Abstract
This article demonstrates the underappreciated import and potential of Peter Winch’s The Idea of a Social Science and its Relation to Philosophy, a classic work published 65 years ago. Its aim is not simply to correct misunderstandings of Winch but to rehabilitate the text as indispensable for understanding past and present woes and cementing the future of sociological endeavour. I reconstruct and defend the claims put forth by Winch and then explicitly draw out their implications, which demonstrate the incoherence of the predominant disciplinary self-image that sees sociology as having a method and/or critical thinking prerogative. This problematic self-conception is jeopardizing the coherence and wider relevance of sociology and is responsible for its perennial difficulties in articulating a mode of discourse that can be seen as cogent by the public. A defensible alternative sees sociology as a second-order study of practices that is premised on a conceptually accurate relation to those practices and on answerability to the criteria and abilities of understanding, description, explanation and criticism they afford. This conception can support the reconfiguration of existing forms of sociological inquiry as well as the development of new ones.
Keywords
Peter Winch published The Idea of a Social Science and its Relation to Philosophy (ISS or Idea) in 1958 (after Winch, 1956), a time when the social sciences were in the ascendant. In the 65 years that have elapsed since, many things have changed. Moreover, Winch’s book has been hotly debated, 1 gaining the status of a classic and some notoriety to go with that. However, sociologists have not been able to appreciate the truly profound lessons that Winch is offering, which are crucial for responding to not only past but also present sociological troubles. This is due to a certain widely held view about Winch, something of a mythology (cf. Hutchinson et al., 2008), comprising several damning assessments. The following are indicative: First, it is believed that his argument is outdated, mainly on account of the positivist or empiricist positions he attacked. Positivism is seen as no longer dominant in social science, and current understanding of natural science sees it as depending on hermeneutics too. So, it is argued, why bother to distinguish between the study of nature and the study of society? Second, Winch’s claims are seen as flawed in the sense of leading to unacceptable implications for sociology, which becomes one-track or is reduced to philosophy, is unable to offer structural explanations, have any critical distance to its subject matter or conduct comparative inquiry. Finally, it is believed that whatever was worth salvaging from Winch has already been incorporated into various sociological programmes (Bloor, 1983; Giddens, 1979, 1982, 1984; Habermas, 1987; Schatzki, 2002). So, once again, it is thought, there is no reason to bother with Winch.
I beg to differ. Even if this latter point were true, the fact that sociological knowledge progresses by frequent loss of memory (Abbott, 2006) would still render it important to repeat arguments in relation to fresh circumstances. In fact, however, the above common assessments, when taken together, preclude a renewed social science drawing on Winch’s key insights. And, as I will show, they are also false individually: the book’s main message is independent from any specific targets and thus perennial. The flaws identified by most critics are far off the mark. And what Winch is really saying remains unappreciated still. What can Winch teach us today? Quite a lot despite the mythology that dominates how sociologists see him. In the currently challenging circumstances as regards sociology’s perception of itself and its uniqueness (Hollands and Stanley, 2009; Manza et al., 2010; Savage and Burrows 2007, 2014; Weinstein, 1997), Winch’s Idea can offer us no less than a new self-image. It can help us responsibly construct a coherent self-conception on which an accountable, intellectually respectable sociology can be built.
In what follows, I will work towards showing what such a sociology should not and what it could look like. I start by offering a reconstruction of Winch’s main claims. They are stated clearly and forcefully. I proceed by showing in detail that common criticisms are misplaced. I then turn to spelling out the most significant implications of his thought, precisely the ones that have not received sufficient attention and which exhibit the incoherence of sociology’s understanding of itself. I end by adumbrating future possibilities suggested by the emerging alternative conception.
A reconstruction of Winch’s key claims
It is not straightforward to see what Winch’s argument in ISS amounts to or implies. Winch’s ‘digressive style of writing’ (How, 1995, p. 63) is in part to be blamed for this. But the main reason why Winch’s thought has not been duly appreciated is that it requires one to break with dominant sociological and philosophical ways of thinking. To address these difficulties, in this section, I spell out the central claims being put forth. My order of presentation is aimed at maximum clarity and systematicity and does not follow that of the book. Although my reconstruction covers most of the book and is premised on its in-depth study over the course of many years, it is inevitably selective. 2
Very many of the past and present troubles of the social sciences start form the following misconception with which Winch begins his book: that ‘the investigation of reality’ takes a singular form, the one that appears in (some model of) scientific investigation. A simple logical inquiry into the sense of this phrase, including what contrast the expression ‘real’ is marking in various contexts, reveals that it varies significantly by activity (ISS, 1.6). Politics, religion, musical expression, street art, sports, science, in every aspect of our social life we take an interest in things that are understood (in different senses) as real, that is, depending on context as not fake, not artificial, not false, not petty, not superficial, not transient, not able to be wished away and so on. Such things can be the presence of political will or party politics, being guided in one’s conduct by the love for one’s neighbour or the virtue of integrity, the intensity of Jackson Pollock’s brushwork and the difficulties involved in making it in the art world or pressure towards using performance enhancing drugs in young athletes. All are realities that are made intelligible, described, elaborated on, ‘investigated’ we might say as a useful shorthand, as part of these social activities.
3
This observation highlights the character of the social life that people lead and what it means to understand that social life. If we reflect seriously on what we consider an understanding of various issues, phenomena, topics and realities to consist in, we will see that we do attach criteria and are able to evaluate various claims to understand them (ISS, Preface to the second edition). These criteria do not only demonstrate what relevant limits might be but vary in accordance with, for example, whether we seek to understand what an author was trying to say in a publication, what a concept like liberalism amounts to, what the issue is between vying parties in debates on abortion, how the past is made use of during modernist conceptions of progress, how our institutions came to be the way they are, how they might (not) work at the moment and so on.
4
Weber’s well-known thoughts on the explanation of meaningful behaviour with reference to motives and reasons is, in part, an acknowledgment of this simple point for the case of behaviour (ISS, 2.2). Because there are criteria of understanding, not anything counts as an explanation. Human actions, for example, may be explained and understood with reference to motives, among other ways. Moreover, and conversely, to answer the question as to the nature of motives, one must refer the concept to the practices in which it has its intelligibility, that is, how it is used to provide assessments of character, articulations of future plans or explanations of action, among other things. The logical picture that emerges reminds us that the concept of motives is not akin to the notion of physiological states or even dispositions but closer to that of reasons (ISS, 3.3).
Further logical reflection shows that notions of motives, reasons or, more generally, rules are presupposed in the concept of meaningful behaviour and the notion of a language. That we possess the capacity to follow rules in a self-evident manner, or ‘blindly’ as Wittgenstein puts is without thereby implying that we must be under coercion, and that the criteria we operate on are public, so that others could conceivably learn them and use them to correct us, are related preconditions to the intelligibility of the notion of a rule. Public rules render intelligible all sorts of actions, events, structures, characters, relationships, cultural ‘objects’ and so on and enable us to issue various judgements to the effect that two things are the same, fall under the same rule, the same concept (ISS, 1.8, 1.9).
Indeed, a huge part of social life is organized via rules of many different sorts (constitutive, regulative, formal, informal, stylistic, procedural, etc.). This simple observation reinforces the point that various activities (or different ‘investigations of reality’) not only enshrine different criteria of intelligibility and understanding but have different aims and pose different problems. Thus, a sociological investigation of social life, however it may be conceived in its details, must necessarily take into account the various criteria of understanding, rules and aims of ‘investigations of reality’ (ISS, 4.2) including the various senses of ‘reality’ that may arise (note that no grandiose claim is being made that there are many realities. Only that there are many uses and thus senses to the word ‘reality’. Indeed in many of these uses there is only one reality). In other words, sociological investigation must be second-order in relation to first-order practices.
It follows that we must reject the rationalist position that sees rules and criteria as singular, monolithic or immutable, as something independent from practice or belonging only to the frame of reference of the investigator. A case in point is the rationalism of the classical sociological point of view expressed in the work of Durkheim (1982) or Pareto (1935).
5
Needless to say that no small measure of (methodological) rationalism is retained in sociology, for example, in so-called structural perspectives such as critical realism, analytical sociology and Bourdieusian sociology, for instance, in the latter’s commitment to omnipresent competition for capitals and symbolic dominance (e.g. Bourdieu, 1986). This derives in no small part from neoclassical economic theory, which has also informed sociological approaches such as social exchange theory. Returning to the case Winch examines in the book, Pareto sees the sociologist’s business as in competition with what people think are the reasons for their behaviour. He hails experimental social science as the paragon of rationality and moves, even if indirectly, to discount entire first-order practices as contrary to reason. This conception sees as the only valid rules, understanding and criteria at play those of the social scientist. But treating social life in this way is deeply incoherent because it removes from view the inherent criteria that are needed for its understanding (ISS, 4.2).
As a result, this point of view cannot be consistently adhered to. Pareto, for example, who thinks that sociology must unearth a deeper level of invariant relationships free from cultural ideas or so-called rationalisations, is forced to make use exactly of the things he has in principle excluded (for instance, notions of purification and asceticism in describing religious practices) to gain any purchase on the phenomena. Winch emphasizes the irrelevance, partiality (which may take the form of ethnocentrism, e.g., in the hands of anthropologists) and blatant misunderstanding that results via this imagined procedure. It distorts social phenomena by using concepts that are inappropriate in various ways, for example, mechanical concepts that turn intelligent action into aimless movement. Another way of making this last point is to say that much like in the case of understanding and explanation, not everything can count as a description of social life.
How much alignment is sufficient is an open question that depends on the argumentative and methodological context. But descriptive accuracy is necessary if we are to be able to identify, understand and explain social life (to give only a very minimal indication of differences in more recent sociological approaches, the ethnomethodological tradition acknowledges the profundity of this point (e.g. Sacks, 2003). On the other hand, approaches such as ANT, by positing a so-called ‘flat ontology’, imply precisely a flat, non-accurate and non-recognizable conceptual landscape: for instance, the principle of ‘free association’ (Callon 1984, p. 200), i.e., that the investigator as ‘observer must abandon all a priori distinctions between natural and social events’ ends up neglecting crucial logical distinctions). Thus, far from being in a position in which they must come up with special rules and descriptions to make something intelligible, the sociologist, if they want to retain the intelligibility of social reality, must become competent or rely on the competence they have achieved in the existing rules, language and forms of understanding. In other words, the relationship of the sociologist to social life takes the form of practical competence in the understanding inherent in their subject matter. There is no room in sociology for the position of an external observer in the sense of someone who has no endogenous understanding of what they are observing but relies on independent theoretical abilities.
In the first instance, the abilities that are required for participation in social life are what allows us to find it intelligible; not sociological theory. If we consider, for example, the case of grammar and logic, we see clearly that our linguistic and rational abilities are a basis for understanding the systematizations or theoretical statements offered therein. To understand whether an ‘is’ is the ‘is’ of predication or the ‘is’ of identity, what a fallacy is, and specifically what it means to beg the question or to commit the ecological fallacy presupposes and requires competence in arguing, telling what point someone is trying to make, what is at stake in an argument, what is the issue and so on. Those skills are provided practically, not theoretically.
6
The same applies to sociology.
And it is this competence-based relationship to subject matter and the understanding and relevance it affords on which any further development of ability and sensitivity or reflectiveness are based. Like knowledge of grammar and logical theory, sociology might make our understanding more sophisticated, more cerebral, more eloquent and perhaps more profound but that will depend on whether we can recognize any distinctions it draws in practice. It is not the ultimate basis or a possible substitute for that practice. In other words, sociological method cannot function as a means of making tractable what must otherwise remain intractable but rather a way of carefully adding to our means of tractability on the condition that they have not been falsified.
These criteria are not reducible to specific beliefs or views but allow the expression of many alternatives. When Winch discusses motives, reasons, rules, concepts and other notions that are central to social life, he is not talking about personal opinions. He is talking about the shared edifice on which liberal, illiberal, anarchistic, banal, contradictory or any other kinds of positions can be put forth, disagreed with and evaluated.
The application of rules is not merely habitual, because it involves the reflective application of criteria to new cases that call for deciding how new circumstances are similar to previous ones (ISS, 2.4). Accordingly, investigating activities, rules and concepts is more like investigating the uses of a musical instrument in the context of composing, radically re-interpreting music, introducing new techniques, joking around, making deliberate mistakes and so on rather than investigating a set musical piece that everyone must regurgitate mechanically. There are open-ended capacities which also include criticism and possible novelties (ISS, 2.5).
We can talk about concepts as ideas that are expressed in and structure social activities (ISS, 5.1). The concept of war structures what belligerent parties are doing and sets limits on how what is taking place can be relevantly described and understood. The fact that social relations are seen to embody ideas means that they are like conceptual relations. Economic relations, so-called material relations within the Marxist tradition, are no less constituted by ideas, for example of labour, wages, property, and after the entry of Marx into politics, exploitation and class conflict. A change in existing concepts or the invention of new ones leads to new understandings and related practices. For example, the recent distinction between Facebook friends and real friends structures social conduct and how people may relate to each other. Because of these interrelations between concepts, relations and activities, it is clear that atomism is a mistaken view not only of nature but also of human life: ideas are constitutively part of behaviour and behaviours are thus tied to one another too. For example, an internal relation between behaviours can be seen in relations between actions in a sequence and between them and their context. In fact, ideas are involved (or are a close analogue) even when behaviours (e.g. gestures, clothing) are not directly translatable into words (ISS, 5.2). To understand social activities and relations requires understanding the involved concepts and expressed ideas. Similarly, understanding in history, like Collingwood (1994) contended, means understanding the idiom of an age and relavant logical relations as opposed to understanding universal theoretical generalizations. 7
Having described the relationship of sociology to social life and what it takes to understand the latter, sociology’s relation to philosophy can also be understood. It is not an external one of sociologists simply drawing on philosophical theories and schemes as inspiration or, say, using the philosophy of science to stipulate explanatory criteria for their own investigations. Sociology and philosophy are analogous endeavours that are joined at the hip around a family of conceptual questions. We may dub these questions philosophical questions, although they are not the exclusive property of academic philosophy. They delineate a common ground between second-order disciplines that must elucidate concepts and practices.
The second-order character of both can be appreciated by considering that there is the philosophy, sociology (and history) of politics, religion, sport, music and many other activities, including of philosophy, sociology and historiography themselves. This conception takes philosophy and sociology (history and other allied disciplines like political science) further away from first-order sciences that do not study other practices. Sociology, like philosophy, is centrally concerned with conceptual accuracy, in the form of the accurate elucidation of the concepts, criteria, rules, institutions, behaviours and problems that the various first-order activities set themselves. This is why Winch invokes the discussion of magic by Collingwood in The Principles of Art and by Wittgenstein in Remarks on Frazer’s Golden Bough (ISS, 4.3). Both produce this elucidation by accurately describing the grammar of behaviour and concepts that make the institution of magic so different to that of experimental inquiry. Winch calls this mode of elucidation philosophical argument. But it is sociological too. This is the first point that ties the two disciplines together.
The second is that both philosophy and sociology concern themselves with a family of questions that call for such elucidation. The occasion for such questions may be multiple, such as when we lack understanding about what we ought to say, or what the significance of practice, institution or piece of behaviour is, or how it differs to another phenomenon or when we wish to draw out and assess the implications of an argument (cf. ISS, 1.3). The form of such questions is ‘not knowing our way about’ and thus calls on us to establish what the relevant rules, concepts and criteria are.
The final fact that ties philosophy and sociology together takes us back to Winch’s opening idea. It is that both must be committed, on pain of contradiction, to elucidating the plurality of conceptions and criteria that make up the various social activities as ‘investigations into reality’, without partiality (ISS, 4.1) but also without any implication that such activities must be endorsed, prioritizing understanding to facile criticism.
Refuting criticisms of Winch’s Idea
Clearly reconstructing Winch’s claims allows us to dispense with many obviously misplaced criticisms.
8
We can also see how the ones that seemingly persist, the strongest of which derive from philosophical and sociological (methodological and programmatic) objections, are ill-founded too. We can start from objections that are based on developments in our understanding of science. Objection 1: Winch’s position is irrelevant because it encourages an obsolete separation between the natural and social sciences.
Recent philosophy, history and sociology of science have discarded (some previous) idealized pictures of science and, via an appreciation of its actual practice and development, have enabled a new understanding of scientific activity as ‘theory-laden’, involving ‘tacit knowledge’, depending on paradigms and disciplinary matrices as well as a background of tradition (Hanson, 1965; Kuhn, 1962; Polanyi, 2015[1958]; but cf. Dallmayr, 1977; Gadamer, 2013 [1975]; Gibbons, 1987; Hekman, 1983; How, 1995; McCarthy, 1973). However, insisting on this fact is irrelevant to the point in question: apart from removing a specific target that Winch legitimately took aim at for historical reasons (while also expressing reservations about its validity (ISS, 5.1)), and which remains more prevalent than is typically believed, it does nothing to undermine the actual differences between the social and natural sciences. They are highlighted in Winch’s claims on sociological description as a matter of accuracy towards first-order concepts that constitute the subject matter under investigation. Whatever our account of the natural sciences, no such relationship can arise (nor any ensuing issues of irrelevance, injustice, misrepresentation, etc.) for it is obvious that there are no endogenous concepts that structure natural processes; only the concepts of different investigators. Objection 2: There is nowadays a conception of causality that is suitable for use in any scientific investigation, including of social life.
Despite the arguments by defenders of the unity of scientific method (Albert, 1988; Popper, 1979), it is easily demonstrable that when we invoke notions of cause in connection with behaviour, decisions, social institutions, social changes and so on, we attach a range of senses that involve considerations, reasons, responses to threats and so on. To say that torrential rain caused people to drown does not employ the same concept as the one involved in saying that climate change caused people to migrate. Singular notions of cause are a result of rationalist commitments rather than an account of how explanations really work within social life. As such, any singular notion of causality (e.g. Bhaskar, 1998[1979]; Hedström & Ylikoski, 2010) is a distortive instrument (see Tsilipakos, 2016). Objection 3: If Winch were allowed to have his way, then that would signal the end of sociology (Gellner, 1960; cf. Bryant, 1970; Palmer et al., 1998; Habermas, 1988; Merton, 1961; Uschanov, 2006) or its absorption by philosophy (Horton, 2000).
The force of the objection is not best seen when critics are taken as advocating the maintenance of strict disciplinary boundaries. After all, ‘the discipline of sociology’ or ‘the sociological canon’ are currently short of supporters and, even when their importance is acknowledged, there are all sorts of interdisciplinary (absorptive) arguments, for example, modelling (parts of) sociology on political theory and legal scholarship (Abbott, 2018). In any case, Winch has a new conception of sociology and philosophy which is not conservative as to traditional disciplinary understandings.
The real issue underlying this objection is that the argued overlap with philosophy is seen as taking away sociology’s empirical component. But all Winch is saying is that a sociology that is incoherent is not worth striving for. And that if many of our questions hinge on careful elucidation, then we have to find a way of responding to them rather than ignoring them. On this principle, several different forms of (empirical) sociological inquiry can be premised. The objection denies this latter point and holds instead that Winch wishes to reduce sociology to philosophy, dismantling it in that way. There are various versions of this argument. One is that Winch and Wittgenstein (and Collingwood), who talk about elucidating the concepts that we are confused about but otherwise are familiar with, are advocating a form of platonic anamnesis or, by equating the conceptual or a priori with sociology, are taking away any possibility of empirical sociology. What they are claiming, however, is that if philosophical–sociological problems are seen as empirical (in some sense), they must also be seen as conceptual in the sense that they are hinging on accurate description and elucidation of criteria. If the operative concepts are not grasped, then any appeal to novel external information, correlations, removed historical origins and other presumable ‘unknowns’ cannot ensure understanding. No part of this argument subscribes to a Humean tenet (Bell, 1967; Louch, 1963) that in the domain of knowledge we must either deal with facts or relations between ideas and that therefore relations between ideas must be already known because they are in our head. No part of it denies the importance of fieldwork or what is called ‘empirical inquiry’ in sociology; but the argument highlights that philosophical–sociological problems are not indifferent about the form that any new information may take; in other words, they require that description be conducted after careful reflection of criteria and elucidation of concepts. Accordingly, the point of fieldwork is to reach such an understanding. Moreover, if a discovery is involved (cf. Winch, 1964a), which is something entirely possible, it is usually a relative one, that is, finding out from a group of people something that was not known to another group of people. It is not a discovery of some piece of information that no human being was aware of. Another version of the above complaint is the following: Objection 4: Winch attempts to reduce social life to concepts and language.
Winch is not a reductionist because he is not in the traditional philosophical business of ontology that sociologists tend to adopt in deed. The ontological reasoning behind this objection, that social life is one thing and language is another, is rejected. It rests on a confused philosophical distinction between ‘language’ and ‘the world’. Alternatively, it may rest on the idea, of equal ontological bent, that the ‘social world’ is made up of various parts, among which there is at least language, on the one hand, and behaviour or practices on the other. None of these distinctions can be maintained. Acting and speaking are not categorially separate: consider, for example, how designations of friend or foe and military action are tied. Such designations can also be seen to be conceptual even if not explicitly articulated because they are presupposed by and expressed through the relevant behaviour. Finally, Winch’s arguments are necessarily made cogent by the fact that reflection on our practices takes linguistic form and that sociological reasoning comprises a set of reflective forms of speaking and writing that are purporting to be relevant to social life. A related criticism follows: Objection 5: Winch is offering an ontology of the social that identifies various restrictions on the use of concepts (Louch, 1963, 1965; Theunissen, 2017, 2020) and ends up recommending a singular way of sociological inquiry and neglecting the many other forms of doing sociology. This is belied by the sheer variety of sociological endeavours (Louch, 1963, 1965).
Once again, the ontological impulse leads to confusion. Winch is in the business of elucidating the presuppositions involved in our endeavour to understand society sociologically. His argument depends on pointing out facts about what we count as understanding in various cases and what we would consider as a relevant analysis, given the questions that we want to ask. Winch is not saying that our interests in understanding social life should be of a certain kind but only that they happen to be of particular kinds and that this fact bears certain implications. Human actions and social life as topics are internally related to certain interests towards those topics. Moreover, Winch’s sensitivity to variety in the multiple forms of understanding means that he is not opposed to multiple kinds of questions or forms of inquiry, only to our misunderstanding of what questions we are asking and the confusion that this may produce. We now reach what are typically perceived as the most penetrating criticisms of Winch; they form a set of interlocking claims. Objection 6: Winch is exclusively concerned with human action-level concepts and the explanation of action via reasons (Fay, 2000; Gellner, 1960; Louch 1963, 1965; MacIntyre, 1964, 1967; Pettit, 2000). He is not concerned with detecting causes that lie at the social structural level (Bryant, 1970). He neglects macro-concepts and focuses on the micro. By fixing his attention to micro-concepts, he simultaneously forbids us to go beyond the understanding of participants. He does not allow us to trace unintended consequences (Bell, 1967; Giddens, 1979). He precludes criticism of social life (Bernstein, 1976; Kemp, 2003a, 2003b; Lukes, 2003; Macintyre, 1964, 1967; Pleasants, 2003; Schatzki, 2003) and pushes us towards conservatism (Gellner, 1960).
This complex set of claims only appears logical because those who express it like to think of social reality as terra nullius. Winch’s point of talking about action is not the alleged one at all. He is not talking about the self-understanding of one, two or even a specific group of participants but about the criteria and rules that make up our practices and institutions. His point is that those criteria are not a mere hindrance to more fundamental phenomena – they are the phenomena– and he wants to warn those who wish to remove them as externally related to ‘society’ or ‘social structure’ that it is not possible to do so and still retain a meaningful sense of ‘society’ or ‘social structure’ (cf. Porpora, 1989). He does not wish to claim either that such criteria are univocal or that we should be bound by particular views.
Winch may have executed his argument as to what counts as relevant description, explanation and understanding by (also) 9 using the example of action. But the point is readily generalizable. It is patent that there are different ways of explaining social life and that many of the ones we ordinarily resort to involve reference to social structure concepts too. People are perfectly competent in making sense of what they are doing in relation to notions such as mentalities, institutions, capitalism, the legal system, what society permits or disallows and whatever other social structural concepts we may like, without presupposing the somewhat artificially sharp sociological distinction between micro- and macro-analytical constructs, which, thus, does not correspond to the logical character of the way people discuss society (see Tsilipakos, 2016, chapter 6). It is this artificial distinction that produces a chasm between action concepts and structure concepts. But the logic of social structure explanations is not only related in many important and intricate ways to social action explanations (presupposing and resolving into the latter in many cases) but is also itself governed by additional criteria of relevance and success. Those criteria are flexible indeed as are all criteria that must be usable for social life. But simply ‘going beyond’ them is not a sensible goal because it leads to irrelevance at best, nonsense at worst.
It is important to restate that Winch does not equate ‘the phenomena’ with individual participants’ understanding or particular intentions but rather with the multi-form understanding that is afforded by the culture. One reason is that there obviously are individual gradations in the abilities and understanding of different actors. For instance, a seasoned writer’s sophistication with insightful description is certainly not equal to that of, say, a dancer.
From the above observations we can see that, first, the objection that action may have unintended or group-level consequences is a red herring because no claim is being made as to the knowledge possessed by (individual) participants. For example, whatever an individual economic actor or state policy official might think about the government imposition of quotas, the resulting ‘unintended’ black market is no less a socially structured practice that must be described in socially accountable ways. The second implication is that cultural understanding inherently includes the possibility of reflection, criticism and debate (see Bottero, 2019; Walzer, 1987). It is thus entirely logically possible to criticize – but criticism, if it is to be genuine elenchus, is bound by criteria of relevance too. The liberty to produce criticisms that are irrelevant or completely off the mark is a parody of the ideal of investigative autonomy, which sociologists rightly emphasize. Objection 7: Winch’s idealism comes back to bite him. Not only has he reduced social life to concepts and ideas but he has also subscribed to the rigid holism of idealism, ruling out comparative (Gellner, 1960) or intercultural inquiry (Clarke, 1970; Horton, 2000; MacIntyre, 1964, 1967; Wilson, 1970). Relativism follows (see Hollis & Lukes, 1982; Lukes, 2000).
Though he makes use of authors who may be seen to lie in the idealist tradition (e.g. Collingwood) and may thus employ particular turns of phrase, Winch is no idealist in any doctrinal philosophical sense. Pointing out the interconnections between our practices and the fact that forms of words when embedded in different social settings amount to different concepts hardly commits Winch to the idea that different societies and historical periods are closed expressive totalities (see Gombrich, 1969; Jay, 1984). In fact, sensitivity to these differences, far from precluding comparative or intercultural analysis, is what makes it possible: if comparisons are not logically appropriate, then simple misunderstandings follow, such as from seeing religion or magic as trying to be science and failing to deliver the goods. It is rationalism and eurocentrism as false universalism which, by enshrining a singular form of rationality, are incapable of conducting any sensitive intercultural analysis. Moreover, considerations about truth and contradiction are only meaningful when propositions are of the same logical type. Propositions that look the same or look like opposites need not be of the same type: ‘witches exist’ when understood as uttered by someone who participates in the Azande way of life, does not necessarily mean the same as ‘witches exist’ in our own society, any more than ‘windows exist’ means the same when uttered by loft conversion specialists and when by pilots landing airplanes during thunderstorms. Before making a determination as to whether two propositions are contradictory or not, it is necessary to accurately portray their sense. To do that one has to explore, for example, the kind of actions they sanction or forbid and the web of practices into which they enter. For another example, the proposition ‘God exists’ and ‘there exist billions of galaxies in the universe’ are not clashing, and therefore it can be the case that both are true, or both false for that matter. Equally, the ways in which we can decide on their truth or falsity are not of the same logical type: the problem of evil is a challenge to the existence of God; it is not a challenge to the existence of the physical universe. These are simple logical observations that result in good sense and in sensitive reflection on our practices and their relation to those of other societies, not suspect metaphysical claims that truth is relative.
Implications for sociology
Having articulated the central claims of Winch’s Idea and having shown how persistent criticisms are misplaced, I now turn to sketching the disciplinary implications of the argument.
Sociology does not have a monopoly on understanding society. It must seriously take into account other sources of understanding
Sociology is not the only means or source of understanding of social life nor is it the original source of criteria as to what counts as understanding. Sociologists tend to forget this. They assume in word and deed that the only understanding that really matters is sociological understanding, and they undertake no serious effort to relate the understanding produced by sociology to other sources or to examine the former's reliance on the latter. The fact that sociology’s audience does not see references to the available understanding or any use of it as an external check (Rule, 1997) is one of the reasons why people fail to be convinced by sociology (Weinstein, 1997; cf. Tittle, 2004). A common piece of sociological reasoning (that is occasionally used in theoretical paper openings) reiterates this ideology in inferring from the fact that no sociological work has taken up a topic to the conclusion that the topic is poorly understood in general and that no systematic exploration is required of the understanding that is available (see Tsilipakos, 2016, 2017), for example, available understanding of emotion concepts before problems in the sociology of the emotions are formulated (Bericat, 2016). This neglect of endogenous sources of understanding and their presumptive replacement by sociological technical concepts (Banton, 2016) or their cavalier treatment in taking various interpretative liberties under the influence of doctrines of radical indeterminacy of meaning (see Tsilipakos, 2020; sometimes also coupled with postmodernist or post-structuralist programmes) is the reason why sociologists may have trouble in cementing a claim to autonomy in front of the public. People see sociological claims as requiring them to ‘abdicate their powers of reasoning’ (Berlin, 2012[1998]) and experience this as a form of unwarranted pretension on sociology’s part.
If understanding comes in the form of abilities supporting a plurality of opinions, then ‘common sense’ is an artificial and (ultimately) incoherent foil or target for sociology
The typical justification for sociology’s imagined monopoly is that society is organized on the basis of common sense as opposed to scientific method. And this is true. But society is organized on the basis of common sense as an ability to reason, as a multifaceted method of understanding, as a component of and not as a second best to scientific method. Moreover, society is not organized solely on the basis of so-called common-sense views, even though particular views might be shared by a majority. Sociologists who distinguish between common sense and enlightened sociological reason, particularly within critical and emancipatory approaches, are thus required to equate common sense with illiberal, right-wing or unconsidered views. This is unfortunate because these views are typically balanced or at least challenged in many Western states by their opposite views: liberal, left-wing, carefully assembled. So the existence or even prevalence of objectionable views cannot serve to distinguish sociology from ‘the society organised on the basis of common sense views’. For, if we are to juxtapose sociological understanding to common sense, we must do so consistently. It is inadmissible to assign the latter to some members of society and not to others. In that case we could just as well claim that sociology tries to understand society like some members do (those who do not espouse common sense views). This reduces the claim of sociology’s opposition to common sense to incoherence. Moreover, it must be remembered that ‘common sense reasoning’ in the twenty-first century is informed by literature and journalism, political commentary and debate and involves a level of competence in handling numerical information, statistical economic claims and so on. This has implications, to be presently explored, for how sociological methods are seen to be related to non-sociological ones.
In seeking its distinctiveness, sociology cannot define itself (as critical) via only selective contrasts
Sociologists are typically selective in their contrasts, mainly juxtaposing the discipline to the world of (mostly) conservative politics, right-wing journalism, xenophobic sentiments and thus highlighting the critical nature of sociology. Accordingly, and after the demise of the natural science model, sociologists typically do not see themselves as ‘problem-solving scientists’ but as critical thinkers. But this is an impoverished and ultimately indefensible conception on which no positive sociological programme can be built. We might indeed be more critical than particular cherry-picked (classes of) actors. But are we on the whole more critical than historians? What about performance artists, or activists, investigative journalism on the left or documentary filmmakers? What about philosophers who with the aid of logic and rhetoric claim a firm grasp on arguments, presuppositions, inferences, fallacies and overall critical thinking skills? Are sociologists more reflective than seasoned writers of novels? When comparisons of skill are made, one has to remember that society is not only made up of illiberal illiterates but includes all sorts of people, many of whom are eloquent, wise, capable, active citizens, and many of whom possess deep understanding of their practices and of the wider society that feeds into those practices. Have we got any solid basis in staking our claim in the way envisaged when we shift to these fuller contrasts?
If not, might not opening up our awareness to the true range of options lead to a revision of our conception of what exactly makes sociology distinctive? It is not that a feature that is heightened with regard to a particular juxtaposition ought to be equally heightened with regard to another. But our conception of ourselves ought to be rich enough so as to contain characteristics that are adequate to the juxtapositions that are relevant, indeed pressingly so.
In sum, like every academic pursuit, sociology has got to find its own unique and distinctive features. If this is done selectively, then no coherent conception can be arrived at, because the advantageous features highlighted in relation to one contrast might end up being shown as not particularly advantageous in terms of other contrasts. Sociology needs to think about its distinctive features not only in relation to the natural sciences or ‘untutored common sense’ but also in relation to other fields of activity. We need to have some realistic conception of the various traditions that differentiate between sociology, politics, history and philosophy and also between literature, punditry, commentary, journalism, documentary filmmaking and so on.
Autonomy of sociological interest is bound by relevance to ‘social reality’
From the claim that forms of social life and various activities within them can be seen as ‘investigations’, it follows that they also contain various ways of taking an interest towards social life. Thinking of sociology, in ontological mode, as trying to get at ‘social reality’ is highly misleading because it carries all the usual predicates of that conception of reality and suggests that the subject matter of sociology is something unorganized and chaotic, mind-independent and external and that sociologists are trying and can only reach it via their methods. Insistence on an ontological point of view of ‘social reality’ as separable from methods and interests obscures from view the realization that the object of inquiry is a form of understanding that is structured precisely in those ways. This is the same as saying that the object of inquiry is a method of understanding that must be reflexively approached, in other words one that is made available via participation and competence in that method of understanding. The questions we ask about society presuppose our forms, ways and methods of understanding and are erected on them.
One implication of this fact concerns the question of the autonomy of sociological problems and their relation to our interests, including to the various social problems. Sociological interests need not be identical but they are continuous with the various interests that we take upon social life. This does not mean that there is no such thing as the autonomy of sociological problems. But such autonomy is bound by relevance considerations, and by the fact that methods, criteria of understanding and practices are co-constitutive. Sociological problems are predicated on such interests. They are also predicated on the condition of conceptual accuracy, which means that even if they are formulated in a technical language, they cannot override that consideration. For a recent example, consider how Reed’s (2020) concepts actor, rector, other presuppose sensitively described historical contexts (see Tsilipakos, 2022b). For a classic example, consider Parsons’ (1951) result that strains in one’s motivation in the form of illness are from the point of view of general equilibrium equivalent to or even constitute a form of deviance. But this will not render gun crime more similar to depression beyond noting that in some sense they are both disruptive. While that may be true, it is somewhat less than fully elaborated and does not undermine the conceptual difference between the two; it rather presupposes it. More generally, questions of social stability and orderliness presuppose explications of these concepts that are embedded in social life. Similarly, the result in economics that a quota imposed on a good or service is equivalent to a tax in the sense that the price paid by consumers is different to the amount received by sellers (who may also receive quota rent equal to the tax in the second case) does not supplant but depends on the distinction in practice between, say, closed professions and value added tax.
The idea of sociology as a method-driven discipline is partial at best
A second implication of the above considerations is that the idea that the social sciences rely on their technical methods for understanding society and thus must see their development as method-driven is a misconception. To gain a purchase on social activities, that is, on forms and methods of understanding, it is necessary to partake and employ these self-same methods. The idea of an expert, method-driven study of society is, in its purest form, a promise of neglecting, replacing or trumping forms of understanding that exist. But, once again, these forms are not dispensable competitor forms of study; they are constitutive of society and indispensable for understanding social life. This means that sociological methods must be predicated on our cultural understanding and must work synergistically with that understanding. A clear explication of how sociological methods build on and enhance that understanding is needed.
Replacing versus augmenting understanding: Antagonistic self-reflection versus enhancing self-reflection
Our methods of understanding (apart from being topics of study) are indispensable resources. If we exclude the incoherent idea of an antagonistic relation of sociology to those methods, it emerges that sociology can seek to offer reflection, careful and accountable employment, extension, elaboration and must otherwise seek to preserve and enhance those methods, while making use of them in asking the whole host of questions that it wants to ask. The possibilities here are open but are largely unexplored because the standard sociological and philosophical impulse has been to see these disciplines as discontinuous competitors with some version of unenlightened understanding.
Normative sociology is not the answer
Sociology has recently seen itself as uniquely placed not only to provide a critique of social life but to put forth normative claims (e.g. Abbott, 2018; Modood, 2022; but see Tittle, 2004) that may be used to reconstruct institutions and social practices. It is imperative to appreciate that to intervene so as to remodel no less than to criticize presupposes that one has a grip on normative resources, that one possesses an accurate understanding of normative problems and practices. Moreover, to claim such a role exclusively, one must be in a special epistemic or moral position so that one has some warrant to not merely add one’s voice to that of participants (e.g. citizens, policy makers) but have some capacity to trump those voices or at least count substantially. If we understand social life as incorporating forms of interest and forms of relevant criticism (Bottero, 2019; Walzer, 1987), then the sociologist is indeed able to mine and rely on a vast range of critical resources. At the same time, however, those resources are also available to many others, who may be more skilled, wiser, morally more sensitive, possess integrity and the capacity to present themselves as moral exemplars. A realistic appreciation of this fact entails that sociology cannot set itself up as moral arbitrator or leader unless it can demonstrate its special position. While this cannot be precluded, it is by no means self-evident in the current disciplinary institutional arrangements under which sociological training, research and overall practice take place.
Decolonizing sociology as openness to other forms of knowing
The conception that has been sketched so far is an open conception that is highly serviceable in the context of attempts to decolonize knowledge (Bhambra & Holmwood, 2021; Go, 2016), precisely because it emphasizes the answerability to different forms of social life within and between societies that may be related in historically complex ways.
Winch (1958, 1964b) took aim at the derogatory attitude towards practices that do not resemble those of western science. As such he originated a conception of what is required for understanding other ways of knowing and coordinating them with ones we are more familiar with, while also pointing to the pitfalls that arise when we are not prepared to cast a candid glance on our own practices and to make an effort to expand our willingness to consider alternatives. This is not to advocate the complete disregard towards disciplinary constitution or the dissolution of the idea of an (expanded) canon. Far from it. In the current context of a heightened interest in decolonizing sociology, Winch is a precious forerunner who offers an eminently sensible and defensible non-relativist version of what an ‘interactive epistemology’ (Connell, 2021) might actually look like, one which is sensitive to differences between varied concepts of ‘knowledge’, ‘wisdom’ and so on and to the logic of various claims understood in relation to their social context. Different pieces of that mosaic are very different indeed, but their interconnection is premised on mutual openness and a commitment to conceptual accuracy.
Conclusion
In this article, I have endeavoured to spell out the claims and ensuing implications that can be extracted from a thorough understanding of Peter Winch’s The Idea of a Social Science and its Relation to Philosophy in light of the body of thought that has reflected on it over the past 65 years. The central lesson to be learned concerns the self-understanding of sociology and consists in the exhortation to sociologists not to continue to labour under indefensible conceptions of what they are doing, premised on misconceptions as to how the discipline is related to its subject matter and to other forms of inquiry. This lesson offered by Winch’s ISS is indeed incompatible with the officially sanctioned views that are defended in theory, enshrined in sociological curricula, repeated among practitioners and used by the profession in advertising itself to prospective members or outsiders. This is not to say that sociological practice is fully governed by these indefensible conceptions. Sociologists are not blind to problems with their discipline and in deed they indirectly show awareness of them and find ways of working around what they might otherwise proclaim. But such awareness escapes explicit acknowledgement and, therefore, does not allow any coherent conception of the genuine possibilities of the sociological endeavour to be assembled.
I have identified two central versions of the self-conceptions that are blocking the way: the technical method prerogative and the critical ability prerogative. According to the first, the sociologist by using their methods makes tractable and analysable the structures and mechanisms, trends and causal relations which govern social life but are not fully accessible to members. According to the second, the sociologist is a highly critical thinker whose training and abilities grant them the ability to see behind mystification and to pronounce on normative questions, for example, on minority group rights, redistribution of income or other political questions.
The first conception involves the assumptions that it is the formal method which makes social life tractable and hence its possession is an exclusive route to understanding, that connections in social life are made intelligible by it, and that social life is governed by a form of causality that can be captured by that method. These assumptions are challenged by all of Winch’s claims, particularly those which concern the fact that criteria of understanding, description and explanation are embedded in social life, are made available via participation and thus pose limits of relevance and accuracy as well as demands of second-order clarification.
The second conception involves the assumptions that critical thinking, by being sociology’s gift to social life, is not made possible by or continuous with social practices but is conducted externally without depending on sensitive use of the resources that are available therein. And furthermore, that it is a matter of demonstration rather than rhetoric and, thus, that it can be made the technical business of social science disciplines, which can be sources of normativity or pronounce authoritatively on what is desirable or what is to be condemned. These assumptions are also challenged by Winch’s claims, particularly those which emphasize open-ended rule use within a social context as necessary to reflection and the possibilities of self-critical self-understanding.
That Winch’s thought shows both conceptions to be incoherent disproves the view that his arguments can be disregarded as constituting a narrow attack on an obsolete idea of social science. His insights are rendered profound and perennial. It is these insights that need to inform a defensible and coherent conception of sociological practice. Exploring possibilities here simultaneously constitutes a response to the question of the relative uniqueness of sociology as compared to other social sciences but also to non-academic (professional and lay) forms of understanding.
The argument I have presented may be seen as in danger of sliding down a slippery slope, so I wish to end by offering some restraining thoughts. If the idea of the sociologist occupying an in principle privileged position must be discarded, then ought that commit one to the idea that the sociologist has no privilege at all, which would then render sociology unnecessary? Let me explain why not. A view that gets close to this latter position is that the sociologist acts as a mere conduit, perhaps for disadvantaged or underrepresented voices which otherwise would not be heard, and is therefore simply channelling others. This view depends on the existence of a profession that is already independently recognized and has the capacity to speak, and is using that capacity to channel the value of something else, but is not itself creating any value, so to speak. The variations upon this theme that highlight the role of the sociologist as emphasizing the knowledge or wisdom of minority groups or as using the sociological position in connection with their own membership in underrepresented groups to obtain critical distance by virtue of double membership (Collins, 2000) cannot get us out of the problem: they do not constitute arguments as to what academic sociology brings to the table. Such views do not help us escape from the aporia that if there is no privilege in sociology then it must be unnecessary.
But if we are to get away from this suspect slippery slope, then we need to acknowledge the difference between absolute privilege and relative privilege and say what relative privilege consists in. What is the distinctive or unique character of sociology and sociological ability? It might be that the two cannot be separated! The difficulty in discussing relative privilege and sociological uniqueness lies in how uniqueness is to be understood. It is not wise to see uniqueness as a uniqueness in methods (a view which cannot be maintained against other disciplines or general developed literacy), or as a uniqueness in ability to grasp the social, to be critical or possess practical wisdom (skills which may be cultivated within sociology, for sure, but not exclusively so, nor necessarily fit to serve as disciplinary aims).
We might opt instead for a uniqueness in terms of a set of second-order questions, endeavours and interests treated under responsive but different standards (explicitness, clarity, descriptive accuracy, conciseness, analytical depth, critical sensitivity, dispassionateness, comparative overview), though not necessarily overall better standards. Such standards may be seen as improvements, and they might enable profound critical reflection and lead to new ideas or heightened awareness much like knowledge of descriptive grammar, although second-order and continuous with linguistic ability, may lead to heightened awareness of language. All this presupposes a firm grasp of what the standards are and what the questions are – and here the relation to philosophy enters forcefully. Moreover, uniqueness need not be defined outside of a historical trajectory and record. Sociology is unique because it is historically unique. It has tended to pose certain questions and struggled to answer them. This shows why sociology is a historical discipline in at least one sense and why uniqueness can receive the form of knowledge of an open-ended disciplinary tradition coupled with participation in an ongoing conversation which constantly relates that tradition to a developing present-future and to first-order practices. But it may also be historical in a second sense, as dealing with human doings or res gestae, as Collingwood understood history to be, while also being philosophical in its concern with conceptual elucidation. There may be uniqueness in this combination too.
The sociologist learns to read and listen attentively, to think in an orderly manner, to write and argue carefully and explicitly in relation to certain academic standards (these are all diffuse skills to humanities and social sciences as they should be), to have a historical sense of the concepts and attempts to study society (which is why self-understanding is important) and to analyse social situations and problems afresh in relation to contingent circumstances (see Turner, 2004). No more. No less. Perhaps more than enough.
Together with the insights in Winch’s Idea, these observations can be incorporated in various ways into a varied range of approaches to social life, be they structural, synchronic, comparative, historical, focused on religious life, politics, culture and so on. They enhance the prospects for an accountable and intellectually respectable sociology.
Footnotes
Acknowdgements
The author would like to thank an anonymous referee for their valuable comments on a previous version of the article.
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.