not have an epistemology or an ontology of its own to propose. For each of these terms ‘ordinary’ means something different and thus each contrasting pair needs to be spelled out carefully. For now I can present a brief explanation as to why ‘OLP’ can be 3 See Ryle (1953). assisting not. mainly empirical ones. in view. and it employs its methods with that end. B: Not quite. for instance. Is this then a possible connection? Might not OLP add another string to our bow. or to put it more accurately. should I also assume that the ‘ordinary language’ part is tricky as well? B: Yes. I might have known.Sociology and Ordinary Language Philosophy: A Dialogue 11 rest of philosophy. Before I ask you any further questions. OLP’s ‘methods’ are not technical procedures but are through and through dependent on our linguistic competence and the ways we already possess of clarifying what we mean. displaying that we understand an utterance. not only in philosophical theories about language. so to speak. A: Right. partly involves a confused relationship to our concepts.3 Although many of these pairs may have some relevance to OLP. since it is a philosophy. OLP attempts to redress the problematic relation to our concepts and thus to settle the philosophical problems that have arisen because of it. sociology is continually adding to its stock of methods. ‘ordinary language’ may be opposed to a number of terms among which are. since there are not two languages involved here. Now I should say something about OLP’s purpose: its contention is that philosophical puzzlement is typically the product of or. . the important contrast is between philosophical and ordinary language. that the philosophical understanding of language evinced. philosophical language or ideal language. not theory construction. so it turns out that the ‘philosophy’ part of OLP is anything but straightforward. but throughout philosophical practices of language use is seriously flawed. technical language. at the very least. scientific language. One might say that it is interested in those questions for the way in which philosophy comes to ask them and attempts to answer them. As far as their nature is concerned. but rather in theory construction? I would certainly not look unfavourably upon such a prospect though it does not sound like this is exactly what you are suggesting. Let me make a first statement on the nature of these methods and the purpose of employing them.

it is important whether those labelled get to decide the label for themselves or whether the label is attached to them by others. Hanfling (2000) and Baz (2012). That is. 6 For example. although were it not for Wittgenstein OLP would probably not have existed. A: I can very well appreciate the need for your words of caution. Consider: ‘The end of the affair’. to those who followed in their footsteps. the movement or the contents of the jar. ‘strawberry jam’. Now. by extension. Austin and. ‘the gay rights movement’. . at least. it has also been lately embraced by those who see themselves as OLPers6 as a way of talking about a movement originating in the work of Wittgenstein.L. Weitz 1953). I am familiar with the temptation of reading too much into a term and more or less oversimplifying. it has an established historical use and for that reason. one should not necessarily suppose that OLP refers to a shared core of methods and outlook among those who are described by others as or call themselves OLPers. Austin and Ryle. To combat such 4 I say at Oxford due to the existence of the term ‘Oxford philosophy’ which is only partly interchangeable with ‘OLP’ (Dummett 1960. other points of view. This is not the case here or. Ryle and Austin – even if embracing the term usually necessitates a number of caveats like the ones I have introduced so far. finally you will be relieved to hear. not to mention those who have developed and applied their work in numerous ways.7 In short. In this respect. the title or label might be thought of as telling us something important about the novel. However. See Hare (1960) and Tomlin (1960) for what the teaching of philosophy at Oxford was like. fn 1.5 Despite the non-illuminating character of the label ‘OLP’. on the differences. I suspect. not as straightforwardly so. by OLP. 5 Uschanov (2006). whether Ryle and Austin would have called themselves OLPers is dubious. I mean to refer primarily to the work of Ludwig Wittgenstein.). in effect caricaturing. Hacker 1996: 148ff. however. who died in 1951 and worked at Cambridge. An even broader term is ‘linguistic philosophy’ (see Rorty 1992[1969]: 3. as many Witgensteinians are prone to do in order to show that Wittgenstein’s account of philosophy was much deeper than Austin’s – which it very well may have been – should not detract from the important fact that Austin’s ‘practice did not conflict with Wittgenstein’s [and that] … there was often little discernible difference between their tactical moves’ (Hacker 1996: 172). as most likely the term was coined by their philosophical ‘opponents’ who were trying to make sense of what was going on at Oxford in the 1950s4 – Wittgenstein. Gilbert Ryle and J. between Wittgenstein and Austin. I think it has to do with the ways we regularly deal with titles and labels which are often taken to be guides to what they title or label. 7 Focusing. and you will be familiar with this phenomenon. most certainly would not have called himself an OLPer. OLPers does not mean that there are not any important differences between them.12 Clarity and Confusion in Social Theory a troublesome term. I hasten to add that calling Wittgenstein. for example.

Ryle and Austin’s thought either acknowledged by those individuals or by others later. for example. B: Perhaps I need to explain myself more. I think I have heard enough about ‘OLP’ and the danger of treating the term as naming a completely homogeneous area of thought.Sociology and Ordinary Language Philosophy: A Dialogue 13 a temptation is partly the reason why I agreed to take part in this discussion: to allow room for detailed considerations. Granted the legitimacy of your objection. purposes and methods of proceeding. Wittgenstein. as you said. Is it not the case that by arguing for. . on the other hand. to grant you the space to teach me what OLP might be about.8 I have tried to forestall treating OLPers as if they were all of a piece. it is those central insights. speaking about the importance of OLP can be just a shorthand way of saying that I consider Austin. A: You seem to be going back and forth between claiming. just like Howard Becker who protested against treating the ‘Chicago School’ as a group unified in terms of thought and research practice. nevertheless. I am not hinging anything that I say on the historical existence of a School of philosophy organised around a piece of doctrine. you commit yourself to the existence of an ‘it’? B: I certainly do not wish to exempt myself from such misunderstandings. for instance as part of the OLP School with a capital ‘S’. etc. that I am interested in and not their historical attribution – which is not to say that they cannot be or have not been attributed thus in the literature for and against OLP. on the one hand. I remember reading. Austin was not that 8 Becker (1999). the importance of OLP and by here and now attempting to exposit it. Now. method. If I may draw an analogy. I think I have made myself vulnerable to it by not going into enough detail. it seems to me that these ‘central insights’ are not significantly different to the core of commonalities you were refusing to acknowledge only a little while ago. Second. or purpose. I will not be claiming any more agreement between OLPers than I can demonstrate in arguing for the importance of certain more or less central insights. that there is not necessarily that much in common between OLPers and. purpose or method. whether you are yourself immune from those misunderstandings. Ryle and others (each for slightly different reasons) to be important thinkers and worth more attention than they are getting these days from sociologists. I have tried to make plain the risks involved so as to make it easier to guard against them. I cannot help but wonder.. regarding your charge that by speaking of OLP I am committed to an ‘it’ I offer the following to my defence: First. If anything. nor on any commonality in Wittgenstein. that although Ryle tried to learn as much as he could from Wittgenstein. that there are certain central insights you subscribe to and I presume those who conceive of themselves as working within OLP do too. Now.

OLP tries to remedy misunderstandings of our concepts and the lack of a clear sense in what philosophers and social theorists say and write. agency. I should stress. say. however. But at a certain level of description one can say retrospectively – and I do want to make precisely that claim – that most of what Wittgenstein. let me latch on to what you have said about doctrines to make clear a fundamental point. it strikes me that you have veered off course and dragged me along too.e. to sociology. we. which could perhaps be translated into a theory of social action or social structure. idealism.9 It is such things that I had in mind when I denied that there is a common core of method. structure vs. These. of course. In fact. for example. it is precisely in what one makes of this insight. to treat it as if it were in the business of debating on either side of. To miss this point is to misunderstand the nature of the challenge OLP presents to the rest of philosophy and. since I presume we understand what the point of using the term ‘OLP’ is. Ryle and Austin did can fruitfully be organised around the common insight that careful attention to how we use words in everyday situations can be extremely instructive in understanding the nature of philosophical puzzlement and can ultimately assist us in dissolving philosophical problems. 10 Contrary. I will avoid repeating any of the caveats I know I have been too eager to insert in our discussion so far. that saying this much is just furnishing oneself with a starting point. reality or knowledge. I want to emphatically say that this is not the nature of the objection it raises. any minor differences aside. as if in the course of its critique its purpose were to deny what is being asserted by. but not ones that will benefit from your attention. it is not to be found in any set of theses. is it not? B: It certainly is. . with your permission. does not mean that whenever falsehoods are indeed 9 Hacker (1996: 172). I will remind you of the question we are after which is what OLP can tell us about sociology. not that we have said something false. So. What OLP can tell us about sociology is not to be found in a theory of language.14 Clarity and Confusion in Social Theory impressed. Except. If I do. A: Let me see if I understand what you are saying. realism vs. eventually. materialism vs. materialists or realists. This. purpose and all the rest. The problem is usually that they. In fact I think from now on. So the question is not whether all these philosophers supported in actual historical fact. for example. that the substance lies. are perfectly legitimate questions.10 No. that is. in order to prove an opposite thesis. essentially the same doctrines nor whether we can synthesise what they said. to David Bloor’s Wittgenstein-inspired work (1983). how one practices OLP. nominalism. That is the question. i.. how one uses it to tackle the problems. It is there and not in a general formulation that one might seek to understand what OLP is about. I take it. have not succeeded in saying something clear and intelligible.

but why force these concerns when I have said nothing that might make them relevant? A: They are recognisable sociological concerns and you are after all purporting to be saying something about sociology. what about the extraordinary. The central recommendation is that this can be done through reminding ourselves in as much detail as necessary of the way we use language in conducting our practical. In fact. it does not demand of those who practice it that they exclude other resources for the appraisal of arguments.e.12 that is. Thus. the sense of the term ‘ordinary’ we are concerned with here is not the sense of the non-technical. I think I should warn you not to presume that all of these concerns apply given the nature of the problems OLP is addressing. B: I will do my best to explain. A: Well what about the non-everyday use of language.11 To summarise. by eliminating the ordinary use . i. especially our conceptual ones. everyday affairs. just bringing certain phenomena of sociological interest to view in order to ascertain how it is that they might be accounted for in your scheme. OLP throws light on the nature of philosophical problems by urging us to understand how they arise and showing us how they can be dissolved rather than solved. A: I was not presuming anything. though. 13 For example. by extension OLP’s. as ‘reject[ing] the purifying process that. Before that. OLP can be seen to demonstrate through its practice how the restriction of resources. is a crucial ingredient in the creation of theoretical puzzlement. OLP does not claim to give us access to a special domain of phenomena at the cost of relinquishing access to other domains. the exclusion of ways we have of talking about something. now is the time to say more about it and to justify it. or nonscientific13 but the sense in which ordinary language is language used 11 This theme is developed further in section 2.Sociology and Ordinary Language Philosophy: A Dialogue 15 asserted – which presupposes that something intelligible has been said – we are incapable of locating them and that therefore OLP is blind to falsehood. whatever one might think of Michel de Certeau’s appropriation of Wittgenstein. nor does it claim to reduce other forms of investigation to its own manner of proceeding. the work required to get a philosophical thesis off the ground. and why should we be guided by them? You said something earlier about the ordinary and philosophical use of language. 12 See Toulmin (2003[1958]: 159–60). the deviant. B: Yes. And I am still waiting for your elaboration on your appeal to ‘ordinary language’. OLP’s critique of philosophy is to be found not in refuting a developed position but in what could be called ‘the preliminaries’. so as I already mentioned. B: Right … Ok.2 et passim.. Well. and what about the moments of revolution? Are not our everyday affairs that give language its ordinariness. he is mistaken in his description of Wittgenstein’s position and.

When we adopt such a theoretical attitude towards language. It is language as it features in our activities be they professional. we are liable to misconstrue our concepts. although you would have to specify what exactly you are calling philosophical and social theoretical language and just how extensive that language is. makes it possible for science to produce and master an artificial language’ (1984: 10). ordinary language. A: I am unsure how to understand this distinction between language ordinarily used and language reflected on. . and is thus. for example in doing philosophy or social theory. does not consist solely or perhaps even fundamentally of this? Rather. is it not the case that sociology. I think that one could very well say that when we write and discuss philosophy and social theory we are also using and not merely reflecting on language and that philosophical and social theoretical language belong to the settings of philosophy and social theory. The point is that scientific language is also the everyday language of scientists. common. This is just a way of repeating the worry I mentioned initially that sociology is not philosophy which is something you seem liable to forget. or treated for purposes of theory construction. uncommon. a claim that the former underpin.16 Clarity and Confusion in Social Theory as opposed to language reflected on. B: This is true. getting clear on them involves looking at the everyday uses of words that constitute these concepts. may I remind you that there is usually a certain kind of claim involved concerning the relation of these conceptual inventions to the rest of our concepts. In so far as these concepts are ones that as speakers of the language we are familiar with and employ in our everyday affairs. as far as OLP is concerned. Although. It is language in the settings where it belongs as opposed to language taken out of these settings and reflected on. it is mainly a form of empirical enquiry. A: I do not see exactly where this is going. are more fundamental or closer to reality than the latter. I agree that it is certainly not philosophical and social theoretical conceptual inventions that one may explicate by turning from philosophy and social theory to their ordinary uses for it is precisely in those settings that they have their ordinary uses. workaday interactions between scientists or scientific journals?) is a matter we need not get into. and why it proceeds on the elimination of everyday language (where from. of language (everyday language). What exactly this artificial language is. even if we grant that it does involve theory construction. the laboratory. scientific or otherwise. lay. Are you saying that one ought to be very careful when in the theoretical attitude or are you making the stronger claim that since the problems in question arise because of such an attitude we should abandon it? I suspect that this fits into some kind of assault on sociology but if that is so.

14 even in our current era of advertised interdisciplinarity. for our purposes. is to say that each discipline consists of a number of activities. ‘biology’.. 17 We will witness this in the following chapters regarding questions of the observability of the social and the fallibility of our concepts.Sociology and Ordinary Language Philosophy: A Dialogue 17 B: There is something to be said about these Aristotelian distinctions. to which are attached logically different kinds of problems. as the latter abound throughout the discipline.16 which do not necessarily call for empirical observation or the accumulation of further knowledge but rather for an elucidation of what we already know yet are failing on occasion to keep in clear view. A: I honestly do not think that any social scientist would disagree with the argument that we should not presume that because one is said to be doing physics. in political and social scientific theory.17 Questions such as ‘what is justice?’ or ‘what is time?’ are prototypical in posing problems whose kind of difficulty is tricky to identify. there are no further ways of differentiating their activities. architecture. One way of doing this. What I am saying here is not that philosophy and sociology are identical – how could they be? – but that they share a form of problems. Even within physics and sociology the forms that enquiry takes can be very different. sociology or psychology. it had arisen because we lacked a clear understanding of our concepts. Obviously we can distinguish between disciplines by using disciplinary categories such as ‘physics’. apart from philosophy. i. 15 Hacking (2002). among others. 16 Hannah Pitkin had already suggested in 1972 that such problems abound. if anybody is aware of this by occupation it is sociologists who are likely to be sensitive to methodological differences. . let us call them conceptual problems. is by talking about ‘styles of reasoning’15 and perhaps showing how different disciplines share a common style. etc. ‘sociology’ and ‘philosophy’ but we may also want to get at some of the commonalities between disciplines.e. evidence. Furthermore. The difference between the forms of problems a discipline deals with is. for example. relying on a generic characterisation such as that sociology is the study of social phenomena. mathematics. B: I would be inclined to say that the fact that many of sociology’s problems are conceptual in character is not often acknowledged and that sometimes we misrecognise a problem so that we treat it as if it were a matter of finding out something we did not know when. in fact. namely our concepts. problems calling for certain kinds of methods. say. rather than. a better way of understanding what it is about. A rather simpler way. And it is also a basis for seeing what it may share with other disciplines. 14 See Lynch (2000). procedures. for our purposes.

youths going clubbing. one can decide on the concepts by going into the field as much as one might need to decide on concepts one is not familiar with. . The following example. before empirical investigation proper can proceed.18 A: I do not like this dichotomy between facts and concepts. (2008). first decide on the concepts and then decide on the facts. Now I am inclined to say – although admittedly in a rather schematic way – that insofar as we are dealing with ‘conceptual problems’ what empirical investigation yields cannot count as a solution since. Luke Yates. a dialectical relationship. there is a sense in which sociology can pose distinct empirical problems although I am not so sure that what passes for empirical sociology does not actually involve what I have been calling ‘conceptual problems’. and social theory on the other. Still. I am not primarily opposing here working with data to working from the armchair. I will not provide any similar demonstration of what I say about empirical sociological research. A: You better had. should give you an idea of the kind of connection I have in mind. one needs to settle on the concepts. that leaves empirical sociology out of the equation. however. I should probably give you an example here. May I remind you that sociologists engage in ‘theoretically informed empirical research’ which implies that there is some contribution made by the theory to the understanding of what is researched and sometimes also by the research to the theory. on our criteria. members of ethnic minorities not 18 For further explication of this point see the introductions to Ryan (1970) and Hutchinson et al. that I am bringing into view and opposing to the empirical. 19 My understanding of this material has benefited from discussions with a number of colleagues on several occasions. for example.18 Clarity and Confusion in Social Theory A: Granted that there might be a connection between these traditionally philosophical questions. which we could call deciding on the facts. I am thinking of the question ‘what is political?’ or ‘what is “the political”?’ as it is posed in the study of subcultures and/or social movements and/or everyday forms of resistance. on the one hand. I am saying that we cannot decide on the facts unless – and not until – we have decided on the concepts. It is the very need to decide on the concepts. I am indebted to Michael Mair.19 The question enters into these studies as the need to come down on whether. Ulrike Flader and. having already decided on the concepts. I would think of it as more of a reflexive or. even better. I wonder if we ever do this. In any case. people recovering food from supermarket bins. especially. B: Before I do I must stress that although I intend to demonstrate in detail how social theory is replete with conceptual problems. nor finding out something new to reminding oneself of what one already knows. B: Well. B: I do not intend it as a temporal distinction but rather as a logical one.

it is a rather uncontroversial remark. are after all doing something political even though they are not engaged with political parties. . Thus the potential discovery.Sociology and Ordinary Language Philosophy: A Dialogue 19 displaying a national flag on a national holiday. (2010). but that they nevertheless are. Other times the idea is that even though they do not conceive of themselves as doing something political. not to acknowledge some sense in which practices can be political.20 for example. when clubbing youths are in fact debating who to vote for in the next elections or the problems of a coalition government in Britain.. i. they nevertheless are. one needs to speak of ‘political’ in the same sense. an extension and not a discontinuous concept. Now. it might be thought important. however. For one. politically important.. say. not being able to see how something is political in any available sense means that it simply is not politics. Let me add that sometimes this is coupled with the idea that to find out whether they are in fact doing something political we need to look hard for they might be engaged in a new form of politics we do not currently recognise as politics. This might warrant treating different senses of politics as ideologically in competition with each other. that. say. 20 Riley et al. one means to say that there are political activities that have nothing to do with any of the senses in which we can understand ‘politics’ then it seems to me that this is an absurd statement in that. they would be flirting rather closely with confusion. intellectually pointless though the idea of the ‘proper’ definition of politics might be for you. and not the product of establishing empirically. elections and the like. etc. A: I think that the example you have chosen features other dimensions which complicate things and need to be taken into account. Youths going clubbing. If. if someone were to pose the question about such activities in the form ‘is this politics or not?’ and insist on a ‘yes or no’ answer. may be doing something political in one sense of the term but not in another. There might indeed be a struggle going on regarding who can use the ‘politically charged’ word ‘politics’.e. given a certain sense of ‘political’. Further. to offer some thoughts. given the different senses of politics. that youths clubbing are in fact doing something political is a product of shifting from a restricted understanding of ‘politics’ to one which acknowledges its other sense(s). if by the idea that we might not be able to recognise what they are doing as politics is meant that they are not engaged in parliamentary activity and electoral pursuit but that they might still be engaged in politics in other senses of the word. I should also add that any purported extension of the concept of politics is precisely that. This means that to be able to claim that they do not perceive themselves as doing something political. such as in fighting for social change or for their right to self-identity and ‘self-determination’.

You are claiming that sociologists think they are conducting empirical research when their research really is not empirical. the worry continues. which are. in the sense of being. But the similarity will depend on the role played in both cases by concepts such as left and right. This is so because deciding whether what any of these groups are doing is something political or not. justice. If these concepts do not apply in the same sense (or at all) calling both activities political will not establish any stronger a likeness. but you are still flattening the problem by focusing on its conceptual aspect. such as clubbing. so you are not speaking of ‘empirical’ in the same sense. In other words. for example. equally political. But note here that the worry is really dictating the comparison between the two cases. power. very few young people nowadays seem to engage in such activity – never mind whether this is true or not. it has just occurred to me that just like those who shift the meaning of ‘politics’ in order to argue that these groups are doing something political you are doing the same thing with the term ‘empirical’. that they are looking at data or going into the field. And this gives us a glimpse into how deep the conceptual aspect of the problem goes. freedom.20 Clarity and Confusion in Social Theory B: I certainly do not wish to discount these problems. for instance. the worry that in contrast to previous generations who were members of socialist or Marxist youth organisations. nevertheless. however. progressive. the fact that youths are not ‘formally’ politicised mean that they are apolitical and conformist? As an answer to that question it may be offered that even if they do not join political parties they can be found to be engaging in other activities. ideology. the kind of problem one is dealing with is misrecognised. Saying that both activities are ‘political’ cannot make them any more similar than they are. I am not primarily speaking of empirical in that sense although what I am calling empirical can be contrasted with what we sociologists do in the following way. I can very well see that the form of questioning I have described may be motivated by a sociological worry. conservative. In fact. anti-conformist. etc. This is not a bad thing in itself but it should furnish us with a reason to be cautious. far from keeping the sense of ‘the political’ as a stable backdrop to the investigation treats it as one of the things in question. Does then. A: Yes. To the extent that it is thought that one must go into the field. deciding the whether turns out to be deciding in what sense . Also. You are not denying. B: As I said. revolutionary. resistance. This is not to say that they are completely dissimilar nor to prejudge that there might be a way in which understanding what goes on in clubs will reveal it to be closer to party politics than previously thought or that some of its consequences might have effects in the sphere of politics.

Sociology and Ordinary Language Philosophy: A Dialogue


something can be called political. This latter question is a conceptual one,
I would like to say, and has no necessary connection to fieldwork.21
The point of the example of sociological investigation into ‘the political’
has been to demonstrate how conceptual problems can be seen to resonate
throughout the discipline. Indeed, if I may generalise somewhat, it is
precisely because the question ‘How should one describe the phenomenon
at hand?’ is implicated in so much sociological work that there is the
potential for conceptual problems to arise and why OLP is profoundly
relevant to that work.22 In any case, I will not come back to the way in
which empirical research is shot through with conceptual questions. I am
here concerned with what OLP can tell us about social theory.
A: Ok, I think I have heard enough. It is a good time to end for today. There
are some things I need to do before I can object to what you are saying in
the way I would like to. See you in a fortnight.

21 I should clarify here that I do not think that sociological enquiry is exhausted in
talk of conceptual questions nor do I mean to imply that fieldwork may not have value in
itself – and a significant one at that. I am only saying that our understanding of concepts
is presupposed in being able to collect data and that although there are cases in which we
need to develop such understanding precisely by going into the field, there are other cases
where we possess an adequately developed understanding which we may readily put to use
in addressing conceptual questions. Furthermore, I am saying that, in doing fieldwork and
collecting data, conceptual questions, questions of what it makes sense to say or, as Ryan
puts it, ‘how we should characterise the facts, what we are to say about them, what we are
to see in them’ (1970: 5) may frequently come up and, therefore, that ‘the inquiry into the
kind of significance possessed by the data … is more like the inquiries of philosophers than
it is like the inquiries of chemists or physicists’ (1970: 145–6) a fact which complicates the
sense in which such work can be characterised ‘empirical’ or ‘observational’.
22 Compare Winch (2008[1958]: Ch. 4) as well as what Hutchinson et al. say in
his defence:
It is not, first, for sociologists to decide what someone is properly said to be
doing. The language they are using, after all, does not belong to them, but is
one that they speak because they belong to the language communities about and
within which they write: ‘washing hands’ isn’t a description that any sociologist
has contrived, and it is indisputable that, whether it is the correct thing to say or
not should be decided by the way in which the language works, is used, within
the activities to which it belongs (to domestic affairs: clean those dirty hands!; to
medical situations: ‘scrub up before surgery’; or to affairs of state: washing hands
as the ceremonial way to recuse oneself) (2008: 77–8).
It should be stressed that this is not the same as saying that what people say they are
doing is what they are doing.


Clarity and Confusion in Social Theory

A: In our last discussion, a while ago, you spoke of OLP as dealing with
‘conceptual problems’, a type of problem which you contended is to be
found in philosophy but also in sociology, and which you opposed to
‘empirical problems’. Your story was that these ‘conceptual problems’
come into existence as we adopt a theoretical attitude towards language
and that they can be – I believe the word you used is – ‘dissolved’ through
conceptual elucidation, that is by reminding ourselves of how we use words
when we are not theorising.
B: I could not have said it better myself.
A: It is a rather memorable story if not, alas, a very sophisticated one. I have
done some reading since last time and my initial impression has been
confirmed that OLP has a very simplistic understanding of philosophy, one
I am sure it will project onto social theory as well. OLP, in effect, reduces
intellectual pursuits to conceptual problems, that is, to not really serious
problems but rather pseudo-problems. Why do these problems occur?
Because social theorists speak differently than the man in the street does,
as if there were no problems arising from how the man in the street speaks.23
The important thing for OLP is not what these thinkers might be after, the
proverbial woods, but the ways they are expressing themselves which need
to be normalised, the proverbial tree.
B: I am not sure exactly what kind of reading you have been doing, but, in
any case, the point of talking about pseudo-problems is that the challenge
such problems pose does not lie in the means of solving the problem but
in its very formulation. For example, when we ask ‘are social structures
real?’ the difficulty is primarily in understanding the question’s sense, what
exactly it amounts to and what could possibly count as an answer to it. That
this can be quite a daunting intellectual task is not being denied by speaking
of ‘pseudo-problems’. What is being denied is that people who adopt a
‘yes or no’ approach and proceed to debate on either side have identified
the problem correctly. I can also assure you that OLPers are as much
concerned with what one is after as they are with the mode of expression.
It is just that the conceptual confusion gets in the way of understanding
what it is that one is, in fact, after. Thinking that we are pursuing something
important or profound does not make it so, nor does it render what we are
saying intelligible.
A: What I am trying to say is that there is a very schematic understanding in
operation which can be seen from the fact that there is always a uniform
diagnosis to be offered. OLPers talk generically about their investigations
culminating at the point where the dissolution of the problem occurs. This
presupposes some sort of faith in the uniform character of philosophical
23 Mates (1958: 164).

Sociology and Ordinary Language Philosophy: A Dialogue


problems and renders philosophy (and social theory) a one-dimensional
pursuit. OLPers are calling philosophy whatever conforms to this
pattern and thus the whole of philosophy is understood as being about
conceptual problems. If something else involves these problems then
it is also philosophy. In cases, however, when a problem philosophers
are concerned with does not fall under the description it ceases to be a
philosophical problem.24
B: The point of dubbing the kind of problem in question ‘philosophical’ is to
tie it to standard philosophical procedures, not to reduce whatever might
happen within philosophy or sociology to it. OLP is not in the business
of legislating on what one should call ‘philosophical’ and it has no stakes
in quibbles such as whether Wittgenstein was a philosopher or an antiphilosopher25 – opponents and commentators can call him what they like as
long as the details are kept in view.26 Nor is OLP pontificating about every
possible form of enquiry, it is simply delineating its object of interest.
A: In doing so, however, it is prejudging the issue, for the moment a problem
has been identified as ‘philosophical’ the outcome has been settled as well,
it is to lie in its eventual dissolution.
B: Granted that some OLPers would say this, I do not think that anybody is
guilty of stacking the deck, for the point is to show exactly and in detail
how the problem is of a certain type. The important part is not to enter
debate over what philosophical problems are like in general, although
one could adduce appropriate evidence for this and although OLP regards
itself as relevant to most of what goes on in philosophy. It is the practice
of OLP that illuminates the nature of theoretical puzzlement (or fails to
do so) through dissecting the particular conceptual problem at hand. This
priority accorded to the practice of OLP is, I think, an important reason
why there is limited value in an exposition of OLP’s methodology in the
abstract. On the other hand I feel some kind of stage setting is required
before particular investigations can be appreciated, given that they are so
very different to what social scientists are used to. These opposing forces
govern what we are currently engaged in.
A: I should think part of the stage-setting would require one to respond to
some salient objections that have been made against OLP. Don’t you agree?
B: I certainly do.

24 Rorty thinks that ‘[t]he more one reflects on the relation between Wittgenstein’s
technical use of “philosophy” and its everyday use, the more he appears to have redefined
“philosophy” to mean “all those bad things I feel tempted to do”’ (2007: 175). Even if we
accept this, nothing of substance hinges on this ‘redefinition’.
25 Badiou (2011).
26 I am here paraphrasing Wittgenstein (PI: §79).

Shall I read the whole thing out to you or would you rather we go over the objections one by one?27 B: The latter. in fact. I have managed to compile a brief list. A: Right … Ok. it does not always make sense to press the distinction. it is between an instrument which is used in many possible different ways – including in producing a set of descriptions – and what this instrument can be used to say things about. to real entities. the study of society involves the study of concepts that we employ in our social life and. there is more to the study of society than the study of the uses of the word ‘society’ or other words that we might say are part of or related to the concept of society. for example. dangerously so. of the logical relations between them. If it were up to OLP. Hanfling (2000). generally speaking. though it may be partly opposed to that of the concept of society. the first objection is related to the theme of flattening out our concerns. . However. 27 For a fuller treatment of possible objections see Ebersole (2003). the study of concepts is nothing else than the study of the way those concepts enter into social practices and of our various kinds of doings with words. In that sense the study of society. B: OLP does not subscribe to the totalising distinction between language and reality which it considers confused. 8) and Baz (2012). 28 As Winch puts it. sociology would study the use of the word ‘society’ as opposed to studying society.24 Clarity and Confusion in Social Theory A: Well. or otherwise they are free-floating. one exemplified in the scheme of reference you invoked: either words refer to reality.28 What is more. Uschanov (2006). You are again imputing to OLP an imperialistic spirit which it does not possess. This is after all a dialogue. I would add. Moreover. Hacker (1996: Ch. In other words. it is to remove the part which ensures that our concepts are held in check by reality and that they are not free-floating signifiers. that we do not confine ourselves to expressions but ultimately seek to comprehend that which expressions refer to. B: Nobody proposed to substitute the study of the concept of society for the study of society. After all. not least in that it involves the idea that there is a single type of relationship between language and reality. but knowledge itself? A: Let me put my point differently. It is an indispensable part of our enquiries in sociology. To chop off the reality part and focus on words is to produce an impoverished version of our enquiry. For example what would it be to investigate not the concept of knowledge. and elsewhere. rather. Obviously. although we might be concerned with language. although in some cases we can distinguish between ‘mere words’ and reality. ‘[i]f we do want to speak of a ‘relation between language and reality’ [which I think creates nothing but confusion] this is not a relation between a set of descriptions and what is described’ (1987: 196). is not opposed to the study of other concepts or. to the study of our concepts. we are ultimately concerned with reality.

Standards may be frequently used but it is not their frequency which allows us to understand them as standards – although frequency can help us locate a potential change in standards. in which one tries to recall what we would say in this and that situation. 30 Consider the difference ‘between being able to formulate statistical laws about the likely occurrences of words in a language and being able to understand what was being said by someone who spoke the language’ (Winch 2008[1958]: 108).32 A: You are not answering my question. 32 Vendler (1971). 31 See White (1985: 13. You seem to forget. and this leads to my second objection: the picture OLP paints is of a deliberately amateurish or unscientific enquiry into language. And why prejudge the issue on the question of whether there is in fact any agreement on what we would say. nor are we describing the statistical distribution of utterances in certain situations. i. How are we to decide between competing standards without any factual evidence we can potentially appeal to? The only alternative I see is the imposition of one standard at the expense of another.30 What we are appealing to is linguistic usage31 but not as a statistical notion. 29 See Mates (1957). Lyas (1971) and Leiber (1999). that there are developed methods of linguistics that deal with these questions and that to speculate on what we would say from the armchair is just not good enough. Statistics are largely irrelevant. the imposition of one’s will on the matter. that you and I disagree on what is to be said. that it is really what we would say.e.. . moreover. Greek. We certainly cannot conclude from how the OLPer thinks a word or phrase is used that this is how it is actually used. In so far as we speak the same language. In other words. in the same way that if it were found that 20 per cent of drivers cross red traffic lights we could not use that fact to conclude that such conduct is law-abiding. 1958: 24) and Ryle (1971[1961]). how we would use words. Fodor and Katz (1963). if not totally biased. English. how are we meant to decide which one of us is right without any evidence to go on? It follows then that the results of this way of proceeding will generally be untrustworthy. B: It is not clear to me that we actually need to decide between different sets of standards. however. knowing that in some situations people may actually produce an utterance does not tell us whether or how it is heard as an intelligible one. etc. really. how are we going to gauge that what seems intuitively right is in fact right. why take it that there is one thing we would say after all? Suppose. B: What we are after is not what people would actually say but what they could say correctly. To give you an example. we are not reporting on the linguistic habits of a community. Coulter (1983).29 For.Sociology and Ordinary Language Philosophy: A Dialogue 25 A: I think that there is much to separate the study of social action as the study of the pragmatics of language from what you have in mind. Herdan (1960). New (1966).

Our use of such words could not. Now. be any more idiomatic than our practice of making change or keeping score in a game (1971: 21–2). If. naturally. 36 See Austin (1979[1961]: 183–4). evidence.26 Clarity and Confusion in Social Theory etc. in general. using words differently. having said that. But this is simply not the kind of problem that causes trouble for philosophers whatever it may do for lexicographers. ‘true’. Friedman (1969). ‘see’. and we may offer explications and further elaborations. The disagreement over ‘voluntary’ did not arise from this sort of ignorance. 34 Bates and Cohen are instructive in describing Cavell’s account of the disagreement between Austin and Ryle and its resolution: Cavell says. when we are in fact diverging from ordinary usage we need to present clearly the rules for using an expression. (1972: 24) 35 Cavell (1971). Rather. being itself the type of source of.33 The point then is to decide what the standards both parties subscribe to are. I am a bit sceptical regarding how disagreements are handled. Cook (1982) and Hanfling (2000: Ch. I see that I don’t say such and such. In order to do this it does not make sense to appeal to any sort of external evidence. It is 33 Henson is illuminating in what he says with reference to the potential reasons for disagreement and their relation to philosophical debate – in this case a debate between Austin and Ryle regarding the use of ‘voluntary’: [S]ome speakers will be familiar with special technical vocabularies unknown to others. ‘We know Austin’s example counters Ryle’s claims’ … How? Not by showing that there is counter-evidence. or rather. not by proving something. or say.36 I do not think the potential for mistakes and disagreements is any kind of drawback or that it significantly hampers enquiry any more than it significantly hampers everyday communication. The point is not to mislead each other or ourselves that we mean the same thing when we are.. ‘explain’ or ‘faith’. And. A: Still. In cases of disagreement what we can do is describe the situation in as much detail as necessary. and no extravagant erudition distinguishes the users of such words as ‘know’. well. for instance. but this is not the rule. our linguistic competence as speakers of the language is already presupposed in the evaluation of any evidence.34 Besides. 4). [and] people at one level of education will use correctly words which people at a different level of education will sometimes use incorrectly. That is. whether there is really any space provided for disagreement.35 This is not to say that we may not be mistaken on occasion or that we may not disagree. it isn’t that I give up S because my evidence for S. in fact. One (Ryle) withdraws his S statement not merely because he is shown that others don’t do. there is no ultimate requirement that we agree on how words are to be used if it is the case that we genuinely disagree on the matter. it is rather the exception. has been outweighed. my abandonment of S consists in my giving up of R. we are to a large extent agreed on the standards in practice. you come up with the thesis that all communication is interpretation you owe me an explanation as to how you are using the word ‘interpretation’. ‘good’. which was R. such and such. ‘prove’. ‘mean’. . ‘think’.

If we are to speak of rules at all [in connection with the use of words]. Strawson: [O]ur practice is a very fluid affair. not having the stamina to run a marathon is a limitation on the part of an individual who might try. their familiar uses. of ‘being able’ to say this or that.F. And I cannot help but wonder what sociology would be like if it were committed to the ordinary use of words and left no room for conceptual innovation? B: I first need to say something about the sense of ‘being able to’. of the fact that ‘we cannot’ or ‘it does not make sense’ to say this. To give you an example. i. but not identical with.38 Whereas in the absence of an intelligible something in the latter case what we need to do is give sense to our words as opposed to construing our inability to observe rules of use39 as a practical inability. It is also belied by the fact that as speakers of the language we have the capacity either to give new meaning to existing forms of the language or invent new forms of our own. that or the other. we ought to think of them as rules which everyone has a licence to violate if he can show a point in doing so. 39 Compare P. if it makes sense to call it that. but. On the other hand. but about a logical impossibility of running a 10-kilometre marathon because of the simple 37 Which is also different from matters of moral propriety. In this case we are not talking about a lack of ability. And may I add that this is not something that only the brightest of us can do.. tact or politeness as Richman (1966: 24–5) and Coleman (1993: 371) are careful to emphasise.e.37 One of the ways that the two differ is that in the former case there is an intelligible something we cannot do. of ‘can’ and ‘cannot’. in their wish to participate in such an event. (1964[1952]: 230) . we do it all the time. B: I agree with you that giving sense to an expression is anything but beyond our powers. applying them to states of affairs which are both like and unlike those to which the words are most familiarly applied … it is in fact hard to frame a grammatical sentence to which it is impossible to imagine some sense being given. In the effort to describe our experience we are continually putting words to new uses. giving sense to an expression is the most ordinary thing in the world. or strength or acumen. telling someone who expresses a wish to run a 10-kilometre marathon ‘you cannot run a 10-kilometre marathon’ is a rather different sort of limitation. The limitation is not of a practical but of a logical kind. on the contrary. This kind of talk does not seem to me to leave much room for options. and on the other hand.Sociology and Ordinary Language Philosophy: A Dialogue 27 patently clear that OLPers speak of ‘what it makes sense’ to say. A: What is then to be made of the ‘cans’ and the ‘cannots’? Where you and OLP seem to locate a limitation I see nothing but capability. to overcome it through training. connected with. 2000) for elaborations on such an understanding. 38 See James Conant’s writings (1991.

or better still. Now suppose that instead you had said ‘Can I have some more water’ and I responded by saying ‘You cannot have more water since you have not had any yet’. would that be incompatible with proceeding to offer you some? I would be rather drawing attention to the fact that it was not clear what you meant by saying you wanted more water. On Carroll’s relevance to OLP see Pitcher (1965). Just do not expect to be understood. So if I say. ‘you cannot say that’ I say it under the presumption that you will recognise the mistake I am pointing out to you 40 I am quoting Tweedledee from Lewis Carroll’s Through the Looking-Glass. B: I could have understood that you wanted water.41 Would that be preventing you from having any water. in fact one you are imposing? You are saying ‘you cannot say that’. To think this would be to misunderstand what freedom of speech comes down to. I would be denying your request and forcing you to remain dehydrated. given that you and I both understand that you have not had any up to the point at which you asked me for more? A: But is this then not a limitation. You are trying to prevent me from saying what you think does not make sense which is what I might want to say anyway. And that involves speaking in ways that both you and I understand or at least explaining what you mean in ways others can understand. In stressing points of logic in this book I will often resort to examples from Carroll’s works (1895. You can indeed speak like Humpty Dumpty if you so wish. But bear in mind that this is not the same as ‘saying what you want’ in the sense in which an oppressive regime might prevent people from saying what they want and believe. yes that is surely the case but is this limitation not a very real limitation? And another thing. That’s logic!40 A: Yes. 1998). it ain’t.28 Clarity and Confusion in Social Theory fact that marathons are (approximately) 42 kilometres long. In any case you seem to have understood what I wanted in your example. 41 The example is adapted from Alice’s Adventures in Wonderland where the March Hare offers Alice more tea. why does one have to be committed to what ‘marathon’ ordinarily means? B: Perhaps another example might bring the question of ‘limitation’ in sharper focus. So if it were a marathon it would be 42 kilometres but as it isn’t. for if you want to be understood then you have to make yourself clear. Am I not free to say what I like? B: Go ahead and say what you like if by that you mean that you would like to utter a certain string of words. Is there any sense to be made of the word ‘more’ here. Do you see the difference? A: This seems to me rather artificial. . Suppose I had with me a bottle of water and after all this talking your throat felt rather dry and you said to me ‘Can I have some water’ and I refused to let you have any. not more water. namely water. even any string of words. and that is not quite the same as refusing to give you any.

. I take it that we can both easily spot the problem with the use of ‘more’ in the example. the state etc. will prove illusory upon closer inspection. however. A: But why does one have to be committed to a specific kind of ‘more’ or. it seems to me that it is the concepts we already have an understanding of that we want to investigate and not some other ones. Now. though changing what it means needs some justification. Our commitment therefore is often located in the very premisses of our enquiry. The fact is that one needs to decide whether he or she is or is not committed to it. however. justice. This sounds more satisfactory to me in that it specifies the conditions under which one is committed. To proceed to change these concepts is to change our subject matter. that it goes without saying. are especially prone to forget that they have committed themselves by what they say.. I would like to say. It is thus not a matter of me speaking one way. the political. For example. unless that is we are somehow involved in an exercise of changing what words mean. The controversial and difficult bit is showing how one is inclined to overlook their commitment or fail to appreciate its extent. to what ‘marathon’ ordinarily means? B: The point is not that one has to be committed to what ‘marathon’ means. A: It seems that one had better add the following clause: given the ordinary concept of marathon one cannot run a 10-kilometre marathon. Social theorists. B: I would not object to such an addition if intended to forestall misunderstanding. but the ‘more’ I described and used to make a point about its being unintelligible in your utterance is not the ‘more’ you have just described. I am inclined to say. But when we want to investigate the concept of freedom. you another. they are hardly clear on what it is they want. but rather by thinking for a moment about how you speak yourself. someone who does not know how long marathons are can hardly be said to be proposing an alternative. and me trying to impose my way of speaking on you. A: Suppose you had another imaginary bottle of water which had been produced by More Ltd and said ‘more’ on it. Would I not be making sense. Now. On the other hand. it may give an appearance of sense which. one might be proposing to call a race 10 kilometres long a marathon.Sociology and Ordinary Language Philosophy: A Dialogue 29 and do so not by accepting what I tell you because I say so. to return to your other example. would I not be speaking English in asking for more water? B: Perhaps you would. But this is hardly controversial. these moves very easily lead to confusion. If one is not proposing to call a shorter race ‘a marathon’ and still wants to run a 10-kilometre marathon. that one is committed to common usage might need establishing. especially when one is not notifying others to the change or stipulating a change in meaning but then not adhering to it. But in other cases it may prove more difficult: a spoken or written sentence may sound or read like English.

Di Norcia (1975) and Benton (1976). ‘what are these examples examples of?’. I still think that Marcuse is onto something more with his objection. the choice of dealing with a dimension purged of social critique. . rejoinders from Rée (1974) and Harrison (1974).42 B: Allow me to explain how Marcuse’s objection misses the point. when we are dealing with 42 Apart from Marcuse (1991[1964]: Ch. verb.. Marcuse misunderstands the logical category of OLP’s analysis. there would not be any point in preferring the sentence ‘the people of Greece elected fascist MEPs’ over ‘Mary and John drove to Manchester’. one stark example being Herbert Marcuse. critical theorists have picked up on this problem. Is this not however a rather conservative enterprise. i. to the special status of situations where ordinary terms do not allow us to decide. rational. Given the prevailing social conditions. In evaluating the kinds of examples used by OLPers it is important to take into account the way in which these examples enter into the discussion. when social critique must also be a critique of concepts at the same time that it safeguards – let me read this out to you from his book – ‘the need to think and speak in terms other than those of common usage-terms which are meaningful. For it is not clear to me that the political dimension need always be relevant. 43 Marcuse (1991[1964]: 178).30 Clarity and Confusion in Social Theory A: Is it really? We can of course approach these concepts reverentially. Mind you. Why should we subscribe to the worldview that shapes them when the point is to challenge that worldview and to propose an alternative one? It is precisely because such a project does not register with OLP that it can subscribe to concepts as they are – and this leads to my third objection. that is to say.. For accepting these concepts as they are amounts to uncritically accepting common sense. and further articles by Fann (1974). it is precisely this orientation that he finds problematic. if one were interested in illustrating syntactic positions such as subject. especially when we are concerned with such politically charged concepts. etc. Similarly. Burke (1975). who in his OneDimensional Man shows how the examples OLPers appeal to in order to establish the application of concepts are completely innocuous ones. namely that OLP ends up being deeply conservative. divested of any radical dimension. A: While I take your point. Marcuse does not ask the simple question. I think.e.43 The need to use such other terms is also related. and valid precisely because they are other terms’. To complain that the second example ignores politics is to misunderstand the kind of considerations that led to examples being offered. 7) see the numerous contributions in the journal Radical Philosophy: Sayers’s (1974) response to the editorial in volume 6. For instance. the common sense ideas that are shaped by ideology or by pre-scientific doxa. the upshot of his argument being that the political dimension is generally absent from OLP. The examples are utilised in their logical dimension with respect to the specific kind of philosophical muddle under consideration.

concerning the idea of social critique as the critique of concepts. ‘though there may be a choice in borderline cases. talking about the need to stop thinking in terms which are commonly used to frame political debate. there is nothing necessarily conservative or radical about. But to say that the world should unite against the workers can be ideologically opposed to saying ‘workers of the world unite!’ A: I am afraid I do not see any inconsistency. on the other hand. irrational and invalid. I must repeat that OLP does not regard ordinary concepts as sacrosanct. cases where there might be reasons to describe things in divergent ways or cases where we would not know what to say. the concepts of worker. see Graham (1977: 17) for . Just look at the passage you have quoted and you will see that he uses common usage-terms to denounce common usage.Sociology and Ordinary Language Philosophy: A Dialogue 31 borderline cases. It is these objections of Marcuse’s that I think carry more weight. however. why exactly Marcuse thinks ‘common usage-terms’ are meaningless. A: Yes. I am thinking here of terms such as ‘race’. am alive or not’ (2008[1958]: 69). He is. but rather to certain assumptions that are made in order to get the debate going. provided we go into enough detail. Marcuse is not talking about English or German. This does not detract from the fact that there are still a large number of instances where. 45 Norman Malcolm (2001: 30ff. It is not clear to me. world and uniting. ‘national character’ or ‘sexual perversion’.45 If. Surely such cases are excluded by OLPers when they focus on circumstances in which the ‘what we would say’ concern applies without particular trouble. One reason for this glaring inconsistency that I can think of is that Marcuse fails to differentiate between concepts and what is said with them: one can equally well express reactionary or radical thoughts in the most ordinary concepts. The latter type of situation can help us elucidate a concept while the former calls more for a decision on our behalf which is not to say that in deciding it might not be useful to draw comparisons with more clear-cut cases. in all likelihood. say. We need to be able to deny using these terms and to say something of the following sort: there is no such thing as ‘race’ or ‘national character’ or ‘sexual perversion’. It is no overall obstacle that a revolutionary or social critic should speak or write in English.) reports an instructive incident between himself and Wittgenstein involving the term ‘national character’. For example. B: The phrase ‘terms of the debate’ does not necessarily refer to linguistic terms and their use. as I write these words. 44 As Winch observes. Also.44 Now. there is not in others: it is not for me or anyone else to decide whether I. German or any other language – in fact this is exactly what Marcuse does himself. B: I do not think that anybody is denying that there are cases which cannot be settled easily. the point is exactly that these assumptions are introduced by the very terms one uses. terms apply rather unambiguously.

A: The fact remains that OLP’s description of the logic of expressions is and has been. however. however. on Wittgenstein rather than Ryle. however. as I did. To explicate the use of such expressions is not to endorse using them in the context of debate on any matter. in practice. under the benign guidance of Gilbert Ryle … if you are young and left-wing. 47 For other examples of Marxists who draw. Cohen. Second. To use such terms is thus to take sides. such terms are so loaded that they leave no one unscathed. and you come to university with a thirst for relevant ideas. for example derogatory ones.46 if taken as an observation about the kind of philosophical problems that OLP sought to address within the analytic tradition. then what he is saying constitutes no objection to OLP. questionable assumptions. purged of any social-critical dimension. and academic philosophy of the Oxford kind is the first system of thought you encounter.A. a student of Ryle’s and a ‘no-bullshit Marxist’. it seems to me that terms of that kind represent quite a specific area of our language. It so happens that I have been reading his book and I have it with me – that is what a reactionary I am – so I will read it out to you. I would like to remind you here of G.32 Clarity and Confusion in Social Theory one accepts these terms then one is taking on board a number of. This much I think is true and also holds when it comes to other stigmatised terms. then it will be hard for you not to feel disappointed or even cheated by it. at the very least. But if. If Marcuse has such terms in mind. Here Cohen is speaking of the impact his encounter with ‘Oxford philosophy’ had on the book. and it will be natural for you to think of Marxism as a powerful alternative and antidote to it.47 What he says in the introduction to his Karl Marx’s Theory of History: A Defence is highly instructive in that it shows OLP under a fairer light. its way of proceeding cannot be taken to preclude social critique either. first. precisely because its purpose was different. 46 Hinton (1973). it is to attempt to resolve misunderstanding regarding their use when they indeed have a use and when that use is in question. so one cannot base a wholesale rejection of common usage on this discursive area. B: There are certainly concepts which are used by or imputed to warring political or moral camps as emblematic of what each camp stands for. But that its way of proceeding therein did not involve social critique I find as no objection at all as its purpose was rather different. B: There is a sense in which this is true. I must stress once more that OLP is concerned with describing the logic of expressions in connection with a conceptual muddle. see Rubinstein (1981) and Kitching and Pleasants (2002). For. though contingently so. He says: ‘The book owes its methodological character to the circumstance that I moved in 1961 from McGill to Oxford. Most importantly. . Wertheimer (1976). you start with ‘sexual perversion’ and for what the ‘there is no such thing as X’ type of objection might amount to.

Do we desperately need to get clearer on our concepts? Are we in fact confused? All this remains to be seen in practice. In fact. not pope’ (1953: 228). and. it seems to me that. . A: I am not sure that this is what we need in sociology. And it will not do to respond to it by saying that it is not a source of politically radical ideas50 for it does not intend to be such a source. namely that it is not a set of doctrines but a way of clarifying concepts and their connections with the purpose of clarifying our claims and responding to conceptual puzzlement. offers no ultimate appeal. it can free our understanding from unnecessary obstruction by counterposing our conceptual capacities to the debilitating theoretical attitude. in the way it did so many of my contemporaries. if I may be a bit provocative. saves no one. hence plays neither dictator. whether what we claim is radical. as best I could. It is concerned with ensuring that what we want to say makes good sense. In practice is where it makes a difference. 50 As Bernstein. in due course. you have demonstrated no part of it. OLP’s concerns are way too radical for most sociologists. OLP does not pretend to be the be all and end all. saint. 49 Morris Weitz makes the point nicely: ‘It coerces no one.48 The use Cohen put ‘Oxford philosophy’ to supports what I have been saying regarding OLP. reactionary. A: We will see. banal or ground-breaking. OLP is powerful enough to change the way in which we debate many of the issues that concern us in the social sciences and to assist us in better grasping what is at stake. protests against Winch (1976: 84). B: To appreciate OLP one has to see it in practice. [I] used it with enthusiasm to clarify and defend. for example. given the theoretical practices prevalent in sociology.Sociology and Ordinary Language Philosophy: A Dialogue 33 Marxism. the central claims of historical materialism’. on the conceptual standards we do and can be shown to subscribe to. I came to Oxford already steeped in Marxism … so Oxford philosophy did not disappoint me. I agree. Most importantly.49 It is not a totalising form of enquiry that needs to exclude other forms of investigation. I doubt though that you are going to convince me … 48 Cohen (2000[1978]): xxi). then it is not difficult to take analytical philosophy on board.

This page has been left blank intentionally .

genuinely explanatory. These arguments can thus be seen as attempts to specify the form of explanation. for that matter. Nowadays. a handful of appropriate types. apart from featuring restrictions on types of concepts (those referring to empirically observable events). in the sense that they are not designed to tell us. Pitt 1988. is based on the assumption that successful explanation is predominantly a matter of form.1 Introduction Social theoretical projects typically depend on arguments about the kind of explanation appropriate to the social sciences. 2000 and in the philosophy of social science Hedström 2005. the explanans comprises the premisses. few philosophers or social theorists would subscribe to the D-N model of scientific explanation or to logical empiricism. for example. criticism and alternative theories (see Salmon 2006[1989]. there have passed more than six decades of debate. Elster . to social structures or to mechanisms are. which social structures or mechanisms we are to invoke. in some sense. Since Hempel’s ‘The function of general laws in history’ (1942) and the more influential ‘Studies in the logic of explanation’ (1948) in which Hempel and Oppenheim proposed the model. The explanandum also occurs in the conclusion of the argument. in Hempel’s Deductive-Nomological (hereafter D-N) model. Machamer et al. within the latter class. at best. to reasons or to causes or. that there is a singular or. largely initiated by Carl G. A. Hempel. The idea that successful explanation in social science hinges on the type of concepts we invoke and. further. Manicas 2006. Thus. therefore B). if we are to be successful in explaining social phenomena.Chapter 2 What Achilles said to Hempel: Unrealistic Restrictions and the Idea of Explanation as a Matter of Form 2. then we need to make use of concepts that are of the required kind. which contain a statement of the initial conditions obtaining at a given point in time and one or more law-stating conditional sentences featuring the initial conditions in their protases and the explanandum in their apodoses. for instance. but rather that. Such arguments attempt to establish that only explanations making exclusive reference. which is constructed after the schema of modus ponens (If A then B. Formulating the assumption in this way allows us to see social theoretical projects as close relations to the philosophical project. also aims at fixing the sentential and argumentative form of proper explanations by the use of formal logical notation. of coming up with a theory of the logic of scientific explanation. The latter project.

Accordingly. In this chapter I will cast serious doubt on the problem of ‘scientific explanation’ and the assumptions on which it is based by examining its moment of inception in Hempel’s work. historically associated with the model. . but have kept the structure of the problem of ‘scientific explanation’ intact. Hempel’s name is. Rather. nevertheless.36 Clarity and Confusion in Social Theory 2007). however. I will further argue that what I have to say applies to philosophers of social science and social theorists insofar as they attempt to fix the form of explanation the social sciences should adopt. 1965) was by no means the only proponent of the D-N model. Most philosophical critics of the D-N model. is incompatible with taking scientific practice seriously 1 Hempel (1942. What follows will throw further light on this claim by showing that Hempel cannot be said to be merely delimiting his object of interest and leaving outside what others mistakenly consider as relevant. 2 The intended sense of ‘unrealistic’ will be made clear in the course of this chapter. and insofar as they show signs of the same confused relationship to social practices that those philosophers of science have to scientific practice. which we consider to be irrelevant to our investigation. whether by making use of the philosophy of science or not. This involves as central the understanding that scientists operate (or ought to operate) on a notion of explanation whose distinguishing feature is that it constitutively applies only to explanations having a certain form. by social theorists as well. The line I will be pursuing is that the idea of a theory of scientific explanation. as we have seen. even more so a formal one. Although in this chapter I will mainly investigate the origins of the problem of ‘scientific explanation’. In the rest of the book we will see that social theorists who dismiss ordinary concepts as irrelevant could be guilty of the same mistake. an understanding which is shared. I have already claimed in the previous chapter that the restriction of conceptual resources forms an important part of the theoretical attitude towards language and that we often fail to appreciate the nature and extent of our commitment to precisely such resources. among others. and by showing that it is a certain species of philosophical problem. I will also try to show that what I have to say applies no less to philosophers who during the past six decades may have criticised the D-N model and proposed improved models. and most critique has been directed at him. the task of the philosophy of science is taken to be to uncover such a form and to express it in a formal theory. which has also been expounded by Popper (1959[1935]). Hempel is excluding what is constitutive of his subject matter and thus imposing incoherent restrictions on himself. take it for granted that the problem of ‘scientific explanation’ is a well-posed problem and that when it comes to ‘scientific explanation’ there is work for a theory to do. Nagel (1961) and Braithwaite (1953). one constructed through the unrealistic2 restriction of resources that can be brought to bear on its solution. It is precisely in the self-evidence of such ideas – and not necessarily in the shortcomings of his model1 which have been widely identified – that the influence of Hempel has really gone unquestioned and is most tricky to expose.

I might perhaps be excused then for following a somewhat tortuous path. They are also logically fundamental in the sense that they possess logical features which are not bestowed on them by the formal logical theory and thus cannot be grounded but only explicated by a theory (obviously by one that does not exclude these very features and which is no longer a theory in the same sense). and try to bring into view Hempel’s motivation for attempting to get to the (metaphysical) logical structure of ‘scientific explanation’. At this point I introduce Lewis Carroll’s version of Achilles and the tortoise and argue that the understanding of logic it evinces leads us from metaphysics back to human practices. on understanding the way in which we are instructed to disregard our ordinary experience. in that sense. flying. To focus on formal specifications as guarantors of explanatory force regardless of the context of scientific practice (either by completely excluding it or by producing some complementary formal account of context. of ‘the pragmatics’) amounts to discarding the complex background of explanatory practice which is necessary if an explanation is to count as such. The main reason for doing so is that realising what is wrong with Hempel’s problem hinges. i. This is 3 Although this is not to say that there is nothing for social scientists or philosophers to do. especially the details of our situated use of language. making a compelling case for the above may prove quite difficult. ‘we first must examine what it is that opposes such an examination of details in philosophy’ (Wittgenstein PI: §52) and. what opposes taking concepts seriously in social theory. In what follows. Obviously. I begin by examining a number of problems (including the well-known paradox attributed to Zeno of Elea featuring Achilles and the tortoise) which treat our ordinary practices in a very special way. Since it is this move which needs to be seen for what it is. I contend. as it does. These conditions are internal to scientific practice and it is only within that context that the presence of a specified form can have explanatory force. I then turn to Hempel’s problem. But this is just to say that there is no room for a theory of scientific explanatory practice.What Achilles said to Hempel 37 for the simple reason that the success conditions of an explanation comprise the whole range of success conditions of a polymorphous action (with the emphasis on both polymorphous and action). for any theory which treats human practices as underpinned by the logical postulates of philosophy or the concepts of social theory.3 Nor is there any room. logically fundamental. in the face of theoretical ‘common sense’. by extension. These variegated examples will yield no necessary and sufficient conditions for something to count as a philosophical problem but they will help us to grasp a pattern. which need to be in place for logic to work and which are. Because this background needs to be in place before one can meaningfully speak of scientific explanation. more generally. . as I have already intimated. any theory is thus bound to be circular: it is scientific explanatory practice which can explain and give sense to the theory and not vice versa. which.e.. conforms to the pattern.

however. further. solvitur ambulando as a road to tackling it is not a real candidate. viz. there is not any meaningful way in which we might try to revise the proposition that the faster runner will ordinarily overtake the slower. by extension. As the problem is posed. namely. an acceptance of this much is necessary to set up one side of the opposed propositions that will generate the paradox. We might try to stage an experiment. What is the point of the paradox then? Inevitably it has to connect to the other proposition and the reasoning that leads to it. the result of that experiment could have absolutely no bearing on the paradox and therefore nothing to do with its resolution.. it is safe to presume that Zeno.2 Setting Up ‘the Problem’ Before examining Hempel’s problem. would concede that under normal circumstances (i. what is ordinarily and empirically the case cannot be the point of the paradox and. by the time Achilles reaches that point the tortoise will have moved to point B. that ordinary experience tells us Achilles will overtake the tortoise.4 It is instructive. Zeno wants to claim that all motion 4 For a detailed account see Salmon (2001[1970]). it follows that Achilles can never overtake the tortoise. as much as anybody else either in antiquity or today. Zeno’s offered chain of reasoning produces the other side of the contradictory pair. for our purposes.e. 2. and so on ad infinitum. One might ask. needs to be revised for the paradox to be lifted. Hence. . First of all. but it is difficult to see what we would be trying to prove by doing so. Ryle (1954) also provides an illuminating discussion and eventual dissolution of the problem. I want to set the stage by considering some problems for the way ordinary facts (and concepts) are positioned with regard to what we are to understand as constituting ‘the problem’ as well as possible solutions. to focus on the preparatory logical work necessary for the paradox to get off the ground.38 Clarity and Confusion in Social Theory the lesson the tortoise teaches Achilles and which he can teach Hempel and. one of the two propositions. given that the tortoise is granted a head start until point A. Let us start with Zeno’s paradox in which it is asserted that the quickest runner (Achilles) can never overtake the slowest runner (the tortoise) in a race between the two because. and what supports them. In fact. philosophers and social theorists following in Hempel’s footsteps. How the paradox is generated might be summed up thus: our ordinary experience tells us one thing whereas reasoning about what will happen tells us the opposite. and if so are they open to scrutiny in the same way? Obviously. that since whenever Achilles reaches the point where the tortoise had been the tortoise will always have moved on. are both propositions equally open to scrutiny. Apart from the fact that we already know what will happen under normal circumstances. in the absence of any exceptional factors) the faster runner wins the race.

Zeno’s paradox is a kind of problem that is designed in such a way so as to teach us. The appropriate sorts of solutions one might look for here might be a mathematical one (a proof that an infinite series [a1 + a2 + a3 … ] can have a sum if the sequence of partial sums [a1. on the use of ‘never’ and ‘all’ in the reasoning that generates the paradox (as those proposed by Waismann. that claim cannot be refuted and the paradox persists. In his collection of essays entitled In Defense of Informal Logic. It is exactly this wish to get behind the phenomena by rendering them irrelevant that I am exploring in the present chapter as a central feature of some philosophical problems and. a distinction involving a hierarchy between the two which is inextricably bound with philosophy throughout its history.. Don Levi discusses logic exercises and argues that they are not the sorts of problems that we encounter in the course of our practical activities. importantly.What Achilles said to Hempel 39 is apparent. focussing. 9). in this case.5 thus. In both cases. Consider in this respect Cora Diamond discussing PI §52 in conjunction with Berkley’s dialogues: Berkeley’s mouse – the one we are concerned with – is the distinction between real things and chimeras … though indeed it is clear on Hylas’ [one of the interlocutors in Berkeley’s dialogues] view that we must use the evidence of our senses in trying to tell what is real. One might take it for granted that it is a product of sophistry and use it as a ‘find the fault’ exercise. In other words. they require schooling and familiarity with the classroom (2000: Chs 8. in getting drawn into a problem. Moving back a step. deny any reality to motion in order to hold that being is unchanging. the kind of problem that it is instructs us how to proceed to its solution. the built-in exclusion of enquiry into ordinary fact remains a stable feature of the exercise. This is perhaps one of the first attempts to introduce a principled distinction between appearance and reality. the paradox functions as a philosophical device instructing us to direct our efforts towards the investigation of ‘reality’ as opposed to the world of ‘mere appearance’.e. i. a1 + a2. from Parmenides onwards. something partly achieved by bracketing off certain routes of enquiry. responses grounded in our practical reasoning. 1954 do).6 but to disregard what ordinary experience tells us because it cannot possibly count as relevant. and that is because however things appear to us. it is not what we actually see or hear or touch that we are ultimately concerned with in such judgements. for instance. A crucial feature 5 The Eleatics. a1 + a2 + a3. In setting up a problem we need to find a way in which what is being said could count as a problem. unless one can find a problem with the reasoning offered. 1997 and Ryle. it is quite another matter how they are (1991: 47). … ] has a real limit. does not tend to infinity) or a conceptual one. Let us consider another example in order to get a better understanding of the sense in which we are being taught by the problem how to go about solving it and of what I could possibly mean by a problem being set up so as to exclude certain responses. . Still. not only to question. 6 Treating it as an invitation to question the reality of what we ordinarily know is not the only way to register the problem. what is important about the above is that it illustrates how. of the problem of ‘scientific explanation’. specifically.

those who feel at home is such settings and have had prolonged exposure to these modes of reasoning (for example. what is at issue in a derivation is that a certain conclusion follows from these specific premisses. 8 Needless to say that objections concerning the obsolete nature of the means of transportation. because they are being required to play a language game in which they are denied (or deny themselves) the conceptual resources to do anything else’ (2009). as far as this book is concerned. that the premisses and conclusion are logically bound together. however. that being taught how to do theory (or philosophy) 7 This can be seen in that (a) the premisses given in a formulation need to be sufficient and (b) adding premisses will not alter the validity of an already valid argument.e. An example of the kind of answer that many who are unfamiliar with such exercises give. according to Levi. are beside the point. or the impossibility of transporting such a heavy load using just a horse and a cart. to the exclusion of all else. let me briefly sketch one of the simple exercises that Levi discusses. Levi underscores just how specialised an activity logic problem-solving is and. Thus. it is an essential feature of logic problems that we take into account only what is stated in the premisses and work through them on the basis that external information is inadmissible. or indeed for any other solution that treats the problem as if it were a request for how-to advice. hence.40 Clarity and Confusion in Social Theory of logic problems is that they cannot be approached using just any means because the manner of proceeding and the tools to be used are fixed. and the logical relations between what is stated. is to misunderstand the exercise. is that X might ask for somebody else’s horse in order to get the job done. philosophers and theorists) might find it very difficult to avoid taking that route. which is not asking for a way of locating a horse. drawing. even though it may lead to the sort of dubious results described by Gavin Kitching (2008).7 To get a sense of this. as they do. It runs roughly along the following lines: p: X wishes to transfer a ton of barley q: In order to do so. Put differently.8 The problem is about figuring out the relation between the propositions given and the conclusion in question and one is supposed to produce an answer after taking into account what is stated. on knowledge that is ‘external’ to what is given in the statement of the problem. That. Kitching scrutinises undergraduate dissertations written by ‘good students of Politics’ (2008: xi) who place themselves in the postmodernist/poststructuralist tradition. how questionable it is to automatically presume that its procedures have validity outside the specialised setting in which it is normally practiced. i. Kitching’s observations point to the central fact. X needs a horse and a cart r: X possesses a cart but not a horse. At the same time.. for the way they handle theoretical language and suggests that in ‘peculiar settings students are left with little conceptual choice but to treat abstract nouns as the names of curious objects. The question posed at the end is whether X can transfer the ton of barley or not. .

to use another example. Consider in this respect Socrates’ (ironical) rejoinder to Thrasymachus in the Republic which may well have been directed at many subsequent cases in the history of philosophy of ending up with an unsolvable problem by restricting the set of acceptable answers: That’s because you are wise. 1). pictorial ways of thinking may be involved in making sense of the epistemological subject which encounters the world in the form of objects or. that is. rigid models for the use of expressions which may end up restricting rather than enlarging the range of our available conceptual resources. in many cases. for. acquiring a set of visual metaphors. but must I say something different from the truth. Such grammatical pictures may mislead theoretical investigation. a number of linguistic operations and the forms of reasoning that accompany those as well as. 10 See Diamond (1991: Ch. . etc.What Achilles said to Hempel 41 involves learning a vocabulary.9 all of which can be used in the formation of grammatical pictures. Thrasymachus. Suppose he had said to you. groups. reason and talk can be rendered irrelevant by the way a (philosophical) problem has been set up. ‘[w]e 9 For instance. In Diamond’s pertinent words. what do you mean? Am I not to give any of the prohibited answers. or what do you mean?’ (Plato Republic: 337 B–C) It is notable that Socrates voices this objection against Thrasymachus but does not consider it as applying to the way he was trying to answer the question ‘what is justice?’ which was arguably not all that different. if the thing really is one of these. in speaking of boundaries and boundary drawing with regard to concepts. and in putting the question warned him: don’t you be telling me. ‘Thrasymachus. for I won’t accept any such drivel as that from you as an answer – it was obvious I fancy to you that no one could give an answer to a question framed in that fashion. The examples provided so far are hopefully sufficient to illustrate the way in which how we ordinarily experience. The above quotation renders perspicuous the danger involved in the posing of philosophical problems and questions whose formulation is ‘unrealistic’ insofar as it excludes from the set of possible answers what ‘the thing really is’. fellow. not even do you mean to say. and so you knew very well that if you asked a man how many are twelve. dealing in abstracted grammatical pictures while refusing to look at the nuance and complexity of relevant concepts not only ensures that they are misunderstood but ends up confusing the order of what explains what. identities. that twelve is twice six or three times four or six times two or four times three.10 The upshot of the discussion is that the restriction of resources can be (at least in some cases) misguided in that what is left to draw on renders the problem unsolvable. One could go on. returning to a previous example. in the same way that treating classroom logical puzzles as forming the basis of everyday practical reasoning is not only to misconstrue the latter but to put the cart before the horse.

. The schema is intended to play the role of a sort of criterion or conceptual tool. accept his posing of the problem. This kind of frustration is expertly dealt with by Avner Baz (2012). refute Hempel or to support a different theory or offer an improved model of explanation. 12 That this problem keeps repeating itself in philosophy can be seen from a statement Oswald Hanfling provides in the closing section of his discussion on Kripke/Putnam on meaning/reference: The remarks just made are claims about our use of the word ‘meaning’.e. except to say that no other approach seems to make sense (1984: 205). i.e. who also notes that the ordinary language philosopher and the ‘mainstream’ analytical philosopher perceive each other as begging the question.11 It will be argued that Hempel excludes precisely that which turns out to be the only ‘realistic’ solution to the problem he poses himself regarding ‘scientific explanation’. . one has to play by his own terms. The fact that Hempel (like other proponents of the D-N model) removes himself from objections13 that appeal to ‘the semantics/pragmatics of explanation’ means that refuting him entails accepting at the same time that the logic of our 11 ‘Similar lines’ are considered lines proceeding by ‘unrealistic restriction’ and not necessarily by embracing the D-N account. Hempel can retort to an early such objection voiced by Scriven (1988[1962]) in the following manner: ‘Explicating the concept of scientific explanation is not the same thing as writing an entry on the word “explain” for the Oxford English Dictionary’ (Hempel 1965: 413). The question was whether the meaning of a word consists in its use. and to indicate what kinds of assumptions and relations may be challenged when an explanation breaks down (Bartley 1962: 17–8). Hempel ensured this by sealing the notion of ‘scientific explanation’ off from what we (lay folk and scientists) ordinarily understand by ‘explanation’. i. What has been sketched so far will be used as a device which will allow us to examine in closer detail the ways Hempel sets up the problem for himself and those who partook in debate on scientific explanation along similar lines. 13 A further line of defence is noted by another defender of the model: [T]he schema is [not] intended as a picture of the verbal form a scientific explanation may be expected to take … only to what might be called the ‘underlying logical structure’ of the process of explanation. and to answer it by pointing out how the word ‘meaning’ is itself used. Apart from the fact that he has been heavily criticised already (Harré 1970 and the essays collected in Pitt 1988). To prevent misunderstanding it must be emphasised once again that the point of what follows is not to. Objections to the model based on any alleged discrepancy between the two are met with the flat denial that such a discrepancy is relevant.42 Clarity and Confusion in Social Theory ask philosophical questions about our concepts in the grip of an unrealistic conception of what knowing about them would be’ (1991: 66). But perhaps it will be objected that this approach begs the question. in order to do so. strictly speaking.12 Thus.. To this objection I have no answer. detailed specifications of scientific practice. is to assume the point at issue.

14 However. In other words. in the same sense that learning modus ponens or the argumentative fallacies is potentially helpful (and not much else) when it comes to realworld arguments. On the contrary. (1965: 412) I want to briefly direct attention to a few things that could be said about this passage. 2. that the combined effect of generalisation and deduction in the model can rule out possible events. 14 The way of understanding ‘the pragmatics’ will be addressed at a later section. i. But to grant this is to grant too much. and that a conclusion which turns out to be empirically false will entail that at least one of the premisses is false too.What Achilles said to Hempel 43 ordinary word ‘explanation’ (in both its lay and scientific uses) cannot have anything to do with his problem. The way this concern enters into his project. he does acknowledge the ‘pragmatic aspect’ of explanation. is of a schema that cannot be said to be a theory or grounding of science but only a possible guide.3 Hempel on ‘Scientific Explanation’ Although Hempel has decided not to consider the concept of explanation as it is used. like Bartley’s. Ryan’s defence. he is after the logical structure of ‘scientific explanation’. . however. Curiously. he is not interested in the actual explanations scientists provide to each other or publish in journals. Hempel seems to be setting up an opposition between what scientists do and what science does.. The construction of our models therefore involves some measure of abstraction and schematization.e. Hempel fails to appreciate the extent of his commitment to the concept of explanation. he fails to take the concept seriously. Their purpose is rather to indicate in reasonably precise terms the logical structure and the rationale of various ways in which empirical science answers explanation-seeking why questions. however. His use of the abstract noun ‘science’ is presumably meant to refer to the standards by which scientists proceed and thus seems intended to preserve the sound idea that scientific standards do not depend on what any one scientist does. Suffice to say for now that I am not advocating a pragmatic theory of explanation. An explicit statement of the intent of Hempel’s explanatory models provides a good starting point to the discussion: [T]hese models are not meant to describe how working scientists actually formulate their explanatory accounts. creates a Ryan (1970) defends precisely such a use of the model in the social sciences on the grounds of its logical merits. I will argue that it is not the case that Hempel can consistently claim to have a well-defined subject matter separate from the pragmatics and thus be able to complain that he is held accountable to what is external to his project. but rather an unrealistic attempt to separate logical structure from human practice. My task throughout this section will be to show that this is not a simple case of delimiting one’s phenomenon of interest.

Before proceeding to explore this direction. I decide to ask the person sitting next to me if they can explain what is going on. In this case I might want to call what I am told by the physicist a scientific explanation and contrast it to the lay explanation I might have gotten had she not intervened. The very fact that we are instructed to proceed thus introduces the situated aspect of explanation giving as a response to someone’s question. if it is science which explains and not scientists. one may wonder. one is also led into the paradox that it is impossible for science (understood as not involving scientists) to explain anything to itself for if it is in a position to explain then it already understands and therefore there is no need for an explanation. The idea that we are concerned with ‘scientific explanation’ and not just any kind of explanation is important here in that it nourishes the belief that there must be something special about the former. In examining this matter we need to ask exactly what the distinction between scientific and lay explanation might hinge on. as if we could take away scientists and still have science. once again. however. Such a use would be consistent with our lay concept of explanation in that the physicist’s response is an appropriate one after our expression of puzzlement. and is it asking these questions of itself?) calls upon us to imagine a situation in which the posing of a why-question is appropriately followed by a ‘scientific explanation’. the opposition between ‘science’ and ‘scientists’ makes possible the contrast between ‘surface formulations’ and the deep logical structure of explanation. Apart from the opposition between ‘scientists’ explanations’ and ‘scientific explanation’. the exclusive appeal to explanation-seeking why-questions. she is given priority. Remember. perhaps some particularly close connection to truth that is simply not there in other kinds of explanations. even though it would not mark the giving of a lay explanation (see below). A physicist happens to be sitting behind me and offers to explain instead. which is what is thought to lie behind the explanations scientists give (and thus can be further said to belong to science and not scientists). that Hempel wants to dissociate his investigation into ‘scientific explanation’ from the everyday logic of the word ‘explanation’ and from the ‘verbal forms’ scientists use.44 Clarity and Confusion in Social Theory tension between science and scientists. The adjectives ‘scientific’ and ‘lay’ are used here to modify ‘explanation’ by standing in the relation: ‘given by a scientist/ a layperson’. Moreover. which is precisely what Hempel wanted to abstract away from. . it is worth inquiring into the use of ‘scientific’ a bit further. the presence of the term ‘scientific’ is presumably meant to mark the further contrast between whatever ‘scientific explanation’ turns out to be and ‘lay explanation’ or explanation simpliciter. to whom is science doing the explaining? To utilise the manifest connections between the concept of explanation and the concepts of audience and understanding even further. In the case above we might also find another reason to call the explanation ‘scientific’. Finally. The question is: can Hempel really do that? Suppose I am on a train passing through the English countryside. Furthermore. apart from suggesting further possible absurdities (is ‘science’ asking these why-questions. It has just rained and a glorious rainbow has appeared – a phenomenon by which I have always been puzzled.

15 Again. both from (the verbal form of) workaday scientific explanations and from workaday lay explanations. apart from the explanation’s content we might wish to single out its particular form. in what follows. The use of formal criteria that bypass ‘the pragmatics’ is presumably intended to (a) (along with other criteria) guarantee a way of recognising proper explanations and weeding out pseudo-explanations. This point also comes out in the comparison he has in mind between formal mathematical proof and super-scientific explanation (which will be discussed below). for example that the premisses comprising the explanans have to be true. there is a consequential difference between them. if syntactically defined. 16 Under a certain understanding of these terms. that they should state something observable and testable. i. However. Indeed. Hempel (and Oppenheim) put forth a whole range of criteria. it is important to note that fixity of form bears the implication that the varying sentential content is not taken to contribute any essential features to explanation and cannot be that which ensures that an explanation is super-scientific. focusing on the criteria dictating proper sentential and argumentative form. . for instance.e. however. based on what we have seen. this is exactly Hempel’s idea and it is formal criteria that he proposes as constituting ‘scientific explanation’. then.. and (c) fix the former as underpinning the latter in the case of science. which is meant to be distinguished. 17 This designation is meant to remind us of the temptation to look for ‘superconcepts’ (Wittgenstein PI: §97). A question 15 Compare van Fraassen: ‘To ask that their explanations be scientific is only to demand that they rely on scientific theories and experimentation. when Hempel speaks of the ‘logical structure’ of super-scientific explanations he is. (b) distinguish the superscientific notion of explanation from the workaday notion of explanation. The idea is. As is well known.What Achilles said to Hempel 45 namely what requires explanation is a natural phenomenon that falls within the purview of science. etc. Since the validity of a formal argument is exactly of this order. If we accept that it is not sentences but statements or utterances that can be logically connected. have reason to call an explanation ‘scientific’ based on features of its content such as the presence of technical terminology and/or mathematical expressions. Since form is Hempel’s hallmark of super-scientific explanation. however. another way the opposed phrases can be understood where ‘scientific’ and ‘lay’ are meant to characterise the explanation itself. There is. Finally. it can hold. We might. not on old wives’ tales’ (1980: 129). the situated giving of an explanation is back into the logical picture. only by virtue of the argument’s form. this conforms to the lay concept of explanation. that there is something peculiar to the ‘logical structure’ of the explanation as a series of sentences or utterances16 which is what makes it scientific. I will use ‘super-scientific explanation’17 to refer to Hempel’s sense. To avoid confusion. formal logic is summoned to fix what counts as an explanation. in all likelihood.

can be neither sufficiency nor recognisability. What is clear. though all instances of the proposed formal sentential sequence are not explanations. which invoke cases where the premisses entail the conclusion without being topically relevant to it. their model is to underpin explanatory practice and not vice-versa) in judging whether something is a genuine explanation. namely. For all the proposed criteria. In other words it is because Hempel (and his critics who have theories of their own to propose) illicitly draw on ‘pragmatic’ criteria of explanation18 – par excellence the criterion of relevance – that they can see a model as capturing explanations in some cases and as inadequate in others (as in cases of asymmetry between explanation and prediction or the aforementioned case of irrelevance. geflunkt would not do because of the possibility of confusion with the word the Churchlands should have employed. But for this to have any purchase (to be something other than pure stipulation) there must be something that the form contributes to explanation. etc. the understanding that an explanation needs to informatively address what is being asked. if this strategy were indeed sound then Hempel’s model would not be open to the well-known objections that have been raised against it. all super-scientific explanations necessarily take this form. . see Salmon 2006 [1989]: 46–50). as has been argued above. requested. 3). one might wonder? Let us take stock of the insistence on this requirement and move on. is that in choosing not to look at scientific practice but rather to stipulate what super-scientific explanation is. strictly speaking. it is to make use of one’s linguistic and praxeological competence. Whence then this preoccupation with form. it seems that we still need to rely on what anybody who wanted to identify an explanation has to rely on. However. And what it contributes. Rather. Hempel (and other philosophers) can appeal to no other criteria apart from the ones proposed (if. Hence.19 It follows then that Hempel cannot hold on to both of the propositions that (a) the super-scientific notion of explanation differs from the ordinary one. according to Coulter and Sharrock. that is. If Hempel really wanted to demarcate his notion from the ordinary one he should have named it geflunkt (Coulter and Sharrock 2007: 4) in order to avoid residual reliance on any of its other senses. and that (b) form is sufficient (jointly with other proposed criteria) for something to count as a scientific explanation. nevertheless. cases where not even Hempel would want to say that the model had produced an explanation. 18 This is just another case of philosophers rejecting ordinary concepts and then going on to rely on them in order to argue whether something can be properly called ‘knowledge’ or an ‘explanation’. which has nothing to do with what philosophers call ‘applying a concept to a case’ (see Baz 2012: Ch. in order to avoid relying on any of the ordinary senses of ‘remember’! 19 This is not to rely on ‘intuitions’. Of course.46 Clarity and Confusion in Social Theory remains as to whether Hempel intends that all groups of sentences that meet the set of criteria are ipso facto explanations. In other words. Hempel’s stipulations on explanation are at best necessary conditions.

specifically. But it is worth examining what else Hempel has to say about the concept of super-scientific explanation. based upon … laws and theories are meant to be objective in an analogous sense. It is this nonpragmatic concept of explanation which the covering-law models are meant to explicate … [T]o call attention to the important pragmatic facets of explanation and to indicate the diverse procedures that may be appropriate in different cases to dispel the perplexity reflected in someone’s quest for an explanation must be hopelessly inadequate. i..e. i. just as analogous arguments concerning the notion of proof cannot show that nonpragmatic models of proof must be sterile and unilluminating. as it were. . The case for scientific explanation is similar … explanations. (1965: 426–8) (my emphasis) Hempel has been careful to address our worries about the notion of explanation and to give us an example we seemingly cannot argue with.What Achilles said to Hempel 47 We have (at least provisionally) concluded that if asked whether Hempel can dissociate his concept of super-scientific explanation from the ordinary notion like he thinks he can the answer should be in the negative: the appeal to whether we would say that something explains is constantly made and the super-concept cannot replace the ordinary one. meta-mathematical theory where there is a firm conception of what will count as a proof and thus 20 I have discussed above how what has been excluded from the problem cannot be brought up as an objection without being seen as question-begging. It is therefore beside the point to complain20 that the covering-law models do not closely match the form in which working scientists actually present their explanations. It comes right after he has acknowledged ‘explanation’ as a pragmatic/ relative notion.e. In the way he positions himself with regard to that aspect we might find an answer to the question about the insistence on formal criteria. Concepts of proof which have this character can be defined once the mathematical discipline in reference to which the concept is to be used has been suitably formalised. the contrary is the case. as well as the predictions. and which does not require relativization with respect to questioning individuals any more than does the concept of mathematical proof. explanation for a person. from the pragmatic one. This ideal intent suggests the problem of constructing a nonpragmatic concept of scientific explanation – a concept which is abstracted. As is well known.. we clearly need a concept of proof … in terms of which it makes sense to say that a given argument Y is a proof of a given sentence (X in a theory) without making any mention of persons who might take cognizance of Y. [F]or the purposes of mathematics and logic as objective disciplines. how he reasons about ‘the pragmatic aspect’ of explanation. The following passage is very illuminating with regard to Hempel’s logic and motivation and exhibits how his reasoning relies heavily on construing the notion of explanation along the lines of the notion of proof.

2. Lewis Carroll provides an imaginary recounting of what happened after the race in order to show the impossibility of including the rules of inference used to draw a conclusion in the premisses of an argument.21 It may no doubt be objected that the same is true of mathematics and that this does not show the impossibility of an objective concept of proof. whose formulation. he thinks that the desired impersonality of a foundational standard whose objectivity will remove it from any possible dispute can only be achieved by ‘abstracting away from’ (understood as eliminating) ‘the pragmatics’.48 Clarity and Confusion in Social Theory where nobody can dispute that something is a proof if it meets certain standards. however. is that just like arguing about philosophy is part of philosophy and dealing with the word ‘orthography’ is part of orthography (Wittgenstein PI: §121). that we are inferring from premisses A and B to conclusion C using the rule of inference ‘If A and B then C’. it is this compulsion which is the key ingredient in super-scientific explanation. and not an external formal reconstruction of the practice of mathematics (which itself is a formal discipline). the rule of inference (call it D) is not part of the premisses. on pain of regress. for example. meta-mathematics is a part of mathematics too. alluding to Zeno’s paradox. Presumably. pedagogy or merely a set of empty signs. be that of proof or explanation. What is ultimately being appealed to. seems to be forgetting that it is only because of the connection with scientific explanatory practice that a formal standard can be said to be a standard of scientific explanation as opposed to a standard of. to use Ryle’s felicitous 21 Compare here Winch’s point (2008[1958]: 83–4) that it is only the relation between economic practices and the social scientific concept of liquidity preference which can render the latter as being about those practices. he needs to presuppose them and a whole lot more besides. It is evident that in Hempel’s reasoning logic functions as a guarantor of the impersonality of scientific standards and human practice is understood as epiphenomenal when compared to those standards. in order to invoke such a standard what Hempel needs to do is exactly the opposite. as he puts it above. then. Hempel. since it carries the promise of the ‘nonpragmatic aspect’ that Hempel wants to bestow upon this notion. The difference. The same cannot be said about formal models of scientific explanation and scientific practice. What this shows is that Hempel is setting himself an impossible task: to provide a foundation of ‘scientific explanation’ while taking away that which constitutes ‘scientific explanation’ as such. Suppose. In fact. will not require any mention of persons. Naturally. Moreover. is the hard logical must which is what compels one to accept a standard. say. I shall now bring Achilles and the tortoise back into the discussion. however.4 What Achilles said to Hempel In his (in)famous ‘What the tortoise said to Achilles’ (1895). . however. it is just an inference ticket.

the drawing of an inference takes us into the domain of human action. nor owning a ticket. The lesson Achilles is meant to learn in the story is that he could never convince the tortoise to accept that C follows from the premisses in that way. ‘[n]either buying a ticket. having learnt the lesson from his encounter with the tortoise. in short. . nor showing or transferring a ticket is travelling’ (2009[1950]: 250). for instance. like. In the D-N model the explanation is a derivation of the explanandum from the explanans (comprising statements of initial conditions and laws). the rule is just like a ticket in the sense that. interpretation are practices. then in order to infer C we would need a new rule. to avoid pressing charges against an attacker) – that show how those doings function in certain circumstances. hence subject to all the considerations that apply to actions’. and the rule does not supply its own interpretation. or rule-like additions to rules’ (1997: 81). ‘makes the same point about signposts. additional mental images. Application. 2). as Ryle puts it. which allows us to write C under A and B. Again.. 24 This is Gilbert Ryle’s term and is meant to direct attention to the fact that ‘to explain’. if we attempted to write down rules of inference as premisses (as the tortoise keeps asking Achilles to do in Carroll’s story).22 This is. i. The only way for the inference to be drawn is for someone to draw it. 22 We need not enter here into the categorial question of whether inference is an action – but see Brown (1955) and White (1971) for a sense of the complexities in categorising ‘inference’ and ‘infer’. But the drawing of an inference based on a rule cannot be the mere inclusion of the rule as a premiss nor appealing to or citing the rule upon request. in Goldfarb’s words. to lend someone money. in other words. in this case if we were to include D as a premiss. they are not further signposts. projection. there is an indefinite number of things that may function as ‘explanations’. Achilles understood well that actions – especially those that are described with ‘adverbial verbs’24 such as ‘explain’ – are polymorphous (White 1967: 58) which means that (a) they can assume many different forms. the mental picture does not determine its projection onto the world. the first thing that Achilles. the explanation hinges on an inference.What Achilles said to Hempel 49 expression. namely E (If A and B and D then C). ‘to help’ and ‘to show mercy’ are descriptions – which can be substituted for more specific descriptions of doings (to describe the properties of alkalis. and so on ad infinitum. to cite only a few.23 said to Hempel: ‘Even if we accept the D-N model. mental pictures. the fact that it requires the drawing of an inference means that it irremediably remains within the field of human action. Winch (2008[1958]: 52–4). and … rules: the signpost does not determine its application. In cases where one does not they may be instructed but (logically) not forced to realise that it does (any more than one can be forced to understand). Now. As a first step in seeing the relevance of Lewis Carroll’s story let us recall the formal structure of the D-N model. Hamlyn (1956: 366) and Wittgenstein who. 23 Also with the help of. The important thing to note is this: the fact that the conclusion follows from the premisses is something one sees. If only Hempel and many of his critics had understood that lesson. Ryle (2000[1949]: Ch. then. Why this is the case might not perhaps be immediately obvious.e.

with (formal) arguments. apart from completely disregarding the fact that. or employed in a different context. In both these cases we may say that the person (or the animal) is not merely disagreeing with us but is rather so removed from our way of life that it is not clear how to make ourselves understood and. abstracted from the context of its use. Accordingly.25 does not necessarily count as the action. unless our ordinary notion of explanation is in place. 25 Context is not a formally specifiable notion: it can include everything from features of a particular setting to the shared biographies and relations that link together particular individuals. make sense of what they are saying. But. Thus. the second thing that Achilles said to Hempel is this: ‘Even if we accept the model as a standard of explanation. In this case Hempel wants to use logic to fix the notion of explanation absolutely but. Ryle gives another more imaginative (and for some macabre) example: ‘To replace the infant’s fingers and feet with pliers and pedals would not be a good plan – especially as the employment of pliers and pedals themselves depends upon the employment of fingers and feet’ (1954: 35). it must be remembered. Hempel is trying to ground ‘scientific explanation’ in logic by treating the latter as something transcendental. our common ways of proceeding. as a key to the discovery of metaphysical truths in which humans need play no part. the microscope is usually only as good as the eye that looks through it. or is otherwise tied to the practice of explanation giving. he is also forgetting that for the explanatory standard to count as such. in turn. there is nothing against which to judge his explanatory standard. a lot more is needed apart from the standard itself.26 This ‘more’ which needs to be in place (and which is missing in Carroll’s story) is our common practices. in this case. is like the person who utters ‘I don’t know if I have ever been on the moon. this presupposes the existence of a whole community which agrees on its actions and reactions (even though it might not agree in its opinions)’. the form by itself. that ‘All men are mortal’ and that ‘Socrates is a man’ but not that ‘Socrates is mortal’. His motivation is very close to the one Goldfarb (1997) identifies in discussing the desire for ‘fixity of meaning’ and involves trying to get to an underlying level which can provide absolute determinations without relying on any facts about us and the world proceeding ordinarily. Therefore.50 Clarity and Confusion in Social Theory and (b) they are connected to each of these forms in a contingent way. What Achilles was getting at is that the tortoise who accepts. 26 The temptation here is equivalent to the one felt by someone who thinks that because the microscope is much more powerful than the eye it can do a much better job by itself. But Achilles did not stop there. unless a formal model is actually used to give explanations. One of the morals of Carroll’s story is that logic cannot work without this shared background and thus that any appeal to (formal) logic is necessarily simultaneously an appeal to such a background. I don’t remember ever having been there’ (Wittgenstein OC: §332). . it cannot be said to count as a model of explanation at all. say.

Hempel considers completeness in the sense of closure. And the explanation itself can be said to be complete provided that what is being explained has been singled out as something in question and separated from what has been presupposed as not’. it contained in turn an explanation. for example. and argues that nothing would be left unexplained ‘if for every fact or law that it invoked. accordingly. Again. if it simply addresses the lack of understanding that motivated it and not any conceivably possible lack of understanding or puzzlement (be that about a concept or a particular natural phenomenon). One might say: an explanation serves to remove or to avert a misunderstanding [or here puzzlement about a phenomenon] – one that is. To this Achilles responded: ‘If the pun may be excused. is a prerequisite for complete explanation.What Achilles said to Hempel 51 The third and final thing Achilles said to Hempel has to do with the notion of explanatory completeness. his views on completeness are actually based on the same picture as the one Wittgenstein criticises in the following passage: [A]n explanation may indeed rest on another one that has been given. there is a perfectly serviceable sense in which an explanation can lead to closure. so that secure understanding [explanation] is only possible if we first doubt everything that can be doubted [or consider that everything requires explanation]. there is a radical sense in which the model is neither complete nor incomplete unless it has been used in the giving of an explanation. Although Hempel is not talking about explanations of meaning and it is easy to get confused here.27 This is just another way of saying that explanations are situated. which according to Hempel (1965: 421). by having all the parts of the D-N model or by the explanandum being entailed precisely as stated by the explanans. of the ways in which a complete explanation might be achieved. the account is incomplete. that would occur but for the explanation. It may easily look as if every doubt merely revealed an existing gap in the foundations. namely. (PI: §87) Thus. Achilles’ point is that although Hempel acknowledges that what he calls closure leads to a never ending regress. . Finally. he does not see that this becomes a worry only given a certain picture that holds that the adequacy of explanations should be absolute and that a proper explanation will guard from any possible confusion or misunderstanding. not everyone that I can imagine. not only in the sense of happening within a 27 This point also applies to the idea of a complete description of some phenomenon. and then remove all these doubts [try to provide these explanations]. Wittgenstein considers this way of thinking in PI where he discusses the completeness of an explanation of the concept of game. Hempel produces an elaborate discussion (1965: 415–25) of the various senses in which one might speak of incompleteness and. but none stands in need of another – unless we require it to prevent a misunderstanding. But completeness in this sense obviously calls for an infinite regress in explanation and is therefore unachievable’ (1965: 423).

there have been. for example. has upheld it by considering language use within the project of a theory of language. Bechtel 2011). This is not the point I have been Completeness of description is only unattainable if we continue to uphold the picture criticised here. the development of the mechanismic approach to explanation. an ordered pair of proposition and type of illocutionary act (1988). as a product of the relation between phenomena (Friedman 1988[1974])28 and thus as “in the world”. His model is now discredited and philosophers have mostly learnt from his mistakes. This is what Achilles teaches us too.52 Clarity and Confusion in Social Theory particular context. an activity (Kitcher 1981) and. as Pär Segerdahl (1996) shows. as Woodward (2011) reports. readers who are well-versed in philosophical debate around ‘scientific explanation’ will say. but also in the sense of having a shared way of going about to hold them in place. however. apart from what Achinstein summarises as “the entity view”. continuous. deriving from explanatory practice in biology (Machamer et al. 30 The case is similar in linguistics where. at the very least. van Fraassen and Achinstein understand explanation as (at least partly) a human action. Furthermore. The first two properties express the desire to get behind the appearances of explanation by seeing it as a relation between phenomena in the world. Finally.e. Thus. The third property is an attempt to acknowledge that the concept of explanation is connected. some philosophers who treat it as an irremediably pragmatic notion. 28 Friedman (1974) extracts from previous candidates the following three properties that a theory of explanation should have: it should be sufficiently general. .29 There is. at least somewhat. philosophers tend nowadays to take scientific practice (more) seriously. to return to Achinstein. 2000. philosophers after Hempel have used examples both from scientific and ordinary settings while regarding the two as.’ In response to this it may be conceded that some of the arguments I have put forth are akin to the arguments provided by those who emphasise that explanation is a ‘pragmatic’ notion.5 Hempel and Since: Taking ‘the Pragmatics’ Seriously ‘This is all very well’. objective and connect explanation with understanding. logically connected. to the concept of understanding. the development of ‘pragmatics’. according to which explanation is a sentence or a proposition which can be characterised independently of an explanatory act (1988: 221). 29 For example. While it is true that some continue to conceive of explanation as objective. as testifies.. for example. they seek to supplement or replace other philosophical theories with a ‘pragmatic’ theory of explanation. there are alternative conceptions which view explanation as an answer to a why-question (van Fraassen 1980). a sense in which both Hempel and pragmatic theorists fail to take ‘the pragmatics’ seriously. 2. i. ‘but philosophy has moved on since Hempel. far from challenging the picture of language presupposed by ‘semantics’.30 Although theorists such as. Achinstein is of course correct in saying that one cannot study the ‘entity’ without studying the action.

Equally.e. scientific explanation comprises the explanatory practices of scientists. is a more defensible philosophical preoccupation. as it does. To understand that explanation is a pragmatic notion is to realise that a theory of explanation is out of place because any theory or attempt at analysis (as opposed to elucidation or explication) of models. This. weather phenomena.31 Thus. although it may be true that philosophers do not nowadays consider lay and scientific explanation as discontinuous. 32 We are here reminded of Bouwsma’s ‘Descartes’ Evil Genius’ (1949) in which he argues that percipio turns out to be cerpicio. making something perspicuous. The idea that there is a difference between scientific explanations of. the origins of the cosmos or the causes of disease and the ones non-scientists are capable of providing creates potential anxiety regarding the status or even the very existence of the social sciences. . at the very least. In the same way that what Hempel calls ‘scientific explanation’ is really ‘superscientific explanation’. i. so the reference to people is built into it. 2. Acknowledging that explanatory practice is the logically fundamental thing entails embracing the fact that the explication and elucidation of scientific concepts. it receives a rather special emphasis.6 Implications for the Social Sciences When social theorists and philosophers of social science approach the question of scientific explanation. it neither grounds nor transcends them. targeting at the same time a specific misunderstanding or puzzlement and not any conceivable one. It is thus done for somebody. has an internal connection with understanding. i. I submit. contexts.e. Responsible for the discrepancy are the unrealistic restrictions which philosophers impose in their urge to look for ‘super-concepts’ that will allegedly ground human explanatory practices.. of practices of language use and of other aspects of scientific practice. is a bootstrapping operation presupposing and relying. etc. sentences. on our praxeological-conceptual competence at the same time that it attempts per impossibile to find an external ground to that competence. for example. What mental cramp or muddle is at issue each time and how specifically it is to be addressed is a specific business which is geared to ‘the pragmatics’. despite appearances. Such enquiry must be conducted in a piecemeal fashion. they must be seen as providing. because elucidating.. the problem lies elsewhere. it is thought. But the logic of explanation is constituted by our explanatory activities. explanations which 31 This point is forcefully made by Sören Stenlund throughout his excellent Language and Philosophical Problems (1990). Descartes’s sceptical thesis is not about our ordinary concept of perception but about a different one. to take this fact seriously is to treat such practices as the logically fundamental thing.. If the social sciences are legitimately to exist. much like – it has been argued – explaining. is the enduring philosophical legacy of Hempel’s problem.32 the discrepancy still remains between philosophical requirements on ‘explanation’ and the explanatory practices of scientists and lay folk. that.What Achilles said to Hempel 53 trying to make in invoking ‘the pragmatics’.

This nourishes the idea that social scientific explanations of the workings of the economy. 33 We will see in following chapters. consider. how this procedure removes any reference to the details of an explanandum which can help us decide which explanatory form to adopt. as both (where perennial philosophical ideas of our models reflecting what is out there can apply).33 Thus.35 The commitment to a special form of social scientific explanation. Manicas 2006. 34 The commitment to fixing explanatory form is also present in more ‘hermeneutical’ approaches such as those emphasising the role of narrative but also subscribing to a programme of causal explanation (Uebel 2012). based on the arguments we have considered in this chapter (see also Reiss 2007. is nothing more than a number of forms of words which. based on what we have seen) as crystallising the criteria used by scientists in their explanatory practice. . One way theorists have tried to fix the form of social scientific explanation is by looking. beyond mere description. These accounts are then used in order to stipulate what explanatory practice in the social sciences ought to look like in the future. Hedström and Ylikoski 2010 for an overview). 35 It is worth noting that there is some ambiguity regarding the understanding of mechanisms (cf. the explanatory models that are thought to constitute improvements upon the D-N model (for example Hedström 2005. part of which involves settling on the form of explanation as a prior step to theory construction. There is a broad consensus on this point amongst approaches with scientific aspirations34 such as Critical Realism (starting with Bhaskar 2008[1975]) and Analytic Sociology. however.54 Clarity and Confusion in Social Theory are improvements on ‘common sense’ explanations. Both approaches hold that proper explanations are mechanistic (or mechanismic) ones (cf. however. as our models of reality. or more generally as a vocabulary in which the explaining is to be conducted. The emphasis on mechanisms starts from the commonplace understanding that social science is interested in ‘how things work’ and treats this interest as carried through by the mechanistic idiom. With regard to Analytic Sociology. Hedström’s claim that ‘mechanism-based explanations are the most appropriate ones for sociological theory’ (2005: 33) and Demeulenaere’s stagesetting intended to lead to the same result: ‘Whenever we start explaining “why” something happens. the overall function of certain social practices or the beliefs pertaining to certain social groups are somehow in competition with the ordinary explanations members of the society in question can provide. for example. serves as a way of validating the (hierarchical) distinction between the two. since those are in rather short supply. Demetriou 2009) which translates into different senses of the fixing of explanatory form. Adherence to the idea of the special form of social scientific explanation. not of course to the unilaterally accepted and successful explanatory practices of the social sciences. especially in Chapter 6. then. Mechanisms can be thought of as a deep reality. Elster 2007) still keep very much alive the Hempelian idea that genuine explanations are those possessing a certain form. we are necessarily led to introduce some type of causal linkage that in turn raises the question of mechanism’ (2011: 2). but to philosophical accounts of the natural sciences which they take (falsely. The mechanistic idiom.

like the difference between explaining and describing.7 Conclusion In this chapter I have dealt with the problem of ‘scientific explanation’ by examining its origins in Hempel’s thought and attempting to expose the mistaken assumptions on which the problem is based. Although Hempel identified his subject matter as ‘scientific explanation’. Even though this claim may receive an uncontroversial interpretation. it often amounts to the idea that what explanations have that descriptions lack is the causal form. 36 This is entailed by the fact that. at the same time he prevented himself from looking at the scientific practices and the concept of explanation informing them. this is to hinge explanatory credentials on a certain form. Yet again. however.36 2. Scriven puts the point nicely when he says: ‘the difference between explaining and “merely” informing. the appropriateness being a matter of its relation to a particular context’ (1988[1962]: 53). rather than being capable of providing a theoretical grounding of such practices ends up presupposing them. but in its being the appropriate piece of informing or describing. these assumptions have proved much more resilient than Hempel’s now largely discredited D-N model. I then proceeded to show how the problem Hempel sought to address excludes from consideration what turns out to be the logically fundamental thing or ‘what scientific explanation really is’. Unfortunately. Anderson and Sharrock 2013). as they continue to underpin a significant part of philosophy and social science. as logically secondary to philosophical theory. the D-N model. The argument bears significant implications for philosophers whose work follows the project Hempel initiated and who thus continue to labour under a Demetriou 2009. . that the formal vehicle of that theory. thinking he could treat them as irrelevant or. but ultimately explanation. especially. are not strongly connected to either the description or explanation of social phenomena. ‘explain’ is an ‘adverbial verb’. I have tried to show. as already noted.What Achilles said to Hempel 55 survives even when there is no direct appeal to philosophical accounts of scientific explanation. Given the tortuous path the argument has followed it might be helpful to restate the main steps taken. does not … consist in explaining being something “more than” or even something intrinsically different from informing or describing. It is a sociological commonplace to say that sociology’s goal is not (only) description. I have first attempted to get us to appreciate a special class of problems whose very set up can involve unrealistic restrictions on what are considered possible solutions. at best. It is only because of our understanding of the concept of explanation and competence in various kinds of explanatory practice that we can see Hempel’s model as a model of the logic of explanation at all. and. It is worth pointing out then that the stipulation of the causal form as explanatory does not capture the distinction between explanation and description which is not a difference in form but a difference in ‘the pragmatics’.

interests. by looking at what each candidate does for our particular queries. for example. Ethnomethodology is concerned with) and the what (cf. activities. In conclusion. events. it has been argued that the (conceivably independent) strategy of stipulating the form of explanation will not work. and more besides.56 Clarity and Confusion in Social Theory restrictive and confused relationship to their subject matter. I want to emphasise how important it is to remove any restriction on what explanation must be and acknowledge the fact that what we call ‘explanations’ are variegated phenomena. as explanation is a contextual and not a formal matter. arguments. Alan Garfinkel (1981) exclusively focuses on – but also address the how (a question. for example. at a general level. I will return to this point more than once in the remainder of this book. In saying this one might stop. persons.. etc. Explanations are called for in cases of puzzlement and in cases where clarification or justification is needed. But it also carries more general implications insofar as it exposes the tendency of both philosophers and social theorists to overlook their commitment to the concepts constitutive of their subject matter. etc. But what makes an explanation better than another is a question to be answered in piecemeal fashion. . even if one were to grant that a reliable account of natural scientific practice is itself a reliable guide to social scientific explanation. Furthermore. The concept of explanation is a concept we (scientists. the arguments provided in this chapter show why turning to the formal theories of the philosophy of science cannot furnish social theory with a reliable account of natural scientific practice. it is not a question it makes sense to want to settle in a general way. among other things. for. If the latter proposition is also questionable then any appeal to philosophy of science of this kind is doubly misguided. Dray 1957).37 37 It might be objected that even if we grant that there are many different kinds of explanations we may still wish to know what it is that makes some explanations better than others (Giddens 1979: 258). natural and social phenomena. students of society and philosophers) are taught to use by means of all these different examples. lay folk. purposes.. and thus explanatory force cannot be captured with the use of formal criteria – it cannot be guaranteed by the use of a certain form of concepts. any more than it can be ruled out in the use of other forms. Explanations may deal with concepts. Finally. explanations need not only come as responses to a why-question – something. there is not much more to be said. As far as the social sciences are concerned.

it has provided insight into social theoretical projects which attempt to fix the form of explanation by portraying them as a continuation of Hempelian thought. investigation into the logic of language. Although there is a set of informal methods OLPers have used and ways of talking about OLP’s practice (logical or grammatical investigation. White (1967: 7) 3. I will provide an explication of the logical2 vocabulary and some informal methods that have been extensively used by OLPers and which I am employing throughout this book. for want of a better term.1 Introduction The discussion of Hempel’s project in the logic of explanation over the previous chapter has served a number of functions: apart from bringing to the fore the relationship between OLP and restrictive philosophical understandings of concepts. . The following important themes have emerged: (a) the problematic restriction of our conceptual resources in the construction of a theoretical problem. A. concept.1 (b) the fact that logical principles are internal to our practices and do not transcend them. nor its vocabulary (deriving partly from the tradition of analytic philosophy but mainly from the vernacular) are to be 1 This theme will be developed throughout the remainder of this book but especially in Chapter 5 with regard to the constitution and application of an ontological scheme. logical role) neither OLP’s methods. which can be conversely formulated as the thesis that activities possess logical features (this is what Winch means.Chapter 3 Ordinary Language Philosophy as Logic It will seem less strange that social relations should be like logical relations between propositions once it is seen that logical relations between propositions themselves depend on social relations between men. In this chapter I will utilise these themes in the process of investigating OLP’s relation to logic. which implies that it does not make sense to attempt to ground them theoretically. 2 It might be said that some of the vocabulary is rather meta-logical. its subject matter (use. P. I have no objection to calling it that as long as it is realised that the vocabulary itself has a logic and that it is not in any sense part of a second-order calculus. I think. the above themes can be reformulated as follows: a. In order to bear on the question of the status of OLP’s vocabulary and method. in speaking of the dependency of logical relations on social relations). kind of word. Winch (2008[1958]: 118) … if we think or say X then we are logically committed. and (c) the fundamental role our practices occupy. elucidation of concepts) and.

White 1967: 7. OLP is the focused employment of any competent speaker’s ways of clarifying what they mean. written sentence. to make a claim or statement that is not an isolated saying. and instances of which we attempt to explicate. etc. Fleming 2004: 42) and I too will speak of the logic of language or the logic of our concepts throughout this book. Weitz 1953: 230–32. open-ended. Ryle 1953: 185–6. It is an exercise of and not a substitute for our mastery of language. 1971[1961]: 60. there is a straightforward sense in which OLP can be said to be logic in that it is concerned with the appraisal of reasoning. theoretically explain. 3. I will emphasise throughout the potential dangers that arise from putting unwarranted stress on the logical vocabulary and from importing a certain conception of logical form from formal logic. do not yield stable or even intelligible results. OC: §56. Strawson 1964[1952]: 230. c. I end by outlining the distinctive attitude OLP adopts towards its vocabulary and method which is very different from the theoretical attitude adopted by a large portion of social theory and philosophy. §82. OLP focuses on the logic of language use. It is natural language reasoning about natural language use. what a claim amounts to. Our linguistic mastery is what we make use of. b. superior to or as excluding other resources. but not for those very reasons. etc. what it implies.3 The purpose of this section is to elucidate the senses in which OLP is logical enquiry. There is no such thing as an explanation of language mastery that does not already presuppose such mastery. in connection with a particular problem. and derive their usefulness from the problem at hand. In what follows I will first detail the senses in which OLP can be spoken of as logic and then proceed to give some examples of its logical vocabulary in use. OLP’s methods and vocabulary are ways of probing that competence. appeal to. they are not part of a technical calculus and.4 but is embedded in a structure of sayings. As such it presupposes an attempt to say something..2 Explication of Logical Vocabulary It is a matter of historical fact that OLPers have spoken of their practice as logic or as an investigation into the logic of our language (for example Wittgenstein PI: §38. 4 By ‘saying’ I mean to cover utterance. Even if such sayings do not constitute a rigid argumentative structure of premisses and conclusion which are 3 Tomlin complains about ‘the habit of linguistic philosophers of retaining the word “logic” in areas where the type of order under examination is rather that of grammar’ (1977: 241). We will see below how ‘grammar’ can be given a sense in which it refers to logical relations. To begin with. .58 Clarity and Confusion in Social Theory generally insisted on as indispensable. if treated so. Those ways are varied.

different signs in mathematical/logical calculi and different combinations thereof symbolise different logical forms. the proper study of logical force in natural language is informal: OLP is a species of informal logic. This introduces considerable potential for misunderstanding in continuing to speak of the logical form of natural language 5 As Bennett and Hacker phrase matters: ‘If the logical relations of implication. ‘application’. in other words it has considered logical relations between propositions which may pivot on expressions such as ‘all’. it may be said. the logical force of propositions is a matter of their form. Since the time of Aristotle. exclusion. formal logic has confined its investigation to inference as a function of topic-neutral expressions (otherwise known as the logical constants). Logical relations may also pivot on topic-specific expressions as well as on a whole range of contextual aspects.7 Thus. understood as amenable to depiction in a suitable notation. differs markedly from formal mathematical notation. valid inferences are likely to be overlooked. there are a number of logical relations which may hold amongst such sayings. a domain held to be in a state of intrinsic disarray. and presupposition that characterise the use of … concepts are not respected. then. comprise the logical force of sayings so that two sayings which have different logical relations differ in their logical force (Waismann 2011). compatibility. The simple and sometimes painful truth. grammatical or any other kind of form but a thoroughly contextual matter in the broadest understanding of context (Coulter 1991).Ordinary Language Philosophy as Logic 59 inferentially connected. analytic philosophy has thus been largely preoccupied with the construction of an adequate notation as a means of making logical form explicit. compatibility and presupposition. i. for example relations of implication. (Ryle 1954: 115). . invalid inferences are likely to be drawn. however.6 Under this conception. ‘and’. thus. Utilising the developments in logic. Since the mathematisation of logic in the late nineteenth century. Natural language. rendering it graspable through inspection of. etc. is that logical force in natural language is not a function of syntactical. exclusion. the latter being consigned to the intellectual dustbin of mere ‘pragmatics’. To speak of a ‘logic of language’ that embraced such putatively disorganised phenomena as ‘use’. if not identical with. with the aim of specifying the valid forms of argument. with the result that sayings which look similar in terms of grammar and/or syntax may have different logical force and vice versa.5 These relations. logicians have built on the idea (familiar to the reader by now) that logical relations and. and nonsensical combinations of words are likely to be treated as making sense’. (2007: 128) Note the two resulting senses of logic: as inference and as meaningful combination. ‘some’.. 6 See Strawson (1964[1952]: 228) for the different sense of ‘logical form’ the identity with a basic notational form produces.e. ‘speech’ (as contrasted to ‘language’) was to be taken to abuse the very concept of ‘logic’ (1991: 32). a chosen notational form. 7 Coulter notes the resistance towards this idea: There was a tendency to maintain that a ‘proposition’ had the logical properties that it did in virtue of its structure independently of its application.

3 Illustration Before discussing the question ‘logical relations between what?’ any further. We say. hence sometimes called logical grammar). and. At the same time the notion of logic implicated here is not one of logical force of a saying or valid inference in argument. His invocation of the logical type of the concepts of Church and Home office is meant to bring to our attention that they are related to each other and to other concepts in such ways. we may proceed to note that ‘logical form’ can be rather innocuously used interchangeably with logical category8 or logical type in the comparison of expressions in relevant ways. but also between sayings. (b) that when logical form is manifest it allows us to decide logical force by inspection of the notational form abstracted from a context of language use. Already. Logically possible and impossible combinations make up the Wittgensteinian notion of grammar (which is different to the one used by grammarians and linguists. (Ryle 2000[1949]: 19) In speaking of the relations between the Church and the Home Office Ryle is not speaking (not directly at least) of the logical relations between sayings we have seen above. We may speak of combinations or of logical relations not only between concepts. between other kinds of things too. compatibility and only in some secondary senses implication and presupposition). for example. but rather of logical possibility (Toulmin 2003[1958]: 157). this is a sense of the term ‘logic’ connected to meaningfulness. sense and intelligibility. . of the combinatorial tolerances and intolerances (Coulter 1983) of our concepts. and (c) that the two forms already coincide for some expressions but not for others. I will come back to this. ‘The British Constitution’ is not a term of the same logical type as ‘the Home Office’ and ‘the Church of England’. Logical relations between concepts can be of the kind we have listed above (exclusion. it may be instructive to briefly illustrate some of the concerns we have raised so far by identifying statements of different logical types. especially if doing so is taken to imply that (a) logical form is a kind of notational form. it is useful to consider the following excerpt: So inter-institutional relations which can be asserted or denied to hold between the Church and the Home Office cannot be asserted or denied to hold between either of them and the British Constitution.Clarity and Confusion in Social Theory 60 expressions or sayings. in other words. as it will be seen below. During the illustration it will 8 Ryle’s The Concept of Mind (2000[1949]) is a detailed argument about the conflation of words from different categories or logical types. For now. For the latter term. that it makes sense to say that the Church and the Home Office both sent representatives to an NGO-led summit but that it is logically impossible or absurd or nonsense that the British Constitution sent representatives too. 3.

9 Consider the following expressions:10 (a) ‘never say never’. It should be borne in mind that the typical way of proceeding for OLP is one where confusion or puzzlement dictates which expressions may need to be investigated. What circumstances would make its use intelligible? We might imagine someone complaining to a friend about the difficulties of finding love in life and saying ‘I’ll never find true love’ to which the friend responds with ‘never say never’.Ordinary Language Philosophy as Logic 61 become apparent that there is a two-way relationship between logic as inference and logic as meaningfulness. 11 We will deal with this temptation in Chapter 5 when discussing the question of the ontological soundness of expressions. 9 We have seen how the concept of logic is connected to both sense and inference. for instance. Although they are of somewhat similar syntactical form it can be shown that they diverge in terms of their logical type. To understand what one is saying when they employ these words. or as we might put it. They seem to be saying ‘don’t give up’. (b) ‘never open the car door without checking for traffic’. then. as Kant puts it. It should be emphasised once more that in understanding these expressions and understanding what their logical types are one is not doing two things. Already we have hinted at the fact that these expressions need to be embedded in circumstances if we are to understand their logical features. to be stressing a more optimistic angle. The reference to logic is made in the observation that the negation of an analytic judgement yields a contradiction. ones which are true by virtue of the fact that the predicate is contained in the subject. Kant dubs a class of statements. Consider how our understanding of what is being said exhibits itself in the different forms of words we could use to perform what the friend is doing in this example. It might also be worth mentioning the philosophical locus classicus of the connection. . What is the expression doing in this exchange? The friend seems to be offering encouragement. or judgements in his terms. Driver instruction comes to mind. ‘someday you will’ all of which are things they might have said instead. Also note that it is not clear what it would mean to insist on one form of words to the exclusion of others or to regard one form of words as the more fundamental. consider expression (a). treating them in abstraction only increases the chances of being misled by their syntactical appearance. ‘it’s not impossible’. Again we want to see in what context it would be an intelligible move.11 Secondly. Firstly. (c) ‘rainbows never appear without the prior presence of rain’. by virtue of the meaning of the terms. a case where the instructor is teaching the learner driver how to keep safe on the road by pointing out to them that they need to make sure whenever exiting their car that they are not hit by incoming traffic or otherwise cause an accident. and to respond to it in intelligible ways is to understand and display understanding of logical type. See Conant (1991) for more on Kant’s conception of logic. all of which can serve very well if we needed to explain what that person was saying. which is Kant’s Critique of Pure Reason. 10 For the purposes of illustration I start with these three relatively random expressions. analytic. ‘there is always hope’. let us examine the expression ‘never open the car door without checking for traffic’.

to designate their logical type is to describe what kind of actions are being performed. the steps we take – the transitions of logical type – will be different. what kind of responses would be consonant with understanding those sayings in the way described and would therefore make sense. in terms of what someone might be actually saying with these forms of words. What is of interest here is not whether this is actually the case but the fact that reporting an instance of the phenomenon without some of the standard observed conditions is a response that makes sense. The response thus counts as a possible contestation of the initial statement. to caution one on how to keep safe on the road and to point to an observed feature of a natural phenomenon. Let us now compare these doings in terms of their logically possible relations. The step from reports of recent tennis-playing form to a predicted selection (or to the statement that a particular player deserves to be selected) is one thing. geometrical axioms and so on. The relation in question. We can paraphrase what the teacher is saying as ‘rainbows have never been observed without the prior presence of rain’. The arguments which we put forward. will be correspondingly various: depending on the logical types of the facts adduced and of the conclusions drawn from them. are. the step from evidence about clues in a murder case to the guilt of the accused party is another. predictions about the future. and the statements of the facts adduced in their support. where ‘appropriate’ here means. one that displays understanding of and is provided for by the kind of saying the initial saying was. as philosophers would say. in the sense we have specified above: it is one of logical compatibility. To summarise our progress so far. we might say. consider the response that there was an observation reporting the occurrence of a rainbow without any prior rain. the above forms of words are used to offer encouragement. We can mark the relation between these two doings by calling one a generalisation (of sorts) and the other one a counterexample: remember. say. it is logically possible to challenge a generalisation with a counterexample.62 Clarity and Confusion in Social Theory Finally. or yet in other words.12 The possibility of providing a counterexample is internal to understanding the saying as a generalisation. is a grammatical one. (Toulmin. a command or a request. of many different ‘logical types’ – reports of present and past events. aesthetic commendations. and the steps which occur in them. Starting from what we paraphrased as ‘rainbows have never been observed without the prior presence of rain’. verdicts of criminal guilt. 12 Compare the following excerpt from Toulmin’s The Uses of Argument where activities and parts thereof are shown to be logically connected: The statements of our assertions. It makes sense to offer a counterexample to a generalisation but none at all to offer a counterexample to. 2003[1958]: 13) (my emphasis) . consider ‘rainbows never appear without the prior presence of rain’. What could someone be saying by this? We might think of a physics teacher instructing a student on how to think of this natural phenomenon by pointing at the history of scientific observations. To witness these relations. it is useful to investigate what responses are appropriate to the above sayings. We have already characterised them in terms of doings.

Obviously. In developing our sense of the differences in the use of these expressions we have thus partly resulted to the use of paraphrase. but stating a rule of conduct. was in order to take attention away from the superficial similarities in syntactical form and not in order to bring some of the expressions closer to their true logical . in other words what the correct way of proceeding is. Suppose the student responded to the driving instructor by citing a case where someone opened the door without checking for incoming traffic and that the student insisted that this proved the instructor wrong. Imagine the complaining party responding to their friend’s encouragement by citing the fact that their friend had once said ‘never’ and thus was in no position to give such advice! Apart from being a good joke that might take care of the emotional tension in the situation. the student might live in a sparsely populated area with no traffic at all. (b) and (c) has been to bring to our attention the fact that some misunderstandings of them are due to their similarity in syntactical form. ‘checking for traffic is pointless’.Ordinary Language Philosophy as Logic 63 Now let us consider whether a counterexample is possible with regard to expression (b) in its specified use. these rules do not depend on a local situation. The purpose of doing so. but are meant to cover the whole range of traffic conditions. This is a curious response given what we know about driving lessons: learner drivers are there to learn the rules and not to question them. Naturally. one which is directed at the point of the rule. consider expression (a) and the possibility of a response by counterexample.e. In such circumstances the student might be saying to the instructor. to object that someone who says ‘never say never’ is using the word ‘never’ is not to expose the fact that they are engaged in a self-reflexive paradox but to misconstrue the kind of doing their doing is. Thus. Finally. One of the points of juxtaposing expressions (a). but to sketch the different options available to us and how our understanding of what is being said and of its logical relations might depend on a whole range of contextual factors. For example. To say ‘never say never’ in these circumstances (and fairly typically) does not imply a wish to ban the use of the word ‘never’. It might be said that we are liable to misunderstand the expression ‘never say never’ precisely because it is similar to ‘never open the car door … ’. such a response would fail to count as that kind of challenge to what the instructor was saying. this hardly stands in an otherwise intelligible relation to the encouragement. to other forms of words which could be used in the specified context more or less interchangeably and which were of a different syntactical form. it is possible to hear the student as producing a different kind of challenge to the instructor’s rule. i. This is for the student to miss the point: the instructor is not making a generalisation to the effect that in all cases where somebody opened their door without checking an accident occurred. It does not make sense for the student to attempt to challenge the instructor’s stating of the rule with counterexamples of drivers who opened their door several times without checking and who were not injured nor caused an accident. The important thing to note is that even if we take it that such a response makes sense it is not any kind of ‘counterexample’.. Furthermore. The point of entering into these variations is not to settle on any one of them. however.

We can thus move from a set of inferences to the sense in which a term is being used and. say. and much could be said against 13 The point that counterexamples and statements of rules do not mix has been forcefully made by Stenlund in the following passage: ‘There can be no such thing as a counter-example to a logical rule for the use of an expression – in such a case the “counterexample” must be another use of the expression.64 Clarity and Confusion in Social Theory form. in which case. that there is no contemporary contract. and is in fact a misuse of the term ‘rainbow’ (it can perhaps be said to be a different use of the term.14 This should not come as a surprise. After all. we can say that in looking out of context at an expression we might expect a certain kind of use and that it is that expectation which leads to potential misunderstanding. it cannot be heard as a challenge to what the teacher is saying). For example.F. Hunter’s ‘The logic of social contracts’ where he is summarising his arguments against the usefulness of social contract theorising: I have now argued that the occurrence of a social contract in.M. If we take it that the physics teacher is providing the student with a definition. The sense in which we speak of logical form (that is. To give yet another example of this fact. subject to other logical rules’ (1990: 51). Also. the student’s response cannot count as a counterexample. either explicit or implied. We deal with misunderstanding by looking closely at the context of use. either with or without resorting to paraphrase. Understanding logical form is a matter of having a clear grasp of what is being said in certain circumstances and it is the logical relations of an expression used in those circumstances that ‘logical form’ is intended to cover. it can easily be seen that the verbs ‘mean’ and ‘imply’ have overlapping uses in the vernacular. 14 Consider Diamond: ‘an explanation of the inferential connections of “someone has measles” may enable a child to grasp the role of “someone’ in it”’ (2004: 157). on paraphrase. conversely. then that spectrum is not a rainbow. we should not be led to the conclusion that expression (b)’s logical form is closer to its syntactical form whereas expression (a)’s is not. A statement of definition would be very different in its implications to that of an observational report: it would follow from it that if no rain or other kind of water vapour has preceded the appearance of a spectrum. as logical type) is misunderstood if taken to depend on syntactical form and. Accordingly. consider the following concluding paragraph to J. therefore. . the relationship between these two senses of logic is what Robert Brandom is exploiting to build his inferential semantics. 1215 would have no contemporary relevance. that little could be said for. however.13 Here we can see clearly that logic as inference and logic as sense are two sides of the same coin: different logical inferences can be cashed out as different senses of a statement and vice versa. It should also be borne in mind that we could well have assigned different logical types to the above three expressions by imagining different ways of using them. At most. from an explication of the sense of a term to the compatible inferences and logically possible relations. the physics teacher might have been stating a definition of rainbows.

scolding. responses to which can be of such kinds. apologising. concepts and contexts. i. We saw previously that the relations that the concepts of Church and Home Office have to the concept of a representative.e. which I cited at the beginning of this chapter. to produce an unintelligible or irrelevant response. To complain in this latter case that there is no evidence to suggest that such a contract took place is to display misunderstanding. With this in mind we can now respond to the question of what the bearers of logical relations are. In the case of social contracts as in expressions (a). it does not make sense to speak of refuting a request. many more. as in Ryle’s example of the concepts of Church. and that there is no way in which the idea of a social contract can be used to settle basic political questions. Certain sayings and doings may not receive refutations as appropriate seconds: for example. Otherwise one is liable to exhibit ignoratio elenchi. whereas nothing of the sort follows from asserting the existence of a virtual or implied social contract. inferential relations vary. implied or virtual. one which cannot refute what has been said. We have already spoken of relations between activities. reading a story. It should be noted at this point that to speak of a challenge or refutation only allows us to cover a range of logical types of statement. we can also speak of the logical type of a concept. the British Constitution lacks. sayings. (b) and (c). Accordingly. To conclude. Home Office and British Constitution. Indeed. (1966: 46) Hunter’s arguments range over the understanding of a social contract as actual. logical relations may be asserted between . We also witnessed above the absence of a certain meaningful relation between the concept of request and that of refutation. The discussion so far may have revealed the depth behind Winch’s claim that social relations are like logical relations. it follows from asserting the existence of an actual social contract that it is possible that there is some record of it.. telling a joke and many. producing an adequate challenge presupposes that one has understood the (logical type of) statement one is responding to. commanding. the simplest way to understand logical type distinctions is to understand them as the logically relevant differences between actions/activities such as asking questions. Apart from the logical type of actions. although we can speak of challenging the sincerity of a request.Ordinary Language Philosophy as Logic 65 any recommendation that we ought to have a social contract. This translates into logically relevant differences and logical relations between concepts. For example. what counts as an objection to the idea that an actual contract took place either in the distant or immediate past is logically inappropriate when the social contract is understood as a virtual one. offering advice. I am not aware of any other angle from which social contract could be said to be a useful philosophical device. Logically relevant differences have to do with the logical relations between an action and other actions both in the sense of logic as inference and in the broader sense of logic as meaningful combination. Depending on how the social contract is understood.

saying X implies something about certain situational features. But they are institutions in different senses of the word (and to that extent are connected to different concepts) since the Home Office can have sub-agencies and an annual budget. Logical relations between sayings (of both the inferential and meaningful connection kind): To say A implies B. nor does it make sense to say that it either has or has not sub-agencies and an annual budget. We examined relations between sayings extensively in this section. it can be overstaffed. efficiently run. 2002a[1979]. has an intelligible connection to response D. what can be generally put as. Obviously.Clarity and Confusion in Social Theory 66 all of the above since they are all aspects of the same manifold. I started from the insight formulated during the discussion of Hempel’s problem. 2002b[1979]). headed by an individual and contacted by the public. b. on the other hand.4 Conclusion As a conclusion to this chapter I would like to summarise what has been done and what may be learnt from it. headed by someone or contactable. 3. the logic of our concepts. efficiently run. neither is nor is not overstaffed. 16 J. It is possible to organise those aspects in the following way:15 a. Logical relations between a saying and its context (of both the inferential and meaningful connection kind): A wide number of features of a situation form the conditions under which we would say X rather than Y. Frank Ebersole’s incomparable imagining of elaborate cases where the details decide between one expression and another is also such a form of investigation (2001[1967]. c. . etc. The most general way of showing these connections is through the kinds of things we can intelligibly say about a concept and the characteristics and qualifications we offer. The British Constitution. both the concepts of Home Office and British Constitution are connected to some concept of institution: we do say of each one that it is an institution. they form complementary (and often mutually implied) ways in which we can investigate. does not imply C. Austin’s phrase ‘what we should say when’ finds its natural application in the investigation into conditions and forms of words. Concepts can thus be distinguished based on their connections. which was recast as the idea that 15 See White (1967) for a lucid discussion of how to do conceptual elucidation and for an attempt at a similar grouping. all of the above are only different aspects of language use. As a result. namely that logic is not something transcendental but is rather embedded in our activities.L.16 In turn. For example. Logical relations between concepts (of the meaningful connection kind): Not all concepts are connected to each other.

‘usage’. Among such expressions are: ‘use’. ‘logic’.Ordinary Language Philosophy as Logic 67 such activities possess logical features. which is flatly opposed to the stipulative or legislative one identified by Stenlund. ‘picture’. to treat such expressions as part of some kind of meta-logical vocabulary with a uniform application across cases. ‘what it means to’. not giving it a use vs. even among some OLPers. the senses in which OLP can be thought of as logic have been specified via the clarification of some of the vocabulary that has been used by OLPers in investigating the logic of our concepts and which I will (sometimes) use in the present book. by stressing the crucial difference between OLP and a large portion of social theory and philosophy. abstract and transsituational “meanings” which can be mobilised for the purpose of constructing highly general theories’ (Lynch 2000: 149). or distinctions such as the one between using a word vs. Throughout the discussion. and only secondarily the vindication of some … doctrine expressed in this vocabulary (1996: 195–7). types or categories does not impose order on what lacks order. misusing it. whose words I paraphrased in the introduction to this book: [The vocabulary] constitutes the linguistic horizon within which problems and questions are expressed and treated … the belief in the traditional ideals of metaphysical philosophy [and social theory] is basically an attitude towards a … vocabulary. it is thus our competence. ‘work’. etc. our practical understanding of language use that constitutes the ultimate court of appeal and not any number of technical categories. In other words. have employed in the course of bringing features of language use into focus. ‘point’. are part of the vernacular and thus are not designed to serve as theoretical terms. ‘grammar’. ‘role’. ‘function’. ‘sense’. ‘concept’. but is a way of reminding ourselves of the orderly distinctions which are part and parcel of our mastery of natural language. among many others. Equally. then. I would like to end this chapter. they apply in rather contextual fashion and are easily stretched beyond breaking point if they are treated ‘as though they had stable. however. caution is advised regarding the status of a number of ways of speaking that OLPers. The temptation is great.18 17 It is only once one has yielded to this temptation that it makes sense to debate in general terms concepts such as rules or meaning. it has been claimed. as some Wittgensteinians are prone to do. ‘what is being said with’. in the way that we speak. ‘criteria’.17 These terms of logical appraisal. We make these distinctions in practice. I have tried to distinguish the way in which logical form or type is understood in informal logical investigation as opposed to investigation that adopts a formal notational perspective on natural language. Furthermore. OLP. consider Ryle’s remarks on how puzzlement concerning mental phenomena is a product of the commitment to a vocabulary: ‘The . I would particularly like to emphasise in this connection that talk of logical forms. ‘meaning’. is a way of explicating the logical features of our activities – of our doings with words – and thus the logical features of our concepts. ‘rules’. 18 To add yet another example. The difference consists in the attitude the former adopts towards its frequently used vocabulary. ‘what counts as’.

mental processes are causes and effects. even further. but different sorts of causes and effects from bodily movements. We will begin in the following chapter with the investigation of a currently very influential idea. “stuff”. it refuses to treat any terms as fundamental or to insist on them across cases. Minds are things but different sort of things from bodies. “process”. We will witness in Chapter 5 the pernicious implications of the idea that we are to account for all concepts with the use of broad categories and. “state”. with the same handful thereof. differences between the physical and the mental were thus represented as differences inside the common framework of the categories of “thing”. “change”. to all cases. In other words. as it were. And so on’ (2000[1949]: 20). “effect”. if at all. fixing the vocabulary in place can only be purchased at too great a cost.68 Clarity and Confusion in Social Theory By contrast to the theoretical attitude. OLP does not restrict its horizon to a number of key terms. that of social ontology. . “attribute”. “cause”. Therefore. that will enable us to proceed to apply OLP to cases of social theoretical reasoning and debate. This chapter concludes the logical groundwork. precisely because it recognises that they cannot be used to pick out stable logical features of linguistic usage and that they do not apply in the same sense. the vocabulary cannot be used in the construction of a technical procedure which will fix the features we want to talk about as it tends to vary together with such features. Whatever terms it might use during general discussion it can dispense with. as will be amply demonstrated in the following chapters.

John Law (2004). as Coulter notes. which either add up to or fall short of making that whole. Varela 2007. as referring to a whole consisting of parts. Harré 2002a. 2007b. one easily ends up treating ‘society’ as always doing aggregate work and. King 2007) which can be seen as recent instalments in debate spanning many decades (for example Porpora 1989. Both additivist and emergentist conceptions rely on this ‘mereological’ understanding. On the other hand it is quite clear that sociology has utilised the many different senses of ‘society’ to bring under its aegis diverse questions. to its sociological programme (Bhaskar 1998[1979]) but asserts that any social theory is necessarily committed to an (implicit or explicit) ontology (Archer 1995).1 Introduction Social theoretical concerns regarding ontology or social ontologies or the ontological status. although. therefore. To get a flavour of some of these consider Talcott Parsons on Durkheim’s use of ‘society’: [T]he source of the sacredness of sacred things is the same as that of obligation to moral rules. of what different kinds of words consequential to social thought are about have proliferated during the last couple of decades. Even regarding the concept of society. which will be examined in the remainder of this book – and. Theodore Schatzki (2002) and John Searle. not only makes use of an ontology as fundamental. Porpora 2007. One salient feature of such debate is that it involves a thicket of issues fundamental to sociological enquiry – issues having to do with the form of explanation and the ascription of causality among others. For example. ontological concerns have surfaced in a number of notable exchanges among social theorists (for example Elder-Vass 2007a. Apart from critical realists.1 While 1 Frisby and Sayer (1986) (cited in Coulter 2001) argue that sociology has unhinged itself from the debate on a suitable sociological concept of ‘society’. if one does not pay attention to the details of the logical behaviour of the concept. namely Critical Realism (CR).Chapter 4 Ontological Confusion in Social Theory 4. is also recognisable as part of the dispute between individualists and collectivists regarding whether sociology should be concerned with the separate and sui generis order of ‘society’. a host of other macro-concepts remain logically obscure. 2002b). whose work (1995) has sparked considerable debate since its publication. as it is sometimes put. It is ‘society’. as such. if not foundational. one of the main theoretical positions in contemporary sociology. 2005). But this position is in need of further interpretation . This synthesis of what had been before regarded as quite disparate aspects of human life was a stroke of genius on Durkheim’s part – of revolutionary importance. examples of theorists who have pressed the question of (social) ontology are Bruno Latour (for example. Furthermore. Varela and Harré 1996.


Clarity and Confusion in Social Theory

the issues involved might then be seen as perennial ones, what is perhaps novel is
that they are debated under the auspices of competing social ontologies. Indeed,
treating traditional sociological questions as being a matter of social ontology is
a novelty of the past couple of decades or so that presumably constitutes our best
current attempt to settle them.
Having said that, it should be noted that this approach springs effortlessly from
(and sometimes incorporates) the two traditional ways in which the relationship
between individuals and society is thought of. One response to the question has
been methodological, in that it asserts the priority or propriety of sociological
explanation in terms of one or the other. The other response has been ‘ontological’,
in that it asserts the existence of ‘something more than’ individuals or, on the other
hand, ‘nothing but’ individuals.2 The first response attempts to settle the form of
explanation a priori and therefore, as we saw in Chapter 2 and shall further discuss
in this and following chapters, must be seen as seriously flawed. The second
response, it will be presently argued with reference to its contemporary form, is
also problematic. The reason is that it treats the relationship between individuals
and society as a matter of ‘what there is’, introducing at the same time ways
of treating expressions which lead to conceptual confusion. The expressions in
question are both those belonging to the ontological form of enquiry, such as real’,
‘exists’ and ‘entity’, as well as those having to do with society, such as ‘the state’,
‘capitalism’, ‘the economy’ etc. I will document both the ontological treatment
of ‘societal expressions’ and the demands made on ‘ontological expressions’ in
what follows. Social ontology projects, far from redressing what is problematic
to clear up the difficulties left by Durkheim’s lingering positivism. Society in this
context is not a concrete entity; it is, above all, not the concrete totality of human
beings in relation to each other. It is a ‘moral reality’. (1968[1937]: 712)
I will return to the issue of the logic of ‘macro-concepts’ in discussing the concept of
organisation in section 6.5.
2 Consider how these two responses are also present in philosophical treatments of
the mind. Stenlund’s remarks are incisive on this matter:
[T]he ‘ontological question’ of whether ‘minds and mental entities exist’ makes
no sense without the philosophical jargon in which ‘mental entities’ are spoken of
as a kind of invisible thing in some place in the world (perhaps under the tops of
people’s heads?). You must be trained in his jargon in order not to be just perplexed
by the question, ‘Are there minds, feelings, intentions?’ – especially when it is
posed in the objective and serious tone characteristic of some philosophers of
mind. What the question really means is something like the following: ‘Shall we
adopt the notions of mind and mental entities as primitive notions in our theories
of human language, action, and behaviour?’ (Stenlund 1990: 22)
I think, however, that theorists also tend to move to the opposite direction to the one
Stenlund sketches, that is, from primitive explanatory notions to a theory’s ontological
commitments. The transition thus can work both ways. We will see with regard to the
problem of the reality of social structures, how there is confusion involved both in rendering
the problem as an ontological one and as one about the primitive form of explanation.

Ontological Confusion in Social Theory


in the above responses constitute ‘more of the same’. Confusion is compounded
by running the question of ‘what there is’ together with the question of ‘what
properly explains’, as does, for instance, CR with the use of the causal criterion
of existence.
Thus, in this chapter I intend to question whether the recent wave of penetration
of sociology by talk of ‘ontology’ constitutes an improvement over the way in
which quite traditional issues are debated and whether, yet another turn to what
are properly understood as philosophical ways of thinking, engenders anything
but continuing confusion. The side-effect of the turn to questions of this sort is
that a perspicuous answer to the issue of what disagreement between different
sociological camps could be about slips further away (Button and Sharrock 2010;
cf. Sharrock and Watson 1988).
Despite the fact that the recasting of methodological and theoretical questions
in ontological terms wears its philosophical origin on its sleeve, it is perhaps
worth rehearsing what the latter entails. Ontological questions are metaphysical
questions in the sense that they attempt to settle what the ultimate reality of things
is, and more specifically – by being ontological ones – they branch out into two
related tasks: that of providing a definitive answer concerning the being or nature
of an entity, and that of compiling a list of the kinds of entities populating reality,
in other words a list of kinds of real things and real things only.
It might be objected to this reassertion of ontology as a metaphysical endeavour
that the theoretical pursuit of ‘social ontology’ in the past decades has successfully
distanced itself from metaphysics and that the term is used rather harmlessly to
signify our chosen categorial scheme. Accordingly, questions about what the latter
allows or entitles us to talk about are thought to be stripped of any metaphysical
implications. This claim, however, fails on several counts:
Firstly, little, if anything, is usually done by way of attempting to disclaim the
metaphysical baggage that the pursuit of ontology carries with it. Consider, for
instance, Theodore Schatzki’s definition of ‘social ontology’ which, depending
on one’s inclinations, can be read either as adding nothing to what social theory
or sociology are about, or, alternatively, as an open invitation to debate ultimate
reality:3 ‘Social ontology examines the nature and basic structure of social life and
social phenomena’ (2005: 465). The resulting ambiguity is a direct consequence
of the fact that diverse forms of enquiry can be described as ontological when the
criterion for doing so is simply the applicability of such forms as ‘investigating X’s
nature’ or the question ‘what is X?’.
Secondly, in cases where an effort to gain some distance from metaphysics is in
fact made (Bhaskar 2008[1975] for example) the result ends up being contradictory,4
3 The same is true for Archer’s account of ontology (1995: 21).
4 As Justin Cruickshank (2004, 2010) has argued, the critical realist commitment
to avoiding the ‘epistemic fallacy’ (see section 4.4) contradicts the attempted redefinition
of ontology as non-metaphysical, i.e., as concerned with the transitive rather than the
intransitive, to use Bhaskar’s terms.


Clarity and Confusion in Social Theory

precisely because there is no genuine intention to give up the transcendental to
begin with. Lack of clarity concerning the commitment to the transcendental thus
compounds confusion regarding the nature of ontological projects.
Ontological projects, we might say for now, are prone to be conceptually
muddled (and hence remain philosophical in a Wittgensteinian negative sense) not
due to a lack of intellectual vigour but due to a certain misconception of language
coupled with a form of questioning that is riddled with confusion. Doing ontology
amounts, in effect, to having a license to subject all relevant expressions to the
question ‘what entity does it refer to?’ and what these expressions are about, in
turn, to the question ‘is it real?’ without any cause for concern. However, not only
is it not true that it makes sense to ask the first question indiscriminately but nor
is it the case that the second question is one determinate question (as J.L. Austin
demonstrated long ago).
To make the case for the present argument I will attempt a detailed examination
of a notable moment in the history of social theoretical debate where one of the
acknowledged progenitors of the now very influential CR turned against it for
being ‘metaphysically misguided’ about what is or can be real. Specifically,
I will scrutinise Rom Harré’s (2002a, 2002b) attack on the idea that social
structures are causally efficacious and therefore real, and debate around it.5 Lest
I be misunderstood, I should like to stress that I have nothing against Harré in
particular. On the contrary, I am even in agreement with some parts of his position.
The matters I attempt to address call for a detailed examination of our reasoning
and Harré’s reasoning is chosen because, first, it forms the centre of the debate in
question, and, second, because although it is conceptually sophisticated in many
respects, it does not break with ontological reason. It thus continues to nourish
an extremely pervasive idea, shared by many social scientists and philosophers,
regarding the way our concepts relate to reality. Because the confusion in ontological
reasoning is deep-seated and tricky to expose, I confine myself to establishing the
latter’s incoherence. I will not be providing a comprehensive criticism of any of
the schools of thought involved (although quite a bit will be said on CR in this
and remaining chapters) and neither will I be providing a detailed account of any
theory of causation (though I will discuss Bhaskar’s causal criterion – more on
causation in Chapter 6).
Finally, I would like to stress that the focus on this particular debate is also
strategic with regard to the position I wish to defend. In dealing with the question
‘Are social structures causally efficacious?’ Harré is interrogating its sense and
might, therefore, be seen as providing us with an OLP angle. However, despite
Harré’s invocations of Wittgenstein and the ascription of ‘Wittgensteinianism’
5 Harré’s critique originated in a Symposium which focused on ‘promot[ing]
inter-disciplinary discussions of how issues of philosophy and social theory can inform
methodological insights and vice versa’ (Williams and May 2002: 109). His published
contribution (2002a) was responded to (in terms of its broader implications) by Bob Carter
(2002) and Piet Strydom (2002) to which, in turn, Harré rejoined (2002b).

e.7 I seize on this comparison with Wittgenstein in the hope of demonstrating that what a (genuinely alternative) OLP move needs to do is expose the ontological project as incoherent by helping us to dispose of one of its deeply entrenched sources: the idea that a language is a ‘conceptual scheme’.S. I want to prepare that discussion by first examining ontology in terms of the demands it makes on the use of ‘ontological expressions’. a conceptual question that is settled by the logical grammar of the language and (b) treating grammatical truths6 that should have the status of reminders about what it makes sense to say as carrying ontological. I am not interested in scoring points against Harré and I will gladly concede that. In any case. It is a distinction between different roles and uses of propositions’ (2007: 18). there would still be significant tension generated by his participation in debate without challenging its ontological/metaphysical aims. the class of grammatical truths is larger than the one of analytic ones and its distinction from empirical truths is not principally an epistemic one ‘even though [the former are] of course a priori. and that to say that a grammatical statement is true is to say that these are the rules. I do not think it is easy to find this description of Harré’s position in what he writes but even if one were to accept it as accurate. metaphysical weight. as P. Typically. This will lead us to the examination of two (critical) realist trademarks: the causal criterion. ontology has not shaken off a certain numbness as to its precise nature. in the explication of ontological 6 Roughly.M. Scare quotes clothe ‘true’ above to emphasise that grammatical truths are not opposed to grammatical falsehoods but to nonsense. My main interest here is with what the debate can tell us about the confusion ontological reasoning brings to social theory and in that respect Harré’s intervention leaves much to be desired. It is an important fact that. Also. part of what he says (even most part of it) is correct. 4. in that he is careful to emphasise. i. in an earlier article (1997) as well.. either completely or at least as involving anything other than conceptual elucidation. which is a means for deciding whether something is real. Rules cannot be said to be true or false but can be said to determine what counts as (not what is) true and false (see Long and Jolley 2010).Ontological Confusion in Social Theory 73 to his views by other parties (Strydom 2002: 132) his discussion is dangerously close to (a) treating the sense of the question ‘Are social structures causally efficacious?’ as depending on the ontological status of social structures. i. he can be arguably said to be rejecting ontology and metaphysics. when in fact it is difficult to see the question of ontological status as anything other than a question about sense. 7 It might be said that I am being unfair to Harré. . and the rendering of the real as the non-observable.2 Fixity of Meaning: The Demands of Ontology on the Use of Expressions Before I proceed to examine the debate around Harré’s critique of the critical realists. grammatical truths are statements that are ‘true’ by virtue of the fact that they state (directly or indirectly) the rules for the use of expressions. Hacker argues. therefore. under a certain interpretation. as already noted..e. that he is offering conceptual reminders and.

Consider the following example of how the form of words ‘what is’ can be used to incorporate questions within the remit of ontology: ‘it is useful to distinguish the ontological problem of causality from the epistemological problem. and that contrary to the constitutive requirements of ontology. that these expressions are only partly overlapping in their sense. among other things. etc. ‘Do social structures exist?’ and ‘Are there social structures?’. it is manifest that there is some relationship between these expressions in that when it comes to asking specific ontological questions we can say seemingly interchangeably: ‘Are social structures real?’.9 The reason is that we need to know the answer to the question ‘a real what?’. nevertheless. conversely. Given these features of the use of ‘real’ we have trouble understanding the characterisation of an object as real simpliciter. say. For to speak of a ‘real object’ is already to suggest that it is real in the sense that objects are real (of course the indeterminacy then comes from the fact that ‘object’ may need to be specified too). to use one of Austin’s examples. questions such as. ‘does X exist?’ become pseudo-technical expressions8 that are guaranteed to apply without restriction if the investigation is to retain its generality which by definition it must (since ontology is concerned with what. J. ‘real’ is (a) substantive hungry and (b) a trouser word. Accordingly. What Austin has in mind presumably is the way of picking out an individual with the use of a variable within a formal notation which does not include any indication as to the kind of thing that individual is. In characterising the behaviour of the word ‘real’ in some of its uses. in treating our ordinary sayings as carrying ontological commitments. Thus ontology can be said to be the study of that which is. the relation between what is real. are conscripted in uses that are seen as constitutive of what it is to do ontology. of what exists. or the study of ultimate reality. The conscription of ordinary expressions and forms of questions results. ‘exist’. everything. Austin (1962) observed that the phrase ‘This object is real’ is quite dissimilar to ‘This object is red’ in that. Now. The former problem concerns the question: what is causation?’ (Kaidesoja 2007: 63). something can be a real decoy without being a real duck. Such questions can also be used to decide which entities are to make up our ontological list. ‘is X real?’.L. what is part of reality can be more complicated than we are inclined to think when theorising. in turn. for example. ‘being’.74 Clarity and Confusion in Social Theory enquiry words such as ‘real’. It needs to be emphasised. A musical instrument can be a real guitar without being a real Fender or. nor thus any indication as to some of the ways in which that kind of thing can be real. there is). of what is real. lead us to the examination of the causal criterion. what exists and. 9 I am quite sceptical concerning whether such a characterisation is possible. of the being of things. The fact that the word ‘real’ – in these relevant uses – ‘hungers’ for a what is why Austin spoke of the word as being 8 The expressions ontological enquiry is tied to serve to turn disparate questions into standard ontological ones on the basis of word form. . ‘reality’. Let us briefly look at ‘real’ and ‘exists’ and then move to consider the differences between seemingly interchangeable expressions which will.

of course. . ‘free’. good at pruning roses. Crucially. In general. are “the same” and “one”. to compile the list of the kinds of things that are or can be real would run into the following complications: we would either end up with a list that would contain all kinds of things or no kinds of 10 Austin notes that: ‘“Real” is not. more or less. If one adopted this style of talking. Finally. say. One needs to know both answers to ‘an ordinary what?’ and ‘ordinary as opposed to?’ (although knowing the what is in many cases knowing. ‘proper’. a real duck can be real in the sense that it is not a toy. ‘exciting’. We might also add that ‘ordinary’ works in the same way – as was hinted at in Chapter 1. In keeping with this style we might say that ‘exists’ refers to an oppositional property: sometimes to not being extinct. but not a good novel. We might say it has a reactionary usage: it gets the content of its position almost entirely from the nature of the opposition (2001[1967]: 293–4). ‘decoy’ it is the negative that ‘wears the trousers’. The force of ‘exists’ can be understood on any occasion only by knowing what it is opposed to. ‘special’. the only word we have that is substantive-hungry. then ‘not-W’ is also a property-word. ‘real’ is dubbed a trouser word. Interestingly. in other words. Other examples. it is worth mentioning Ian Hacking’s (1999) observation that statements of the type ‘X is socially constructed’ exhibit the two features Austin identified with respect to ‘real’. and plenty of others: ‘It’s a real y’ – what exactly are you saying it isn’t? ‘I wish we had a proper stair-carpet’ – what are you complaining of in the one you’ve got? (That it’s improper?) ‘Is he free?’ – well. in one of its aspects. Frank Ebersole has pointed to one of the salient features of the use of ‘exist’: It is a common practice of philosophers to speak of ‘red’ and ‘not-red’ as both referring to properties. ‘unusual’.’ (1962: 69–70) 11 Austin adds in a footnote: Compare. a body of troops may be one company and also three platoons. what have you in mind that he might be instead? In prison? Tied up in prison? Committed to a prior engagement? (1962: 15).). Trying. etc. perhaps better known ones. but not good at mending cars. however. whether the opposing term is ‘extraordinary’.Ontological Confusion in Social Theory 75 substantive hungry. then one would presumably have to classify ‘exists’ with the ‘not-W’s’. or a fake. in this respect. the use of ‘real’ serves so as to exclude some way in which what is talked about is not real.11 In a similar vein. the above independently arrived at observations on the uses of ‘real’ and ‘exist’ have very important implications for the pursuit of ontology which. or a hologram or a picture of a duck.10 Moreover. somewhat counterintuitively. in the positive-negative pair ‘real’ and. sometimes to not being mythical. hence. or a decoy. ‘Good at what?” – a good book. Therefore. The same team may not be the same collection of players. and so on. the sense in which the duck is real depends on the kind of contrast that is being made and. perhaps. ‘real’. where ‘W’ is a property-word. involves pronouncing once and for all on which kinds of things are real and which are not. Then what about “good”? We have here a variety of gaps crying out for substantives – “A good what?”.

i. It may be noted. It stands to reason then that there can be as many kinds of real things as substantives real can devour and as many senses of ‘real’ as opposites can slip into the trousers. that the shift from ‘real’ to ‘really real’ is as empty as the shift from how we tell that A is real to how we tell that anything is real (Diamond 1991: Ch. therefore. We are tempted to forget that each word has a complex use of its own and in serving our ontological purposes we may portray that relation in the form of what seems a trivial and obvious truth: we may say. are 12 I am thinking. is like. someone were to ask us whether reality consists of what is unreal. a direct consequence of Austin’s point that ‘real’ is not like ‘red’). that real things are part of reality. we understand what real as opposed to counterfeit money is. with all ‘really real things’ (which is not contrasted to ‘what appear to be real things’ for that is opposed to ‘real things’!).e.. both cannot be subsumed under a single criterion. therefore. Apart from the complexities just sketched. under physical objects there will fall tables some of which might not be real dining room tables. They conceive of the quest for the ultimately real as imposing an overriding criterion which is supposedly universally applicable. for example. Moreover. for instance.76 Clarity and Confusion in Social Theory things.1) when how we tell a real piece of paper is quite different from how we tell a real €100 note and. for example. there are also complications concerning the way ‘ontological expressions’ are related. one caused by the drug. Bouwsma [1949] for an attempt to dispel an analogous misconception). but what would non-counterfeit social structures be? Naturally. given the special contrast being made in every case. That is. described by Howard Becker (1953) as depending on an instructional context in which one learns to recognise the effects induced by the drug and to identify what a real hallucination. for example. it is dubious whether it makes sense to apply ‘real’ in one of its senses without restriction. this claim does not make the sense philosophers and theorists think it does is a formidable task which I cannot undertake here (but see. 13 One of the perennial (indeed Socratic) ways we are led to confusion is by assenting to truistic assertions or dissenting to false-sounding ones when we are not sure exactly what is being said. for example. of learning to smoke marijuana for pleasure. to make plainer the fact that there is an additional intended contrast involved. nevertheless. Socrates’ truistic or flagrantly false formulations would be designed to solicit the assent or dissent of his interlocutor upon production of which he would then proceed to demonstrate incoherence – the interlocutor would then find themselves where they did not expect. since we can either make a kind out of all things which can be said to be real (for even hallucinations can be real ones)12 or admit only kinds whose individual members cannot be said not to be real in any sense (per impossibile since. Showing in a convincing way that this contrast is empty and that. for example ‘real’ and ‘reality’. Suppose. Surely we would protest at the self-evident nature of the question and we would . those who see promise in the ontological mode of thinking are not bothered by these facts but attempt to render them irrelevant by professing their concern not with ‘all real things’ but with ‘all ultimately real things’ or.13 But what about ‘real ancestors’ and ‘real hallucinations’.

Turning this assimilation into a question we get the idea that the question ‘does it have a meaning?’ can decide the matter (or at least be posed) in each case. but the capital was never in motion between them. The question about the meaning of red is a question about what the colour symbolises. starting. linguistic theory. observable or non-observable. in other words. that appearance is misleading. if at all. Wittgenstein PI: §62). it may be suggested. But. the moment we demand a yes or no answer in each case. cannot work because it amounts to trying. with the idea that everything is either meaningful or meaningless.’ (Z: §202) 15 It might be said that the conscription of ordinary words in the service of ontological projects does not require that they all be thought to ask the same ontological question. to give a minimal counterexample. we create difficulties for ourselves. not about what the word means. If all meaningful things have a meaning. meaningful or meaningless. but as argued. the question calls for clarification (cf. or in the case immediately above. We are now in say that it is obvious that reality cannot consist of what is unreal. It is only a short step from there to concluding that reality must consist of what is real. not everything which [has moved] has been in motion’ (1989: 43) (my italics and alteration from ‘moves’).Ontological Confusion in Social Theory 77 they part of reality?14 To the extent that this question encourages us to give a quick ‘yes or no’ answer it is misleading. per impossibile. Furthermore. and words are meaningful. my friend whom I see every day. perhaps. to substitute objective for indexical expressions (Garfinkel and Sacks 1970) in order to guarantee a uniformity or fixity of meaning that is required by the universal scope of ontology (applying to all entities)15 or linguistic theory (applying to the whole of language). therefore all words have a meaning – accordingly the expression ‘what is the meaning of X?’ is predicable of all words. or to mention other exhaustive dichotomies. for example. but ‘the meaning of the word “red”’ is also awkward. Putting the word in quotation marks. it is an important fact that we do not say ‘what is the meaning of red?’ when asking about the word ‘red’. Surprisingly. But no “act of recognition” takes place. ‘Consider well how we use the word “recognize”. for the moment we try to apply the dichotomy exhaustively. regardless of whether the words in fact apply in the same way. What these complications show is that the conscription of forms of words to do ontology. Asking about it could elicit the response ‘red is a colour’ but we would not say that the word ‘red’ means a colour. I recognize the furniture in my room. Wittgenstein also emphasised how we could be led astray by neglecting the nuances in seemingly tautologous expressions. might solve the problem of exactly what we are asking. 14 Some further examples might be instructive. It is thus ill-advised to say that everything is either real or unreal. This might be true for ontology loosely conceived but not for ontology as a comprehensive list . It seems an easy task to divide what is real from what is unreal on the basis of general principle. For as it stands. theorists can then reason with regard to what is meaningful by aligning cognate expressions: ‘To be meaningful is to have meaning’. Champlin observes that: ‘A cold moves as a capital moves … there are many rivers between Aix and Paris. nor that the colour red is the meaning of the word ‘red’. which the king and his courtiers had to cross.

In the following passage Stenlund describes this equivalence when it is put to use in order to ensure the uniform understanding of ‘refers to’. 16 In parts of philosophy retaining the linguistic ideal of formal logic. namely the causal criterion: the causal criterion turns on the capacity of the entity whose existence is in doubt to bring about changes in material things.3 The Causal Criterion and Realist ‘Non-Observables’ also Require Fixity of Meaning The purpose of the discussion so far has been to sensitise us to some of the facts about the use of words as a preliminary for the issues to come. Notice that a magnetic or gravitational field satisfies this criterion. Thus. but not a criterion of perceivability. supposedly. however. as we will encounter this conception of reference in examining the understanding of language ontology depends on: It is a typical feature of this grammatical scheme as well as of others that it is connected with similarities in the mere linguistic forms or paraphrases of ordinary language.16 4. also emphasised in the previous chapter. to be is not to be perceived. X denotes Y are acceptable paraphrases (1996: 205). that the price to pay for pressing certain forms of expression as applicable in all cases is too high. ‘is a name of’ and ‘denotes’. . It is precisely this procedure which. as Stenlund notes. but rather (in the last instance) just to be able to of what exists or what is real. equivalence is asserted between expressions based on their formal similarities. via their mapping onto the calculus as paraphrases. If we are to understand the question whether social structures are real then we need to know what critical realists wish to oppose ‘real’ to. it also implicitly requires that the classes of what exists and what is real are co-extensive. it is nothing less than conceptual confusion. Furthermore. the conscripted words are often thought to derive their fixity of meaning from the mathematical-logical calculus. The latter requires that all members of the list exist or are real in the same sense. Even without the help of the calculus. ontological expressions constitute ways of reading formulae containing the existential quantifier. for language to refer. For example. It seems that ‘causal inefficacy’ is the intended contrasting term as can be seen by the universal criterion proposed. X is a name of Y.78 Clarity and Confusion in Social Theory a position to appreciate the fact. the seeming tautologies that are generated by connecting expressions based on their etymological or morphological connections while neglecting their use are bound to lead to confusion. The scheme is isolated by focusing on the surface grammar of language … the idea behind the grammatical scheme for ‘reference’ is of course that it is meant to be applicable and valid in all cases where expressions of the form X refers to Y. The reader should keep the following in mind. I will expand in Chapter 5 on the forcing of a mode of expression in the process of operating with a ‘grammatical scheme’. (1996: 205) (my emphasis). as what it is. On this criterion. cuts ‘across several conceptually different forms of use’.

The standard hermeneutical fork. Winch 1958[2008]: xii). the concept of cause. a question that undermines the idea of a single causal criterion. This means that any criterion intended to fix the sense of ‘real’ ends up being no single criterion after all. But there are only certain kinds of cases where this is true. His lifelike hologram brought tears to my eyes.17 Now.e. Furthermore. be careful it will bite you’. Consider some simple examples: • • • • • The thought of losing the case made me shiver. Apart from displaying the variety of things that can be causes. generated by the conceptual/ perceptible dichotomy of classical empiricist ontology. or a hermeneutics drained of causal import and impervious to empirical controls. for doing so produces a most natural way of speaking (cf. so producing either a conceptually impoverished and deconceptualizing empiricism. The question can only be exactly in what ways various causal concepts can apply. 18 Consider also what Archer has to say: ‘[O]ntological status needs to be accorded to such aggregate (and emergent) social properties precisely because they are mechanisms facilitating or frustrating various policies’ (quoted in Elder-Vass 2007a: 27) and ‘The existence of structural properties and powers is established by the causal criterion.18 It may be granted that the equation of being real with being a cause or having effects in a certain sense may have its home in certain activities. 17 See Pleasants (1999: 193 n10) for a different complaint against the causal criterion. he stipulates the following form as criterial: to be real is to be causal or to have real effects. Their determination brought shame to my red face. (Bhaskar 1998[1979]: 12) We have already considered that fixing an opposing term to ‘real’ does not allow us to claim to have universally applied the expression in a single sense. which [is] invoked by Winch. Thus both parties [that is ‘interpretivists’ and ‘positivists’] to the naturalist dispute have assumed that the social must be either merely empirically real or in effect transcendentally ideal. provided that it does.. Catching with my eye what I thought was a ghost caused me to jump. that is in terms of generative effects’ (quoted in Varela 2007: 204). This fact made me realise that I was being excessively apprehensive. . For example: ‘Is this snake real?’. is not any more uniform itself than the concept of the real is and thus also requires clarification as to its application in each particular case. ignores precisely those possibilities opened up by a causal criterion for ascribing reality. in each particular case. For any universal application will inevitably require an immense amount of clarification as to the way in which the expression applies. Besides. i. ‘Yes.Ontological Confusion in Social Theory 79 do. there is a family of causal concepts applying equally to what ontological enthusiasts would be rather reluctant to class as real. Bhaskar does exactly this. what these examples show is that nobody should be sensibly against applying causal concepts to human affairs as a general matter. utilised as a presumed anchoring point.

Dichotomies. as we have seen. William Outhwaite. are. the concept of ‘empirical world’ arises from an illegitimate reduction of an ontological question to an (empiricist) epistemological one. however. which is limited to (what are considered) observables. once again. in effect. non-observable and where the notion of observability or non-observability have no application (or none yet). the terms ‘observable’ and ‘nonobservable’. But if the latter is true. Starting with (a). to what they do and what eventuates when they do … [t]he empirical is defined as the domain of experience’ (2000: 12). in this case. Although some critical realists apart from the term ‘unobservables’ also use ‘nonobservables’. a reason for holding that the real is more extensive than what we can perceive. unobservables and non-observables which are understood as non-unobservables as well. . a way of stipulating the unrestricted application of a pair of terms. explains: Implicit in the … critique of empiricist ontology and the discussion of laws of tendency was an attack upon the concept of an ‘empirical world’. the actual refers to what happens if and when those powers are activated.21 For example. It is thus better to avoid giving a ‘yes or no’ answer to the question whether the social is observable or non-observable. This constitutes. it entails that we have to take into account things that can be said to cause even though they might not be observable. then there is no such thing as not observing society either and we could equally well say that society is not unobservable. Consider. Bhaskar’s stratified ontology (empirical. nevertheless.80 Clarity and Confusion in Social Theory As is evident from the quotation above. the question of the observability or non-observability of society or the social is a case where the connection is unclear and thus the question is not quite intelligible as it stands. Bhaskar (and others) commit the conceptual error of turning a distinction between observables and non-observables into a dichotomy. 20 As Sayer. Now it would be extraordinary if ‘the real’ just happened to be exactly coextensive with the limits of our sensory powers (1998: 133). the ‘non-observables’ Margaret Archer casually lists: 19 Sayer’s explication of the ontology is useful: ‘Whereas the real … refers to the structures and powers of objects. 21 The distinction could also be made by saying that there are observables. real)19 extending beyond what we know and can observe is (at least in part) motivated by (a) the identification of what is real with what has real effects and (b) the realisation that a causal power cannot be identified with its exercise.20 In advocating this position. at the most general level. There is what is observable. then. wants to argue that we must accept that ‘Society is not observable’ because ‘[t]here is no such thing as observing society as such’ (1998: 284). actual. they are not employing it in this way. But there are concepts in the language which allow for the connection with the question of observability or non-observability and concepts where that connection is not provided for. Interpreting ‘empirical’ here as ‘that which is observable’. further. The improvement this conception offers is presumably most pronounced when contrasted with an empiricist ontology.

and which relegates everything that does not conform to the picture of a directly given physical object to the category of non-observables. The fact that she does not lose the power when she is not driving does not mean that the power is present somewhere but impossible to observe. one that derives from some version of empiricism. Observing powers or abilities might not be the same as observing colours or shapes but this does not render the former non-observables. is what observing her power to drive amounts to. but frequently we just are dealing with non-observables which remain that way (i. ecological imbalance. for example. are they in principle or only contingently so and what would it mean to observe them? Is there a clear sense in which we are indeed restricted in our enquiry due to (an intelligible) something we cannot do? And. Furthermore. The mistake here lies in overlooking that what it means to observe something depends on the something. in open systems. third dimensional power or ideological mystification.Ontological Confusion in Social Theory 81 Some of the non-observable social factors which concern us may be open to humanistic hermeneutics (my motives in writing this book or yours in reading it). exercised unrealised. kleptomaniacs possess the tendency to do so.’ (2008[1975]: 222) . the fact that causal powers are not exhausted by their manifestation. powers (or what he sometimes22 interchangeably calls) ‘tendencies may be possessed unexercised. it is (perhaps for this reason) hardly clear what it means to say that the above items are non-observable. they are distinct from a psycho-analyst’s attempts to help clients to awareness of something unconscious). (Archer 2000: 48–9) In reading through this passage. and realised unperceived (or undetected) by men’ (2008[1975]: 175) also fails to support their non-observable status because it 22 But only sometimes. In this category would go international finance markets. is it not possible to specify ways of observing some of the above items? We might take stock of these probing thoughts for now and move on to consider (b). that is. This. It is plain that Smith’s power to drive a car is something one can observe when one sees her driving. are institutional contradictions nonobservable. least of all in some transcendental realm. finally. as in ‘All men (living in certain kinds of societies) possess the power to steal. It is noteworthy that Varela and Harré (1996: 318) seem to think that (b) is the only reason that Bhaskar has for being committed to a stratified ontology that features non-observables.e. In the appendix to A Realist Theory of Science Bhaskar acknowledges that he uses two concepts of tendency. Moreover. While this thought is correct in itself (see Hacker 2007: Ch. it cannot warrant cataloguing powers under non-observables. Why. institutional contradictions. Bhaskar’s programmatic idea that. one interchangeable with power and one distinct from it. Nor does being unable to tell whether someone can drive just by looking at them render such a power non-observable. 4). after all. we might suspect that Archer adheres to a very restricted idea of what can be observed.

they only need to have V-ed. having the tendency to win cannot be possessed unexercised. and most importantly. to say that someone has the tendency to win (as opposed to the power to win) is not to report what he will do given certain circumstances (once. unless necessarily leading to a change in policy. i. Secondly. whether policy changes in the end or not. we do just manage to keep our tempers … they both make the mistake of seeing the fulfilment of a tendency a condition of its exercise’ (2008[1975]: 89–90) (my emphasis). the tendency to do so cannot be possessed without ever having been ‘exercised’. But here Mill is right and Geach is wrong. Yet again. . say. This. If we look at the examples he invokes. as Ryle (2000[1949]) observed. although the power to lose one’s temper can be possessed without being ‘exercised’. nothing can be in motion or really going on. Equally. stimulus conditions are also fulfilled) but (since it is not entirely up to him) what the usual outcome of him competing is. not what further end they may be used for.82 Clarity and Confusion in Social Theory is deeply incoherent. not X-ed. Finally. otherwise that person would be faking losing their temper). does not count as exercising that power. running fast is. so it does not make sense to speak of some winning going on but not reaching fruition. for example.23 it is clear that none of them warrant his conception: Firstly. we would rather say that one manifests that tendency or power (because it is not clear that it makes sense to say that one exercises it. the construction ‘exercise of a power or tendency’ is an ‘achievement construction’ which signals success with regard to the power or tendency in question.. To say that one ‘exercises’ the tendency or power to lose his temper without realising it is unintelligible for reasons we will shortly come to.e. Moreover. when in a particular occasion they respond angrily to.24 Accordingly. If the power in question is the power to alter policy then ‘exerting pressure on officials’ by itself. Although one can fail to win once or even many times and yet not lose the power to win. besides enabling conditions. towards explaining which the exercise of the tendency goes some way. 25 The verb ‘to win’ is an achievement verb. winning is not the realisation of the tendency to run fast. To say of someone that they have exercised their power to V in order to X. 24 A monarch exercising their power to exert pressure on government officials with the aim of altering policy can be said to have exercised that power if s/he manages to exert that pressure. unless someone has frequently won in the past. say. powers and tendencies are individuated by what they are powers or tendencies to do. a mild provocation. this is not so for their tendency to win – which a person does not ‘possess’ in the same sense since ‘tendency’ here is a statistical notion. if anything.25 When someone is running fast it is not 23 Now when a tendency is exercised unfulfilled two things are not in doubt: (a) that something actually happens. … Mill’s mistake here is to suppose that whenever a tendency is set in motion the effect must be in some sense (or in some realm) occurring (as if every time we ran fast we had to be in some way winning). Balaam’s ass is pulled in two ways. But Geach (and following him Ryan) in ridiculing this position make the converse mistake of supposing that whenever no effect (of a given type) occurs. is also to realise that tendency or power.

institutional contradictions. it is important to realise that the pervasive analogy in CR with the postulated entities of natural science (for example. one can witness the depression in one’s neighbour and at the same time the effects of the economic crisis on Greek society in one’s neighbourhood.27 The illusion that there is such a problem owes its existence to a false analogy with natural science. We have seen throughout this section that the mistaken idea that expressions can be universally applied with fixity of meaning is one constitutive of critical realist theory – its ‘causal criterion’ and postulation of ‘non-observable’ entities. or the mind. ‘society’. We will encounter this analogy again in Chapter 6. ‘the social’. the procedures of which are constructed on this basis. we see that there is nothing in principle non-observable about mental things or about ‘large scale processes’. they are just not running fast. For now. not new entities’ (1998: 300). To bring the discussion of non-observables to a close. no ‘nonobservability problem’ which requires us to postulate transcendental entities or powers. that society. 2005: 39). once we specify what would count as observing society. 26 The analogy serves. For instance. Moreover. There is. for example. whose pernicious influence is compounded by the idea that ‘observable’ and ‘non-observable’ possess a fixity of meaning. as Archer does for example. More ‘ontological’ conceptual procedures will be scrutinised in Chapter 5. including. Scientists routinely assume the existence of unobservables depending on them both to predict novel facts and to intervene successfully in the world’ (Shapiro. for example the resources provided as part of a research project. one can also witness the latter in a large number of communities around the country. to distinguish realism from empiricism: ‘… realists also differ from empiricists in the status they accord to unobservables. with the possession of the required means. By closing the supposed gap between a power or tendency’s exercise and realisation. what is exercised is their tendency to run fast. any space for ‘non-observables’ has also evaporated. This fact is obscured by thinking of ‘non-observables’ as matters of potentially very large (society) or very small scale and hidden nature (the mind).Ontological Confusion in Social Theory 83 their tendency to win that is being exercised but which remains unfulfilled if they happen to lose. we can proceed to examine Harré’s critique of CR. But again. financial systems and the rest are nonobservable and not (before we can do that) to find a technical means of detection (which is arguably what is lacking regarding the particles postulated in physics). once again. with the above in mind.26 is completely out of place. . Nor does it make any sense to say that they are exercising but failing to realise their tendency to run fast when unsuccessfully trying to run fast. ‘mental events’. that is. hence. the graviton). as the difficulty for social theory is to specify what we mean by saying. 27 Nicholas Capaldi puts it boldly and simply when he says that ‘in “social science” … we merely discover a new language. most importantly. ‘the economy’ and ‘the state’. ‘positrons’. of ontological enquiry. that they can be predicated in the same sense and without need for clarification to all expressions. powers and tendencies – and.

(2008[1995]: 250) I will avoid getting into how the above (seemingly inappropriate conception when it comes to the ‘social world’) might apply to social science. Descriptions belong to the world of society and of men. objects belong to the world of nature. and understand something about our bit in it. economic systems and mythical things of that sort. he assumes a position much closer to the. The focus of Critical Realists and other such well-meaning but metaphysically misguided people being on constituted authorities. Harré’s stated motivation in engaging with CR is: [to] identify what it is that bothers me about the enthusiastic and well-intentioned claim that social structures are causally efficacious and therefore if we want to ameliorate our lives we should change them.4 Harré’s Critique Harré frames his argument as an intervention in the debate between critical realists and social constructionists where. broadly conceived. viz. social structures are postulated as (one of) the proper objects of social science by way of a ‘transcendental deduction’ (1998) – argument that is directed from the actuality of science to the conditions of its possibility28 – and their reality is established by appealing to the causal criterion. is the systematic attempt to express in thought the structures and ways of acting of things that exist and act independently of thought.Clarity and Confusion in Social Theory 84 4. the above excerpt is instructive in that it provides a clear formulation of the standard against which SC is found wanting. Nevertheless. but we can only know them under particular descriptions. We express [our understanding of] nature in thought. Social Constructionism (SC) camp. The following passage from A Realist Theory of Science exhibits the logic of arguing this: Things exist and act independently of our descriptions. despite Roy Bhaskar having acknowledged him as an intellectual progenitor of CR. it is useful to note that CR is designed to avoid what Bhaskar (2008[1995]) terms ‘the epistemic fallacy’. then. Based on the general conception of science articulated in Bhaskar’s reasoning. I claim that these exist only as discursive categories. the former is not exhausted by the latter. in fact. (2002a: 121) In order to understand what the direct target of Harré’s attack could be. The world is structured and complex and not made for men. After all. the implication for the study of society. that is what Bhaskar’s The Possibility of Naturalism sets out to demonstrate and indeed qualifications peculiar to the varying subject matters of different sciences are offered.. It is entirely accidental that we exist. Having established an allegedly ontological basis. Science. . equating what there is with what we can know when. as further spelled out in Margaret 28 See Kemp (2005) for some problems in applying this method to the social sciences.

as Carter (2002: 134) registers. but also in that it is seen as pulling the rug under the feet of sociologies that wish to talk about something more than individuals. namely ‘event causality’ and ‘agent causality’. according to Strydom’s understanding. somewhat more subtly. that is. Harré’s attack is seen to be devastating not only in that. It is instructive to consider at this point exactly what his contention is. roles together with relationships between roles. Harré’s argument takes the form of a reductio ad absurdum in that he starts with two conceptions of causality and by examining what they can apply to ends up with the thesis that social structures are not the kind of thing that could be causally efficacious and that. The rules by which they manage it must not be reified into a transcendent realm from which they exert their benign 29 Indeed. or acts (from which follows a static or dynamic conception of social structure. Although Harré’s attack is directed at the above conception (especially as it relates to the way in which it is possible to effect social change). (2002: 125) . Strydom in his response quickly seeks to restore the basis that has been taken away: ‘As I see it. that social structures are not the kind of thing critical realists imagine them to be. He argues that: social reality is exhausted by what people do. is that ‘both agency and structure have real causal powers and properties sui generis’. therefore. I now pick up the thread of Harré’s argument again. Social structures are understood by Harré to be secondary products ‘of the activity of people acting according to rules. they are not real ‘in any substantive sense’ (Harré 2002b: 147). Having briefly sketched what is perceived to be at stake. the two problems can be avoided only by means of an interactionist theory that is based on a broader ontological foundation’. as ‘examin[ing] the interplay between these two sets of powers and properties’ (Carter 2002: 134). Firstly.Ontological Confusion in Social Theory 85 Archer’s work. as removing the necessary ontological foundation29 by proposing a more restricted ontology. customs and conventions’ (2002a: 115). i. Harré is not flatly claiming that ‘there is no such thing as’ social structures but. are the only moulds in which social structures will have to fit if they are to be thought of as causally efficacious. He then delimits structures as being either institutions. the attack is not limited to the work of critical realists as such – Strydom himself is not a critical realist – but has broader and devastating implications for social science (2002: 124). he judges. then.. respectively).e. it is seen. which serves as the foundation to a methodology defining the objective of social science. given the critical realist ascription of reality based on the causal criterion. it should be emphasised that. Harré makes the (controversial) move of importing two conceptions of causality from the physical sciences (and in this he is no different from the critical realists). Harré adds the proviso that we should be cautious in our conceptualisation of rules. I will go on to briefly sketch his argument but I will only elaborate on the parts relevant to the problem of ontology. it denies the viability of a methodological programme which purports to investigate the interplay between structure and agency. These. Presumably.

in short does the question that the referents of social structure expressions could be causally efficacious make sense. (2002a: 116) (Harré’s italics) Taking stock of the fact that the ontological project has already started taking its toll in placing undue emphasis on statements of the sort ‘social reality is exhausted by people and what they do’. The question of the question itself is a critical one. What is the category of efficacious agent that brings them into being? Only persons. Harré frames his argument in the following way: If we try to come to a judicious conclusion about the question it will not be a simple yes or no answer. why the comparison with Wittgenstein is a clue to what remains unchallenged but deeply problematic in both Harré’s and critical realist accounts. it is useful to return to Harré’s initial statement about how he will address the question of whether social structures are causally efficacious. namely discursive acts. and to claim that neither social structures (conceived as acts or role structures) nor rules and conventions are such: Now we have to ask whether any of the social things we have identified could possibly meet these conditions for ascriptions of causal efficacy. I would like to move on to the next (and final for our purposes) move in Harré’s argument which is to lay down the condition that causal efficacy is to be ascribed only to powerful particulars. many of which are discursive. The conclusion based on this is that only persons are causally efficacious agents and that our ontology should contain persons and their actions but not social structures. by showing that social structures do not meet the causal criterion. that is whether it makes any sense at all.86 Clarity and Confusion in Social Theory influence. But both rules and acts are discursive products. also. however. In short. It is the grounds and status of this conclusion that I wish to question. What reality do they have? Again we must distinguish a mode of being as immanent in practices. and a mode of being as concrete instructions. that is.5 A Wittgensteinian Move? In this section I attempt to show that what Harré says has a potentially questionable Wittgensteinian character and. 4. Before doing so. rules and the acts that people jointly performed within the frames of possibility that they determined. the only reality norms and rules have is of the same ontological status as the activities they ‘govern’. No rule or convention is the kind of thing that could be an efficacious agent. There were roles. which are real as instances of discourse. At the end of the day I hope to show that such referents are not the kind of entities that . (2002a: 117) One can claim that thus far Harré’s attack has been in tune with CR in that he has sought to refute it by its own account.

. is not an event’ (2002a: 117) seems to invoke the picture of social structures as entities being made of something. That is. the Wittgensteinian thing to do when confronted with ontological reasoning is to expose the underlying sources of confusion but Harré does not seem to be taking any issue in this respect. In sum.Ontological Confusion in Social Theory 87 could be causally efficacious. though composed of events. instead of speaking about the concept of social structures. an old philosophical idea P. 30 Harré’s claim that ‘a social structure.e. by making a grammatical picture out of one of the many uses of ‘refer’ (for example ‘I am referring to this piece of furniture’ and not. I would. or entities we use social structure expressions to refer to. understood simply as whatever we use expressions to talk about. It seems to me that Harré’s penchant for expressions referring to entities which are powerful particulars is not too distant a relative of this idea. Furthermore. in that he is challenging the question itself. From what Harré says here it is clear that he is addressing the question in a rather special way. but they’re not the right kind of thing to do the sort of work that ‘some people’ would like them to do (2002a: 112) (my emphasis). and treats it as depending on a consideration of ‘what kind of thing something is’ which is itself understood not as a question about the logical status of a concept (another way of talking about what it makes sense to say) but as a separate question about the ontological status of an expression’s referent.30 and/or (c) the requirement that expressions be referring is understood as a condition of their having any meaning. about what it makes sense to say. say.F. In examining whether the question makes sense or not. (b) ‘entity’ is thought to be applicable without restriction but is really understood along the paradigm of physical things (is a bad habit an entity?). however. he submits. i. like to point out that speaking in this way does nothing to loosen the grip (and likely falls prey to) the ontological fascination which ‘entities’ and ‘referents’. he more or less equivalently speaks about what things. and for all of Harré’s objections to the fallacious and reificatory metaphysical talk of real structures (2002b). then I can have no quarrel with this part of his argument. that Harré takes a question about logical grammar (to use Wittgenstein’s term). where (a) ‘refer to an entity’ is understood in a restricted way. Strawson (1950). the crucial test upon which the sense of the question depends is. ‘what kind of thing’ social structures are. he seems to be presuming that social structure expressions are ‘referring expressions’. Notice. I’m not saying that there are no such things as social structures. has forcefully argued against. ‘I am referring to a bad habit’). instead of performing the more radical move (which I wish to argue is the required one) that would challenge the sense in which social structure expressions can indeed be said to refer to any entities. however. among others. It may be objected that in the quoted passage Harré is merely switching between the formal and the material mode. If this is indeed the case. This is because these expressions are typical sources of confusion during ontological discussion. .

it would. Williams and May characterising Harré’s position: 31 Another way of making this distinction is to say that the relevant response is not ‘it does not go anywhere’ but ‘it does not go anywhere’. it is clear that to show that the question of the causal efficacy of causal structures is senseless entails rejecting the view that they are not causally efficacious as senseless too. are under the impression that Wittgenstein denies that we can know what we think and feel … But the ‘can’ and ‘cannot’ in these remarks are grammatical. and to show that this is more generally the case. First.31 If this is the force behind Harré’s remarks on the question whether social structures are causally efficacious then we are in agreement as far as that goes. it needs to be established in somewhat more detail what the Wittgensteinian move regarding the problem of the sense of an expression could be. he seems to be treating what it makes sense to say as picking out some ultimate reality captured by the parsimonious ontology he is substituting for the critical realist one. the proper response to the question ‘Does music go anywhere when we do not listen to it?’ is not ‘No. however. I stress that the reason I have placed such emphasis on the ‘Wittgensteinian move’ is because Harré’s characterisation of what he is doing is clearly meant to follow a Wittgensteinian line and because the comparison with Wittgenstein helps us examine the status of Harré’s conclusions that only persons can be causally efficacious. At the same time. it does not go anywhere’ but ‘No. Harré’s claim that in doing sociology we should keep Ockham’s razor in mind is ample evidence that it is an overpopulated ontology he is against and not the idea of ontology per se. (if we are still tempted to say ‘No’) because the question is senseless and equally senseless is the response that it does not go anywhere’. equally make no sense to say of me that I do not know what I am thinking. therefore. (Cavell 1971: 189) Based on the above as well as on a number of examples discussed throughout this book. . The implication is not that I cannot know myself. The following passage discusses Wittgenstein’s response to the question of whether I know I am in pain: Other philosophers. or that I do not know I am in pain. To illustrate this point by using an example of patent nonsense. I believe. but that knowing oneself – though radically different from the way we know others – is not a matter of cognising (classically ‘intuiting’) mental acts and particular sensations. they mean ‘it makes no sense to say these things’ (in the way we think it does). by pressing the question of whether they are reminders about what it makes sense to say or metaphysical theses. Stanley Cavell provides some guidance concerning the difficulties in responding to a nonsensical question.88 Clarity and Confusion in Social Theory Again. In order to bring into sharper view the fact that the idea of ontology remains intact. consider the following remarks on the concept of class (my emphasis in what follows). Nevertheless.

is a taxonomic term that has no causal powers. (2002a: 114) . Once this kind of loose talk is allowed we are well on the way to the mystification of macro-politics. (2002b: 146) The people are like the magnetic poles we invoke to explain the pattern of iron filings through which the structure of the field is manifested. (2002: 108) The above remark refers to an earlier article of Harré’s where he also argues against: [the] temptation to use ‘power’ for the way macro-entities can exert influences of one sort or another. I include some additional material regarding explanation which I will utilise shortly: I am not sure what is left of the position of someone like Margaret Archer. This and other macrocollectivities are viewed as rhetorical devices with clear implications for a social theory. Harré is quite explicit about this. let us take a look at Strydom objecting to Harré’s position: It is obvious that such a psychological. such as critical realism. race and gender. It no more explains the predicament of [a] pensioner [who could not present his situation in a way that a welfare agency could register and act on] than does being a baron explains why some people were so worked up at Runnymede. for example. so that a class that is said to have power is thereby implied to have responsibility. agency-theoretical understanding of social structure leaves no room at all for what is social scientifically accepted as central structural features of social reality and organizing principles of social positions. problems of equal access do not derive from structural variables but are rather attributable to the incompetence of the disadvantaged. namely class. According to him. which has transformational intentions. They are a necessary foundation on which the possibility of social structures depends. They are the underlying generators of the structure. Class. For social structural constraints. It is not a property that may simply be ascribed to an individual in terms of their positioned capacities. (2002: 128) And the final excerpts are both from Harré’s initial contribution and from his rejoinder to Carter and Strydom. In fact.Ontological Confusion in Social Theory 89 the only powerful particulars which are held to exist in the social world are persons. are no more and no less than ‘just story-lines’. We must dismiss any such concept as ‘someone’s class position’ as no more than a façon de parler. (Harré 1981: 157) Next. I take this usage to be transparently rhetorical since it is clearly involved in exploiting the responsibility implications of that use. but one that exists in the mind of the classifiers and so is not a property of the social world. in his view.

structures. to speak tongue-in-cheek. this is strongly reminiscent. the other has to do with explanation. if only because it is all too easily forgotten. (2002b: 143) There are two salient. (Archer 1995: 12) Now. To state my objection bluntly. the fact that to speak about those things is not in any way extraordinary and that as speakers of the language we usually understand what is being said when people make use of these expressions should give us pause for thought. For given the above characterisations it seems that to speak about ‘changing the economy’ or someone acting in a certain way ‘because they are middle class’ (or to use concepts Elder-Vass [2006] lists such as discourse. However. 4. or to speak loosely or rhetorically. You do not explain why a lion prefers meat by telling us it is a carnivore.) is to speak in a special kind of way. society or with employing a host of other expressions that are used in the description and explanation of social affairs. etc. a picture which is foisted upon us by the idea that we must think ontologically. although in an oblique way. then we are likely to presuppose that picture. capitalism. especially when it is claimed that these expressions are somehow inadequate. then there can be no social theory without an accompanying social ontology (implicit or explicit).6 A Distortive Picture of Language (and Explanation) In this section I confront the idea that there is something wrong (i. as can be witnessed in the following exemplary passage where Archer is attempting to establish ontology as an inescapable concern: Since theories are propositions containing concepts and since all concepts have their referents (pick out features held to belong to social reality). One has to do with language. but form part of the vernacular that (usually) both sociologists and other members of the society speak. They only become one given a certain picture of how language relates to ‘what is out there’.e. The reification of the concept of ‘working class’ and similar macro-concepts is a case in point. which will turn out to be not too unfamiliar. that these expressions are not sociological inventions (see Rose [1960] for evidence that sociology is through and through dependent on common terms). interlocking conceptions exhibited in the quotations above.90 Clarity and Confusion in Social Theory Mistaking a taxonomic category for a substantive entity. the state. If we accept that we must do so. money. institutions. values. groups. I will start with the conception of language and work towards the one of explanation. It must be remembered. there is nothing loose about these ways of speaking. that is to say they are not technical terms. nor are they a mere façon de parler. ontologically unsound) in some way or other with talking about classes. of the idea in early analytic philosophy (adhered to by Bertrand Russell and other philosophers) of a .

Once language and ontology are uncoupled. however. Cook dubs the misconception at work the ‘picture of language as a conceptual schema’. and be meaningful by virtue of the fact that they refer to the entities that populate ultimate reality. too. 8). ontological or of any other kind (cf. and mistaken. Winch 1987: Ch. in thinking that there is an ontological terrain to be mapped. The reason is that under the picture of ‘language as a conceptual scheme’ the existence of a certain concept in the language is interpreted as ipso facto committing the speaker to the existence of an entity falling under that concept and hence entailing the ascription of a belief about existence to that person. this involves using language to express different beliefs which often are conflicting or even contradictory ones. either a good one or a poor one. involving the central idea that ‘ordinary language is a map of the ontological terrain’ and that it can be evaluated on the basis of whether it does justice to it or not. But her conception of a natural language is dangerously close to Russell’s. namely in the idea that every language comes with an ontology built into it and therefore speakers (be they sociologists or lay folk) must have certain implicit or explicit ontological commitments. Finally. more recently. she can only have in mind the ordinary concepts that comprise the natural language sociologists speak. that is. not vice versa. an ideal language in which referring expressions would have a distinctive logical structure.Ontological Confusion in Social Theory 91 logically proper language. It can then be seen how Archer’s idea originates in the same place as the philosophical idea that a logically proper language would show unequivocally what the ontological commitments of its speakers are. But the language itself is neither of these things. true or false. Elder-Vass’s [2006: 3–4]). It is clear that Archer is not talking about any ideal language. and these descriptions can be accurate or inaccurate. not only . but the language itself is not a set of beliefs. Similarly. (Cook 1980: 36) The key point in this passage can be brought out in the following manner: one can use language to describe what the world is like. Cook goes on to argue that: [N]ot only have philosophers not shown ordinary language to be a ‘poor map of the ontological terrain’. we can employ concepts to make true and false assertions and the making of both kinds of assertion presupposes the meaningfulness of our concepts. This includes assertions about the existence or reality of something which depend on our concepts being meaningful. We can now appreciate why the fact that Cook puts together the idea that concepts embody the ‘common sense view of the world’ with the idea that language is an ontological map is a crucial step towards understanding Archer’s conception (this is Bhaskar’s conception too [see Cruickshank 2010] and. John Cook both illuminates and rebuts the implications for natural language that the ideal of a logically proper language carries in his superb article on the fate of OLP (1980). This idea is supported by the related one that concepts ‘embody the “common sense view of the world”’ (1980: 36). In short. our language is not a ‘conceptual schema’ at all. instead. but they have been mistaken all along in thinking that our language is any sort of map.

Harré keeps the picture intact and separates the literal talk which is ontologically sound from the rhetorical talk which is a mere façon de parler. This is a common intellectual trap and. least of all metaphysical beliefs. (2002b: 145) In this passage Harré is attempting to expose the reificatory impulse critical realists give into with the result that they turn social structures into concrete (and therefore manipulable?) things. There is no such thing to be changed. What I am denying is (a) that those referential uses are what is fundamental about language. What can perhaps be said is that the sense in which one changes discursive practices is different to the sense in which one changes the social structure (although in many cases changing discursive practices is changing the structure or might result in a changed structure) in the same way that the sense in which one changes a light-bulb can be said to be different to the sense in which one changes the light level in a room (a case in which the former might actually count as the . it seems that. I want to return to Harré. instead of questioning whether we must adhere to this picture of language. Consider. (b) that there is or should be an ‘ontological basis’ to meaningful language use and (c) that it is sound or possible to infer uniform referential uses from the concepts without looking at what is being said with them. in order to distinguish between ontologically proper and improper ways of speaking. but the predominantly vernacular concepts (or second order concepts which presuppose the former) that might feature in these propositions do not commit us to anything by themselves. I should emphasise that I am not denying that we can and do use language in referential ways. he wants to extract more out of the fallacy of misplaced concreteness than the latter permits.92 Clarity and Confusion in Social Theory does it become possible to entertain the idea that the ontological consideration is both external to language and dispensable but one can also see the very idea of ontology as arising out of a confusion concerning what can be said about language as opposed to what can be said about our doings with language. indeed. There are widely shared discursive practices through which the social world is constituted and reconstituted each day. However. thinking in ontological mode. The theoretical propositions Archer mentions in the excerpt quoted above might commit us to the existence of something (but we will have to look at the propositions and what a theorist might be saying to decide this). which attributes causal efficacy to powerful particulars. This is a clear indication that he does not take issue with the attitude that drives us into confusion but finds the solution instead in using his scheme. Lest I be misunderstood. It has hopefully been thrown into sharp relief that. for example. what he says in the following excerpt: Trying to ameliorate the quality of human life by trying to change the social structure is similarly mistaken. it is difficult to see how one might quarrel with Harré on this point. Bearing this in mind.

. instead of looking at words as they enter into the doings of speakers who. This is an ultimately suspect way of accounting for the grammatical facts. Toulmin likens the theoretical attitude to an aesthetic attitude whereby propositions are being treated as ‘frozen statues of statements’ (2003[1958]: 167). is that one of these two uses is somehow ontologically superior because it is based on referring to (real?) entities whereas the other is ontologically empty. what contrast is being marked by the insistence on changing social structures (for example. That is all’ (2002b: 145). or a façon de parler.e. it is otherwise merely rhetorical (or metaphorical). ontologically invalid. 33 I would like to seize this opportunity to exploit a number of figurative ways of describing the theoretical attitude towards language: Henson compares it to ‘the distortion of one’s normal perceptions which may occur when one stares at a familiar object – or repeats a familiar word over and over – until it becomes strange’ (1971: 210).32 Thus. there are people performing discursive acts and there are material poles and charges. might be making the same point by saying either. ‘we have to change the way the buying and selling of things takes place’ or ‘we have to change the economy’ the way of proceeding demanded by ontological projects is that we stare at the terms33 and ask instead whether there is something real that they name and whether ‘the economy’ is reducible to ‘the way people buy and sell things’ because how can people ever say anything meaningful. for example. as discussed in Chapter 3. in this universe. etc. to use a relatively recent example. although shootings and killings are different kinds of things. however. that it is not only individual cases we need to change or..Ontological Confusion in Social Theory 93 latter or the latter might result from the former). the facts concerning the use of expressions. for instance. when someone is tempted to confuse one use with the other (as presumably are the critical realists according to Harré’s charge of reification) the appeal can be made to the way the two uses are related. i. literal while the other is metaphorical. It is one thing to point out to someone that they should not think of changing the structure as changing something additional to social practices and it is quite another to claim that ‘[t]o put it bluntly. as far as Harré is concerned. There is evidently room in our language and our social life for phrases such as ‘changing the economic system’. we can and do say that this shooting is a killing. Instead of paying attention to. 32 The point I am trying to make here is equivalent to the one White (1979) makes that. that is. ‘abolishing capitalism’. The reason these phrases can cause trouble is because it is easily forgotten (especially when theorising) that we need to pay attention to what is being said with them. In other words. that later developments proved that it was not enough that Egyptian President Hosni Mubarak was ousted by the Egyptian people) the assumption is all too easily made that unless an expression names some kind of thing which is material (or occupying another ontological status – what are the options really?). What cannot be concluded from this. ‘ontologically sound’ or. a manner of speaking. for example. in saying that one of them is.

‘object’). or (b) refers to a non-material thing and therefore the mind must be an immaterial entity. Hence. the proposition ‘he has pains’ could be false in some other way than by that man’s not having pains. Our problem must be to try to find out what they could possibly mean’ (2002b: 147) it is not clear that he sees this as an eradication of the preoccupation with entities. 35 Amie Thomasson (2009) calls them sortal and covering uses. . referents and ontological soundness. is potentially wrong not because what they report is inaccurate but because the form of words they use is ontologically unsound! This is as glaring a prejudice as prejudices go and one which Wittgenstein put his finger on in the following remark: [W]e’re tempted to say that our way of speaking does not describe the facts as they really are. based as it is on the indiscriminate employment of the word ‘entity’ as one of the recognisably ontological ways of speaking (‘object’ is used in the same way as is the ‘is it real?’ question). As if. not ontologically sound English for there is no such thing as ontologically unsound English. In other words. However. for example. (PI: §402) It is the pursuit of ontology which fuels these misconceptions. whereas the latter are simply place-holders for a sortal term. even when the proposition faute de mieux asserted something true. ‘thing’. Although Harré says in concluding his rejoinder to Carter and Strydom that ‘[t]here is a vocabulary of social structure words. To use one of Edward Sapir’s examples (cited in Cook 1978) to illustrate the reasoning involved even further. The former have what Thomasson calls application and coapplication conditions. this is just the prejudice that the fundamental thing about language is that it names entities. where those other things are still understood to be like material things in most ways (apart from the fact that they are somehow not material). it is as if someone who reported on a rock slide in a different language by using the grammatico-syntactical form proper to that language. to do ontology is to speak about what exists and what exists are ‘entities’. Thus. we must conclude that ‘mind’ (by virtue of being a noun) (a) must also refer to a thing which is material and thus most likely the brain. or (c) because ‘the mind’ does not refer to something material minds do not exist 34 The issue here is of speaking intelligible. In the latter case (but not the former) virtually anything can be an entity (cf. in using this simplified scheme we are liable to confound two uses of the word ‘entity’35 and to end up generating a criterion of what it means to exist or be real that is modelled on physical things. As if the form of expression were saying something false. the supposition is frequently made that there are material things and there are other kinds of things.94 Clarity and Confusion in Social Theory explanatory with those terms if they are not referring to something real?34 But. ‘it stones downwards’. these other ‘entities’ end up as mysterious ones! To give a prominent example from the philosophical disputes on mind. for example. a lot of confusion has been generated by the implicit reasoning that because ‘body’ may be used to refer to a material thing.

on which. Once we have exposed the idea that ‘all concepts have their referents’ or that all expressions must refer to entities as a misconception. To fill in the complex picture at work somewhat further. consequently. such a response would fail to take into account that we can and do use causal concepts in connection with social structure expressions. perhaps Harré’s critique of the critical realists could be summed up in the following way: ‘What you are doing is incompatible with my theory of causal powers which you frequently invoke. as we will shortly see. In other words. is equally misconceived as thinking that to change social structures is to change something out there hovering over our heads – a mistake of the kind Harré attributes to the critical realists. then. Based on the above. On rejecting this conception. . Instead. as something transcendental. however. This will hopefully throw some light on how the outlawing of certain forms of expression on grounds of causal inefficacy – a move which may be misleadingly presented in the form of not admitting entities into our ontology – 36 One might think of logic here in ontological terms. I will briefly sketch the conception of explanation evinced in the quotations provided at the end of the previous section. (c) capable of being used in explanations. Still. much like Hempel thought of it. it can be seen that Harré’s way of confronting the critical realists by asking whether social structures can be (a) causally efficacious (or not) and therefore (b) real (or mythical) and. apart from the arguments provided in Chapter 2. Harré’s conclusion to the effect that only persons can be causally efficacious cannot be seen as a profound discovery that forms the basis of a parsimonious ontology. is based on the mistaken presumption that social science should exclusively utilise such a concept. Divested of any ontological implications. hangs the question of whether it makes sense to want to change social structures. Moreover. scientific causal concepts cannot apply to what is not a powerful particular’. these options cease to be seen as the only ones available. is to be purged of any purportedly metaphysical significance. it seems to me.Ontological Confusion in Social Theory 95 and only brains do. The point is not to judge these uses by appealing to causal concepts as employed in the physical sciences. in order to establish whether they should be admitted into our ontology. does that mean that the former way of speaking is rhetorical or somehow loose because it is not covered by the concept of cause as it applies to the powers of particulars? Harré’s critique. in turn. that is. Does it not make sense to say that capitalism causes people who run businesses to either become competitive in the market or perish? Does capitalism cause people in the same way that a billiard ball causes another to move? And if that is not the case. it is best seen as a logical and not an ontological point36 (in anything more than a trivial sense having to do with the application of expressions such as ‘what persons are’) and. The possibility opens up then that ‘the mind’ does not refer to any entity but is a concept used to talk about the abilities and capacities of human beings. it could be acknowledged as a ‘grammatical remark’ and therefore be treated as a reminder about what it makes sense to say. see Conant (1991) and Putnam (1994: 247). but to get them right in the first place.

4. as we saw. regarded explanation as a matter of establishing nomic relations between events. Harré. In one sense this was a welcome advance in the philosophical understanding of what science is about. I began by detailing the incoherent demands that ontology places on the use of expressions and proceeded to show that the same demands are made by the critical realist programme in its employment of. Thus. first.Clarity and Confusion in Social Theory 96 is supported by the idea that locating causal efficacy is the only proper form of explanation. this conception remains just as rigid and stipulative as the one it sought to correct. whose behaviour. causal efficacy. . Social constructionists relocate the efficacy of the social fact in the changed pattern of discursive acts through which the everyday life of working people is constituted. in order to suggest that ontological reasoning brings nothing but confusion to social theoretical debate. The final step is to relocate the efficacy of acts in the people who perform them. Instead. the claim by Harré based on an analogy to the way that the beneficial effects of red wine on the human heart are explained by biochemists locating efficacy in molecules. there is ultimately one (legitimate) form of question which requires explanation which. As a result. as far as Harré is concerned. then. I have identified some further sources of confusion in distortive pictures of language and explanation. in turn. Consider.7 Conclusion In this chapter I have taken a first step towards discrediting the recent preoccupation with ontological matters in social thought. reality and explanation are made to fit together. is explained by physical chemists ‘relocating the efficacy in electrons’ (2002b: 144): We say that universal suffrage improved the condition of the working class. In another sense. however. emphasised the central role mechanisms play in accounting for the relations between events and identified causal mechanisms as the proper object of scientific understanding and as what holds a properly explanatory role. There is no efficacy in the structures themselves. (2002b: 144) The above exhibits rather nicely Harré’s position which was based from its inception (Harré 1970) on an attack on the Hempelian conception of scientific explanation which. Finally. the dichotomy between observables and non-observables in order to postulate transcendental powers or tendencies. the causal criterion in order to establish whether something is real and. is thought to be ‘which powerful particular caused X?’ I will come back to the question of the form of explanation in Chapter 6. I have then tried to exploit some Wittgensteinian elements which are invoked but not given due weight in Harré’s critique of CR. Both Harré and the critical realists have substituted equally monolithic conceptions of explanation for the one the logical positivists held before them. second. thus overlooking the polymorphic character of explanation as a human action. for reasons we have extensively developed in Chapter 2.

Ontological Confusion in Social Theory 97 One moral to be drawn from the discussion is that it is well worth thinking twice about whether casting the issues that divide opposing sociological schools in terms of ontology does anything to help make their differences clearer. the problem becomes clear. distinct from each other. If the exemplary ontological questions whether social structures are real or whether they exist are not questions about the existence of an entity but about in what sense something can be said to be real or to exist (Sharrock 1987) then the ontology (i.37 And if ontological talk is a prerequisite of grandiloquent disagreement then perhaps the disagreement is a product of the transposition. who are tangible. He begins by going down the familiar material/immaterial path. at some point. Stripped of the concern with ontology. Collectivities. which means that its being said to be ontologically unsound is even more obscure a claim. albeit in a slightly different way and by using another piece of reasoning which will help us appreciate the pervasive nature of the conflation. we stipulate that ‘social structures’ is somehow an ontologically unacceptable expression.e. To witness the transition.38 Here 37 Such a move would be doubly dubious for the additional reason that ‘social structure’ is used as a formal concept to collect many kinds of expressions. . it is more useful to think of the disagreement as methodological as Button and Sharrock suggest (2010: 33). and very material. they act.. other than in the actions of their members (Jenkins 2010: 147). it is perhaps worth asking whether critical realists or social constructionists really mean to regard as contentious whether it can make sense to want to change social structures and whether we can understand what one might be saying by this statement. I will try to make perspicuous one final time a fundamental distinction. unless. Doing so further points us in the direction of giving up the idea that one requires an ontology on which to ground one’s methodology and empirical investigations. he claims that: ‘Collectivities have a distinctive ontological status: they do not exist in the same way that individual humans can be said to exist’ (2010: 147). They possess agency. however. I will further discuss the question of ontological grounding in the next chapter. ‘What are collectivities?’ Once we take this question seriously. the one between ontological and logical (or conceptual) investigations. At this point I have only attempted to shift the burden of proof concerning such a conception by questioning its perceived inescapability. Perhaps. consider an excerpt from Richard Jenkins’s recent polemical piece on the sociological uses of the notion of structure. In our everyday lives we participate in a world populated by embodied individuals. In conclusion. with appropriately authentic naïveté. 38 What immediately precedes this quotation (beginning with a question I have argued we have no reason to ask in the way suggested) reads as follows: We need to ask. by contrast. the picture of ‘entities’ and ‘referents’) drops out. that is. are less visible or tangible – certainly visible and tangible in different ways – and they do not ‘act’. three-dimensional. where he attempts to provide a rationale for ascribing a different ontological status to collectivities as opposed to individuals. The reason is that it transposes the disagreement to a philosophical plateau where the preoccupation with ‘entities’ and ‘referents’ is dominant.


Clarity and Confusion in Social Theory

Jenkins is unknowingly giving the game away, for the question has now become a
matter of settling the sense in which collectivities can be said to exist, a question
not about whether there is some entity out there, but about the logical behaviour
of ‘collectivities’ (or collectivity expressions) and ‘exist’ which, given that ‘exist’,
like ‘real’, clamours for an opposing term (Austin 1962), can be adequately
addressed by specifying what their existence is being contrasted to. Equally, the
question ‘in what sense are social structures real?’ need not be a matter of ontology
any more than the question about the sense in which a bargain is a real bargain
need be a question about which ontological domain real bargains occupy.

Some discussion of collectivities or their members acting is provided in Chapter 6 and
in the Postscript, with reference to organisations.

Chapter 5

Continuing Ontological Confusion:
The Constitution and Application of an
Ontological Scheme
Imagine someone’s saying: ‘All tools serve to modify something. Thus a hammer
modifies the position of a nail, a saw the shape of a board, and so on’. ―And what
is modified by a rule, a glue pot, and nails?― ‘Our knowledge of a thing’s length,
the temperature of the glue, and the solidity of a box’. ―Would anything be gained
by this assimilation of expressions?
L. Wittgenstein (PI: §14)
For if a philosopher allows himself only an artificially limited set of logical
cupboards and pigeon-holes, then when he finds that a lot of his belongings will
not readily fit into them, he is tempted to solve his storage-problem by highhanded devices. Either he deems his articles to be of the required sizes and shapes
or, if intellectual scruples forbid this, he deems his cupboards and pigeon-holes to
contain just those secret shelves and interior compartments which will afford the
required fit.
G. (Ryle 1950: 147)

5.1 Introduction
These two quotations from Ludwig Wittgenstein and Gilbert Ryle describe
potential problems associated with some of the procedures for doing philosophy,
but also, and no less recognisably so, with the ways of doing social theory, for the
latter is equally concerned with the use of categories and classificatory concepts.
It may be pre-emptively objected that there is no reason why this should not
be the case, as is there no reason why associating these ways with problems does
anything to undermine them. Categorisation procedures leading to conceptual
systematisation, the objection continues, are standard, respectable scientific
procedures for organising materials in, at the very least, methodologically and
theoretically useful ways. It may be conceded thus without much consternation
that such procedures are not foolproof and that they may potentially misfire when
not carefully carried through. Any arising difficulties are, nevertheless, rather
sporadic, well-understood, and could in no way discredit the general validity and
potential fruitfulness of these procedures.
While these remarks may be commendable as cautious, careful reflection
shows that they are not, in fact, appropriate when intended as a description of what


Clarity and Confusion in Social Theory

philosophers and social theorists are up to and/or when thought of as an adequate
response to the concerns that Ryle and Wittgenstein might be raising. Why this
is the case is difficult to establish through argument abstracted from the detail
and rationale pertaining to the application of such procedures in particular cases,
no less so because those who employ such procedures wish to associate them
with science in the strongest sense, so that to attack them must imply attacking
science. Through this association, if not identification, such ways of proceeding
recommend themselves as the ones one must adopt, on pain of irrationality.
That such procedures do not, in fact, constitute bona fide scientific methods is
then difficult to establish for the additional reason that so much is invested in
them, indeed sometimes the legitimacy of a whole discipline hinges on their own
legitimacy; in short, they often appear as unassailable.
In this chapter I will continue the onslaught against ontological reason and the
conceptual procedures it is based on. To that end, I shall examine some materials
that might serve to bring Wittgenstein and Ryle’s remarks into sharp relief,
but, somewhat more importantly, the materials will hopefully be illuminated as
well. Specifically, I will be focusing on one of the latest and more sophisticated
versions of CR, one propounded by Dave Elder-Vass (2006, 2007a, 2007b, 2010)
who promises to place the social sciences on a sound ontological footing by
demonstrating that a general emergentist ontology, allegedly derived from the
natural world, is also applicable to society.
As I have already had reason to note, CR legislates that scientific explanation
should be in terms of causal mechanisms (Bhaskar 2008[1975]; cf. Demetriou
2007) where particular events are held to be explained once the mechanisms and
causal powers at play are identified. The rationale for this derives from the belief
(see Chapter 2) that science explains or should explain natural phenomena in this
way. In turn, this spawns the idea that a social ontology which correctly identifies
the components of the mechanisms and the powers producing social phenomena
can increase the explanatory power of the social sciences.
Under these auspices, what Elder-Vass is essentially aiming at is a foundation
for social scientific explanation in the methodical subsumption of regional concepts
(ones relevant to the study of society) under the causal mechanism components
sketched in his general emergentist ontology which comprises entities, relations
between their parts and emergent properties. As a first step to cataloguing the
kinds of causal powers which are operative in the social world, Elder-Vass focuses
on social structures (specifically on organisations and what he calls ‘norm circles’)
and identifies structures as entities with parts whose organisation gives rise to
emergent causal powers. Elder-Vass is thus attempting to defend a version of the
causal efficacy idea, whose coherence we saw Harré attack in the previous chapter.
In understanding and evaluating this latest version of CR we are presented
with two possibilities. One has to do with the project’s procedures. The other has
to do with its overall payoff, i.e., our being able to explain what we could not
explain before (as thoroughly or as comprehensively). In this chapter I wish to
focus on ‘the methodology’ of the ontological project, i.e., on how the general

To forestall potential misunderstandings. charged with the task of scrutinising their explanatory practice and its implicit ontology.2 The Ontological Scheme and its Alleged Grounding in Science I would like to begin by reiterating that for CR philosophy serves as an underlabourer for the sciences. in the first instance. lies elsewhere. not (the preservation of) scientific understanding. one cannot but be impressed with the systematic way he applies these methods. one which stresses the fallibility of the ontological scheme. continuously with Chapter 4. This position is also applied to the social sciences. I argue that it is incoherent to call the ontology fallible when it is in fact aprioristically derived and legislatively applied and. is to be found in the method of identifying causal powers and . I examine in detail the method of applying the general emergentist ontological scheme to discourse about society. moreover. that it is not possible to justify it on scientific or a posteriori grounds. by Bhaskar himself (1998[1979]). we might proceed to examine the materials. So much for the radical overtones of the argument for now. that the general and regional ontological project. I am trying to show in this book. it ‘must consist of enduring and transfactually active mechanisms’ (2008[1975]: 20) whose operation (in open systems) is responsible for actual events and our experiences thereof. and could not in fact be grounded in science: ontology is a philosophical pursuit whose result is generality.Continuing Ontological Confusion 101 emergentist ontology is constituted and applied to society. The ontology is discoverable through philosophical enquiry which. and later by Archer (1995. Before concluding. those discussed in the previous chapter). Bhaskar’s philosophy of science thus holds that the world must be stratified. His contribution. and another emphasising its pragmatic justification. it needs to be stressed that I do not intend to imply that Elder-Vass fails to meticulously follow the standards and methods he is propounding. focuses on ‘what the world must be like for science to be possible’ (Bhaskar 2008[1975]: 36). that is. and projects of a similar kind. and thus cannot serve as a foundation to social scientific explanation. I start by questioning the alleged grounding of the general emergentist ontology in scientific examples and argue that it is not. that of the real. starting with social structures. and at its deepest stratum. where I will also discuss additional methodological and conceptual procedures. While highlighting further problematic procedures that ontological enquiry brings into play (besides. while the latter route will be considered in more detail in Chapter 6. On the contrary. The problem. It is contended that the scheme distorts rather than illuminates what it is applied to. at least in part. namely in the ontological methods and procedures themselves. Dave Elder-Vass builds on that work by setting out to catalogue the causal powers which are operative in the social world. for example). 5. I am thus arguing. I attempt to answer two possible objections. thus claiming the right to revise its components. constitutionally rely on incoherent methods. as we have seen.

organised by particular relations between the parts. what one might call. Tables are entities. But there would be no point in visiting her to find that out. The elements of the general ontological scheme and how it feeds into social theory are concisely articulated in the following passage: [A]n emergentist ontology identifies a number of structural elements that we would expect to find in any object of scientific enquiry: entities. causal powers that belong exclusively to that whole. a thing. I said four. One can eat off a . The dinner table in Smith’s house is an object. ‘What about the screws?’. Entities have parts. And once we are equipped with these elements. i. by showing how the emergent properties or causal powers of the entities concerned interact to co-determine these events.. I will indulge in the juxtaposition of some thoughts in an uncommon key. The focus is on talking in terms of the scheme and on its relation to scientific enquiry: Smith has a wooden dinner table in her house. we need to identify the mechanisms by which the parts and relations lead to the properties. the morphogenetic causes that bring this set of parts into this set of relations in the first place and the morphostatic causes that keep them so. One can pay her a visit to find out.e. Not all tables have the same number of parts. it is an old piece of furniture. Tables are wooden things. In order to get us to reflect on the categorial structure and as a preliminary to the points I wish to raise. an entity. Well. The first day Smith moved into the house she did not have time to put the table together so she just ate dinner off the tabletop. All tables have the same kinds of parts. Elder-Vass proposes a general emergentist ontology involving a metaphysical position on causality couched in. and perhaps event regularities or partial regularities. mereological terms. Do they? Tables have parts. and possessing emergent properties in virtue of these relations. Did you say Smith’s had three legs?’ No. relations and emergent properties (which he uses interchangeably with causal powers). The core idea is that the interacting parts of a whole produce emergent properties. When Smith moved into this house the table had been stored up in the attic. It lay there disassembled into five parts: four legs and the tabletop. relations. ‘Jones’s table does not have any screws. Tables are objects. yes the screws were there too. What follows is meant to exercise our ear for sense and for the distinctions we make between different kinds of questions. and properties. In order to explain these entities. Although it is quite worn out one feels it is an entity dominating the kitchen space. we can go on to explain events. so I ask that the reader excuse its unconventional character.Clarity and Confusion in Social Theory 102 the mechanisms in which they are implicated. one might say. (Elder-Vass 2010: 68) The ontological scheme Elder-Vass is employing consists of entities and their parts. made up of parts (which are themselves entities).

one might say. Tables do not have parts. Smith’s table is quite an ordinary table. ‘The wind is an event?’ The blowing of the wind is an event. If one is up for a challenge perhaps they should try making it so that a table cannot have any parts. ‘If only we knew where to look’. §60) for they have served their purpose. A tropical cyclone has parts. ‘But it is made of wood and wood is made of molecules and molecules are made of atoms’. When the former kinds of statements are placed in the . the tabletop did not have any parts. ‘Did the tabletop come in one part? Mine comes in two parts. especially statements of the sort ‘entities have parts’ and ‘cloud formations are entities’. Are particles the parts of the wind? What about water vapour.Continuing Ontological Confusion 103 tabletop. the wind is an event. The wind is not an entity. Wait. Science is not interested in her table. I did not say it was not made of wood. is an operation of quite a different nature. these reflections have tried to bring into view a number of different kinds of statements we might be making in talking about entities and their parts. Tables produced by advanced manufacturing methods do not have any parts. Things do not have parts. Is its eye then an entity? A cloud formation is an entity. Meteorology seeks to understand the weather. as long as we find something! A weather system is a collection of entities. The weather is an object of scientific enquiry. at least not of the kind their members are. If only the weather were an entity with parts then science could understand it. and of the potential violence this may do to our ordinary (lay and scientific) ways of speaking. Entities cannot have parts. One can use a tabletop as a table without attaching four legs to it. Secondly. I just said it did not have any parts. Wait again. The wind is made of particles and those are entities. we can look anywhere we like. But I am tempted to say. Science is not interested in tables. but sets of entities are not entities. Engineering might be interested in tables though. A table might consist of a tabletop and four legs but people eat at the table. Tables are not an object of scientific enquiry. they hopefully have given us a sense of the directions we might be led into when speaking in terms of entities and their parts. held together by hinges’. a part of it is what we call its eye. it is in no way special. of course. How can something be a collection of entities but not be itself an entity? Entities make up wholes and these wholes are themselves entities. One should remember Elder-Vass’s words: ‘a number of structural elements that we would expect to find in any object of scientific enquiry’. Wittgenstein PI: §49. air pressure and wind shear? Are these entities? If not are they then either relations or emergent properties? We need not pursue these musings any further (cf. But then again the wind is not. No. they do not eat at a tabletop and four legs. The weather is not an entity and it does not have parts. Firstly. ‘What if one were to use a saw?’ One could certainly make it so that the tabletop had two or more parts if they wished to. This.

viz. not because the latter are somehow mysterious. in other words. Upon careful inspection. would the implication be that these are the categories that also have to work for the social world?1 But let us avoid moving too fast and consider instead the claims that the emergentist ontological scheme derives from and/or should reasonably apply to any object of scientific investigation. The point then is that emergence is a blindingly obvious truth when it comes to the organisation of matter. relation and emergent property are used by many or most sciences or. seems to provide for the application of the categories of the scheme to chosen concepts. part. But much like the social and natural world. Moreover. but because they are conceptually discontinuous in many important respects. relations and emergent properties? Moreover. natural phenomena and social phenomena are not two species of the same genus. We need to keep this in mind when we employ the language of ‘world’ and ‘phenomena’. H2O has different properties to its constituent elements. Is it the case. there is hardly anything controversial as far as the basic claims of emergence are concerned. Emergence seems to make perfect sense when we think of water or. the semblance of identity or at least consistency of the emergentist scheme with the concepts employed in scientific 1 To talk about the natural and social world or about natural and social phenomena is to go some way down the road of modelling the latter on the former. that is concepts that scientists use to refer to entities. for example. that is the constitution of the scheme and its application. The examples work by invoking a setting where the talk of properties and component elements has its home and thus rings true by virtue of scientific common sense. In order to reflect on these two concerns. The latter type of statement. At first glance. by virtue of its molecular structure. of H2O and its composition.104 Clarity and Confusion in Social Theory context of the general emergentist ontology they work as rules stipulating the interrelations between the concepts. it will pay off to examine in more detail the origin of the ontological scheme and especially Elder-Vass’s conviction that the structural elements comprising the scheme are to be found in any object of scientific investigation. however. that the concepts of entity. as they appear to be backed by scientific enquiry. to switch to a chemical register. we might go one step further and wonder: were one to grant that this is the case. the chemical properties of hydrogen and oxygen that make up the water molecule are not those of the latter. H2O has the properties that it has because of the way its components are organised. that we have a good indication that the concepts in use in a large number of scientific disciplines are of the kinds (and only of those kinds?) allowed for by the ontological scheme. The example of H2O and other wellunderstood examples are frequently used in illustrating emergence (for example in Elder-Vass 2010: 17). on the other hand. . in the same vein but perhaps somewhat more accurately. they are responsible for the constitution of the ontological scheme. They are both questionable.

properties may range from being of a colour. (Pleasants 1999: 111) 3 It is worth noting that. we understand very well that one of the paradigmatic uses of ‘property’ is when talking about boiling point at different levels of atmospheric pressure: boiling point is a property of H2O. they are entirely independent of the ‘transitive’ realm of explanation and meaning. then how are we able to understand that they emerge in the first place? If it is a contingent claim. does it mean just that we cannot (as yet. The difference between the philosophical use of ‘property’ and its use in chemistry can be approached through Thomasson’s (2009) distinction between sortal and covering uses we have previously invoked. so that almost anything that. But the temperature at which water turns from a liquid into a gas state is not a power of water – although the fact that it can do this can be said to be one.Continuing Ontological Confusion 105 enquiry quickly disappears. the difference between what properties wholes have and what properties they can have is obscured. to having the ability to do something.2 Consider. in philosophy. the emergentist identity of properties and powers: [W]hat is meant by a property or power[?] A property is some intrinsic aspect of an entity that can have a causal impact on the world.3 Further problems arise in the criterial use of the identity asserted in this passage when we attempt to settle the question whether something is a property 2 Consider as a preliminary comparison what Nigel Pleasants has to say about the work that emergence does for Bhaskar: Bhaskar’s ‘ontologisation’ of the concept ‘emergence’ is riven with perplexities that are just as puzzling as those produced by the ontological pictures that he rejects. For in talking about the emergent properties or causal powers of wholes of so many different kinds. and will be discussed in the following chapter. constituent elements and powers. Moreover. consciousness and agency? Bhaskar’s answer … is that emergence and irreducibility are ontologically real features of the intransitive domain. at least) explain precisely how such emergent phenomena emerge? (an answer that is perfectly acceptable to the Humean compatibilist). from the definition of properties. such as ‘larger than x’. What does it mean to say that certain phenomena ‘cannot be reduced to’ the phenomena from which they emerge? How is this ‘cannot’ to be understood? Is this a necessary or a contingent truth? If it is a necessary truth that the phenomena are ‘irreducible’. For example. Properties and powers may therefore be regarded as synonyms. Or is it just a ‘linguistic reminder’ that talk about brains etc. I use intrinsic in order to exclude purely formal relations with other entities. does not have the same meaning as talk about minds. for instance. . This distinction is fundamental for understanding Elder-Vass’s idea of exclusive causal powers of social structures. the distinction between external and internal properties or properties something possesses by virtue of its ‘intrinsic nature’ does not make matters any easier. (2010: 17) This move removes the context of intelligibility that the setting of chemistry provided to the talk of properties. can be predicated of a subject philosophers can call a property. syntactically speaking. or size to being disposed to react to something. shape.

It is. Thus. produced by conceptual incoherence. that is. At least that is the discernible intention at the starting point of this exercise: that all concepts should reasonably fit in this manner. For. although the scheme might be said to derive (in the sense that it may pick some of its terms) from one or more scientific disciplines. Wittgenstein’s words on the ‘universal function of tools’. one. do not (and could not) do is settle the range of application of the scheme (the use of examples is treated more fully in the following chapter). I will say more in the next section. . In Sören Stenlund’s polemical words ‘the function of the simple and “intuitively appealing” examples given to illustrate a certain scheme is not constitutive evidence and support for the correctness and the general validity of the scheme. the latter would not be able to do the work that they are recruited to perform. As doing so. Still. are very pertinent here: in the same way that ‘modifying’ is stretched in order to apply to anything that can be called a tool. whether. For what these examples. on why the criterial use of stipulated conceptual connections is distortive and/or inadequate. The initial feeling of self-evidence of emergentist claims quickly gives way to uncertainty regarding their scope which is in fact left unspecified but apparently presumed to be unrestricted. its purpose is rather different. we are led to ask ‘Can water’s boiling point have a causal impact on the world?’ and it seems that the answer must be in the negative. even if there were no incoherence between the definition of the emergentist ontological scheme and the scientific examples which serve to illustrate it. that we tend towards understanding ‘boiling point’ as not being a causal power is really symptomatic of the fact that we are at a loss when it comes to answering a question we do not know what to make of. we are to completely relinquish or retain our understanding of properties in the way that chemistry or physics understands them but then eliminate those that cannot have a causal impact on the world. of course.4 When we apply Elder-Vass’s criterion of causal impact it appears that boiling point might not be a property of water. however. which is what philosophers want us to believe’ (1996: 206). What is gained by this ‘assimilation 4 It is unclear whether the stipulation is meant as a restriction or an expansion. that is. the concepts of entity. However. To see that the scheme’s use is not scientific but rather philosophical through and through consider what it does for us: its abstract categories do not enable us to understand and explain a particular range of phenomena.106 Clarity and Confusion in Social Theory by relying on the stipulated conceptual relations. embracing the definition. relation and property – which are already quite broad as things stand – are stretched even further to potentially encompass most scientific concepts. this cannot warrant its ontological use which is of a rather different character. with which I opened this chapter. extends beyond the offered criteria it cannot serve to recommend them. We are faced with an unintelligible or yet unspecified question. which can serve as constitutive of what Stenlund calls the grammatical scheme. This crucially presupposes that we have some independent way of telling what can have a causal impact. for a general ontology is by definition meant to apply without restriction. possible to give a sense to the question.

and the conceptual procedures its attempted application relies on will be seen to be even more so. where a property of the whole is strictly of that whole and such that it arises from the way the parts are organised. For. partially extended by the ontology itself. Concerning the former. whether any significant payoff is to be expected from doing so.3 Employing the Grammatical Scheme Having seen that the ontological scheme’s relation to science is more tenuous than might be thought. we have good reasons to question whether it is a good idea to apply the scheme to concepts used to talk about society and. 5. in turn. It has already been argued in Chapter 4 that embarking upon the quest for a (social) ontology brings into play a number of generic philosophical procedures that lead to confusion. but then the crux of emergentist claims is the question of emergent properties. lack of content together with the failure to preserve intelligibility. that is. To say that X would not have property Y if its parts were not organised in that way is to say that X would not have the property Y if it were not X. and does it not also follow truistically that these parts stand in some relation to each other? Perhaps it does. appears to be empty but its emptiness. be retorted that any property which is strictly of the whole will be a property that the whole has by virtue of what it is and wholes are what they are by virtue of the way their parts are put together. We may now add to those procedures that of operating on a simplified grammatical scheme and stipulatively employing its categories to reorganise our concepts. given that ‘an entity may be defined as a persistent whole formed from a set of parts that is structured by the relations between these parts’ (Elder-Vass 2010: 17) does it not follow by definition that if something is an entity then it can have parts in some sense which are themselves other entities. In fact. it is not clear that more is needed for the scheme to apply than this first step. given the extensive range of application and the many uses of the expressions ‘entity’ and ‘part’.Continuing Ontological Confusion 107 of expressions’ is being able to perceive the world in the greatest generality (as Wittgenstein answers his own question a little further down in PI: §104). that is. given the definitions that bind the elements of the ontological scheme together. even further. But this generality comes at a multiple cost. thus. where again in this respect emergentists have not managed to escape tautology. it might be considered at this initial stage whether upon application of the general scheme to a specific domain it would be a great discovery if one were to find some of the concepts employed in a scientific field to be concepts that can be used to refer to entities which have parts. one may reply. In Elder-Vass’s ontological method we can witness the use of a grammatical scheme providing us with a number of logical boxes in which to insert our concepts – Elder-Vass himself speaks of ‘pop[ping] the concepts into the relevant ontological . To this it can. The scheme. does not rule out the potential violence it can do upon its application to concepts of different kinds. The presumption that the scheme must apply appears questionable. as we shall now find out. that some concept might be described as an ‘entity-concept’.

we can see Elder-Vass trying to make the case that the social sciences are somehow missing what the natural sciences have. consider money.108 Clarity and Confusion in Social Theory boxes5 (2010: 70). non-human physical objects. ‘logical’ invokes the use of words. which is telling in more than one ways.e. In the following excerpt. . the point of the example is only to illustrate the need for real work to answer such questions. namely a scientific ontology. but the dichotomy between language and reality is itself part of the misconception and not a remedy for it. well-defined and locally consistent scientific ontologies. and frequently in the social sciences. For an example of the latter. But coins can be money. But it would be misleading to flatly assert that what is problematic is that Elder-Vass projects a scheme taken from language onto reality. But we 5 Characterising these boxes as ontological rather than logical invokes the picture of language that was scrutinised in the previous chapter. organised in a characteristic set of relations. the task of determining what type of structural element any given concept might represent is often far from trivial. ElderVass puts the procedure to use with regard to the concept of money: Yet the social sciences are at least one step further removed from the status of ‘well-developed science’ – they consistently lack plausible. For the social sciences. As sketched above. thus stressing language. we might resolve this by concluding that money is not a property of material things like coins but rather a property of the social institution itself. It may be further noted that one perhaps obvious difference is that ‘ontological’ is somehow meant to convey the intended grounding in reality. that the suggested elements somehow are the elements of reality. it would have to have a characteristic type of parts. a type of entity. relation and causal power/emergent property.. cheques can be money and electronic transfers generated by swiping credit cards can also be money. In a sense this is true. social structures and cultural or conceptual systems in the social world make it extremely challenging to disentangle the entities and properties involved… (Elder-Vass 2010: 70) (my emphasis in bold) It is instructive to follow the reasoning exhibited in the above passage. For money to be a type of entity. First. the scheme that is meant to generate the social ontology features entity/part of an entity. One response … is that being money is a property possessed by a variety of different types of entity… Alternatively. The complex inter-relations between human beings. in our terms. i. then. I do not mean to suggest that one or the other of these paths actually does offer a viable solution to the challenging problem of the ontological status of money. it becomes clear that this cannot be so. But as soon as we start to examine the ontology of money. as if money were a thing. One of the pitfalls of the social sciences is that we may assume that they do have such ontologies and accept unthinkingly the sorts of ontological categorisations that appear implicit in social theories. The word is constantly used in everyday life. On the other hand. or even in our everyday language about the social world.

without any ontological foundation (and a fortiori without an externally provided one) then this throws into doubt the need for and contribution of a foundational scheme. These are not facts that can be cited as correctives to the ordinary conception of money. a number. it is obvious that most members of society are quite familiar with money and with the forms it can take. a possibility. which has come to occupy a prominent position in analytic philosophy. the aims of philosophy and science are categorially different. the pain in my leg. physics. whereas philosophy is concerned to establish that material objects or events exist (pace Davidson). a plot to depose the king). what it means to say that such-and-such exists (e. 6 If Elder-Vass’s claim is that his general ontological categories derive from science then this is as evident an expression of a commitment to naturalism as one might get. for there is no such thing as a theory of everything that exists. as Peter Hacker has convincingly argued (2010). from one domain to another. renders money as something mysterious. Here it is important to note that the puzzlement regarding money does not exist independently of the procedure just highlighted. Puzzlement is generated due to a perceived lack of fit which. a substance. Returning to the above passage and to the application of the general ontological scheme to the social world. metaphysics onto science. if one is interested in preserving the label.g. a law or legal system) (Hacker 2006: 232–3) (my emphasis). nor is any such scheme presupposed by science. we can witness how Elder-Vass.Continuing Ontological Confusion 109 are now in a position to appreciate that ‘scientific ontology’ is a rather dubious construction. and quite successfully so. a complementary conception to the naturalised epistemology Quine pioneered. if the natural sciences are capable of operating. That is. opening bank accounts and managing them online. For one. projecting.g. making transfer payments and using a credit card when buying commodities. a property. . e. Mannerist style. conjures up the ontological problem of money by attempting to fit the concept of money into one of the slots provided by his ontological scheme. It is not as if physics is concerned to establish that mesons or quarks exist. in effect. no such ontological scheme could inform science in any fundamental way. Hacker’s own conception of what ‘philosophical ontology’. might plausibly do is also instructive: In no ordinary sense of ‘science’ is science the sole and final arbiter on what exists (e. There is no specific science that offers us the best theory of what exists. The task of ontology is to clarify. the Romantic movement. No such understanding is involved in members’ competence in handling coins and notes. In light of this it is worth considering Hacker’s cogent remarks on the conception of a naturalised ontology. and the mere fact that they use a noun to talk about it. writing cheques. does not imply anything of the sort. in turn. far from committing them to conceiving of it as a thing. international law. despite Elder-Vass’s attempt to saddle ordinary understanding with alleged confusion. Philosophical ontology is not concerned with determining what exists in the sense in which biological taxonomy is concerned with determining.6 For another. as it does. if. the meaning of a word. tabulating and classifying what living things exist. Russell’s childhood diaries. Nor is it differentiated from a science. by generality of categories. nor do the sciences collectively do so. a concept.g.

not being able to perform categorial accounting need not imply that money is ontologically intractable.7 If it is not even that then it is bound to be something mysterious. There is certainly no reason to think that there must be. Nor. i. Is good a relation? No.110 Clarity and Confusion in Social Theory as Elder-Vass contends. However. (Hacker 2004: 21) If Hacker is correct in his assessment. Our concepts evolved to meet the needs we have.. by subsuming the concept under a general category. category–hungry. on the contrary. i. rather it derives from ElderVass’s scheme. After all. non-human physical objects. social structures’ and so on as some purportedly ontological feature of the social world. . so it must be a relation. In his 1950 piece titled ‘Logic and Professor Anderson’. one might add. but are. Consider Hacker’s concluding remarks in his detailed investigation of the concept of belief: Having a belief is not a feeling or a mental state or a disposition. nor need it imply that we must remain in the dark unless some category can be found or constructed. ironically. in the first instance. Thus. our lay and scientific concepts already have a logic of their own which is much more nuanced than any such simple scheme will allow. ‘is X a quality or is it a relation?’ and ‘if X is not a quality then it must be a relation’. they constitute the ordinary concept and conception of money. the presumption that all concepts can be accounted for in this way is unwarranted. in this case. Furthermore. which I believe he is. and hence his reasoning takes a form similar to Elder-Vass’s. Ryle offers an incisive depiction of the reasoning entailed by the employment of the ontological scheme: there Professor Anderson is employing qualities and relations as his logical boxes. which is predisposed to read its own requirements into any concept it attempts to subsume. But. His regular touchstone is the question “Quality or Relation?” Is knowing (or willing) a quality? No. it is not categorial subsumption itself that can provide us with understanding. from which I have quoted in the beginning of this chapter. a product of trying to account for nuanced concepts with a crude set of categories. it is looking at all the ways in which money enters 7 Ryle argues: ‘Anderson seems to be oblivious of any logical differences save the difference between qualities and relations. The described use of the ontological scheme can be further illuminated by considering Ryle’s insight into the procedures it depends on. the defective understanding of money as an entity with parts does not derive from ordinary practice. to repeat. It is neither a behavioural tendency or proneness nor a liability to behave. Moreover.e. What then is it? It is not at all obvious that there is any categorial term under which belief can be illuminatingly subsumed. not to satisfy the classificatory demands of a concept-classifying. Linnaeus.e. so it must be a quality’ (2009[1950]: 249). money is only valid if understanding the nature of money necessarily involves being able to account for it categorially. are the difficulties in the task of popping concepts into boxes due to the ‘complex inter-relations between human beings. the idea that a potential lack of fit shows that we are missing something about the ontology of.

etc. it might be said. throws into doubt the idea that an ontological scheme can allegedly act as a foundation to or improvement upon our conceptual understanding. For. for example regarding its history. Elder-Vass is willing to allow that there are cases where the method will partly or wholly fail us. In such cases the recommendation is that we either revise the mapping of the concept onto the ontology or if that is not possible. and that this remains to be discovered – unless. puts in place. Indeed. Reflection on Elder-Vass’s accounting of concepts based on his ontological scheme has shown that his project is a species of the unrealistically restrictive problem we encountered in Chapter 2. however. of course. does not preclude the possibility of there being other kinds of concerns about money. In unmistakeable Hempelian manner. as for example in the case of ‘mental entities’. None of these. a multi-step iterative procedure9 to ensure that there is in fact a fit between the concepts and the structural elements comprising the ontological scheme.4 Objection 1: What About Fallibility? It might reasonably be objected at this point that I have been doing Elder-Vass an injustice.8 that somehow people have been using money for millennia while ignorant of what it is. understanding the ontological status of money is restrictively defined as discovering a suitable general category under which to include it. for example. its function in an overall economy. the iterative procedure uses the criterion of whether the remaining logical boxes can be filled as an indication of how well a concept fits in one of the boxes.Continuing Ontological Confusion 111 our practices. The repeatedly evidenced incoherence of this procedure. be argued that it is more sophisticated than I have led us to believe by mentioning Professor Anderson. ElderVass constructs his method in such a way so that our ordinary conceptual resources are rendered irrelevant (but he selectively relies on them when the details need to be filled in) and ends up distorting our understanding of the concepts with the use of the scheme’s categories. then: we may have to abandon the suggestion that all entities can ultimately be connected back to the natural world in this way. that is. where ‘the ontological structures that seem most consistent with the actual social world simply do not fit the general ontology outlined’ (2010: 74). Indeed. what he calls. then. at all the things we do with money and say using the concept of money. therefore. how strange when we come to think of it that there should be a general problem with the ontological status of money. . the regulation of its flow as a handle on the economy. his method for social ontology does not consist only of the parts I have sketched so far and it may. Elder-Vass. 9 Crudely put. (2010: 75) 8 This. and allow for the possibility of conceptual systems that have a different kind of ontological structure. besides emphasising that the ontology is a fallible one and thereby reserving the right to revise the set of categories. are of the type raised by Elder-Vass. 5.

crucially. under which some and only some of our concepts can loosely fit. Upon careful consideration. and indeed why presuppose that a set of categories is adequate to do justice to our concepts. Consider also Lawson’s description of retroduction: To pursue causal explanation as interpreted here. why not 100. Rather than constructing the scheme on the basis of an examination of our concepts and their commonalities and differences. even further. the ontology is a requirement of the investigation and not its result. end up producing more of it. it also becomes directly accountable to those structures and to our understanding of them. This is retroduction. Yet again. these categories are to become fewer (ElderVass 2006: 17) rather than proliferate dramatically to accommodate conceptual complexity (in which case of course one might ask why we must have categories: we might as well stick with the existing concepts!). the important point is that.112 Clarity and Confusion in Social Theory Although when considering the above passage Elder-Vass’s procedure strikes us as rather more sophisticated. i. the identification of mechanisms.10 here also the relevant facts about the kind of concepts involved. Furthermore. In this case the scheme consists of the categories entity/part of an entity. is not something that can be improved on by calling the scheme fallible. It takes us from a recognition that ‘this metal before us expands when heated’ to a conception of the metal’s intrinsic structure (or whatever) in virtue of which the metal has the power to expand when heated (Lawson 2003: 80). relation and causal power/emergent property. contrary to fallibility requirements.e. The fact that a scheme with very broad categories is deployed. any use of the scheme in a legislative way creates significant tension with conceiving of the procedure as informing and being dependent on retroduction. . For if it is possible to have an understanding of the ‘ontological structures’ independently of the scheme then the scheme becomes somewhat redundant. starting from the facts. one finds that there are no mysteries why it is possible for the 10 Elder-Vass (2010: 72–3) holds that the application of the ontological scheme and retroduction depend on each other in the sense that the ontological scheme describes the components of the identified mechanism and in that the identification of an entity with causal powers hinges on the identification of the mechanism that makes it so. which requires. why start with those three categories? And why three. More importantly. like Professor Anderson. or more generally from phenomena lying at one level to causes often lying at a different deeper one. That the concepts making up the scheme could be prioritised and used to reorganise some of our other concepts is not independent evidence that supports the scheme but. and that. a non-negotiable feature of the scheme’s application. The coherence of the notion of a fallible ontology is undermined by the use of a fallible ontology. why not 10. This is why Elder-Vass’s pronouncements that the ontology is fallible. while being an attempt to save the procedure from incoherence. he too uses the scheme legislatively and aprioristically. we require a mode of inference that takes us behind the surface phenomenon to its causes. But one might wonder. the above passage casts doubt on the contribution of the procedure to our ‘ontological’ understanding.

even at a first level. the application of the scheme to regions of discourse will need to demonstrate its effectiveness in terms of enhancing our understanding of the described concepts. for example. it is not as if philosophy were science only at a more general level. in some sense. is incoherent as it conflates philosophy and science by portraying ontology as a matter of scientific choice and as capable of being justified by the science. in some sense. then it would have to be possible for it to arise from specific domains where it has proven to be fruitful. to derive the scheme in an a priori way and then claim that its justification lies in its application is to refuse to follow one’s own admissions and/or to delegate the work to others. This also bears the implication that the ontology of. it might be contended. But this will not do. But if one resorts to that criterion. Such a move shifts the weight of justification to the indeterminate future and away from the procedure that has. then what we have seen so far. its emptiness and conceptual crudity.Continuing Ontological Confusion 113 scheme to guarantee its own applicability. Moreover. entities with. parts is not in any sense a natural world ontology. if the validity of the ontological scheme really hinged on its ‘scientific’ or empirical application. This idea. does not actually use the fact that these categories can be treated as master categories as independent justification for the scheme. that something is an entity or a part is to say very little. Like Harré. To echo Hacker (see fn 6). To say that something is an entity or a part in the most abstract way is to make an empty statement. Rather his position lies close to Harré’s who claims that: sciences are created by choosing an ontology. for it is only the understanding of the kind of entity or part it might be that will enable us to embed this piece of knowledge into any argument or explanation. 5. Elder-Vass holds that ‘any social ontology must ultimately be validated by empirical application’ (2010: 205). pace naturalism. i.. To say. namely. Elder-Vass.e. that the scheme can improve our comprehension. however. been used to come up with the ontology which itself must undergo scrutiny (cf. through the use of which phenomena are to be identified and ordered and explanations are to be constructed … in the end the choice of ontology is largely justified pragmatically – how much of the phenomena of interest does it enable us to comprehend in a fruitful and constructive way? (1997: 178). Thus. it is practically to say nothing. If the proof of the pudding is in the eating then. provide evidence to the contrary. in fact.5 Objection 2: Ontologies are Justified Pragmatically Yet another objection might surface at this point. or as if philosophical positions (the various ‘-isms’) or . then. Kemp 2005).

are completely and utterly vacuous. 5. For an argument directed at critical realism see Cruickshank (2010). as a result. Philosophy is a priori. first. right here and now. football is played from a football player’s point of view. Ultimately.e. involved point of view. and have chosen instead to examine the conceptual procedures relied on. positioned. no replacement or foundational work. is necessary for ‘ontological’ understanding. the explanation of social affairs once what falls under the elements of the social ontology has been catalogued.’ Each of us can only have our own actor’s point of view on anything we do – a living. there is no such thing as the scientific justification of the ontological categories of entity. Such categories could then be justified if shown to correctly describe (parts of) the structure of concepts in question but they can do. This is consistent with the pragmatist insight that science should be seen as just one more tool for us to cope with our environment. interested. which is conducted on the presumption that.114 Clarity and Confusion in Social Theory ‘ontologies’ were necessarily presupposed by science. which were found to be flawed on several counts. (2004: 239). however. For us. this does not make much sense: it seems nonsensical to claim that someone kicking a football is assuming the ball to be ‘existentially intransitive. Nor can it serve as an underlabourer for science. . what I have called categorial accounting. second.11 Philosophy’s aims are categorially different.6 Conclusion In this chapter I have continued the dissection of ontological reason by focusing on the constitution of Dave Elder-Vass’s ontological scheme as well as its application to the ‘social world’. I have said little about the projected sociological payoff of the exercise. they must apply and. and social scientific inquiries are carried through from a social scientist’s point of view.. These are best thought of as formal conceptual categories. I began by dissociating the ontological scheme from science and the scientific examples invoked to illustrate it by showing that it comprises a set of crude inter-defined categories that neither do nor are capable of capturing scientific understanding of anything. i. and as such cannot invoke scientific investigation in order to prove its points. 11 Kivinen and Piiroinen are also good at making this point in relation to the tendency to see commitment to philosophical positions everywhere: Realists seem to assume that any activity could be meaningfully described as someone taking a realist ontological stance. i. relation and emergent property.e.. not as a picture of reality. I then proceeded to examine the application of the categories. that subsumption under such categories. in the midst of our activities – and so metaphysics is practiced from a metaphysician’s point of view. The categories are aimed at greatest possible generality and. categories that can correctly or incorrectly describe types of concepts and can legitimately be invoked in the course of elucidating the conceptual apparatus employed in science or any other human activity.

for example. however. the ontological procedures we have examined neither do nor can provide any such grounding. even further. . are operations that restrict our movement by excluding. as the case of applying the scheme to the concept of money showed. whether there is any point in lamenting the absence of any ontological grounding for the social sciences (or for any form of investigation for that matter) when. it must be externally introduced (in this case by using some scheme allegedly derived from science). in fact. but that they also serve to distort whatever concept is considered in terms of the scheme and to create rather than resolve ‘ontological’ puzzlement about it. provide any understanding. it has also been argued that the philosophical procedures that are used to produce and apply the scheme make nonsense of the idea that it either is or could be pragmatically justified. Nothing I have said is meant to rule out such a possibility. he (and not only he. that Elder-Vass is in large part driven by the systematising formalism. while implicitly relying on. With regard to a second objection. In doing so. Returning to the words of caution at the beginning of the chapter. in actual fact. that is. of applying the grammatical scheme of emergence as a way of conceptualising and accounting for the social world. and. of course) is merely enforcing unrealistic restrictions (à la Hempel) in employing procedures that. especially if systematisation is understood not as seeking to replace or ground concepts that are constitutive of social life but rather to elucidate them in a methodical way. but also rejecting the possibility of any conceptual systematisation. I do reject the idea of systematisation if it is based on an underlying assumption to social theory which supports and is supported by the theoretical attitude towards language. calling the scheme fallible only produces a caricature of fallibility. guarantees its own applicability by its vacuity. Instead of assenting to the seeming self-evidence of ontological projects. I am aware of the fact that for some theorists the argument forwarded in this chapter may involve potential overkill. namely that there is no conceptual order in the social world and that. not to question why we need to subject ourselves to mystification. One can see. therefore. the resources available to us as competent concept users. On the other hand. it may be asked whether I am not only condemning ontological schemes. crucially. we have no reason.Continuing Ontological Confusion 115 I have sought to demonstrate that these presumptions not only render irrelevant our conceptual resources that can. then. far from being capable of enhancing our understanding. In response to a possible objection it has been argued that since the scheme starts from the categories and not from the concepts to be subsumed and. by the idea.

This page has been left blank intentionally .

rely on a host of logically flawed procedures. which I will be discussing in considerable detail. as a result.Chapter 6 Some Problems with ‘The Problem of Structure and Agency’ 6. for example. recognisably one of the central problems in social theory. I will argue that the issues are posed and addressed in ways which. Since success has been questionable. once again. Even though the problem has been made famous under this appellation during the past 40 or so years of social theorising. theorists have proposed solutions that offer some kind of reconciliation of the terms (for example. one finds that currently employed formulations of the issues involved go back to the time of Marx and. then. debate will remain irresolvable and the perceived problem perpetuated. . These are the vices of social theory. is unavoidable if we are to appreciate just how little care is taken in describing cases and examples. just how much confusion there is. especially if speaking of ‘a problem’ invites us to reasonably expect ‘a solution’. Once again. I will be focusing on materials from (critical) realist theorising. it is a continuing feature of professional sociological reasoning that sociologists frequently distinguish between (and criticise) theories based on whether ‘structure’ or ‘agency’ is emphasised at the expense of the other term.1 Introduction In this chapter I will provide a culmination of the argument of this book by turning to examine ‘the problem of structure and agency’. Giddens 1984). some of which we have already encountered in previous chapters. they will take us far beyond ‘the problem of structure and agency’. his famous saying about men and history (1977[1852]: 500). that if no acceptable reconciliation of ‘structure’ and ‘agency’ has been reached it is not for lack of trying. acknowledging defeat. It seems. The fact that the problem can be seen to have persisted for as long as it has presents us with a reason to suspect that not all is well. I submit. The problem has often been treated in such a way: for one. understanding it as a singular problem about ‘structure’ and ‘agency’. I shall judge it necessary to offer some observations on the nature of the problem and to characterise the host of issues it serves to collect as well as the kind of debate it encourages. But I will also outline some of the virtuous ways of proceeding I think social thought should strive for. and that as long as such procedures are presupposed. From this one may conclude that the problem is indeed so demanding that we will either have to try harder or simply give up. just how much our understanding is restricted. Detail. just how little attention is paid to concepts and. Rather than attack the problem straight on.

The a priori adherence to an ideal of explanatory form and to rigid conceptual models. not only stands in the way of adequately dealing with chosen theoretical issues.1 What needs to be said straight away about ‘the problem’ is that it is no single problem at all. Porpora 2007.118 Clarity and Confusion in Social Theory However. not all of which surface in the versions that individual theorists seek to address. and the systematic lack of care in distinguishing between the empirical and the conceptual as well as between different types of concepts. 2007c. The treatment of conceptual issues as subject to empirical verification/ falsification. my aim in this chapter is to connect them to ‘the problem of structure and agency’ and to build further understanding of just how they lead to confusion. Moreover. understandings thereof and proposed solutions but. 2007b. it will be argued. I will propose that if we are to adequately deal with ‘the problem of structure and agency’ we need to realise that it involves a number of unrealistic restrictions and. a number of procedures that transcend any one version and. if we are to make any headway what is called for is a shift in both the understanding of the problem and. rather than take it that the problem is a well-posed albeit demanding one. which I do not discuss. Although there are other possible versions stressing different aspects and concerns. b.3 onwards) I will examine these procedures in the order stated above. when accurately described. Not dispensing with them. the particular angle I am adopting in the following discussion is meant to challenge not only version-specific collections of issues. discussion of ‘the problem of structure and agency’ draws on a pool of many different issues. are uncontroversial and cannot (either exclusively or at all) support the scheme in question. most importantly. In this chapter I will be arguing my case through the detailed examination of a recently proposed solution (Elder-Vass 2007a. I hope that the version I have chosen captures a sufficient breadth of issues to tell us something about ‘the problem of structure and agency’. In the ensuing sections (6. c. as should the idea of unrealistic restrictions. The illustration of a proposed theoretical scheme through examples which are either misdescribed when described in terms of the scheme or. The procedures I will be discussing are the following: a. in this case of causal concepts. but also tends to render debates between theoretical schemes irresolvable for theorists can always (i) insist 1 The idea that a way of understanding a problem should go together with certain procedures should hopefully be familiar from the discussion in Chapter 2. as has been repeatedly suggested in this book. Varela 2007). 2010) and of ensuing debate (King 2007. Rather. . especially. form a taken-for-granted part of social theoretical reason. Some of these procedures will be readily recognisable to the reader. the procedures employed in order to address it. I would like to pursue the idea that there is something wrong with the problem itself and with the kind of debate it encourages. therefore.

apart from the existence of individual causal powers.e. theorists need to discharge the burden of proof concerning (iii) and (iv). Thus. is what Archer attempts to accomplish via her account of the mediating role of ‘internal conversation’ and. together with the critical realist commitment to a form of explanation in terms of underlying mechanisms. such debates are bound to remain irresolvable as long as understanding of what it takes to demonstrate a theoretical scheme as superior remains poor. I remain quite sceptical regarding the premium placed on theoretical schemes. In the context of structure and agency this amounts to advocating. and (iv) the application of the theoretical scheme and model of explanation does not make a demonstrable difference in comparison to commonly available understanding of what it is applied to.2 and that it also illuminates the perhaps smaller scale mechanism which gives rise to such powers. as long as (iii) the attachment to a theoretical scheme does not translate into a contentious factual claim that is necessarily connected to one of the competing schemes. and (ii) select different areas of the causal conceptual terrain as models of what counts as causal.Some Problems with ‘The Problem of Structure and Agency’ 119 on the a priori propriety of a model of explanation. Sharrock and Tsilipakos 2013). Even though it is unwise to rule out the possibility that a scheme will meet such requirements and thus prove its worth. In this vein. 6. Furthermore. for example..2 Elder-Vass’s Attempted Solution to Structure and Agency Within the critical realist paradigm. It will hopefully have become evident by the end of this chapter that it is doubtful whether any such scheme can serve as an adequate response to the host of concerns involved in ‘the problem of structure and agency’. more generally. by showing the process to be governed by his general emergentist ontology. The process is captured in the core dictum that the relations between the parts of an entity give rise to emergent causal powers which the individual parts may not possess. i. itself part of an open system. Dave Elder-Vass’s version (2010) of CR is designed to account for the way individuals with causal powers of their own interact to produce structural causal powers. namely that both social structures and individuals possess causal powers but of different kinds. one irreducible to the other (Archer 2003: 15). the 2 This. requires that social theory demonstrate that structural and individual powers are connected as parts of a causal mechanism. for if the same facts can be recognised by anybody they cannot be invoked to make a difference. In other words. I believe that once we understand the nature of the issues comprising the problem we will see that it is not a theoretical scheme that is needed but rather conceptual clarity. This thesis. reflexivity (cf. in order to demonstrate the superiority of their proposed theoretical scheme. the problem of structure and agency is treated in the language of causal powers (deriving from Harré and Madden 1975) as is evident in the claim we have seen Archer put forth. .

the fact remains that. despite the appearance of schematic elegance. the ascription of causality or the possession of causal powers. to argue that settling (a) can provide an answer to (b). that the causal criterion is a good way to establish whether some kind of thing is real (see section 4. the fact that they are posed together. Involved are. is presumably indicative of the implications one is thought to have for the other. An integral part of Elder-Vass’s argument has proved quite controversial. third. I will have reason to reiterate some of these points in the course of the following sections. most of the participants in the debate (with the exception of King) presuppose. all of the above issues have histories of their own. suggesting this configuration of issues means that one is committed to positions in yet further controversial topics. for example. Throughout this chapter I will reference and discuss. and the involvement of both in the production of social events (2010: 4).120 Clarity and Confusion in Social Theory existence of exclusive causal powers of social structures. Let us proceed to unpack some of its features in the hope that they might prove illuminating as regards the nature and persistence of such debate: Firstly. and is therefore informed by and meant to fit together with other aspects of an overarching programme. (b) how collectivities are related to individuals and whether the former are reducible to the latter. . spring from philosophy. Others. Perhaps emphasising the above amounts to no more than saying that the stance taken is a realist one (cf. I have already gone to some length in previous chapters in order to show the depths of confusion that ‘ontological’ questions about the reality or existence of social structures bring to the discussion. second. apart from creating the potential for confusion. b). and (c): if collectivities possess causal powers which cannot properly belong to individuals. Indeed. Fourthly. some of these contributions. and Porpora (2007) testifies. and (c) which social ontology to adopt. that something like a theory of causal powers is required to guarantee their sound ascription. the issues of (a) how to decide the location of agency. such as the question of ontology. King (2007). as the debate between himself (2007a. Some are tied to the emergence of sociology as a discipline. Be that as it may. where appropriate. then the former cannot be reduced to the latter and we can admit into our ontology only those ‘entities’ possessing irreducible causal powers. as well as to expose the problems associated with attempting to fix the form of explanation. To see the issues as interrelated is. that the identification of causal powers and their proper bearers is a necessary (at the very least) requirement for explanation. first. Secondly. Kaidesoja 2007). Thirdly. apart from (and as part of) the relationship between structure and agency. the debate ranges over a number of controversial issues.3) and. The debate is extremely instructive because it stands as an example of what controversy around structure and agency looks like. Varela (2007). treating this set of issues in the manner just sketched produces such an intractable mix of problems that attaining a perspicuous view proves very difficult.

The issue. i. my purpose is not to disprove the MC (neither. it is perhaps a fact to be thankful of that the debate does not feature any more issues that have been treated under the rubric of ‘the problem of structure and agency’. with the idea of causal powers and their proper bearer. however. The account does contain the. in what follows we must proceed with the utmost care to separate sources of confusion and to avoid producing a response that will consist of ‘more of the same’. then. further. familiar by now.. which is here taken for granted. to prove it) but to use it as a vehicle with which to examine the procedures I have sketched in the introduction to this chapter. is primarily concerned with the consistent ascription of causal powers based on the scheme ‘power of a particular’.3 In any case. which by itself provides a plausible explanation as to why such debates continue interminably. the reader may be guided by the headings to the following sections which reference both the aspects of the MC and the conceptual procedures in question.e. As already noted. of free and therefore morally responsible action. As I am sure is clear by now. examining the aspects of the MC will serve as a way of developing detailed understanding of what exactly is at issue. . is most directly related to the philosophical problem of free will.Some Problems with ‘The Problem of Structure and Agency’ 121 Given the theoretical morass. This scheme of the ‘logic of causal powers’ establishes what kind of thing the bearers of causal powers may be. which imposes the criterion that what can have such powers is an entity with parts. That this is the case as far as the debate is concerned can be seen from Varela’s position. Elder-Vass’s emergentist account lays stress on the second aspect of the MC. of course. The next section investigates the use of modal 3 One might wonder what the presence of such a perennial philosophical problem in social theory implies about the relationship between the two and. of course. that is. On the other hand. idea that causal powers are a product of the way the parts of a whole are organised. The MC can be analysed into two claims: one about the possession of causal powers and one about the exclusive possession of causal powers. allow the determination of the exclusive possession of powers by one of two or more candidates except in the limiting case where the other candidate(s) turn(s) out to have been wrongly identified. Starting. It does not. like Harré (as we saw in Chapter 4). In the following three sections. but the emphasis is placed on the core idea of emergence that the parts taken separately do not or – understanding the nature of this restriction is crucial – may not possess the causal powers in question. I will discuss it as an example of adherence to a priori stipulations on explanatory form and (here) causal concepts. I am referring chiefly to the question of the very possibility of human agency. Elder-Vass is the main contributor to the debate in question and responses are clustered around his thesis that social structures as ‘higherlevel entities may possess causal powers in their own right’ (2007a: 28) – call this the Main Contention (MC). To navigate the discussion. in cases where the first condition is not fulfilled because an allegedly ascribed power lacks a proper bearer. Varela. by what means the latter could ever hope to settle it if not by the means of conceptual elucidation I am proposing in this book.

where causes are to be located. This is the engine’s power we might say. A good place to start is the concept of power in one of its paradigmatic uses that we all understand. along the pattern we have already identified in Chapter 2. Some simple observations will help us see this thought as confused and will also prepare the consideration of the concept of causal powers. We say that an engine produces 100 horsepower. second. it can cause . which originates in Harré and Madden (1975) and finds its way via Bhaskar and Archer into Elder-Vass’s CR. shifting terms slightly. Finally. Consider some examples: The engine’s power can be a cause for pride for the car’s owner or the engine’s manufacturer if such power is extracted out of a relatively small cubic capacity. the attempt to fix explanatory form by stipulating. What kind of consideration can help our search? Well. we might ask.122 Clarity and Confusion in Social Theory expressions such as ‘may’ in the claim that structures may possess powers together with the issue of misrepresenting conceptual issues as empirical. has to do with establishing what can have causal powers. The engine’s power is what (among other things) enables a car to reach a speed of 100 miles per hour. I will argue that this conception rests on two mistakes: the idea that it makes sense to debate the proper bearer of causal powers without specifying what kind of causal power is at issue and. we need to find out what kinds of things can be causes or. but also in the one we encountered in Chapter 4. is that power a causal power? How can we decide this? Perhaps we have in mind the way in which the engine’s power can be said to cause something. But before we can do that we need to know where to look.3 Causal Powers and their Proper Bearer: Fixing Causal Concepts and Explanatory Form I have already mentioned on more than one occasion critical realists’ chosen form of explanation in terms of underlying mechanisms. the question of the exclusive possession of causal powers by social structures is examined together with the problematic ways of relating examples to the theoretical scheme. that citing the powers of proper bearers has explanatory force without any reference to an explanandum in question. Its disassembled parts do not produce anything. Taking a step back. the thought concludes. I will first look at the question of the proper bearer of causal powers before proceeding to trace the latter mistake in Harré and Madden’s work. It is the following temptingly simple thought: if we are to produce any account of the shaping. We have seen that one of the issues that surfaces in the debate we are concerned with. Such emphasis is programmatic and rather straightforward to discern. what kinds of things can have causal powers. I would like to bring to our attention some of the potential reasoning that might have led theorists to consider this as important. Perhaps more tricky to unearth is the conception of explanation presupposed by the realist account of causality in terms of causal powers. we need to look for causes. influencing or determining of social phenomena. 6. Now.

the diversity of senses encountered above in which the engine’s power can be said to cause something is an important fact in this consideration. when we are guided by the reasoning I have identified above. states. under the latter interpretation the engine and the engine’s power have different causal powers. Shared commitments do not end here. and to that end they cast powerful particulars or entities with parts as exclusive occupants of the role. but also extend over the assumptions underpinning the use of the concept of causal power. Accordingly. Still. The question now is not what the causal powers of something are but rather what those somethings that can have causal powers can be. This is also true for the question ‘what can be a cause?’ when we have failed to describe what kind of cause we have in mind. the causal power(s) one has in mind need to be specified. But the general question ‘what can have a causal power?’. a proper bearer for causal powers. more generally. among others. instead of trying to fix causal powers on a single category of what. where ‘causal power’ is unspecified. Harré. Is then citing these facts. in other words the causal powers of the engine’s power and not the causal powers of the engine. however. in other words. what it takes to establish that the engine’s power is indeed a causal power? Perhaps we are on a completely wrong track and a causal power is not thought of as what a power can cause but as what something can potentially cause. events. insubstantial agents. it can be meaningfully asked what can have causal powers. Varela 2007) and. We can perhaps answer the specific question of what has the power to corrode some metals by offering that oxygen does. including in the family of causes substantial agents. But the role is not one it makes sense to fix on a category of occupant as the occupant needs to vary with the causal powers in question. Borrowing Hacker’s (2007: 87) words: ‘our causal discourse incorporates [a significant] degree of conceptual diversity. properties. The stated . Let us now turn the question around into the form that it receives when treated by the realist theory of causality (Harré and Madden 1975. is an empty question. What it shows is that the question in its present form does not make much sense.Some Problems with ‘The Problem of Structure and Agency’ 123 an opponent in a race to feel intimidated if the engine in their car is a lot less powerful. Harré and Madden’s book comprises both an assault on the Humean theory of causality and the simultaneous effort to avoid the latter’s shortcomings by constructing ‘a conceptual system capable of accommodating the actual intellectual practices of science … [and] common sense’ (1975: 1–2). Elder-Vass and other realist theorists are committed to the idea of finding a suitable what that can be a cause. our best bet in responding to such a question is to remind ourselves of the sheer variety of whats we can connect with causal concepts. I now turn to examine the rationale for introducing the concept in Harré and Madden (1975). that is. one might wonder. it can cause a lot of wheel spin during acceleration. Before. This entails that so far we have been identifying the things that the engine’s power can cause. conditions and facts’. Flying in the face of what these observations on the lack of intelligibility show. processes. and can cause fuel consumption to increase dramatically at higher revolutions. and the question of what has the power to cause embarrassment by answering that one’s deeds do.

for example how substances act and react with other substances. Porpora (2007) is eager to point out that there are cases where we can say that something which is not a powerful particular or an entity. or even in certain contexts.5 Another kind of 4 As Hacker insightfully points out. nevertheless. as. well-adapted to the conditions of our life. multifaceted and frayed at the edges as our other characteristic categorial concepts. ‘[t]he salient pitfall in the philosophical scrutiny of our concept of causation and of the forms of our causal statements is to suppose that usage is conceptually homogeneous. a state of affairs. this does not mean that its outcome cannot be expressed in the form of necessary relations between theoretical concepts – an example Harré and Madden cite in this respect is the concept of acid which is internally related (at least at the time they wrote the book) with the power to turn longwood solution red (1975: 8). by the authors focusing on only one of the centres of variation (Hacker 2007: Ch. is a powerful particular. however.’ (2007: 89) It should be noted here that some critical realists show signs of resisting any straight-jacketing of causal concepts. In the following passage this fact is acknowledged by Harré and Madden who. when the proportion between people working and those receiving pension may render a welfare system unviable. For one. But what may be singled out as the cause may be an event. In this sense. causation always involves a material particular which produces or generates something. We show how in terms of that notion the citation of the presence of a particular has explanatory force … Our conception of causality is deliberately in keeping with one of the commoner ways in which this concept is employed. Kaidesoja (2007) is prepared to give up the idea that there is only one ‘ontological analysis’ of causal concepts. After the formation of such theoretical . Lewis (2000) attempts to broaden the critical realist conception of cause by bringing in Aristotle’s material cause. Finally.124 Clarity and Confusion in Social Theory aim of the book is vitiated. undermines their aim of doing justice to our actual practice of causal concept use and exposes their project as one of reconstruction or revision rather than elucidation: Our most central idea is that of a powerful particular. sets of activities where causal concepts are employed to different ends.4 This manoeuvre. 3) of our causal concepts. the crucial element. For another. try to circumvent the difficulty by promoting their chosen employment as the ground of all other cases of causal talk. of course. and hence to seek to reduce all forms to a single form. (1975: 5) (my emphasis) Instructive in the above for our present purposes is how Harré and Madden distinguish between the powers a particular has to generate or produce something and the citation of the presence of a particular or singling out of a cause in an explanatory context. in short. the presence of which makes the action causal. but obviously related in the context of science. a material substance … [nevertheless] in any specific application of the notion of causality. for instance. We might present this distinction as one between two different. we argue. But our general category of causation is as unruly. One set of activities is the investigation of natures and powers of stuffs. It is. 5 Let it be noted for future reference that although such enquiry can be empirical. namely a proportion is the cause.

but it is the unexpected presence of a nail. 6 This point is also emphasised by the pragmatist tradition which. the question whether acid turns longwood solution red is no longer an empirical one. is not limited to picking ‘powerful particulars’ or ‘agents with powers’ as causes. . If Achilles has managed to teach us something. the poor state of the tires or the wrong type of tires. but a category that guides enquiry toward solving problematic situations’ (2004: 233). Having sketched the provenance and rationale behind the use of the concept of causal powers.6 It thus makes no sense to be constrained in our explanations by the search for powerful particulars. it all depends on what we are trying to explain. but rather that it is not possible to guarantee that it will independently of what it is we wish to explain. serve to classify such substances as acids. ‘entities with parts’. Thus.Some Problems with ‘The Problem of Structure and Agency’ 125 activity focuses on the identification of a particular cause (or a number of causes for that matter) in the context of an investigation or explanation of a particular phenomenon. still want to guarantee that powerful particulars serve as explanatory causes and this is exactly the service that their concept of a causal power attempts to provide by running together these two logically distinct activities. the lack of a roadside barrier. bad road conditions. of course. can on occasion be disjunctively – or in some cases conjunctively – cited as the cause(s) of the accident. What this means is that finding out which powers make acids acids becomes a conceptual investigation. as Harré and Madden acknowledge. In other words. The second type of activity. As noted. Harré and Madden. when one is operating under the auspices of the relevant theory. it can now also be understood how it functions in Elder-Vass’s project. the level of alcohol in the driver’s blood or poor driving skills which. the concept is meant to fix the transition from having a power to do something to having the power to be identified as (one of) the cause(s) in the context of an explanation. a large nail has the power to pierce a car tire and may thus be cited as the cause of a puncture. The powers that particular acids have can still be decided empirically but not the ones that. being a powerful particular (or an entity with parts) is neither a necessary nor a sufficient condition for something to be cited as a cause (cf. excessive speed. defeats the purpose of having a general and explanandumindependent guarantee. Establishing that (some) social structures are entities with parts and thus concepts. Adequate specification of the explanandum needs to precede any such guarantee which. in other words. although not classed as ‘powerful particulars. understands that ‘causation is not an ontological concept. For example. which is not to say that some of our explanations may not cite precisely such particulars as the cause(s). after all. by definition. the former power is neither necessary nor sufficient for the latter ‘power’ which is not a power of the particular in question but a function of the explanandum and the circumstances in which an a explanation is required. as Elder-Vass would have it. it should be this. however. ‘agents with powers’ or. however. the power to have explanatory force. as Piiroinen and Kivinen point out. Hart and Honoré 1985). So it is not that citing a causal power can never have explanatory force.

for example ‘does BP have the capacity to purge the Gulf of Mexico of the adverse effects of an oil spill?’. speaking of causal powers only in connection with such bearers.e. Instead the question operates at the level of kinds and. can be phrased in the following way: is there at least one kind of social structure that can or may have exclusive causal powers that its parts do not possess ‘causal powers of the structures and not of the individuals’ (2007a: 40)? Although. pivoting on the modal ‘may’ (or ‘can’) as it does. allows Elder-Vass to add these bearers to the catalogue of ‘entities’ operative in the social world which. to stick with the same example. as already mentioned. it is concerned with the causal powers that organisations can exclusively possess. . The question Elder-Vass is asking is not ‘does this organisation have the causal power to V?’. The following section will attempt to show that conceiving of conceptual matters as subject to empirical verification/falsification and. are to be legitimately invoked (along with the relevant causal powers) in order to explain particular events.4 The Causal Powers that Structures May Have: Conflating Conceptual and Empirical Matters Elder-Vass’s MC states that social structures may possess causal powers in their own right. We have examined some of the work done by ‘causal power’ but the word ‘may’ also merits attention. more generally. the question still operates at the level of kinds. In this section I have examined the realist deployment of the concept of causal powers and the idea of their proper bearer and have argued that the procedures of reducing the bearers of causal concepts to a single category are misconceived and that so is. i. ‘did BP cause on its own the oil spill in the Gulf of Mexico?’. under the conception which has just been sketched. 6. This section will deal with the question of how modal expressions such as ‘may’ or ’can’ are to be understood. thus producing a rigid conceptual mould which can then be used to guarantee that citation of such bearers or powers will have explanatory force. conversely. which excludes from consideration the particular causal powers of a particular social structure. nor is the question ‘did this organisation exclusively cause X?’ or. Elder-Vass tackles this question by taking organisations (and what he calls ‘norm circles’) as examples. and what their use implies about the kind of claim the ascription of causal powers is.. lack of care in distinguishing between conceptual and empirical claims further add to the confusion.126 Clarity and Confusion in Social Theory proper bearers of causal powers. But this is exactly the mistaken entailment that is being obscured by the fusion of two logically distinct activities in the concept of causal power as it originates in Harré and Madden’s account of the natural world. Enquiring into the exclusive causal powers social structures may have introduces a distinct logic to the MC. I will next consider problems that arise in the investigation of causal powers in the social world. especially when such investigation is conducted in general categorial terms and/or concerns concepts that are not social scientific inventions.

7 It will now be contended that.Some Problems with ‘The Problem of Structure and Agency’ 127 It may be contended that observing differences between kinds and particulars is. Confusion is compounded by misunderstanding the particular nature of most concepts in the context of the social sciences. It is not clear to me that they do. ultimately. that is. even further. apart from not respecting the contrast between the powers of kinds and the explanation of particular cases. does not necessarily apply to human affairs. The notion of powerful particulars also operates at the level of kinds. To see this one may look at Elder-Vass’s response to Varela’s challenge. as was argued in the previous section. which he thinks.e. there is also a failure on Elder-Vass’s part to appreciate the differences between establishing that a particular thing has a causal power and that a kind of thing can have a causal power. Varela (2007) holds that the means critical realists use for ascribing causal powers are generally unprincipled as they do not follow Harré and Madden closely enough. appealing to his own method for social ontology as what enables him to make systematic ascriptions. . the identification of causal mechanisms by accounting for the way that the causal powers of a kind of entity come to exist based on the organisation of its parts. as claims about some transcendental domain which escapes us. The feeling of déjà vu in the reader is warranted. Lawson 2003: 80). consistent with distinctions (critical) realists tend to make such as the one between retroduction.. i. like Harré and Madden. if anything. however. Varela supports the ‘power of a particular’ scheme deriving from Harré and Madden’s causal powers theory which he considers as the only means of ‘principled ascription of causal powers to entities of the right sort’ (202) and as such. Elder-Vass’s response to the challenge posed by Varela consists in denying the coherence of Harréan criteria and. is used to run together retrodiction and retroduction by restricting what can be cited as a particular explanatory cause to kinds which are organised in the form of a generative mechanism (entities thus organised so as to exhibit emergent powers). he too is committed to using a singular legitimate mould for causal concepts. the explanation of a particular event with the identification of the multitude of causes that brought it about (see Elder-Vass 2010: 72–3. and retrodiction. differences between questions pertaining to particulars and questions pertaining to kinds are already potentially confounded in the concept of causal power which. in speaking of powerful particulars. Archer and Bhaskar’s causal criterion is found too open-ended and Elder-Vass’s own relational emergence angle is thought to ‘violate the cardinal principle of causal powers theory’ (Varela 2007: 208) in locating causal power in social structure (which under Varela’s conception does not constitute a powerful particular) thus adding this power to the particular (individuals) rather than predicating it of a particular. Harré and Madden avoid this problem. What he says 7 It might be said that. However. This amounts to a conflation of the empirical with the conceptual which is manifest in the fact that both claims are unhappily glossed as fallible knowledge claims and.

lends some plausibility to CR’s insistence on a stratified ontology. The point of the ontological seriousness produced by this sleight of hand is to maintain the impression of unknown entities beyond our knowledge frontiers which. for example. . in fact.. Where controversy can and does arise. issuing a claim about knowing. None of these things we know are entities nor can claims to know them be spoken of as existence claims after the paradigmatic postulation in physics of. however. in this case with ‘the logic of causal powers’. say. gravitons and dark matter. the mass of Pluto (or what Pluto’s mass is). is when we consider what sort of criteria one may employ and whether the fact that not anything goes implies that ascriptions of causal powers to kinds are correctly represented as fallible knowledge claims. it is sure to falsify investigation which is concerned with concepts.3. (2007b: 468) (my emphasis) Elder-Vass’s epistemological discussion has an air of ontological gravitas (originating in Bhaskar’s attempt to avoid the epistemic fallacy by postulating nonobservables) with the result that – by running together ‘knowing X’ (in the sense of being acquainted with) and ‘knowing that p’ – knowledge claims are spoken of as (involving) existence claims: to know something or someone presupposes that that same thing or one exists. of course. as being about hidden entities the existence of which we cannot conclusively prove. pace Elder-Vass and others’ ontologising of the question. that it does not exist).e. understanding. in turn. i.8 Some examples will help us appreciate a number of different kinds of knowledge 8 Accordingly. the assertion that there are indeed criteria by which we may assess ascriptions does not turn on any ontological claim and is rather uncontroversial. and as all good realists know. Thus. and so we need criteria by which to assess such ascriptions. all such claims are fallible since the thing we purport to know exists (or does not) independently of what we purport to know about it. our concepts. But such knowledge is mostly irrelevant to scientific investigation where most of the ‘things we purport to know’. As I have argued in section 4. nothing could be more misplaced than this picture of a hidden ontology in guiding our thinking about society. the grammatical criteria that govern the logical relations between our concepts. On the contrary. in the way that knowing John presupposes that John exists. He claims that: when we ascribe a causal power to an entity we are making a knowledge claim. the length of the gestation period in mice and that Greek sovereign debt exploded after 2009 cannot be said to presuppose that what we know exists (nor. far from making an empirical claim we are. Indeed. the only criteria one may defer to when one is concerned with ‘the nature of causal efficacy itself’ (2007b: 468) are. It will be argued that when we say that a kind of thing can have a causal power we are not making a knowledge claim whose substantiation can lie beyond our current knowledge frontiers and which can depend on currently inaccessible information.128 Clarity and Confusion in Social Theory in the process evinces his conception of this form of enquiry and of the logic of ascriptions of causal powers. Ascription of causal powers is therefore distinct from their existence.

Some Problems with ‘The Problem of Structure and Agency’


claims, the backing required and how the question of fallibility arises or fails to
arise in each case.
First, let us return to one of the examples already used and consider the claim
that a specific organisation, BP, has the power to purge the Gulf of Mexico of an
oil slick. Obviously, knowing this hinges on one having a rough idea of the extent
of the problem and the resources available to the organisation. It makes sense
to say that one may be mistaken in that estimation, if for example the extent of
the damage caused has not been assessed properly or one has miscalculated the
executive and financial capabilities of the organisation. All in all what is needed to
back the claim is a sound understanding of the particulars of the case.
Second, consider the general claim that organisations have the power to make
their employees redundant. What kind of backing does this claim require? One way
we can understand this is as an assertion about the legal status of organisations,
the law governing the relations with their employees and, for example, the latter’s
compensatory rights in case of redundancy. Accordingly, the claim can be backed
by an appeal to statutory law. The application of the law in particular cases where
disputes arise can be anything but straightforward, but as a question of what the
law generally says the claim can be easily settled (cf. Hart 1997[1961]: Ch. 7).9
Note that once this is done there can be no further question of fallibility: a claim
to know the law is something fallible yet settled by consulting what the law says.
Moreover, although it is not clear what it would mean to say that the law is fallible,
one could talk about the law being wrong, but that is clearly not what is at stake
here. To consider one or two additional relevant expressions, ‘questioning the
law’ is not a matter of consulting any other law unless there is a question of it
contravening the constitution, etc. and this is hardly a matter of fallibility. And
‘knowing that the law is right’ is typically not an empirical knowledge claim but
an acknowledgement and/or expression of agreement or conviction.
Finally, consider the assertion that organisations have the power to mobilise
their members in pursuit of some goal. Is this a fallible knowledge claim? Sure
enough someone who said that organisations have the power to lay eggs would
not be right. However, in the same way that the person who asked how much
capitalism weighs would be questioned about whether they have understood
what capitalism is, saying that organisations cannot lay eggs is not something
that it makes sense to portray as a fallible knowledge claim. Equally, saying that
organisations have the power to mobilise their members in order to accomplish an
objective is a way of talking about the concept of organisation, and in so far as we
understand the concept no question of fallibility arises.10
9 Suppose that we came across a case where an employee had immunity due to special
reasons. What should we conclude? Would this mean that our former claim is mistaken?
Only if we meant to say that the law stipulates that all organisations of all kinds have the
power to fire any of their employees. But no such thing need have been involved.
10 Compare here Stenlund’s point with regard to logical rules for the use of
expressions: ‘There can be a problem about the validity or the correctness of [such rules] …


Clarity and Confusion in Social Theory

It is instructive at this point to return to Harré and Madden’s example of the
concept of acid in order to compare it with the concept of organisation. I noted
previously (fn 5) that the relation between acid and the power to turn longwood
solution red is an internal one. Turning longwood solution red is a criterion for
being an acid and, accordingly, we may use it as an inference licence to pass
from the observation that a substance does not turn longwood solution red to the
conclusion that it is not an acid. Hence, to be able to entertain the thought that
something might be an acid and not turn longwood solution red, the latter effect
needs to be dropped from its criterial role.11 But that would entail a change in the
concept of acid (call the new concept acid*) so that the claim that it transpires
that acid* does not turn longwood solution red is not the negation of ‘acid turns
longwood solution red’. We can express the difference between the two claims as
a distinction between physical and logical possibility. In this case the difference
between physically possible powers and logically possible powers is decided by
the way being an acid and the power to turn longwood solution red fit into an
investigation: are they internally related and used as an inference licence, that is to
say are they criterial, or do we use a concept of acid that does not depend on such
a criterion and, accordingly, it is an intelligible empirical matter whether it has
the power in question? The ‘can’ or ‘may’ that befits the investigation of criteria
is one of logical, not physical possibility.12 The crucial point then is that if one is
interested in what causal powers acids can have then one is interested in exactly
those causal powers that are used as criteria for something to count as an acid.
The chemical concept of acid, precisely because it is a theoretical concept,
i.e., it is part of a theory and stands or falls with it, is designed to classify and
account for a number of phenomena, enabling the drawing of certain inferences,13
in a certain situation only in the sense that it is unclear what the logical rules are for that
use, but as soon as they are made clear there is no additional problem about whether they
are satisfied or valid or correct’ (1990: 51).
11 Indeed, according to Glock (2009: 663), the criterion for being an acid has changed
to ‘being a proton-donor’.
12 Speaking of an ‘empirical matter’ as not being a question of logical possibility
may elicit the objection that logical possibility ultimately hinges on physical possibility.
Such an objection might be based, for example, on Harré and Madden’s claim that logical
necessity is a reflection of natural necessity. In this context, one may argue that although
the issue might be one about the definitional powers of acids, nevertheless, those powers
feature in the real definition of acid. This might be thought to show that our concepts are
not impervious to reality. Whatever it means to claim such a thing, and whether or not the
notion of a real definition is a coherent one, the point here is to clarify that taking an interest
in the definitional powers that make something what it is, is to take an interest in concepts,
not to say that those concepts are arbitrarily what they are.
13 Toulmin emphasises this point about scientific classification in that gem of a book
of his:
What is the point of the physicist’s reclassification? To see this, recall that it is his
aim to find ways of inferring the characteristics of phenomena from a knowledge

Some Problems with ‘The Problem of Structure and Agency’


and is based on a set of criteria that crystallise and systematise experimental work.
However, to repeat, the examination of the concept of acid is not an empirical
investigation nor are statements laying down the criteria which constitute concepts
fallible, unless we are prepared to change the concept and modify/jettison the theory
that it plays a part in. The criteria for knowledge are the criteria for understanding
the concept of acid, and those criteria are neither true nor false although we can be
wrong or right about them and, naturally, statements we make using the concept
of acid can be true or false.
On the other hand, if the concept in question is not a theoretical concept,
i.e., is not tied to any scientific theory but is rather part of ordinary English,
then knowledge is not knowledge of a theory but knowledge of the language.
This crucial observation removes further grounds that might serve to make talk
of fallibility plausible. Concepts which are part of scientific theories allow for
Harré and Madden’s idea that necessary relations between concepts reflect natural
necessity, as well as for a sense in which one can speak of the concept’s fallibility
as attached to the fallibility of a theory with regard to some kind of ontological
beyond, misleading though such a way of phrasing matters might be. But when
it comes to the concept of organisation there is no equivalent beyond or sense of
fallibility of this kind. The concept of organisation is not part of any scientific theory
nor is it formed after experimentation in order to serve theoretical explanation
(cf. Bittner 1965; also section 4.6). The concept is what it is, it has its lay and legal
uses, and if the investigation is to be concerned with that concept – for there is no
case to be made that Elder-Vass is concerned with a different concept (see ElderVass 2010: 144) – then it is not up to one to change it or to portray it as if it were
a concept of the same type as the concept of acid.14
Thus, it is problematic to say that a claim attributing a causal power to a
kind is a fallible knowledge claim when that attribution is the expression of a
rule constitutive of the concept in question, especially when one is engaged in an
investigation focusing on precisely these constitutive rules. For although ElderVass does not specify what he is willing to count as a causal power and whether
any causal power will do, there is nothing to suggest that what is being sought after
are some contingent powers that a particular organisation might have and not the
conceptually related powers which enter into the concept of organisation or other
related concepts. That the latter is in fact the case is manifest in the concern with
the powers kinds and not particulars may or may not, can or cannot have, steering

of their circumstances … To speak of something as a ‘blackboard’, for example,
implies hardly anything about how it will behave. No doubt, if it explodes, or
crumbles into dust, or vanishes without warning, we shall be very much surprised …
but it cannot be said to be implied by one’s description of it as a blackboard
(1953: 52).
14 This might involve taking into account that there may not be clear criteria as there
sometimes are in the case of a scientific concept.

etc. we will witness how the examples that are conscripted as supporting Elder-Vass’s theoretical scheme are. that are not part of the finger. is it the finger or the person that actually does press it? The causal power question is easily answered by considering the counterfactual question of whether the finger would be able to press the keyboard if it were not part of the whole human. 6. when they are accurately described.e. misdescribed by the imposition of the scheme onto them and. are uncontroversial and cannot (exclusively) support the scheme. that organisations can have certain causal powers that their parts (i. or detonating a bomb. I have tried to show that speaking of the knowledge of (statements about) concepts as corrigible in light of scientific progress is mistaken. Clearly it would not: the ability to press the key depends upon a larger configuration of bones. Hence the causal power to press the key is a power of the whole human and not of the finger. as exemplified in the sense in which such ascriptions are spoken of as fallible knowledge claims. In this section I have considered Elder-Vass’s response to Varela with regard to the issue of the logic for ascribing causal powers to kinds. Would we not rather say more straightforwardly that someone was pressing a key on a computer keyboard (with their finger). at best. is it the finger or the person of which it is a part that has the causal power to press the key? And secondly.132 Clarity and Confusion in Social Theory interest towards what makes organisations organisations and what is logically possible for them to do.5 The Exclusive Causal Powers of Social Structures: Problems in the Use of Examples This section will provide a detailed discussion of the most controversial aspect of Elder-Vass’s MC. for example.? . managers and other employees taken as individuals) may not possess. I have argued that it harbours confusion of conceptual and empirical matters. even more so when these concepts are not technical or scientific ones and that the issue here is rather of understanding our concepts and portraying conceptual relations accurately. muscles and the like. namely the idea that social structures may possess exclusive causal powers or ‘causal powers in their own right’ (2007a: 28). 15 Note that this is already a strange way of describing the situation. or even better that someone was typing. Let us start by examining how Elder-Vass uses examples to introduce the ideas which he will then apply to social structures: [T]ake the example of a human finger pressing a key on a computer keyboard15 … First. While engaging with this thesis at some analytical depth. and upon the brain’s ability to send signals to the finger through the nervous system. their members such as administrators. at worst.

people do. one may contend by the same token that were one’s finger cut off they would not have the power to press the key. The most plausible answer. when a person acts. the question is not one that could be settled by the counterfactual test.Some Problems with ‘The Problem of Structure and Agency’ 133 The second question is more open. it seems to me. also features a number of conceptual difficulties. I would want to say. The response to the second question. about the logical issue of whether it makes sense to say that a human and a finger can do the same thing.16 In fact. This case is perhaps a step more challenging than the key press case. it can be objected that the counterfactual test is not sufficient to establish the exclusive possession of powers. which appears to work by removing a part of the functioning causal mechanism (for example: would the wheels of a car turn if we disconnected the transmission?). for having the situation described as ‘a human finger pressing a key’ we are puzzled about what could possibly be going on and. The finger does so very directly and the person does so through the finger. ‘who is it that actually presses the key?’. Thus when the finger presses the key it acts both as a finger and as a part of a larger whole. whereas human individuals. we may even be led to wonder whether one’s finger was severed and struck a key while falling to the ground. This is why the initial description of the situation is suspect. In such cases. because it is difficult for us to conceive of the finger as having causal powers of its own. the two elements cannot be empirically but only analytically disentangled: there is no event ‘finger pressing key’ that can be empirically distinguished from the event ‘person pressing key’ and yet using the counterfactual method we can distinguish between the two corresponding causal claims. In such cases the structure acts through the person and the person implements the structure’s causal power. in an attempt to find some situation which makes the description intelligible. The finger. is the part of the person that implements the person’s causal power to press the key. (2010: 27–8) There are a number of dubious moves in the above illustration that are consequential as regards the claim to apply the argument to social structures. a consideration applying before any mechanism removal test can gain a foothold. is that both the person and the finger press the key. Again. … I will suggest that this argument can be generalised to people and social structures: that sometimes. both have causal powers of their own and implement causal powers that belong properly to higher social entities. in fact. they do so both as an individual and as a part of a structure..e. As far as the question ‘who has the causal power?’ is concerned. the seeming interchangeability of ‘person pressing key’ and ‘finger pressing key’ is only 16 And it will not do to say that they could then press it with their nose for fear of losing that too! . I will argue. which is one of its parts. Far from being concerned with physical possibility the question is. i. What actually makes the difference in this case is not the counterfactual test but the conceptual point that fingers do not press anything of their own accord.

one we can and do take note of in describing them. I can be said to press a key directly when I do so without using some kind of tool. either that the finger is a severed one. Still. ‘human finger pressing key’ and ‘person pressing key’ are not interchangeable descriptions. even if one were to allow that these two descriptions are interchangeable. or as if I am somehow encaged in my body so that I can only touch the key indirectly. inadvertently. section 4. he notes that the sales assistant does not own the TV and does not sell it on behalf of herself. who arranges for it to be delivered to my home in a few days’ time’ (2007a: 33). It is nonsensical to describe the situation as if I am at a distance and need to use some device. This familiar situation serves as the scene for a number of no less familiar and rather uncontroversial observations. To these observations then Elder-Vass forces the questions which were invoked in the finger-key example (my formulation): ‘who is it that sells me the TV?’ and ‘who has the power to sell me the TV. my finger in this case. the assertion that the two descriptions are empirically indistinguishable operates on a very impoverished sense of what is observable that critical realists share with empiricists (cf. as squash player). because the former can in some cases imply. etc. Elder-Vass notes that I ask the person at the shop in their capacity as a sales assistant (and not.e. say. it would not follow that there are two empirically indistinguishable events going on simultaneously.3. even though such a difference need not be one that can be captured by means of still photography – which seems to be the constraint Elder-Vass has imposed on the case. Accordingly. or that I pressed the key by accident. Instead. to help me reach the key. people do not press buttons through their fingers. These conceptual moves on Elder-Vass’s part are meant to lend support to the scheme of emergence. especially not in contrast to my finger itself which is doing the job directly. which will allow us to decide which description to apply. a role the person assumes as part of the organisation. also Stenlund 1997). First.134 Clarity and Confusion in Social Theory possible due to the lack of an adequate specification of the situation. i. they do it with their fingers. the sales assistant or the organisation?’ He responds thus: . it is as part of the organisation that she accepts the payment and can arrange for the TV to be delivered to my home address. Furthermore. Finally. but they fail because they misdescribe the chosen example. it follows that there is an observable difference between such situations.. It will now be argued that the attempt to generalise the scheme to individuals and social structures makes no less problematic a use of the examples it employs. it would be the case that one event or action admits two or more descriptions (see Anscombe 1979. White 1979). Although when I am pressing a key my finger is pressing against the key. In so far as we use these two different forms of words to describe different situations. Elder-Vass illustrates his argument with the example of a person buying a TV which he introduces in the following way: ‘I walk into an electrical shop and purchase a TV from a sales assistant. Moreover. I cannot be said to be pressing the key indirectly when I am using my finger.

we might reasonably ask. yet we understand perfectly well that it is the dog that is primarily causally responsible for the bite. First. Smith’s pin factory) if we say that the productive capacity of the organisation depends on both the workers and the relations between them that exist when they are organised as they are in this organisation. this is necessarily equivalent to saying that the productive capacity is a causal power of the organisation and not of the workers. so to say that the people plus the relations have a power is the same thing as to say that the higher level entity has the power. does this uniformity derive from the examples or is it rather the outcome of an imposition of the emergentist scheme and its requirement of consistent application? In fact. so we must accept that organisations are causally responsible for the behaviour of their members or employees when that behaviour is motivated by organisation policy. if we carefully look at the examples we can see that the comparison between the dog and the organisation (or the division of labour) is misleading because in the case of the dog. Certainly we must accept that the individual agency of the sales assistant co-determines the outcome. for example. But. in an organisation that practices the division of labour (e.Some Problems with ‘The Problem of Structure and Agency’ 135 It is the organisation that sells me the TV. This is perhaps more difficult to accept when the parts have a mind of their own. and not the teeth. The key points in this argument were (a) that if the people concerned were not organised into such an entity. who is one of its parts … if a dog bites me. although the dog bites with its teeth in the same sense that it sees with its eyes. (2007a: 33) (my emphasis in bold) In its discussion of organisations. unlike the case of the organisation or sales assistant selling the TV. . (b) the co-determination of the outcome. those powers would not exist. in another variation of the canine example. it is not conceptually open to suggest that the teeth of the dog bite. something considered as displaying the virtue of consistency for the principles of emergence. that the lungs of the dog bark. Just as we accept that human beings are causally responsible for the behaviour of their parts when it is directed by their decisions.g. human beings and their parts. it does so through its teeth. windpipe and mouth of the dog barked. but the principle is similar. though it does so through the sales assistant. nor. let us take up the contention that there is a similarity of principle involved between the dog and its teeth. (2007a: 38) (my emphasis in bold) There are three pivotal conceptions displayed in the above passage: (a) the similarity of principle between cases. and (b) that the people plus the relations are the higher level entity. unless somehow the dog barked without meaning to (if we can get that to stick as a description of its behaviour). but it only co-determines it. this paper has already stated the positive case for seeing people plus relations as constituting a higher level entity with causal powers. and the organisation and the sales assistant. Thus. nor would we say that the lungs. and (c) the correlated idea that there is a higher-level entity which is the proper bearer of the causal power and the main co-determinant of the outcome.

In order to warrant the transition to the case of organisations and their members Elder-Vass uses the converse path of formulating the relationship between human being and limb in the way we would describe the relationship between organisation and member.). We might as well take after Lewis Carroll’s Alice and consider gifting our feet a pair of boots during Christmas to ensure they take us where we decide we want to go (1998: 16). on the counterfactual test and the ‘redescription principle’ that ‘people plus relations equals higher-level entity’. not following the alleged similarity of principle does not lead to causal myopia (Elder-Vass. on pain of inconsistency. then. according to Elder-Vass. etc. where decisions taken can be said to direct member behaviour. if using the ‘same argument’ is tantamount to riding roughshod over conceptual distinctions. attention will 17 On controlling and being in control of one’s body see Hacker (2007: 276ff. This. Since the first segment has been already partly examined. Opposing this violation of reason. However. The argument turns. it leads to good sense. . Obviously. to use this strategy in order to establish a ‘similarity of principle’ is no argument at all as it amounts to fabricating the evidence one is appealing to.136 Clarity and Confusion in Social Theory Similarly. then it does not follow that rejecting what Elder-Vass says concerning one case commits one. 2010: 199).18 Systematic application. but to say that one is responsible for the behaviour of his parts when directed by his decisions is to force a way of describing organisational life onto humans and their limbs. as for example when someone is failing to perform an intricate dancing routine. is the nub of the disagreement between Elder-Vass and King and the central aspect of the former’s theoretical scheme. one can intelligibly speak of the behaviour of one’s parts when these are somehow not under his (desired degree of) control.17 and of directing the behaviour of a person by their decisions. mouth. The importance of preserving sense militates against the idea of applying the scheme systematically. Elder-Vass sees himself as a non-reductionist who – guided by his scheme – can always preserve the higherlevel entity: ‘Emergentism is therefore recursively consistent: it ascribes causal powers at each level on the basis of the same ontological argument’ (2007a: 37). The ‘similarity of principle’ idea is also used as ammunition against the individualist opposition (Elder-Vass has King in mind although King himself disavows the label) which is deemed inconsistent and seemingly reduced to absurdity because it is held to entail that it is not the dog that barks but its lungs. It remains for us to examine the justification Elder-Vass offers for accepting the claim for the higher-level entity. after all. 18 Thus. There is thus nothing to be judged as similar (or dissimilar for that matter) in terms of emergentist part-whole relations. can only be guaranteed by conceptual gerrymandering which is intended to support the unrestricted application of the tests (for example the counterfactual test) and the principles they purportedly yield. windpipe. to denying what he is saying in another case. One can potentially deny that it makes sense to speak of a higher-level entity in the organisation-sales assistant example. while also denying that the teeth or the lungs of the dog can be said to bark.

it must be made clear that even if it were the case that the causal power is properly only the organisation’s causal power it would not follow that any additional entity is involved. mainly Dick’s doing. on the one hand.19 It may be objected. and rightly so. Elder-Vass claims that the ‘individual agency of the sales assistant co-determines the outcome’ but the main responsibility lies with the organisation.Some Problems with ‘The Problem of Structure and Agency’ 137 be primarily directed to the ‘redescription principle’ which says that the causal power of people plus relations is the causal power of the organisation. Despite the common sales policy of an organisation they are unique individuals each bringing something of their 19 Ryle famously used the equivalent example of. is that speaking of a higher-level entity enables him to argue. and. they are both determinants but the main one is the organisation. A boat engine may have the causal power of propelling a vessel but it is not an entity additional to the engine parts. The reason is that entities are not individuated by their exclusive possession of causal powers. that a TV was sold) is some kind of resultant event. much like in mechanics the sum of forces acting on a body can yield a component force vector whose direction represents the direction of the acceleration the body will experience. administrative offices. at least on first inspection. Alternatively. One option is to read it as an expression of the critical realist understanding of multiple determination where the idea is that a multitude of causes determine the event taking place. although the former may be paid 20 per cent of the value of each sale. it seems that. This will be considered together with the idea of co-determination. viewed through the lens of codetermination. in the same sense that eating a cake was partly Dick’s doing who ate eight slices and in part the cat’s doing who ate the remaining two but. on the other hand. This may be granted. The point. Indeed. in part the sale-assistant’s doing. The idea of co-determination is a crucial element in Elder-Vass’s scheme as it serves to reconcile structure and agency. Although this picture seems to work well with the co-determination idea it cannot apply to the example in question: it is not the case that the sales assistant sells one-fifth of the TV while the organisation the remaining four fifths. This is what the idea of codetermination is meant to allow for. scientific departments. We must thus try our best to understand it. colleges. for example. that Elder-Vass does not speak of an additional entity but rather of a higher-level entity. It is no less questionable to portray the selling of a television as in part the organisation’s doing. This conception would seem to suggest that the outcome (presumably described as ‘what took place’. etc. . Elder-Vass may have in mind something of the following sort: Salespersons are quite different to each other. Rather one is dealing with descriptions of different logical categories. First. however. that there is an additional causal power to be taken into account. therefore. We will see that understanding (the individual agency of) the ‘sales assistant’ and the ‘organisation’ as simultaneously acting separate causes is problematic. taking place once component causes have been computed. ‘The University’ which is not a further member of the previous class but rather belongs to a different logical category (2000[1949]: 17–18).

In elaborating the idea of co-determination Elder-Vass appeals. but it is hard to see how apart from any such specifics ‘the outcome’ is a result of the organisation and the sales assistant as individual being related as major and minor co-determinants. it need not do so as an additional cause. it is certainly ill-advised to speak of ‘the outcome’ in such an underspecified manner that it becomes difficult to decide which factors are indeed relevant.e. it seems that this is not always respected. some organisations are more liberal than others in the procedures they lay down regarding employee behaviour. Finally. were it not for an operational banking system and the earth continuing its rotation ‘the outcome’ would not have occurred either. This allows him to tie his scheme at will to the various ways we would use to describe the relation between ‘wholes’ and ‘parts’. the specification of the job description will not tell us what it is that every person does unless we look at how the individual implements this. i. Co-determination construed thus would more likely come to bear on the way a particular sales assistant makes a sale. However. ways which do not necessarily . talking about purchasing a TV from a sales assistant. etc. as Elder-Vass does not take care to distinguish when he is using the organised-part and when the unorganised-part sense of being a part of something. It is patently absurd to say that all these things are involved as co-determinants. ‘the outcome’ would not have taken place. he hopes. organised parts) equal higherlevel entity. While it is obviously true that the organisational background needs to be taken into account if we are to understand what took place. it fails in that role too. It seems acceptable to say that were it not for someone interested in buying the TV and for a sales assistant who was capable at her job and part of an organisation having the power to sell TVs. will yield causal co-determinants. in the sense of an additional cause.. if Elder-Vass wants to separate the individual’s causal powers from the organisation’s causal powers then using the description ‘sales assistant’ is a choice that creates confusion. What accounts for the ideas of co-determination and higher-level entity is the fact that action takes place within an organisational setting. unfortunately. Although the emergentist account Elder-Vass puts forth is one of relational (or weak) emergence and therefore states that wholes have exclusive powers when compared to their unorganised parts. nor is the corollary that parts and relations (i.e. Thus. Thus the individual qualities of the sales assistant may enter as explanatory factors when some feature of the making of a sale cannot be accounted for by organisational procedures. Describing the situation in one of the ways we normally would. to the counterfactual test which. by the same token. Moreover. for to say that people are organised in a certain way does not imply that they are organised into any entity. Thus. Moreover. attire. In other words. once again. which – let it be noted – is also the way Elder-Vass introduces it.138 Clarity and Confusion in Social Theory own to the job. the counterfactual test also fails if it is meant to establish the equivalence between ‘people and relations’ and ‘higherlevel entity’ as a further determinant. is what settles such a background. the understanding of acting within the organisational setting is typically involved in the description ‘sales assistant’.

The reason is that ‘the organisation’ and ‘the sales assistant’ can be said to be selling the TV in different senses and thus in this case we are dealing not with descriptions of two alternative events but rather with two alternative descriptions of the same event. can perhaps be answered by either ‘the sales assistant’ or by ‘the organisation’ but here one does not serve as an (even minimal) contrast to the other. should have the power to sell TVs: that’s what their job is after all. however. to offer a claim logically unconnected to the higher-entity thesis: the fact that Tom. Sales assistants. that the connection . it is far from clear what it means to assert causal relations between the two. deriving from emergentism. to claim that the organisation has exclusive powers of selling compared to the sales assistant is either to support something different to relational emergence or. such as in the term ‘sales assistant’ which. when adequately specified in other occasions. activities bound to the category and hence invoke the capacity of ‘sales assistant’ as relevant to the action in question. by understanding ‘sales assistant’ to mean the person divested of their capacity as sales assistant. besides distinguishing between acting qua individual and acting qua sales assistant. that the organisation is causally responsible for selling the TV. for instance. we find it sits awkwardly with weak emergence. ‘buying from’ and ‘selling to’ are. is typically an organised-part descriptor. Crucially. any more than it is clear what it means to assert causal relations between ‘speaking’ and ‘moving one’s mouth’. precisely because they are ‘organised parts’. the sales assistant committed murder and this need not involve that she acted in that capacity much as we may say that the prime minister had chicken pox when she was 10 years old (at a time she was not serving as prime minister).20 Thus when we read that ‘[i]t is the organisation that sells me the TV though it does so through the sales assistant who is one of its parts’ and. further. 21 It is the ideal of a mechanistic explanation which provides for Elder-Vass’s understanding that structural powers and individual powers need to be connected as different causes and which is coupled with the idea. unlike the term ‘individual’. There are indeed some cases where the distinction is potentially recognised. Dick and Harry as individuals do not have any such power is not an observation in favour of bestowing the power upon the organisation in contrast to the sales assistant. there is a further distinction to be had between the sales assistant selling the TV qua sales assistant and the organisation selling it.21 20 Although we would say in a relevant circumstance that. The interrogatives ‘who sells me the TV?’ and ‘who has the power to do so?’ which are used to establish the contrast. We are not correcting someone who said ‘the sales assistant sold me the TV’ if we respond by asking whether it was mainly or also TVs’R’US who did so. although the organisation can be said to sell the TV by having one of its sales assistants sell it. Speaking of ‘the sales assistant’ and of a ‘higher-level entity’ also invites confusion because it makes it appear as if. but rather misunderstanding what they said.Some Problems with ‘The Problem of Structure and Agency’ 139 reflect any distinction of this kind. Finally. in this particular case it seems hard to find an intelligible contrast between ‘sales assistant’ and ‘organisation’. as Harvey Sacks would say. Hence.

Apart from the fact that individuated beliefs are not mental events and cannot be said to be located in the brain. To take the paradigm case of the engine and its parts. These two powers. We could. ‘president’.. ‘wholes’. It is imperative then that the logical features of such expressions be brought into view. as is to think that there is any causal mechanism which can include the parts of the engine and the engine. This is equivalent to constructing a set comprising the left-hand glove. Completeness does not end there. ‘being a member of’. however. ‘wholes acting through parts’. to use Ryle’s example. however. the right-hand glove and the pair of gloves. ‘acting in the interests of’. ‘acting in the name of’. simply does not align with our language. much in Hempel’s manner. ‘parts and their relations equalling wholes’. must raise serious questions about the presupposed distinctions. that descriptors referring to unorganised parts be clearly distinguished from those referring to organised parts and that individual and social structure expressions be also clearly distinguished. what these difficulties show is that the condition for any meaningful discussion of relational emergence and ‘structure and agency’. This renders the connection paradoxical. as the emphasis on connecting causal powers leads. the connecting rod and the crankshaft but there is no causal connection (in any similar sense) between these three and the engine. that of the natural world. no more a proliferation of causes which co-determine anything than there is a proliferation of things we can wear on our hands. connected to brain activity (2010: 92. To that is itself causal. ‘individual powers’ and ‘structural powers’ glosses over expressions such as ‘organisation’. .140 Clarity and Confusion in Social Theory In truth.. ‘acting on behalf of’. i. ‘employee’. It might also be noted. via a rampant naturalism. that Elder-Vass’s explanatory account. to channel it through someone’s beliefs which are. ‘following instructions by a superior’. What this neglects is that explanation depends on/addresses a question pitched at a certain conceptual level. There is. among other levels. in passing. to the idea that a complete theory requires us to connect the causal power of organisations to. Coulter and Sharrock 2007). in turn. are thought to be at different levels as the former are emergent powers of ‘higher-level entities’ (since individuals can be amongst their parts). when of the conceptual level of beliefs. This. The engine parts are instead logically connected to the concept of the engine. in a manner that obscures their logical behaviour. Despite Elder-Vass’s temporal account of ‘downward causation’ (2010: 59–62) wholes and parts cannot straightforwardly be said to be themselves causally connected. however. to ask what beliefs led to an action makes talk of the brain irrelevant. It is thus to commit a category mistake to speak in the same breath of causal relations both among the parts of the engine and between the engine and the parts.. typically precludes any appeal to brains or biology (cf. ‘manager’. in turn. etc. ‘shareholder’. 198). adopt a way of speaking under which wholes can cause (in a special sense) changes in their parts. ‘director’.e. This is all Elder-Vass’s argument can amount to: a change in idiom. provides for the notion of completeness of an explanation (2010: 177–8) when the whole mechanism will have been sketched. thus. This fact creates significant problems for any attempt to tie weak emergence or ‘structure and agency’ to common descriptions of discussed examples and to make sense of both the descriptions and the examples. there is a causal connection between the pistons.e. i. The language of ‘parts’. Adequate specification of the explanandum. ‘representing’.

most of . It is precisely because selling TVs is what the business is about that the verb ‘to sell’ can apply both to the sales assistant and the organisation. as well as a range of supporting activities. can be said to sell electrical and electronic equipment. to buy its products from other organisations. Although. the services and products it provides to customers. aims or goals. (or that it is the organisation’s power to do these things) is to describe an organisation which is engaged in different business activities to the one in question. a(n) (un)reliable image for investors and a positive or negative one for its customers. then. clean. can have property and other assets.. for instance. through which various groups of people coordinate and make sense of their concerted activities (Sharrock 1970: 178). for example. as a legal entity separate from the natural persons that are its members. These powers are legally conferred on the organisation. The organisation in Elder-Vass’s example. stack and connect TVs to the power supply but these activities are not predicable of ‘the organisation’. to transport and stock them in warehouses which it leases or owns.e. This does not detract from the understanding that when sales assistants do all those things they do so precisely because they are required to by the organisation. a number of different departments. a balance sheet. It is the organisation that may also be said to possess or have. In speaking of some of the organisation’s central and supporting business activities I have already mentioned some of the things the organisation can do which correspond to rights that are tied to its constitution as a juridical person. As a juridical person. Some of these powers are or include powers to have or possess and thus allow for speaking of things the organisation has or possesses. are to help move.e. can sign and enforce contracts and can file lawsuits and have lawsuits filed against it. i. can hire employees. let us consider the kinds of things the organisation can be said to do. what goes on within such an organisation is much more than selling TV’s or electrical items. in different senses. The expression is used in a number of ways. not all things members of the organisation do as part of their everyday affairs are things the organisation can be said to do. I would like to briefly consider the expression ‘the organisation’. as we have seen. of making use of ‘the company car’ and in a related sense of ‘the company credit card’.Some Problems with ‘The Problem of Structure and Agency’ 141 end. First.. namely corporations. stack and connect TVs to the power supply. I will not discuss how the expression enters into the activities of different kinds of organisations but will focus instead on organisations of the kind implied by Elder-Vass’s example. not conferred on it by ‘the interaction of its members’. Obviously. and are not powers legally ascribable to ‘members and their relations’. i. the organisation can have a treasury. It may be organisational policy that sales assistants. clean. to display such products in its stores and provide several buying options to customers. To say that the organisation helps move. such as property and assets: members may speak. A large number of cases have to do with using the expression ‘the organisation’ in conjunction with verbs describing an organisation’s central business activities. which is of central importance for understanding the difficulties with Elder-Vass’s rendering of the example.

Dissenters may say that although the person sent to the meeting represents the organisation. that the contract is not between the directors of two organisations. .. unless they are implementing what they themselves have previously decided. The question is very complex as there are potentially different kinds of responsibility at stake (causal. for the whole organisation. it is between two organisations. another organisation. although members of the organisation. i. Decisions taken by the organisation cannot be said to be taken by ‘members and their relations’ nor usually are they taken by the whole organisation but rather by the board of directors which decides for the organisation and. lowlevel managers and other employees implement such decisions. Note. both can describe the interaction as one between the third party and the organisation. cannot be said to be implementing any decisions taken by the organisation.e. This description is much better understood as depending on the representational powers members have rather than the causal powers the organisation as some higher-level entity has to determine what its members do.22 22 Space precludes discussing the question of the responsibility of the organisation or its members. say. since some members may have different views about the desirability of a deal with the manufacturer in question. criminal and civil. The board of directors. To finish the brief logical picture of ‘the organisation’. The organisation may also be said to act through its members in the sense that its members implement what the organisation decides. In such a meeting the representative seeking to seal a deal may represent the organisation though not necessarily the whole organisation. for example. For example. as when. by a senior manager) in a business meeting with a TV manufacturer. social). A sales assistant can represent the organisation in engaging with a customer seeking to buy a TV but not (unless specifically granted the power by the organisation. one of its directors needs to sign the contract. I will try to address some of these aspects in the Postscript to this book. one or more of its members need to do that something. ‘The organisation acting through its members’ in the case of the board of directors deciding can only mean that what the directors decide counts as what the organisation decides – here ‘organisation’ cannot be changed to ‘board of directors’ or to ‘the management’ as in the case of low-level employees implementing decisions – and not that what the organisation decides (the higher-level entity) counts as (or causes) what the board of directors (the members) decide.142 Clarity and Confusion in Social Theory the things the organisation can be said to do are things some of its members can be said to do. also. it is they who represent the organisation’s real interests. as in Elder-Vass’s example. The organisation then can indeed be said to act through its members in the sense that it is typically the case that for the organisation to do something. in order for the organisation to sign a contract with. it might also be observed that when members of the organisation deal with third parties. however. which is the second reason why we may say that ‘the organisation sold me the TV’ (the first being that selling TVs is its main business).

which mainly sells me the TV albeit through the sales assistant who nevertheless does not really have the power to sell but rather implements the organisational power. But the facts concerning the examples used can be commonly acknowledged by both theorists (and by anyone else for that matter since. second. All in all. nevertheless. as we saw. for social theoretical practices. We have witnessed in this section a significant part of Elder-Vass’s attempt to present the emergentist theoretical scheme under a plausible light. end up in confusion and thus make significantly less rather than more sense of the example. We saw that in many instances the examples are misdescribed and subordinated to the legislative use of the scheme itself thus constituting illustrated versions of it. I would like to draw out some implications from the above discussion which have relevance for ‘the problem of structure and agency’ and. by conscripting a number of examples that are meant to lend support to it. more broadly. and. Elder-Vass’s description of his disagreement with King with regard to the example of the division of labour (the same could be said about the TV example) is phrased in the following way: ‘[King] recognises the same facts of the case as Archer and myself … where we differ is on the question of whether these facts entail that the group as such has causal powers in its own right’ (2007a: 37–8). they are utterly banal) precisely because they do not necessitate the MC. nor do they show that it should be extended to other cases as well. Elder-Vass’s claims.Some Problems with ‘The Problem of Structure and Agency’ 143 The picture that emerges from looking at the uses of ‘the organisation’ and other relevant expressions in a way that ‘respects the logic of social situations’ (Coulter 2001: 46) is light years apart from the descriptions generated by ElderVass’s theoretical scheme and the understanding they afford. what is at stake is . In fact. It is important to note. 6. namely the organisation.6 Conclusion: Beyond ‘The Problem of Structure and Agency’ By way of conclusion. first. Let us first soberly reflect on Elder-Vass’s claim that social structures may possess exclusive causal powers and consider what is at stake in the contention and what would somebody who denied it be denying. This is not because there need be a general problem with induction as a procedure but because there are no grounds to justify induction in this case. continuously with the arguments provided in Chapter 5. Examples could potentially warrant an extension or generalisation of the scheme but that depends on the scheme constituting an improvement over the commonly available understanding of chosen cases. that there is also the sales assistant as individual who minimally sells it to me and thus is a co-determinant. most importantly. In the absence of any such improvement there is not much to recommend widening the scheme’s scope. that even in instances where the examples are accurately represented they still do not warrant the unrestricted application of the emergentist scheme. that there is a higher-level entity.

. 24 Contrary to Elder-Vass but also King who speaks about ‘conceptual convergence’ and ‘ontological divergence’ in his response. On the contrary.25 In other words. of more 23 Thanks to Wes Sharrock for this point. it explains nothing.144 Clarity and Confusion in Social Theory not an additional factual issue but rather a notational one23 which means that the ‘denial of structure’ Elder-Vass attributes to opponents is really the denial of the critical realist notation. the cases are commonly understood and adopting the proposed scheme does not enhance our understanding of them. buying. They are not such that can be dispensed with. It changes nothing. it adds nothing to our understanding of each case and. as I have repeatedly stressed throughout this book. many other social theorists is placing the cart before the horse when he says that ‘[o]nce we have understood the various entities and powers at work. some might worry at this point. but what we do not need to say. Moreover. it is not because of social theory that we can understand society any more than it is because of logical theory that we understand logical relations in our discourse. that we can see some sense in the theoretical scheme Elder-Vass is proposing. i. 99). given our ordinary techniques and information’. it should be stressed. the notation is dependent for its sense on the commonly available. they constitute our very phenomenon of interest (see Hutchinson et al. as well as. 2008: 77. there is not an explanation missing from the described cases that is provided or facilitated by the notation.e. by converting the way we do talk to the form proper to causal talk. recognisable ways of describing the situation (that both Elder-Vass and his critics resort to) and the understanding those ways can afford. The proposed emergentist notation is language idling. in other words. it is only because we can employ concepts such as sales assistant. (1966: 108) 26 These ways are implicated in the very identification of the explanandum. of the stipulation that we should talk about a higherlevel entity with exclusive causal powers (cf. in this case there is no actual gain from the long and arduous process of redescribing examples we already understand and possess the means to talk about in the unrealistically restricted terms of the relational emergentist ontology. 25 Compare this to Alfred Louch’s immortal remark that so much of psychology ‘is a matter of what we could say. following Winch (2008[1958]: 125–7). what is to be made of structure and agency. Elder-Vass like. despite the conceivable gains that an informal logic and a form of social enquiry that can do justice to our forms of reasoning and understanding can bring. the portrayal of the issue as ontological will not do: there just is no ontological question distinct from the conceptual question. If all this is true. thus. we can then begin to understand social events by examining how these different powers of different entities interact to produce them’ (2007b: 475). broadening the scope of the argument somewhat. But as I have tried to argue in various places in this book. selling and so on. organisation.26 Thus. Moreover.24 We can indeed decide that from now on we will talk in the manner favoured by Elder-Vass. Wittgenstein PI: §402). on the contrary. But there is no reason why we must talk like this for there is no factual question at stake that hinges on adopting the new way of speaking. .

the pervasive . then this is what is called for. On the other hand. But it does call upon us to rethink whether the issues really amount to what we thought they did and whether our ways of dealing with those issues are adequate. Similarly. 3). is the above meant to show that it is impossible for social theory to constitute an improvement over commonly available understanding? And what about the issue of exclusive causal powers. Clarity is needed regarding which claims can be decided empirically and which are rather conceptual in nature. These ways of elucidating concepts have also worked well with ethnographic approaches (for example. of this book has not been to flatly deny that what theorists hold dear is important.. So how are we to deal with the issues? As a first step to answering that question. how we practice theory and what could be wrong with it. it is always possible to choose some part of the causal conceptual terrain and argue for its propriety as a model of causality. by extension. The evidence we have seen does not force us to abandon the issues that puzzle us or we think are important.Some Problems with ‘The Problem of Structure and Agency’ 145 generally used social theoretical procedures of redescription? Are we to believe that theoretical debates have been in vain? Even worse. I would like to sketch what it would take to move beyond some of the problematic features that render theoretical debates irresolvable. Returning to the procedures that were identified in the introduction to this chapter and examined in subsequent sections. it is always possible to insist on the a priori propriety of a form of explanation. its point has been to avoid pronouncements about the importance of structure and agency or the importance of theorising and to show us instead in detail how we argue about structure and agency. does it not make a difference.. Suffice it to say for now that none of these conclusions are necessitated by the arguments provided. In so far as these latter claims centre on existing concepts (as is the concept of organisation) they can be addressed via conceptual elucidation of the kind OLP engages in. Social theory should avoid starting from a fixed ideal of what the form of explanation ought to be and start instead from a thorough specification of the explanandum. with regard to corporate responsibility? And is that not important? I will take up some of these worries in the Postscript to this book. b. i. i. but to get us to reflect on what we do without prejudice.e.e. I would like to offer the following as alternatives: a. of what it is that we do not understand and which aspect of the phenomenon requires explanation. If this is what is called for. In other words. But the mere fact of the diversity of causal concepts which makes such an operation possible entails that it might be more reasonable to replace the idea of a model with the idea of an overview (see Hacker 2007: Ch. for example. Ethnomethodology). This makes sure that there is a background that guarantees intelligibility and a basis from which to judge the question of explanatory form. through detailed examples and accounts of how the concepts are used – this does not mean that somehow we are concerned with words as opposed to reality. The point of this chapter and. Otherwise.

it constitutes what is going on (2008[1958]: 119). of course. If the empirical facts are equally acknowledged by competing schemes they cannot be meaningfully appealed to. not about society. Furthermore. then it needs to be shown that each scheme is connected to opposing sides of a factual question. That is. A systematic absence of understanding cannot be presumed any more than can a systematic inadequacy of available concepts.) can do the job.. it is not a concept invented to explain what is going on when societies are in armed conflict. more importantly. the above alternatives. readily and radically generalisable. If exemplary cases are to be employed to support a scheme and to recommend a theoretical notation over our common ways of speaking or an opposing notation. The point is. for methodological problems need 27 As Winch insightfully points out with relation to the concept of war.146 Clarity and Confusion in Social Theory strategy of legislating on the matter does not work for conceptual issues because it reinstates the differences between parties and. Despite their innocuous ring. Finally. the use of already well understood cases may be a good way of illustrating the scheme but it cannot demonstrate its explanatory power. if a theoretical notation is to be recommended as necessary. I submit. it needs to be shown that after the chosen case has been described accurately and in detail in available ways there is still something left unexplained. but when it is genuinely illuminated by the scheme. the concept in question. i. This fact cannot be claimed to be simply a feature of a problem which is properly registered as methodological. then their use has to take into account the following constraints: Firstly. Clarity is also needed concerning the use of examples in relation to theoretical schemes. c. Our encounter with ‘the problem of structure and agency’ has shown it to be about social theory. . it is reasonable to suppose that an example might support a theoretical scheme not when it can be redescribed in order to seem consistent with the scheme.e. Let me explain. This presupposes that the example was previously (somewhat) obscure. At this point it needs to be remembered that although there is nothing sacred about our common ways of speaking and thus no need to rule out possible changes. are far more radical than social theory is currently prepared to handle as they entail a deep shift in practice and orientation. these ways are not theoretical27 and thus are not overall competitors to a proposed theoretical notation in the way that another notation is. because it pretends that the question of clarity of concept use can be solved by changing the subject. it had mostly to do with the possibility of sociological explanation and with projected theoretical understanding and not with understanding anything which we might have reason to think we misunderstand. if one theoretical scheme is to be preferred over another on the basis of empirical fact. If it is not empirical fact which is to decide then it is likely that conceptual elucidation (see (b.

The alternatives that have been outlined can then be used as stepping stones to a form of social thought that does not begin with or necessarily aims at theory. I have tried to show that once these issues are treated with some conceptual sensitivity they stop making the sense we thought they did. We have thus no reason not to start elsewhere. it is but a direct result of the order of proceeding in social theory. and having relinquished the possibility of any meaningful and detailed discussion of examples that we find ourselves in the grip of the problems making up ‘the problem of structure and agency’: we puzzle over where we are to locate causality. but one that is lucid. form of taken-for-granted part of social theoretical practice. If we begin in this way then it seems to me that we will not end up where the usual order of proceeding had begun. having adopted a theoretical attitude towards language. to begin. and we tie answering these questions to the very possibility of giving an explanation of social affairs. as I have suggested above. with a thorough description of what it is we do not understand while making use of detailed examples and of the unrestricted range of our conceptual resources. even further. if the problematic procedures identified in this chapter are constitutive responses to the issues which are thought to make up ‘the problem of structure and agency’ and. In sum. For what we will have started with are the kinds of things I have used to dispel the perceived need to propose any such scheme in the first place. Instead. what entities there are in the social world with what powers potentially escaping us and whether structures as well as individuals exist. intellectually honest and satisfying.Some Problems with ‘The Problem of Structure and Agency’ 147 to have demonstrable import for substantive issues. We think that we must start by replacing ordinary ways of speaking with a theoretical notation which will relieve us from paying detailed attention to our concepts as they are used. namely in the construction of a theoretical scheme and in debate between opposing schemes. . It is only once we are well down that path. then giving them up reveals radical possibilities.

This page has been left blank intentionally .

causal and explanatory concerns. which depended on a conception of logic as something transcendental and separable from social life. comprising ontological. I have opted not for summary discussion of numerous examples from different kinds of social theory. After a preliminary exposition of OLP and its relevance to sociology. cause. I began this book with the contention that doing social theory consists primarily in assuming a theoretical attitude towards language and only derivatively in vindicating any proposition about social phenomena. only critical realists are susceptible to. In the remaining chapters I have tried to trace the problematic pattern of relating to concepts through the various procedures and ways of reasoning that make up the convoluted core of social theory. I then proceeded to explicate what seeing logic as embedded in social life entails for the elucidation of the logical features of our concepts in a way that will replace confusion with clarity. a way which is distinctive to OLP and depends on its eschewal of the theoretical attitude towards language. among others.Conclusion Throughout this book I have been documenting and detailing the social theoretical reasoning that follows from the adoption of a theoretical attitude towards language. . followed by responses to various objections. I tried to locate the pattern of unrealistic restriction and of systematic refusal to pay attention to concepts in Hempel’s problem of scientific explanation. the theoretical attitude is embedded in the very premisses of social theoretical enquiry. This is why. hoping that in this way the pattern of logical incoherence will be understood by readers. who will also be able to spot it elsewhere too and to adequately respond to it. as I have claimed. but rather with confusion that relates to the fundamentals of social theoretical practice. that is. but for detailed examination of a narrow range of theorists. My main purpose has been to render explicit the resulting logically deficient pattern of relating to our concepts as well as the alternative pattern OLP can help us establish. the conceptual confusion it creates regarding. We are now in a position to understand more fully why doing social theory is primarily adopting the theoretical attitude. the present book has not been concerned with countering confusion that. the concepts of explanation. for example. existence and observation. In one sense. it is adopted as a first step and built into any subsequent investigation. The theoretical attitude is accompanied by the insistence on the introduction of a ‘technical’ vocabulary which was seen to be completely misguided as it directs attention away from the real problem. debate or substantive proposition. for instance. reality. Since these issues are interrelated. complex and thus difficult to untangle.

study it and practice it themselves. I hope to have done enough in this book to show that concepts need to be taken seriously and handled appropriately and to have presented OLP in such a way that social scientists will be able to understand it. though profound. instead. the problems can be dissolved. a number of social theoretical problems would cease to exist. have not been given sufficient attention by social scientists. then. Finally. To avoid potential misunderstanding. I should like to stress that at no point have I claimed nor wish to claim that what goes on within the social sciences should be exclusively identified with ‘conceptual questions’ or ‘matters of logic’. further. it is hoped. however. Thus. Moreover. Accordingly. Moreover. not least in that it presupposes distinctions that cannot be connected nor do justice to the conceptual distinctions reflected in how we intelligibly and intelligently speak about social life. . but only that since the social sciences must relate to a ‘social world’ that is already conceptually and logically structured and whose intelligibility is endogenous – to allude to Harold Garfinkel’s way of putting it – such matters are bound to be extremely consequential ones. All of the above. Careful attention to the concept of explanation shows that there just is no intelligible question about what general form can be explanatory. in a position to realise that ontological schemes are not the kinds of things that can augment or systematise our understanding. it stands to reason that research which claims to be ‘theoretically informed’ will make use of ways of describing its materials and findings which rely on theoretical schemes that encourage the adoption of precisely such an attitude. upon removing the obstacles and restrictions imposed by the theoretical attitude and. also able to see that the ‘problem of structure and agency’ is a spurious one. can be said to demonstrate the radical impact that OLP can have on the ways in which we understand and debate social theory. we are. careful examination of the theoretical attitude as a source of misunderstanding has revealed the possibility of a lack of any controversial proposition underlying theoretical debate: controversy can be seen to have arisen not because of disagreement over matters of fact or method but due to conceptual confusion. The question whether social structures exist or are causally efficacious loses its grip on us once we understand where to look in connection with ‘social structure’ expressions and expressions such as ‘real’ and ‘exists’. The consequences they may have. and once we understand the varieties of causal concepts. moving freely by making full use of our linguistic mastery in elucidating the logic of our language. For it is this very attitude which brings into being the problems I have scrutinised and bestows on them their peculiar character.150 Clarity and Confusion in Social Theory In another related sense. the attempt to decide which concepts should be used in explaining social affairs before an explanandum has been specified in detail appears pointless. were it not for the adoption of the theoretical attitude. We are. although I have not set out to document how the theoretical attitude plays a part in empirical research.

I will suggest. that some of these forms lead us to misunderstand the nature of these kinds of questions as well as the connections that social science (and philosophy) can have to those questions. so I want to end this book by indicating in somewhat more detail how the line of thought developed herein. It should be obvious. This form of liability has been gaining momentum in the past 20 years or so as a way of responding. The temptation here is to presume that there is a straightforward connection. to use the importance that may attach to moral or political worries to argue for the importance of social science. the response identifying the issue as important is not itself an important response to the discussion. it is only a superficial understanding of what Elder-Vass is proposing that sees it as necessary. after discussing debate on ‘structure and agency’ and Elder-Vass’s treatment of organisations and their members. for example. on the one hand. especially in connection with the question of holding organisations criminally responsible. that to dispute the perceived connection to the issues is not the same as disputing the importance of the issues. or even very relevant. I will begin with the idea. Since I cannot discuss various possible connections at any satisfactory length. . my aim in what follows is limited to providing some material for reflection. ‘that social and political criticism requires a foundational theory of individual and social “ontology”’ (1999: 2). their striving to be a source of critique. however. can be extended to cover social scientific and philosophical forms of reasoning about legal. is not the question of corporate responsibility important? We may certainly be justified in claiming that being able to hold corporations accountable for their conduct is important. In other words. I will try to give a brief characterisation of the issues around corporate responsibility in order to help us appreciate what they are. This allows social scientists. namely. Conversely. discussed much more fully by Nigel Pleasants. political and moral issues. that social theoretical forms of reasoning produce a misunderstanding of our concepts and a confused relationship to our subject matter.Postscript It may be said that the preoccupation with logical matters in this book leaves out one of the central preoccupations of the social sciences. somewhat symmetrically. to any arguments in favour of corporate responsibility. However. I have spoken of the compatibility of logical matters with political concerns in Chapter 1 but it might be objected that compatibility is not enough. namely. we encountered the temptation to ask: but are not the issues important. At the end of Chapter 6. a morally and politically progressive force in society. it is only a superficial understanding of corporate responsibility that sees its possibility as depending on social science’s identification of causal powers operative in the social world. to corporate involvement in work-related ‘accidents’.

on the other. such as those included in the Health and Safety Act (1974). Greece) and common law (England. as Gobert (2008) argues. Before applicability of criminal law to legal persons. along with the form of organisational activities and objectives are what determines the type of sanctions (or penalties) that can be applied. where apart from civil law there may also be a role for criminal law (Eijsbouts 2011). to soft law and to hard law. Moreover. Legally conferred rights and duties. one which has a sociological flavour to it in that it puts the blame on. Khanna 1996). the question of CCR does not have to do with whether it is possible to hold organisations themselves responsible for something as opposed to their individual members. what may be described in the UK as regulatory offences. In the UK. even further. to the complex structure many organisations operate under. Furthermore. United States) traditions as well as among individual countries that have to do with enforcement. which may prevent locating an individual or set of individuals as responsible or. that there is not any strong sense in which criminal law can treat corporations as ‘social structures’ in the way social theory understands them. for instance. for example. can be considered as a matter of regulatory effectiveness. the difference between civil and criminal liability is not. Treating the issue of introducing or extending CCR as a function of the effectiveness of criminal law serves to connect it to the wider discussion of how to get corporations to adopt Corporate Social Responsibility (CSR) polices. but whether it is desirable. But. there are important differences between civil law (Germany. in a sense. from voluntary self-regulation.152 Clarity and Confusion in Social Theory large-scale environmental. one might say. a difference in type of sanction but rather in its characterisation as civil or criminal. which also bear on questions of effectiveness. nevertheless: it may be that this development is not an especially progressive one. financial and other kinds of disasters and. Apart from fines and other financial sanctions. This may seem like a progressive development and. Appearances can be deceptive. where questions of effectiveness dictate the variety of types of regulation needed. questions about the very nature of criminal law and its suitability to cover legal persons need to be . the offence of Corporate Manslaughter (in England and Wales) and Corporate Homicide (in Scotland) has come into existence with the relevant Act in 2007. more generally. there has been a wide ranging discussion throughout many jurisdictions concerning the introduction or extension of Corporate Criminal Responsibility (CCR). It is already well-enshrined legal practice to hold corporations responsible through enforcement of civil liability with regard to. when that is possible. administration and procedure (Wells 2001: 128. and. social structures. such as corporations. forcing it to restructure its operating policies and subjecting it to adverse publicity (Wells 2001: 37). non-financial ones include dissolving the corporation or restricting its actions in certain ways. as already intimated. may point the law only towards low-level employees. Wales. effective or consistent to hold organisations responsible for criminal offences and what range thereof. such as those mentioned in Chapter 6. Given the existence of civil liability for corporations.

in the sense in which a human being is punished. moral disapproval and remorse. as opposed to its members. This raises important conceptual considerations. for example. This observation. that ascribing criminal liability to the organisation entails (Kopanidis 2013).). captured in the principle nullum crimen sine culpa which. is parasitic on the sense in which these notions apply to human beings. benevolence. extended to members of the organisation conceivably having nothing to do with the offence in question. Central operative concepts in criminal law. being found culpable and being punished. also explains the reservations expressed by legal scholars in countries such as Greece regarding the unacceptably diffuse responsibility. and with the establishment of individual culpability. given that many corporations have brand images under which they may appear as having moral virtues. There is no such thing as punishing ‘the organisation’. The legal form ‘the organisation’ is a suitable logical subject for such predicates only to the extent that its members can be substituted for it. punishing. While it may make sense to speak of directing moral disapproval at the organisation. as do supporting concepts such as blame. and for a different range of concepts. Accordingly. such as act. a form of criminal punishment can be said to be produced only to the extent that the organisation’s members can be said to feel the moral force of the punishment.5 in which we speak of ‘the organisation’. The discussion so far has tried to exhibit the nature of the different issues around corporate responsibility. culpability and punishment apply to natural persons in a paradigmatic manner. criminal offences are predominantly mens rea offences. Pieth and Ivory (2011) include the following: (a) attributing liability vicariously to the organisation through the actions of its members – usually for strict liability offences. The issues are misunderstood if taken to furnish social theory with the task of demonstrating that . which correspond to some of the ordinary ways I discussed in section 6. the conceptual parasitism at play. guilt. Despite differences among legal systems. (c) ascribing liability to the organisation by aggregating the states of mind and actions of its members or by establishing that it has ‘inadequate organizational systems and cultures (corporate culture. in turn. apart from removing any grounds for talk of including ‘social structures’ as the ratione personae of criminal law. which do not require mens rea (Wells 2001: 84ff. it is clear that criminal law is strongly connected with liberty-limiting penalties. requires the establishment of the mental element in crime. for example. with emphasis on criminal responsibility. have arisen to facilitate responding to the variety of cases that may be brought before legal attention and which may call for pressing charges against the organisation and/or for separate charges brought against its members. corporate (dis)organization)’ (Pieth and Ivory 2011: 22). an organisation in the sense in which punishment is understood in criminal law needs to reflect as a conceptual matter the punishment of its members. These models. (b) identifying the organisation with its senior managers and their deeds and states of mind. Hence. The models used across jurisdictions for attributing liability to the organisation show under a different light. and so the sense of an organisation itself acting. such as imprisonment.Postscript 153 addressed.

are more likely to be made at times of economic prosperity when xenophobic sentiment is also likely to be rather subdued. enforcement and administration are important givens to be taken into consideration. for example. This helps us to see what role the facts of the contribution of immigrants to the economy can have. One of the ways empirical social science can be used to shoot down such campaigns is by attacking their bases in claims about the contribution that immigrants allegedly do not make to the society in question. human rights. depending on the case. Having examined a perceived connection between social theory and issues we might consider politically important. anti-immigration campaigns. right-wing Eurosceptic parties have made gains by waging xenophobic. humanity. as is the current period for many European countries. their economic non-contribution. morally important. the richness that can be found in different cultures. that is.154 Clarity and Confusion in Social Theory either the organisation as a social structure or structures within the organisation or the organisation’s culture exist as causal factors to which responsibility can attach. Even so. Compare now the case of immigration during a period of economic crisis.? And does not the fact that we have to rely on such notions even in relation to possible economic facts allow that it will be possible to read the significance of these facts differently: to say that the wealth produced by immigrants is an appropriation of the wealth that rightfully belongs to the country’s inhabitants or to say that what really matters is that foreign culture can corrupt our cultural forms? Readers may ponder the following questions in this connection: do anti-immigration politicians and their . across a number of countries. is the absence of a positive argument something that could morally justify anti-immigration policies any more than its presence at a different time could justify pro-immigration policies? Would we not have to resort in both cases to other ways of justifying our attitudes and policies towards immigration. although there is space for clarification of various conceptual issues regarding how exactly they can be said to be so. solidarity. But it need not be. The variety of possible cases and the differences among legal traditions and jurisdictions when it comes to substantive questions as well as questions of procedure. to appeals to. That organisations can be responsible in various senses is something commonly understood. which might help to address some people’s anxieties about immigration. I am writing these lines not long after the European Elections in 2014 where. etc. They complicate matters. I want to turn to the empirical dimension of social science – granting for the sake of argument that social science can be characterised as a strongly empirical form of enquiry – and examine how that dimension may be connected to issues we might consider politically or. and by arguing that far from taking away from the country’s economic prosperity they are adding to it. especially. For in this case we would have to entertain the possibility that not only might there not be a positive argument to be made about that contribution but that there could even be a negative one. and that they may in fact be in particular cases is a matter for the courts. as they should. and deny any strong connection between arguments for corporate responsibility and social theory. It is true that such arguments.

presuppose such conceptions and do not establish them. that is not something which depends on empirical reason. to business people. What may be hoped from studies of this kind is that they bring us in contact with marginalised voices and that in hearing such voices we find that they speak to our moral sensibilities. at the cost of some repetition. under some of which what goes on may not even be appropriately described as ‘giving reasons’. Empirical reason does not go very far in moral matters. To develop this point somewhat further I will briefly consider a form of social scientific reason which draws on moral philosophy in seeking to theorise moral points of view and conceptions of the good. But we have to be clear about what it is that speaks to us and how it does so. Moreover. Studies conducted under such a conception may strive to make the experiences of marginalised groups available to us. namely conceptions of what it means. which might be better viewed as calling for a wholly different range of forms of reasoning. at whichever level of social life. I would like to point out. if we accept that academic social science is indeed accorded a higher status when it comes to its contribution to governmental policy or political debate than is accorded. To say that empirical social science cannot show a moral point of view to be wrong or right is not to say that in practical terms it does not or cannot play a part in influencing policy or political debate. as I will suggest. Empirical social science. can provide support for a moral point of view in the sense that it can produce empirical arguments that are articulated from and used in favour of the moral point of view social scientists adopt. is it because they get their facts wrong that we might think that they are wrong? And. I should stress at this point that there are a number of ways in which social science studies may combat xenophobia and place themselves in the service of tolerance that may be removed from a picture that sees fact-finding as the strength of social science and rather closer to one which sees social science as related to literature. in what sense can we speak of using social scientific methods to prove them wrong? The existence of relative moral agreement among social scientists obscures what the moral load-bearing elements really are. because we accept empirical statements as true. Whether such status is legitimately accorded is. for example. Social science cannot show a moral point of view to be the right one for. The point I have been trying to make is that a conception of the social sciences as a form of fact-finding empirical enquiry necessitates that moral points of view or conceptions of the good are presupposed rather than established. if we do. Rather. arguments for factual connections between immigration and certain societal benefits.Postscript 155 supporters who express such views have their facts wrong? Even if they do. but that it can be said to do so by trading in empirical ways of supporting moral causes. it can support a moral cause in the additional sense that it may bring whatever status it is seen to have by politicians or the public to back the relevant cause. to be a xenophobe or a racist and of what it means to see immigrants as fellow human beings. We do not accept these conceptions. a different question. journalists or members of the public. of course. not by trading in moral causes. for example. that the very idea of a theory of these matters is completely . then.

like talking. then. 15) and Raimond Gaita (2000. which she then used to articulate and ‘justify’ a moral point of view: I saw that education was a road down which every child had to walk. are manifestly problems of this kind. was where the concept of ‘special needs’ came from. 2004) shows. Far from achieving anything of the kind they do violence to our thinking about moral issues. but there was only one road. in effect. often to do things. whether to let a runaway trolley crash into a group of five people or redirect it so that it crashes into one person. who in 1977 chaired a government enquiry into providing education for children with ‘special educational needs’ (1998: 14). I would like to cite the words of Mary Warnock. 2). enjoy and control. (1998: 15) I do not want to suggest that one has to find this picture or metaphor morally compelling but only that it is the kind of thing that might speak to our moral sensibilities. The implication. Some would emerge fully equipped to understand. thinking and writing about moral matters is something which has much closer connections to art and literature than to the academic forms of writing and reasoning utilised . At the end of the road lay the common goals: understanding of the world.156 Clarity and Confusion in Social Theory confused. the kind of thing we might use to cultivate those sensibilities as well as communicate them to others with moral force. ‘Trolley problems’. To provide a contrast to this picture of what it means to reason about what is right or wrong. As Allen Wood (2011) insightfully points out. trolley problems work by stipulating a range of acceptable responses to a morally challenging situation while leaving out information about aspects of the situation which we might consider as morally relevant thus. she describes a vision she had. enjoyment of it. which other children learn to do without effort. Trolley problems are made use of in moral philosophy and political theory in order to elicit our ‘moral intuitions’ about the right thing to do. not only conceptual violence but moral violence. too. In their general form they require a decision regarding a morally challenging situation. importantly. with a common destination. As I think the work of Cora Diamond (1991: Ch. Others would progress very little towards the shared goals. and the independence that would empower the child to control it. presupposing what the morally relevant responses and facts are. These children would have to be helped to surmount the obstacles one by one. but for others there would be formidable obstacles. and the need for help. for example. named after the one originally introduced by Philippa Foot. in other words. that we have to choose between the options offered and that we ‘have been given enough information to settle the question is offensive’ (Wood 2011: 74) and produces corrupt forms of moral thinking (Gaita 2004: Ch. as it induces us to begin by adopting a theoretical attitude towards our moral concepts and to produce unrealistically restricted problems along the pattern we have seen throughout this book. the kind of thing that ‘reasoning’ about moral matters might amount to. Below. But in their case every step counted. Some children would have a smooth and easy passage down the road.

and that is something pertaining to persons. is of course not limited to the kinds of theoretical concerns that I have dealt with in this book. and whether the forms of reasoning we resort to are well-suited to these worries. For the moment the best I can do is to invite you to carefully consider what the intellectual. moral and other kinds of worries are that may lead us to take part in the social sciences. Rather it is the antithesis between understanding the subject and what most people want to see. important – is not that it would take some special instruction about abstruse things to understand it. are we more knowledgeable. to be able to reason about moral matters with any authority requires that one’s experience and the way one lives one’s life be brought into the picture.Postscript 157 in either the social sciences or philosophy. then I have tried to assist the will by showing that what we thought as necessary only produced more confusion and could thus not deliver what we wanted. Because of this the very things that are most obvious can become the most difficult to understand. at least in large parts of the English-speaking world. are we any closer to addressing what troubled us. Besides. And to ask in honesty: are we satisfied with our enquiries and what they have yielded. This implies that as things stand. neither the social sciences nor philosophy can claim to be particularly well-equipped to either produce or understand moral thought or otherwise claim any particularly strong connection to it. what I have tried to do in this book is to go against our wanting to turn away from our concepts and the ways they enter into our social life while fixing our gaze on theorising in the hope that it will bring understanding. (BT: §406–7) Seen in the light of Wittgenstein’s remark. By way of conclusion I would like to consider a remark Ludwig Wittgenstein made regarding the kind of difficulties philosophy presents which I think is also very pertinent to the social sciences: What makes a subject difficult to understand – if it is significant. but of the will. I hope to be able to come back to what those are in the future. not disciplines. For better or worse. What we wanted. are we any wiser? . what we call ‘the social sciences’ provide a shelter for a variety of concerns and interests. what social scientists may want from or see in their subject. If turning our attention back towards the logic of our language is a matter of the will. What has to be overcome is not a difficulty of the intellect.

This page has been left blank intentionally .

Harvard: Harvard University Press. T. S. . Radical Philosophy 13: 1–6. Bernstein. Social Research 32: 239–55. R. A.co. Noûs 13(2): 219–33. London and New York: Routledge. New York: Columbia University Press. M. (1998[1979]) The Possibility of Naturalism: A Philosophical Critique of the Contemporary Human Sciences. Bhaskar. Baz. Badiou. (2012) When Words Are Called For: A Defense of Ordinary Language Philosophy.uk [Accessed: 10 June 2014]. Cambridge: Cambridge University Press. Archer.L. (2010) ‘Routine.L. Wittgenstein and Marxism’. R. Archer. T. R. (1976) The Restructuring of Social and Political Theory. Bhaskar.J.References Anderson. and Searle. and Hacker. and realism’. H.S. Becker. (1979) ‘Under a description’. Available at: http://www. Bennettt. Anscombe. Sociological Theory 28: 272–303. so called’. Metaphilosophy 3: 1–24. London: Routledge.. D. 3rd ed. Cambridge: Cambridge University Press.E. Neuroscience & Philosophy. (1972) ‘More on what we say’. Oxford: Blackwell. (1999) ‘The Chicago School.J. The American Journal of Sociology 59(3): 235–42. R. Qualitative Sociology 22(1): 3–12. Archer. J. (1979[1961]) Philosophical Papers. A. Oxford: Oxford University Press... J. Bates. (1976) ‘Winch. J. M. (1965) ‘The concept of organisation’. Hacker. (1953) ‘Becoming a marijuana user’. 127–62. London: Verso. Oxford: Clarendon Press. E. Dennettt. P. (2000) Being Human: The Problem of Agency. Bittner. M. Archer. and Sharrock. (2011) Wittgenstein’s Antiphilosophy. M. Benton. Agency and the Internal Conversation. (2008[1975]) A Realist Theory of Science. (2003) Structure. G. 3rd ed. reflexivity. M. (1995) Realist Social Theory: A Morphogenetic Approach. in Bennett. Austin. W.M. (2013) The Methodology of Analytical Sociology: The Challenge of Creating a Robust Empirical Social Science [Unpublished]. (2007) ‘The conceptual presuppositions of cognitive neuroscience: A reply to critics’.sharrockandanderson.S. Becker. Austin. P. pp. Cambridge: Cambridge University Press. and Cohen. M. H. (1962) Sense and Sensibilia.

pp. Human Agents and Social Structures. (1978) ‘Whorf’s linguistic relativism I’.A. J. G. E. C. 166–89. in Martin. Cambridge: Cambridge University Press. (2002) ‘People power: Harré and the myth of social structure’. Philosophy and Linguistics. Manchester: University of Manchester. and Read. in Lyas. (1989) ‘Head colds and thoughts in the head’.). Dordrecht: Kluwer Academic Publishers. and the Tractatus’. London: Penguin. J. J.B. and Sharrock. Capaldi. London and Basingstoke: Macmillan. Braithwaite. D. L. (1953) Scientific Explanation. (2000[1978]) Karl Marx’s Theory of History: A Defence. (ed. J. O. (eds). 20–50. European Journal of Social Theory 5(1): 134–42. A. Brown. B. (eds). Coulter. Burke. (1998) Alice’s Adventures in Wonderland & Through the Lookingglass. N. in Crary.S. (1983) ‘Contingent and apriori structures’. The New Wittgenstein. Coulter. J. Mind 14(4): 278–80. (1991) ‘Logic: ethnomethodology and the logic of language’. Kant. Cook. Cook. Carroll. Human Studies 6(4): 361– 76. (1982) ‘The illusion of aberrant speakers’. Carroll. (1993) Language and Social Life [PhD thesis]. L. (1975) ‘Wittgenstein’s conservatism’. (1949) ‘Descartes’ evil genius’. in Button. Conant. J. S. and Dennis. Cavell.K. J. (1955) ‘Misconceptions of inference’. expanded ed. Cohen.J. Champlin. Manchester: Manchester University Press. A. Ethnomethodology and the Human Sciences. Philosophical Investigations 5(3): 215– 26. Cambridge: Cambridge University Press. T. (1971) ‘The availability of Wittgenstein’s later philosophy’.). pp.160 Clarity and Confusion in Social Theory Bloor. Conant. P. (1998) The Enlightenment Project in the Analytic Conversation. . Radical Philosophy 10: 27. Frege. New York: Columbia University Press. pp. Button. R. (2010) ‘The structure problem in the context of structure and agency controversies’. Carter. 174–217. Cook. R. Bouwsma. Coleman. pp. Analysis 15(6): 135–44. G. (1983) Wittgenstein: A Social Theory of Knowledge. 17–33.G. W. (1895) ‘What the Tortoise said to Achilles’. Philosophical Investigations 1(1): 1–30. Philosophical Topics 20(1): 115–80. (1980) ‘The fate of ordinary language philosophy’. (1991) ‘The search for logically alien thought: Descartes. The Philosophical Review 58(2): 141–51. W. Philosophical Investigations 3(2): 1–72. (2000) ‘Elucidation and nonsense in Frege and early Wittgenstein’. Philosophy 64(247): 39–48. G. (ed. Oxford: Oxford University Press. London: Routledge. D.

(2004) ‘Saying and showing: An example from Anscombe’. in Stocker. Available at: http://faculty. Cambridge: Cambridge University Press. Available at: http://www. Ebersole. Philosophy and the Mind. De Certeau. 37–49. Bloomington.) (2011) Analytical Sociology and Social Mechanisms.com [Accessed: 17 March 2011]. D. Elder-Vass. and Von Savigny. . Diamond. F. J. and Sharrock. J. Dray. Eijsbouts. Oxford: Oxford University Press. F. B. KnorrCetina.htm [Accessed: 15 June 2014]. W. Elder-Vass. (2009) ‘The realist approach to explanatory mechanisms in social science: More than a heuristic?’. Diamond. (2003) ‘Who killed ordinary language philosophy’. Lewiston: The Edwin Mellen Press. MA: The MIT Press. Cruickshank. Philosophy of the Social Sciences 39: 440–62. K.. Bloomington. pp. Radical Philosophy 12: 25–9. Cruickshank. Cambridge. F. (1984) The Practice of Everyday Life. C. V. pp. (2004) ‘A tale of two ontologies: An immanent critique of critical realism’. London: University of California Press. Journal for the Theory of Social Behaviour 37(1): 463–77. F. Demeulenaere. W. (2006) ‘A method for social ontology’. M. beyond voluntarism: Regulatory options to reinforce the licence to operate’ [Inaugural lecture]. The Sociological Review 52(4): 567–85. 2nd ed. Maastricht University.frostburg. 2nd ed. (1991) The Realistic Spirit: Wittgenstein.R. The Practice Turn in Contemporary Theory. Mind and Human Behaviour in Contemporary Cognitive Science. Philosophy of the Social Sciences 40(4): 579–602. Indiana: Xlibris Corp. 2nd ed. (2011) ‘Corporate responsibility. (ed. Post-analytic Tractatus. in Schatzki. eldervass. Elder-Vass.References 161 Coulter. T. P. London: Routledge. (1975) ‘Ordinary language and radical philosophy’.php [Accessed: 5 June 2104]. (1957) Laws and Explanation in History. J. Demetriou. D. J.l4bb.). (2007a) ‘For emergence: Refining Archer’s account of social structure’. Journal for the Theory of Social Behaviour 37(1): 25–44. E. New Blackfriars 41: 74–80. (eds). (2007b) ‘Social structure and social relations’. D. (2001[1967]) Things We Know: Fifteen Essays on Problems of Knowledge. Coulter. Ebersole. (2002b[1979]) Language and Perception: Essays in the Philosophy of Language. Ebersole. (ed. Di Norcia. Bloomington. M. (2002a[1979]) Meaning and Saying: Essays in the Philosophy of Language. (1960) ‘Oxford philosophy’. Available at: http://www. Aldershot: Ashgate. (2007) Brain. Dummett. J. Ebersole. Indiana: Xlibris Corp.org/pages/Resources.edu/phil/forum/OLP. C. 151–66. (2001) ‘The observability of the macro-social’. Indiana: Xlibris Corp. C. (2010) ‘Knowing social reality: A critique of Bhaskar and Archer’s attempt to derive a social ontology from lay knowledge’.

(1977) J.T.C. London and New York: Routledge. J.A. J. (1981) Forms of Explanation. H. J. R. Lewisburg: Bucknell University Press. W.M. Fodor. Oxford: Oxford University Press. M. and Katz. A. London and New York: Routledge. W. Structure and Contradiction in Social Analysis. Hacker. H. A. Gaita. Radical Philosophy 8: 24–7. Friedman. (2004) Good and Evil: An Absolute Conception. Theories of Explanation. reprinted in Pitt. (1986) Society. Cambridge: Polity.S.C. and Tiryakian. in Tait. Gaita. E. 2nd ed. J. (1974) ‘Wittgenstein and bourgeois philosophy’. IL: Open Court. Fann. Elder-Vass. 75–89. D.162 Clarity and Confusion in Social Theory Elder-Vass. Berkeley and Los Angeles: University of California Press. (1963) ‘The availability of what we say’. (1969) ‘The availability of ordinary-language philosophy’. (2009) ‘Concepts. Garfinkel.A. Brighton: Harvester Press. Gunnell. Graham. J. (2007c) ‘Reconciling Archer and Bourdieu in an emergentist theory of action’. Cambridge: Cambridge University Press. (2004) First Word Philosophy: Wittgenstein–Austin–Cavell. Philosophical Review 72: 57–71. J. Oxford: Blackwell. Frisby.G. Writings on Ordinary Language Philosophy. 337–67. K. H. (1984) The Constitution of Society: Outline of the Theory of Structuration. (1970) ‘On formal structures of practical actions’. New Haven and London: Yale University Press. (1997) ‘Wittgenstein on fixity of meaning’. Sociological Theory 25: 325–46. pp.W.L. D. Friedman. and Sayer. pp. (2007) Explaining Social Behaviour: More Nuts and Bolts for the Social Sciences. Man and World 2: 410–22. pp.) Early analytic philosophy: Essays in Honor of Leonard Linsky. P.-J. 2nd ed. 188–98. New York: Appleton-Century-Crofts. Elster.L. Fleming. (1979) Central Problems in Social Theory: Action. New York: Palgrave Macmillan. in McKinney. D. K. Giddens. (2011) Political Theory and Social Science: Cutting Against the Grain. . Theoretical Sociology: Perspectives and Developments. Philosophia 37: 653–68. Cambridge: Cambridge University Press. P. Structure and Agency. conceptual schemes and grammar’. Chicago and La Salle.J. Giddens. Austin: A Critique of Ordinary Language Philosophy. Goldfarb. and Sacks. Garfinkel. R. A. (2010) The Causal Powers of Social Structures: Emergence. (2000) A Common Humanity: Thinking about Love and Truth and Justice. (1998[1974]) ‘Explanation and scientific understanding’. (ed. R. (eds). (1996) Wittgenstein’s Place in Twentieth-century Analytic Philosophy. Glock. New York: Tavistock. D.

(1956) ‘Analytic truths’. London: Routledge. 139–60. P. K. Hart. Harré.L. 219–54. Mind 65: 359–67. 185–222. O. European Journal of Social Theory. Philosophy 81: 221–43.M. R.S.S. Hacker. European Journal of Social Theory.M. Cambridge: Cambridge University Press. pp. (1960) ‘A School for philosophers’. Hedström. E. in Siebel. London and Basingstoke: Macmillan. Radical Philosophy 8: 38–9. and Cicourel. (2004) ‘Of the ontology of belief’. Philosophical Investigations 7(3): 187–205. (2007) Human Nature: The Categorial Framework. 5(1): 111–23.). Harré. Hart. I. Hanfling. R.M. (1999) The Social Construction of What? Harvard: Harvard University Press. (1984) ‘Scientific realism and ordinary use’. 16 January 2001’. R.References 163 Hacker. The Principles of Scientific Thinking. and Honoré. in Anthony O’ Hear’ (ed. (eds). P. R. Cambridge: Cambridge University Press. Hare. Hacking. 27(2–3): 173–91. Hanfling. Hamlyn. Oxford: Clarendon. . pp. (2006) ‘Passing by the naturalistic turn: On Quine’s cul-de-sac’. 2nd ed.A. Oxford: Blackwell. Harré.M. Hacking. H. T.M. 5(1): 143–48. Harré.L. The Nature of Philosophy. (2000) Philosophy and Ordinary Language: The Bent and Genius of our Tongue. (2002a) ‘Social reality and the myth of social structure’. (2002b) ‘Tilting at windmills: Sociological commonplaces and miscellaneous ontological fallacies’.S. B. in KnorrCetina. (1981) ‘Philosophical aspects of the micro-macro problem’.S. H. Frankfurt: Ontos Verlag. Ratio 2: 107–20. M. London: Routledge and Kegan Paul. M. Harré.W.H. but to human understanding’. not to human knowledge. O. Hacker. (2010) ‘Philosophy: a contribution. (1974) ‘Ordinary language philosophy and radical philosophy’. (1985) Causation in the Law. Semantik und Ontologie. D. P. (2005) Dissecting the Social. (1975) Causal Powers: A Theory of Natural Necessity. (eds). Economy and Society 31(1): 1–14. R. Oxford: Blackwell. R. Harré. Oxford: Oxford University Press. A. (1970). (1997) ‘Forward to Aristotle: The case for a hybrid ontology’. pp. Journal for the Theory of Social Behaviour. (2002) ‘Inaugural lecture: Chair of philosophy and history of scientific concepts at the Collège de France. Harrison.A. Royal Institute of Philosophy Lectures. Hacker. and Madden. and Textor. P. P. (1997[1961]) The Concept of Law. Advances in Social Theory and Methodology: Towards an Integration of Micro and Macro Theories. R. I.

P. (1966) ‘The logic of social contracts’. P. Available at: http://digital. and Sharrock. London and Basingstoke: Macmillan. O. J. (ed. C. Faculty of Law. T. (eds). European Journal of Social Theory 8(2): 171–91. (2004) ‘The relevance of ontological commitments in social sciences: Realist and pragmatist viewpoints’. R. . N. S. May 8. Human Agents and Social Structures. and Ylikoski. Kopanidis.. C. Read. Crows Nest: Allen and Unwin. (2009) ‘Student theorists: Remarks on the educational costs of postmodernism and post-structuralism’. P. A. Kemp. [Master’s Dissertation].F. and Dennis. (1965) Aspects of Scientific Explanation. W. Dialogue 5(1): 31–46. Hutchinson. C. (2010) ‘Beyond structure’. Khanna. (1960) ‘Linguistic philosophy in the light of modern linguistics’. Manchester: Manchester University Press. Annual Review of Sociology 36: 49–67. Journal for the Theory of Social Behaviour 34 (3): 231–48. J. (2007) ‘Why I am not an individualist’.lib.). G. (1981) ‘Explanatory unification’. G. (eds) (2002) Marx and Wittgenstein: Knowledge. Kitching. (1973) ‘Linguistic philosophy. (2007) ‘Exploring the concept of causal power in a critical realist tradition’. in Martin.164 Clarity and Confusion in Social Theory Hedström. P. Journal for the Theory of Social Behaviour 37(1): 63–87. Journal for the Theory of Social Behaviour 37 (2): 211–19. Jenkins. London and New York: Routledge.gr/ record/131649 [Accessed: 1 June 2014]. T.M. Hunter. King. Philosophy 48: 381–5. Philosophy of Science 15: 135–75. Kaidesoja. Philosophy of Science 48: 507–31. P. C. R. and Oppenheim. 109(7): 1477–534. and Piiroinen. (2008) The Trouble with Theory: The Educational Costs of Postmodernism. Aristotle University of Thessaloniki. New York: The Free Press. (2005) ‘Critical realism and the limits of philosophy’. R. Kitching. (2010) ‘Causal mechanisms in the social sciences’. Hempel. A.M. Morality and Politics. Philosophy and Linguistics.J. Hempel. (1942) ‘The function of general laws in history’. pp.auth. A. (1996) ‘Corporate criminal liability: What purpose does it serve?’ Harvard Law Review. in Lyas. Hinton. Herdan. (1948) ‘Studies in the logic of explanation’. V. (2008) There is No Such Thing as a Social Science: In Defence of Peter Winch.S. presented in ‘Issues in the Philosophy of Society and Social Science’ at Manchester Metropolitan University Cheshire. The Journal of Philosophy 39: 35–48. Kitching. Kitcher. Henson. and Pleasants. (1971) ‘What we say’. (2013) The Criminal Liability of Corporations. 133–51. Language and Speech 3: 78–83. pp. empiricism and the left’. G. 204–22. Kivinen. Department of Criminal Law and Criminology. Hempel. Aldershot: Ashgate. P. G.

Dordrecht: Kluwer. (eds). Bhaskar. and Ivory. New York: The Free Press. C. Corporate .T. London and Basingstoke: Macmillan. and Ivory. and Jolley. Cambridge: Cambridge University Press. (2005) Reassembling the Social: An Introduction to Actor–Network Theory. causality and the problem of social structure’. Parsons. Marcuse. Nagel.. New York: Harcourt. L. pp. 2. Critical realism: Essential Readings. Law. and Craver. (1958) ‘On the verification of statements about ordinary language’. (2006) A Realist Philosophy of Social Science. C. badly reinventing oxford ordinary language philosophy’. J. P. E. (1977[1852]) ‘The eighteenth Brumaire of Louis Bonaparte’.References 165 Latour. W. Mind 75: 368–84. (2000) ‘A new disease of the intellect? Some reflections on the therapeutic value of Peter Winch’s philosophy for social and cultural studies of science’. or. Oxford: Oxford University Press. Journal for the Theory of Social Behaviour 30: 249–68. London and New York: Routledge.). Synthese 120: 193–211.. P. (2003) Reorienting Economics: Economics as Social Theory. N. Berkeley and Los Angeles: University of California Press. Boston: Beacon Press. (ed. Louch.T. Machamer.G. 2nd ed. Mouzelis. T. J. B.S.. M. Manicas. K. Lawson. D. Philosophy of Science 67: 1–25. (2000) In Defense of Informal Logic. 300–25. B. (2000) ‘Realism. C. (1966) ‘A plea for linguistics’. in Jolley. H. (ed. (1998) ‘Realism and social science’. Long. R.) (1971) Philosophy and Linguistics. K. (2011) ‘Emergence and convergence: Corporate criminal liability principles in overview’. Lynch. Karl Marx: Selected Writings. T. R. Malcolm. (2010) ‘Grammatical investigations’.D. Oxford: Oxford University Press. Lyas. (1999) ‘Language without linguistics. Wittgenstein: Key Concepts. Darden.D. Marx. (2001) Ludwig Wittgenstein: A Memoir. New. P. Inquiry 2: 161–71. Durham: Acumen. (2004) After Method: Mess in Social Science Research. A. Pieth. R. M.F. Collier. in Archer. Lewis. History of the Human Sciences 13(1): 140–56. Brace and World. (1991[1964]) One-Dimensional Man.R. K. (1995) Sociological Theory: What Went Wrong? London and New York: Routledge. N. (2000) ‘Thinking about mechanisms’. Lawson. London and New York: Routledge. Outhwaite. (ed. R. in Pieth. A. (1961) The Structure of Science: Problems in the Logic of Scientific Explanation. London: Routledge. (eds). and Norrie. in McLellan. (1966) Explanation and Human Action. Mates. T. Leiber. pp. M. vol. 282–96.). Oxford: Oxford University Press. D. M. (1968[1937]) The Structure of Social Action. Levi. A..

. using electronic dance music culture as an example’. Rée. Y. Shorey. 245–63. C. (1989) ‘Four concepts of social structure’. J. (2007) ‘Do we need mechanisms in the social sciences?’. Ryan. Farnham: Ashgate. Pleasants. Plato (1937) Republic. IL: University of Chicago Press. Pitcher. London: Routledge and Kegan Paul. (1999) Wittgenstein and the Idea of a Critical Theory: A Critique of Giddens. The Massachusetts Review. MA: Harvard University Press. (ed. Ryle. Popper.V. (1966) ‘On the philosophical significance of what we say’. (2010) ‘The case for “everyday politics”: evaluating neo-tribal theory as a way to understand alternative forms of political participation. Ryle. (2012) Wittgenstein among the Sciences: Wittgensteinian Investigations into the ‘Scientific Method’. J. Richman. 3–60. Read. . Philosophy of the Social Sciences 37(2): 163–84. Cambridge: Cambridge University Press. R. (ed. 6(3): 591–611. Dordrecht: Springer. G. J. Pitkin. G. K. Rorty.E. D. 1. S. Basingstoke: Macmillan. American Sociological Review XXV (April): 193–208. N. pp. pp. (1960) ‘The English record of a natural sociology’. The Philosophical Review 62(2): 167–86. Journal for the Theory of Social Behaviour 19(2): 195–211. 37(2): 195–200. Pitt. (1981) Marx and Wittgenstein: Social Praxis and Social Explanation. Oxford: Oxford University Press. Rorty.166 Clarity and Confusion in Social Theory Criminal Liability: Emergence.C. Rose. nonsense and Lewis Carroll’. and Morey. and Risk. Rubinstein. (2007) Philosophy as Cultural Politics. R. Sociology 44(2): 345–63.). (1972) Wittgenstein and Justice. P. (2007) ‘On Elder-Vass: Refining a refinement’. (1954) Dilemmas. (1965) ‘Wittgenstein. D. A. (1959[1935]) The Logic of Scientific Discovery. (1953) ‘Ordinary language’. Griffin. G. Habermas and Bhaskar. vol. Cambridge: Harvard University Press. Putnam. Convergence. D. H. H. Australasian Journal of Philosophy 28(3): 137–53. G.R. in Conant. (trans. (1974) ‘Ordinary language philosophy and radical philosophy’.V. Riley.). (ed.) (1988) Theories of Explanation. R. E. (1950) ‘Logic and Professor Anderson’. Cambridge.C. (1970) The Philosophy of the Social Sciences. Radical Philosophy 8: 38. Chicago.) (1992[1969]) The Linguistic Turn. Berkeley and Los Angeles: University of California Press. J. R. New York: Basic Books. Words & Life. Porpora. Ryle. Porpora. London: Routledge. Cambridge: Cambridge University Press. Journal for the Theory of Social Behaviour. (1994) ‘Rethinking mathematical necessity’. Philosophical Studies 17: 21–6. Reiss.

Manchester: University of Manchester. W. 244–60. M.. Scriven. (1987) ‘Individual and society’. Searle. (ed. London: Macmillan. W. London and New York: Routledge.J. in Button.. Segerdahl. 120–43. (1995) The Construction of Social Reality. pp.sharrockandanderson. W. (2001[1970]) Zeno’s Paradoxes. vol. and Slaney K.). (eds). (2005) ‘Peripheral vision: The sites of organisations’. Shapiro. P. (1988[1962]) ‘Explanations. A. (1971[1961]) ‘Use.A. W. New York: Free Press. Hughes. (2006[1989]) Four Decades of Scientific Explanation. (1970) Concepts in Action: A Sociological Analysis of Ideological Activities [PhD thesis]. T.C. J.. M. R. (1998) ‘Abstraction: A realist interpretation’.L. Schatzki. (1991) ‘Epistemology: professional scepticism’. T. (1988) ‘Autonomy among social theories: the incarnation of social structures’. W.R.G. Sayer. Sayer.. C. Sharrock. L. Salmon. (1974) ‘Ordinary language philosophy and radical philosophy’. Cambridge: Cambridge University Press. and Tsilipakos. in Fielding. and Anderson. W. pp. J. W.). T. pp. London and Basingstoke: Macmillan.J. co. 50–76. Classic Disputes in Sociology. London: Sage. Ryle. S. Theories of Explanation. in Archer. in Pitt. A. J. Hughes. in Collected Papers. R. Bhaskar. pp. Oxford: Oxford University Press. (eds). I. R. Available from: http://www. (ed. Salmon. G. .R. A. (2000) Realism and Social Science.uk [Accessed: 5 April 2014]. R. 2: Collected Essays 1929 –1968. 56–77. N. Pittsburgh: University of Pittsburgh Press. Sharrock. Pennsylvania: Pennsylvania State University Press. pp. Schatzki. Sharrock. Sayers. (ed. (2013) Wittgenstein. in Racine. London: Sage. Basingstoke: Palgrave Macmillan. London and New York: Routledge. in Lyas. G. Indianapolis: Hackett. usage and meaning’. Sharrock W. and Watson. (2005) The Flight from Reality in the Human Sciences. Critical Realism: Essential Readings. and laws’. pp. (2009[1950]) ‘If.. pp. (2000[1949]) The Concept of Mind. in Anderson. Sharrock. (2002) The Site of the Social: A Philosophical Account of the Constitution of Social Life and Change. Organization Studies 26(3): 465–84. A.J. 126–56. A Wittgensteinian Perspective on the use of Conceptual Analysis in Psychology. pp.). Woodstock: Princeton University Press. Ethnomethodology and the Human Sciences. Winch and Sociology [Unpublished]. and Anderson. 51–74. (eds).. (1996) Language Use: A Philosophical Investigation into the Basic Notions of Pragmatics. T. Ryle. G. Radical Philosophy 8: 36–8. London: Hutchinson. 54–60. Actions and Structure: Research Methods and Social Theory. London: Penguin Books.References 167 Ryle. predictions. and Norrie. Collier. R. Sharrock. W. G.R. 195–213. and Sharrock.A. and Because’. Lawson. (2013) ‘A return to “the inner” in social theory: Archer’s “internal conversation”’. J. So. Philosophy and Linguistics.

pp. (2001) Corporations and Criminal Responsibility. (1997) The Principles of Linguistic Philosophy. 5(1): 124–33. (1960) ‘Mr Hare’s paper: A rejoinder’. (2011) ‘The logical force of expressions’. pp. A. S. S. and mind’. Theoria 62: 187–211. and Wasserman. Cambridge: Cambridge University Press. (1996) ‘Conflicting varieties of realism: Causal powers and the problems of social structure’. Available at: http://www. C. Metametaphysics: New Essays on the Foundations of Ontology. (ed.168 Clarity and Confusion in Social Theory Slater. (eds). (1977) ‘Linguistic philosophy forty years on: A general statement’. B. Vienna Circle Institute Yearbook 15.M. Strawson.E. Philosophy and Linguistics. S. in Chalmers. and Harré. (1980) The Scientific Image. Journal for the Theory of Social Behaviour 26(3): 313–25. Stenlund.F. Tennessen. (1953) ‘Oxford philosophy’. S. Philosophy of the Social Sciences 42(1): 31–67. P. (2009) ‘Answerable and unanswerable questions’. Waismann. (2012) ‘Narratives and action explanation’.E. Stenlund. Cambridge Quarterly 7: 231–43. Z. (1965) ‘Ordinary language in memoriam’.helsinki. Inquiry 8: 225–48. European Journal of Social Theory. T. 2nd ed. Toulmin. Waismann. Mind 59: 320–44. (ed. Uebel. (1997) ‘Language. and Erneling.. (2006) ‘The strange death of ordinary language philosophy’. Varela.).W. Weitz. Wells.P. Freidrich Waismann: Causality and Logical Positivism.F. (1950) ‘On referring’. 444–71. M. F. (2007) ‘Elder-Vass’s move and Giddens’s call’. (2002) ‘Is the social scientific concept of structure a myth? A critical response to Harré’. C. S.fi/~tuschano/writings/strange/ [Accessed: 18 June 2014]. . D.J. C. P. T. (2003[1958]) The Uses of Argument. 245–65. Strawson. London: Macmillan.W. D. Philosophical Review 62: 187–233. (1990) Language and Philosophical Problems. Philosophical Investigations 9: 205–15. London: Hutchinson. D. 185–95. E. (1998) An Intelligent Person’s Guide to Ethics.F. (eds). Ratio 3: 1–8. R.E. Vendler. (1953) The Philosophy of Science. Warnock. M. in Lyas. 2nd ed. Tomlin. Toulmin. London: Methuen. Manley. London: Duckworth.F. The Future of the Cognitive Revolution. Journal for the Theory of Social Behaviour 37(2): 201–10. Varela. 302–16.F. Oxford and New York: Oxford University Press. B. Strydom. in McGuinness. Oxford: Oxford University Press. (1964[1952]) Introduction to Logical Theory. Thomason. F. E. C. Uschanov. Tomlin. Oxford: Oxford University Press. London and Basingstoke: Macmillan. London: Springer. H. C. Stenlund. in Johnson. B. P. pp. R. (1971) ‘Summary: Linguistics and the a priori’.). action. (1996) ‘Language and metaphysics’. Oxford: Oxford University Press. (1986) ‘A defence of linguistic philosophy’. Van Fraassen. pp. London and New York: Routledge.H.

A. 58–82 . ‘Shooting. (1976) ‘Is ordinary language analysis conservative?’. (1987) Trying to Make Sense. White. A. 2nd ed.References 169 Wertheimer. 4th ed. (1967) The Philosophy of Mind. White. Oxford: Clarendon Press. P. (1971) ‘Inference’. 2nd ed. (2009[1953]) Philosophical Investigations. (1969) On Certainty. London and New York: Routledge and Kegan Paul. Oxford: Blackwell. M. L. (2002) ‘Rom Harré on social structure and social change: An introduction’. On What Matters. (1985) Grounds of Liability: An Introduction to the Philosophy of Law.. Oxford: Oxford University Press.E. A. A. New York: Harper and Row. T. Wittgenstein. L. (1981[1967]) Zettel. L. in Parfit. (2005) The Big Typescript: TS 213. Wittgenstein. White. Winch. Wood. L. European Journal of Social Theory 5(1): 107–10. New York: Random House. Williams. pp. vol. (2011) ‘Humanity as end in itself’. Wittgenstein. D. Oxford: Basil Blackwell. Moore: A critical Exposition. The Philosophical Quarterly 21(85): 289–302. Political Theory 4: 405–22. White. (1979). White. Winch. Oxford: Blackwell. (2008[1958]) The Idea of a Social Science and its Relation to Philosophy. Oxford: Wiley-Blackwell. 2. Oxford Blackwell. Proceedings of the Aristotelian Society 80: 1–15. and May. (1958) G. killing and fatally wounding’. A. A. P. Wittgenstein. A.

This page has been left blank intentionally .

35–6. 121n. 76n Bhaskar. 114–15. 26n. see also polymorphous actions Analytic Sociology. 134–5. 151 emergence. 89–98. 69. 119–45. 75 Elder-Vass. 125 adverbial verbs.Index Achilles and the tortoise. 147 systematisation. 40–41. 140n. 92. 85. 26n. 80–85. 71–4. 55n. 12–14. 145–6 issues/problems. 132–3. 105n. 29–32 completeness of explanation. 18. 106. 106. 7. 80–83. 7. 17–23. 127. 78–85. 153–4 procedures. 48–50. 7. 78–9. 79n. 66n. 77. 9. 108. 119. 100–101. 144n. 94 Corporate Criminal Responsibility (CCR). 6. Dave. 104–6. 91. 69. 80. 16. 72n. 36. 140n of description. 99–115. 88. Jeff. 143 Austin.. 49. 54 Anscombe. 80. 7. 102. 84. 100. 121 resources. Gertrude E. 121. 143 counterfactual test. 146n relation to. 6. 52n concepts constitutive of social life. 22. 94n categorial accounting. 152 Coulter. 1. 86. 70. 115. Roy. 136. 151 Cavell. 24n. 96. 128 examples. 100. 127 causal efficacy. John. 69. 89–92. 127. 89. 118. 137–40 Cohen. 69. 134 Archer. 57. 6. 132–47 . Howard. 74–6. 145–6. Frank. the. 7. 15. John L. 46n. 96. 48n. 32 common usage. 135. 152 Corporate Social Responsibility (CSR). 42–56 Demeulenaere. 122. 102–8.M. 127–8 Carroll. 114. 100–101. 98 Becker. 107. 138 Critical Realism (CR). ix. 156 dichotomies. 13. 101. 7. 150. 83–6. Gerald A. 110. 37–9. 90–91. 7–8. 96. 126–32. 66n. 64n. 72. Bob. 39n. 79–84. 156 elucidation. 4. 105n. 25n. 125n. 25–9 Ebersole. use of. 85. 122. 71. 79n. 11. 3. 144n. 108n disagreement. 137 Deductive-Nomological (D-N) model. 20. 111. 91. 136.. Cora. Lewis. ix. 122. 128. 97. 115. 71n. 71–2. 136 Carter. 5. 52. 54.. 122. 73. 36. 100–102. 37. 41–2. 89. see also causal power(s) epistemic fallacy. Stanley. 52n. 149–50. 66n. 88 co-determination. 3. handling of. 33. 151 conceptual confusion/incoherence. 26n. 112. Pierre. 56. 83. 76. 5. 54. 104–5. 100–102. 2–4. 114 causal criterion. 99. 112. 138–140 emergent properties. 101. 7–8. 122. 69n. 54 Diamond. 120. 78. 51. 118–44. Margaret. 1–2. 115 conservatism. 48–51. 119–20. 128n causal power(s). 56. 28n. 22. 30–31 Cook. 82. 95. 59–60. 117–18. 73n. 72. 115. 30.

106. 50. 62. Raimond. Garfinkel. 67. 35. 37. 53. 150. 131. 39–41 logical form. John G. 42n. 112n. Michael. 37. 87–8. 35. Harold. 80–83. 117 Goldfarb. 7. Edward H. 57n. 124n Kemp. 69n. 30. 25. 35. 76. Herbert. 149 ontological boxes. 70n. 20. 114n methods. 67 Madden. 5 Hacker. 73. 32. Rom. 55 form of. 150 Giddens. Peter. 37. 125. 129n. 62. 140n. 35–6. 145 mechanistic/mechanismic. 95–6. 32–3. 114. 57–8. 155.. 156. 102.A. 109–10. 74n. 84. 120. 58–9. 64. 134 observation. Karl. 67. 88 mechanisms. 87. 104. 144 of language. see also conceptual procedures mind. 72n. 35. 21 fixity of meaning. Tim. 91 Louch. 54n. 122. 4. 128. 92. 125. see also real/exists scheme. 77. 95.. 74n. 5. 45. 153–7 Mouzelis. 96. 50. the. 97. 81. 96. 45–8. 48–9. 81–96. 47–8 May. 73. 67 logic. Carl. 1 Hutchinson. 128n. 140n.. 59–60. 111–12. 119. 57–68. 100–102. Anthony. 123–4. 7. 143 Harré. 144n Kitching. Tuukka. 144n Lynch. 130 Jenkins. 39. 64–5 Husserl. 109. 77–8. 111n commitments. 136. 54 Hempel. 21n. 1n. 41. 144–7. 110. 3 observable/non-observable. 57–8. 79n. 130. 157 logic exercises/puzzles. 69. 99. 93–5. 24 formal. 10n. 139n scientific. 14n. 48. 121–31 Hart. ix. Anthony. 115. 21n. 95n. 69–70. 2. 46n. 17. 57. 38–9. 39–40 linguistic/conceptual competence. 110. 64. 101. 73n. 50 grammar/grammatical. 40–41 Levi. 101–2. 35–56. 113.M. 83.. 127. 78n. 17n. 24n. 52. 121. 96. 39. 18n. 122. 127–32 fieldwork. 7. 73–83 Foot. 57. Peter M. 50. 43–5.. 106–7. 58–60. 3. 94–5 modus ponens. 59–61. 66. 119. 140n metaphysics/metaphysical. 7. 30–2 Marx. Stephen. 67. Alfred R. 89–90. 111. 70. Gavin. 10. 28. 17. 35. Edmund. 143. 17. 56n. 12n. 96. Herbert L. 67 logical structure. 45. 80–83. 64–6. 35.. 100–101. 115 Gunnell. 108n. 5. 11. 10–14. 118. 149–51 and social activities. 54. 5. 58n. 6. 115. 72–3. Phil. 53 fallibility. 144n immigration. 112. 154–5 inference/inferential. 149 Hunter. 120. 78n informal. 82 Gaita. 5–6. 95. 156 frequency/statistics.. 100. Warren. 43n morality. 73. 107–15. 32. 151. 149–50 and description. 87–8. 17n. 54. 56. 117 mathematical proof. Nicos. 5. 113 King. Philippa. 10n. 42. 133. 129 Hedström. Don S. 78n.172 Clarity and Confusion in Social Theory explanation. 100. 84n. 45. 149 super-scientific.S. 100–101. 113. 119. 59. 5. 119–20. 121–31 Marcuse. 107–8. 136n. 50. 45. 35–56. 26. 59n. 49n. 71–3. 91 expressions. 70n. 97–8 Kaidesoja. 57. 27n. 6. 96. Richard. 128n grammatical scheme/picture. 150 . 25. 58. John F.

138–40 unorganised. 46n. 113–14. 60. 101. Harvey. 85. see also causal power(s) terms of the debate. 141. 59n Read. 118. 41 Pleasants. 80–81. problem of. 35. 147. 18n. 36. 79n. 107–8. 149–50. 64n. 89. 73. 78n. 97. 129n. 6. 134–45. 144. Gilbert. 1–6. 41. 7. 145. 42–3. 21n. 14. 151 Strydom. 127 retroduction. 39. 100. 144. 24. 87 structure and agency. 73. 94n tendencies. 79. 16. 150 reality. 86–7. 76–7. 48 Stenlund. 2. 117–21. 48. 132. 55n. 9–33. 35. 107–8. 101 trolley problems. 136–7. 156 theoretical scheme. 108. 129–32. 110. 136. 60. Nigel. 81. Michael. 7. Peter F. 76n. 96. 40n. 71 scientific explanation. 128. 67. 59–60. 77. 25n. 111. parts. Alan. 143–7. Theodore R. transcendental deduction. 98. 2. 82n Ryle. 132. Sören. 130n. 93. 37. 69. 14. 80 Oxford philosophy. 1. 71–2. 10–11. see also logic physical. 70.. 58. 12–14. 97. . 118–19. 144n similarity of principle. 78–9. 18–21. 104. 105n. 72. 18n. 114n. 7. 84–6. 10n. 69. 115. 132–4. 125–7. 127 stratified. 137. 31 theoretical attitude. 22. 127. 53. 152–4 Socrates. 110. 58n. 72n. 140. 26n. 94. 133 powerful particular(s). 24. 54n. 95–6. 42n. 97. 33. 134 Strawson. 151. 55. 140n organised. 115. 92. 137n. 27. 102–3. 62–4. 151 political. 78n. 11n. 130. 48–50. 106. 84–6. 25. William. 143–4. 89. 10n. 90–95. 62. 50. 93n. 12n. 96–7 social structures. 108–9 social. 43n. 84. 154–6 polymorphous actions... 151–4 Outhwaite. 127–8. the. 111 ontology. 21n. 72–4. 135–6 Social Constructionism (SC). 4. 110. 96. 52–4. 85. 109n. 29–33. 100–101. 16. Wes. 78. 69. 15n. 119–20. 143. 81–3. 120–21. 99–115. 70n. 53n. 6. 32n. 140. 140n Sacks. 42. 80n. 100. 140n. 82n. 149 naturalised. 130. 110. 120. Rupert. 150 Thrasymachus. 8. 113. 144n real/exists. 39n. 125n. 121–7. 67n. 32–3 paraphrase. 79n. 41 Toulmin. 93n. 84–9. 146–7. 55. 156 Varela. 49. 45. 86–7. 57–68. 22. 145 refer/reference/referent. 68–71. 35–6. 88–91. 111. use of. 109n ‘scientific’. 105n. 90. 108. 72–3. 76. 71. 1n. 37. 55n Sharrock. 107. 43. 150. see also emergence and wholes. 120. 27n. 65. 92–8. 70n. 7. 127 Ryan. 156 unrealistic restriction. 128 Ordinary Language Philosophy (OLP). 124n. 131. 149 Scriven. 70n. 119. the problem of. 25–6. Charles. 59n. 55. 126–7. 32. 149–50 organisations. 133. Piet. 74–6. 99–100. 78n. 5.Index status. 123. 100. 96 possibility/impossibility logical. 62–3. 7. 58. 119n. 38n. 145 redescription. 68. standards. 108n. 39 and words/language. 123–7 pragmatics. 140 173 retrodiction. 33. the. 69–98. 36. 62n. 6–7. 91. 86–7. 139n Schatzki. Stephen E. 18n. 4–6. 5. 21n. 149. 119–22. 132–7. 130. 150. 41–3. 138–40 Plato. 112. 74. 122. 149 and appearance. 118.

58. 67–8. 48. 67n. 96. 144. 65. 23. 48n. 21n. Allen. 12–14. 146n Williams. 41. 24n. 157 Wood. 103. 32n. 49n. 49. Friedrich. 57–8. 156 White. 94. 50–51. 60. 99–100. 31n. 56 Winch. 25n. 91. Mary. Malcolm. 37–9. 15n. 72n. 33n. 2. 156 Zeno’s paradox. 57.174 Clarity and Confusion in Social Theory vocabulary. 52. 144. 31n. 45n. 79. 49n. 54n. 106–7. 39. 1n. 86–8. 66n. 59 Warnock. 2–4. Ludwig. 149 Waismann. 77. 5–6. 88 Wittgenstein. 43–4. 134 why-questions. 25n. 10n. Alan. 93n. 4–5. 72–3. 57–8. 37. 48 . Peter.