Using Wittgenstein to respecify constructivism
Francis D. (2005) Using Wittgenstein to respecify constructivism. Human Studies 28(3): 251–290. Available at http://cepa.info/2932
Table of Contents
Constructivism and the Theoreticist Tendency
Lineaments of Constructivist Threoreticism
Radical Constructivism: Knowledge as Subjectivity
Some Conceptual Problems with Radical Constructivism
Against Linguistic Subjectivity
Escaping Constructivist Dualism
Bloor’s Social Constructionism: Knowledge as Consensus
Wittgenstein against Rule Scepticism
Bloorian Social Constructionism: No Place to Stand and Not Social Enough
Epilogue: A Note on Realism
Taking its orientation from Peter Winch, this article critiques from a Wittgensteinian point of view some “theoreticist” tendencies within constructivism. At the heart of constructivism is the deeply Wittgensteinian idea that the world as we know and understand it is the product of human intelligence and interests. The usefulness of this idea can be vitiated by a failure to distinguish conceptual from empirical questions. I argue that such a failure characterises two influential constructivist theories, those of Ernst von Glasersfeld and David Bloor. These are considered in turn. Both theories seek to give a general, causal account of knowledge: von Glasersfeld’s in term of cognitive subjectivity, Bloor’s in terms of social agreement. Ironically, given that both writers cite Wittgenstein as a source of theoretical inspiration, assumptions of both theories run counter to key Wittgensteinian arguments. To show that Wittgenstein’s views offer no solace to the realist, the article closes with a brief consideration of John Searle’s theory of knowledge.[Note 1]
In his book The Idea of a Social Science and its Relation to Philosophy (1958/1990), Peter Winch made the controversial claim that much that passes for empirical inquiry in the social sciences is concerned with issues that are not empirical ones at all. The problems addressed, he asserts, are actually conceptual in nature. As such, they are to be settled not by research but by conceptual analysis, that is, by gaining a clearer understanding of the “grammar” of our language. Inspired by the writings of Ludwig Wittgenstein, Winch argued that such an understanding would reveal many of the central issues of social science to be non-issues, at least from the point of view of empirical science. This claim guaranteed the book’s notoriety, a notoriety that it continues to enjoy today almost half a century after its publication.
In what follows I will argue that, properly understood, Winch’s claim is correct, at least in relation to some currently influential theories in the human sciences. A Wittgensteinian approach to the questions answered by such theories reveals precisely what Winch claimed it would, namely that those questions are misconceived. Empirical solutions are proffered to puzzles that have no empirical substance, at least as they are theoretically formulated. Thus the answers provided to these puzzles are not so much wrong as incoherent. The two qualifications just indicated are important. Winch has been misunderstood by numerous commentators down the years as arguing that there are no empirical questions in the social sciences, that their inquiries can be wholly subsumed within philosophy. A more careful reading, such as is provided by Colin Lyas (1999), reminds us that Winch’s point referred to general questions concerning the nature of the social. Social scientists, he argued, go astray when they regard these questions as ones demanding empirical answers of the kind provided by scientific investigations. It was no part of Winch’s case to claim that there is no room for empirical work in social science and that empirical questions are never matters of legitimate social scientific interest. Instead, the lesson to be taken from Winch is essentially this: Social science goes astray when it fails to remember that conceptual and empirical issues are like oil and water; they don’t mix and failure to distinguish the one from the other has the capacity to produce catastrophic results.[Note 2]
Constructivism and the Theoreticist Tendency
Nowhere is this point more clearly demonstrated than in “constructivist” thinking in contemporary social science. Constructivism is arguably the most influential and wide-ranging approach in the human studies. It is a many- headed beast[Note 3] with limbs that are as varied as they are numerous. Over the past thirty years or so social scientists have offered constructivist explanations for an enormously diverse range of phenomena, from social problems in general (Spector and Kitsuse, 1977) to specific ones such as alcoholism (MacAndrew, 1969), child abuse (Gelles, 1975; Hacking, 1991) and female crime (Ngaire, 1987); from mental illnesses such as schizophrenia (Barrett, 1988; Gilman, 1988) to physical ones like leprosy (Waxier, 1981) and diabetes (Posner, 1977); from identities in general (Michael, 1996) to racial identities (Frankenburg, 1993), sexual identities (Kitzinger, 1987) and identities of disability (Goggin and Newell, 2003); from the mind (Coulter, 1979) to the emotions (Harre, 1986); from media news (Tuchman, 1978; Altheide, 2002) to authorship (Woodmansee and Jaszi, 1994); from technologies of missile accuracy (Mackenzie, 1989) and fluorescent lighting (Bijker, 1992) to technological systems in general (Bijker, Hughes and Pinch, 1987); from mathematical reasoning (Bloor, 1982) to, most controversially, scientific facts and findings (Latour and Woolgar, 1979; Knorr-Cetina, 1981; Pickering, 1985).[Note 4]
What, if anything, do these studies have in common? I am tempted to answer “not much” yet at the heart of constructivism is an orienting idea, the deeply Wittgensteinian notion that the world as we know and understand it is a creation of human intelligence and interests, via practices in and through which that intelligence is realised and those interests defined. This view has radical consequences: it “historicizes” rationality and undermines the claim that knowledge in any field of human experience derives from methods and procedures that transcend the circumstances of their use. It thereby opens up to empirical social scientific investigation what Michael Lynch (1993) terms “epistopics.”[Note 5] In so doing constructivism shifts the focus of attention of investigations into the character of human knowledge. In a kind of reverse- Winch, constructivists like Lynch seek to show how forms of rationality, the study of which has traditionally been regarded as the province of philosophy, can profitably be respecified as practices and thus topics for social research.
However, just as Winch insists on a proper demarcation between philosophical questions and social scientific ones, constructivists must be careful not to lose sight of the differences between conceptual problems and empirical ones. As Wil Coleman and Wes Sharrock (1998) point out, the founding text of social constructionism, Peter Berger and Thomas Luckmann’s (1966) The Social Construction of Reality, is notable for specifying its sociological project in a way that distinguished that project from epistemology. In Berger and Luckmann’s own words, the book is a manifesto for the sociological study of “everything that passes for knowledge in society.” In other words, they define the sociological task as accounting for how ideas acquire the status of knowledge and the consequences of that status for the roles such ideas come to play in social life. They are clear that this task is distinct from the one epistemology sets itself: to assess the logical justifications for attributing such status.
If Berger and Luckmann were conscious of the need to respect the conceptual/empirical divide, more recent approaches have been far less so. The sheer diversity of constructivist studies has seemed to some to demonstrate the need for a general, unifying theory. There is, discernible within constructivist thought, what Wittgenstein (1967) famously called a “craving for generality.” While writers like Lynch eschew foundationalist concerns and are clear that their focus is upon knowledge as practice, producing studies of a resolutely empirical kind, among some constructivists there is a wish to address a larger agenda. This desire for general theory has lead them towards foundationalist theoreticism.[Note 6] They aim, in effect, to provide social scientific solutions to philosophical problems. Notwithstanding the appearance of philosophical profundity such conflation gives to this kind of work, such projects are characterised by confusion and contradiction. As I hope to show, these constructivists should have read Winch more carefully.
Two influential constructivist approaches that manifest this theoreticist tendency are those of Ernst von Glasersfeld (1989; 1995a; 1995b) and David Bloor (1976; 1983; 1996; also Barnes, Bloor and Henry, 1996). The former goes under the title of “radical constructivism,” while the latter is more associated with the label “social constructionism.” For ease of reference I will refer to them hereafter as “GRC” (Glasersfeldian radical constructivism) and “BSC” (Bloorian social constructionism). In what follows I will examine these theories in turn. In relation to both GRC and BSC the discussion will seek to show that, notwithstanding the fact that constructivists/constructionists frequently cite Wittgenstein’s writings in support of their views (von Glasersfeld, 1995a; Bloor, 1996), a proper understanding of key Wittgensteinian arguments fatally undermines any claim that these theories present empirically relevant insights into the nature of knowledge.
Lineaments of Constructivist Threoreticism
Before getting to specifics, let me begin with some general comments about the differences and similarities between GRC and BSC. The most obvious difference is that the two theories have different disciplinary origins and locations. GRC has its home in psychology and draws upon the ideas of key figures in 20th century psychology, most notably Jean Piaget. BSC resides in sociology and involves ideas taken from major thinkers in that discipline, especially Emile Durkheim. Notwithstanding these disciplinary differences, however, GRC and BSC have several important features in common.
The first and most obvious similarity concerns the cognate fields in which GRC and BSC exert their influence. The former has had considerable impact within science education, the latter within the sociology of scientific knowledge (SSK). This is no coincidence, given the second common feature: the intellectual ambition they share. Both GRC and BSC offer a general constructivist theory of knowledge. In each case the goal of the theory is nothing less than to explain the foundations of human knowledge in terms that refute rationalist philosophical analyses of such foundations. Given the special status accorded to scientific knowledge within that philosophical tradition, it comes as no surprise that GRC and BSC take science as the main empirical application of their explanations.
A further similarity is that the two approaches understand their theoretical task in essentially the same way, since both take the general foundations of knowledge to involve a certain sort of belief, or rather, meta-belief. This is our belief in the objective nature of reality. According to GRC and BSC, it is this meta-belief that underpins science and all other empirical human knowledge. Thus explaining the foundations ofknowledge requires answering the question: “How is it that we (human beings) believe in an objective reality, about which it is possible to have scientific knowledge?”
This question of the possibility of science is one which has traditionally been the province of philosophy. Indeed, it could be said to be the defining question of the Western philosophical tradition since Rene Descartes. The fourth common point between GRC and BSC, then, is their critique of that tradition. Philosophy, both approaches agree, has failed to provide a satisfactory answer to the question of the possibility of knowledge. The reason for this failure is its mistaken assumption that the answer is to be found through philosophical analysis, whereas a correct solution to the puzzle is only to be arrived at via theoretically informed empirical investigation. GRC and BSC conceive themselves as empirical sciences of human knowledge. If our belief in objective reality is a necessary one, GRC and BSC insist that this necessity is of the causal rather than the logical kind. The problem of the possibility of human knowledge is an empirical one, no different in kind, though greatly different in scope, from other problems of the human sciences. Thus both GRC and BSC aim to give theoretical shape to empirical research programmes. Science education and SSK are fields in which empirical studies proliferate, many of which are carefully argued and substantively informative. However, the methodological and theoretical differences displayed by such studies means that they offer little basis for any systematic theory of scientific knowledge. In SSK, for example, the conceptualisation of the social character of scientific knowledge varies so widely across the field that its studies, taken together, give little prospect of a systematic sociological explanation of science. The same holds true with respect to science education. Whether GRC and BSC are capable of bringing theoretical order to their respective empirical fields depends, among other things, on the coherence of their assumptions.[Note 7]
The fifth and final similarity between GRC and BSC is the fact that in their respective critiques of philosophy they identify the same basic adversary. This is the so-called “copy theory” of knowledge, the view that knowledge exists when ideas reflect or correspond to those things in the world to which they refer. This theory is associated with the philosophical tradition of empiricism, though whether any significant empiricist thinker in the history of philosophy ever held this theory in any straightforward way is open to question.[Note 8] Regardless, GRC and BSC together view the copy theory of knowledge as both the main legacy of the Western philosophical tradition and as entirely misconceived. An adequate conception of knowledge sees it not as a copy or reflection of an independent, pre-existing reality but as a process of construction whose product is reality. The question is: what kind of process is this and what are its elements?
GRC and BSC, of course, offer very different answers to their common question. GRC views knowledge as essentially an individual, psychological construction. Knowledge is theorized as a mental state, the task of the theory being to explain how this mental state arises. BSC, on the other hand, conceives of knowledge as a social phenomenon, as the product of social relations and activities. The possibility of knowledge thus resides in shared understandings among the members of a group or society. Before I consider these two approaches in more detail, however, it might be useful to indicate, in a rough and initial way, some grounds for thinking that the question they seek to answer may not be as intelligible as they suppose. On the surface it appears a properly formed empirical question that should allow of an empirical answer. After all, if we hold a particular belief, it should be possible in principle to account for why we do so. Thus, if it is true that we believe in an objective reality then it is not unreasonable to ask why and how we hold to this view. Yet this belief does seem a rather curious and distinctive one. For one thing, to refer to it as a “meta-belief” marks the fact that it is not the sort of belief we are familiar with in daily life. It is not like, say, someone’s belief that they know the best Chinese restaurant in the city, or that they could have made it as a professional cricketer if only they had been given more opportunity. Neither is it much like socially shared beliefs with which we are familiar, such as the beliefs of Christians in the risen Christ or Marxists in class struggle. It is not simply that it is not the sort of belief that can be specified in terms of identifiable believers: who believes it and who does not. It is also that, unlike the examples above, it is a belief in relation to which it would make little sense to ask people why they hold it.
But what evidence is there that we do hold it? For proponents of GRC and BSC there is no difficulty in providing such evidence; indeed, it is available in just about everything that human beings do. Just look around you. Whether you are an Marxist or a Christian, a lover of Chinese food or a thwarted cricketer, you conduct your affairs on the basis of a distinction between ideas and the world. The ideas we possess are one thing, the world to which we apply those ideas is something else. This distinction seems to be one that is universally made and taken for granted by human beings. Its validity is confirmed for us on a daily basis, such as every time one mislays the house keys, scalds one’s tongue on a hot cup of coffee or misses a train through misreading the timetable. The very notion of “knowledge” presupposes that what we think – individually or collectively – may or may not fit with what is really the case, since “know” is an achievement word and where there is knowledge there also can be ignorance and error. One can suppose, therefore, that the concept of knowledge involves commitment to a general assumption of the objectivity of “reality.” In respect of the many and varied ways we attempt to establish whether our ideas to correspond to reality or not, this assumption does seem to operate as a “meta-belief.” Or, at least, so it seems to GRC and BSC.
Arguably, there is something odd about the idea that a universally necessary meta-belief underpins all human knowledge. To anticipate the argument I will develop later, does it actually make sense to suppose that “the real world” is something about which persons might have such a belief? Is it intelligible to conceive of a phenomenon called “objective reality” that functions as the object of human knowledge and to regard all knowledge as unified in its assumption of such a phenomenon? What do these propositions mean? If they possess intelligible meaning, then the projects of GRC and BSC cannot be dismissed a priori. But, conversely, if there is no sense to these propositions, then GRC and BSC can make no contribution to our understanding of human knowledge, since their answers are solutions to a non-existent puzzle.
Radical Constructivism: Knowledge as Subjectivity
So much for preliminaries. In the remainder of this paper I will consider these two approaches in turn, beginning with an exposition of the key ideas of the approach, and then present some critical observations from a Wittgensteinian viewpoint. The stated aim of radical constructivism is “to show that a relatively stable ’experiential reality’ can be built up without presupposing an independent world-in-itself” (von Glasersfeld, 1995a: 88). This reality is built up by the human mind progressively over the period from birth to adulthood. The work of building it is not straightforwardly cumulative or linear but involves distinct phases, in which not simply the “content” of experiential reality changes but so also do the structures that organize this content. Underlying these developments, however, are certain “master processes,” general mechanisms and principles that are differently realized at each stage of reality construction. The role of these processes in GRC theorizing is to account for how it is that cognition emerges from pre-cognitive activities. Clearly, scientific and mathematical knowledge is massively different from the physical gropings of the young infant. The claim that these constitute the poles of a single developmental process can only have plausibility if some sort of continuity is posited between them. Such continuity is provided by Piaget’s (1929/1997) concepts of “assimilation” and “accommodation.” The basis of knowledge comprises “action schema” that arise when the individual (or “organism” in GRC terminology) acts towards things in its environment. Such actions generate payoffs for the individual, and these payoffs establish predispositions to repeat the action, an action schema. Each further action in response to some stimulation then results in some modification of the patterns that have been built up.
These modifications are of two kinds. The first process, assimilation, involves the extension of an action schema to new stimulations (“experiences”). Since such schema amount to a template or model of the experience-action connection, assimilation resembles the process of applying such a template to new objects. The second process, accommodation, occurs when an action schema fails to result in the expected outcome. It involves an adjustment of schema in the face of “perturbations,” stimulations which obstruct and thereby “disequilibrate” existing action schema. New schema are created which restore equilibrium, thus moving development on in a qualitative way. The culmination of a series of such qualitative changes is the emergence of structures of “objective knowledge” consistent with (and making possible) scientific understanding of the world.
I note that this theory has strong similarities with behaviourism. Like the behaviourist, von Glasersfeld wishes to account for intelligence in terms that do not pre-suppose mental states. Both approaches conceive cognition as patterned behaviour (or “action” in GRC) and both view such behaviour as the product of causal processes. A key difference, of course, is that while behaviourism is committed to the redundancy of mental states, GRC wants to show how the cognitive structures that define intelligence develop out of pre-cognitive behaviour. Another important difference is the epistemological point of view each approach adopts and that defines its theoretical stance. Behaviourism’s rejection of subjectivity involves the adoption of a “third person” epistemological stance; phenomena are described strictly from the observer’s point of view. Only that which is describable from such a point of view – “behaviour” – is allowed within the theory. GRC adopts the opposite stance, rejecting the notion of objectivity. This notion presupposes “ontic reality,” reality as-it-is-in-itself. In denying such a reality, GRC rejects the assumption that phenomena can be viewed from any point of view other than the subjective, first person perspective.
In a reverse image of behaviourism, then, GRC attempts to expunge any terminology that assumes objectivity or smacks of a third person perspective. Thus instead of stimulus and response, it speaks of “experience” and “action.” It insists upon the ’fundamental subjectivity of experience” (von Glasersfeld, 1995a: 72). This means, for example, that the concept of “environment” has a quite different sense in GRC than in behaviourism. For the latter, this concept signifies those conditions whose capacity to affect the individual is available to independent observation and description. For the former, it refers to the “experiences” the individual is capable of having, given the cognitive structures she or he possesses. In other words, that which comprises the individual’s environment at any point is subjectively constituted. Thus the development of knowledge is to be conceived in wholly “internal” terms. The individual builds such knowledge through an internal, self-regulating process in which “external reality” plays no causal role and has no independent explanatory function. GRC also rejects the concept of truth as irremediably objectivist. Ideas are not adopted into the individual’s cognitive structures because proved true, but because they are “viable.” The viability of action schema, and later of concepts and conceptual operations, consist in their usefulness in relation to the individual’s goals or purposes. GRC assumes all action to be purposive, from the first attempt of the baby to suck at a nipple to the astronomer’s attempt to measure the distance between two stars. A single underlying principle governs such diverse actions: any construct that facilitates the attaining of a given purpose will be retained and strengthened, while those that fail to do so will be revised or abandoned. This principle of “adjustment” governs all human activity, both pre-cognitive and cognitive. The development of knowledge, therefore, is not to be conceived as involving constructs that “correspond” progressively with a mind-independent reality, but as the emergence of increasingly systematic structures to support ever more complex forms of purposeful action.
The concept of viability occupies a central place in GRC thought, since it provides an account of how knowledge differs from “mere thought” without assuming the correspondence conception of knowledge said to be characteristic of traditional epistemology. Truth, it is argued, presupposes that ideas can be assessed against an independent reality, with only those that correspond to such reality being granted the status of knowledge. This traditional view is rejected as incoherent. How, asks GRC, is the individual to access objective reality independently of the cognitive structures she or he possesses, the very structures that make experience possible in the first place? Yet abandoning the correspondence view of truth does not mean that “anything goes” and that cognition is unconstrained. It is simply that the constraints must themselves be understood with reference to the same cognitive structures that they limit. Michael Hardy and Peter Taylor (1997: 139), explicating von Glasersfeld’s views, summarize the process by which judgments of viability reorient cognition as follows:
The learner re-presents and compares experiences in an effort to determine what was unique about the perturbing experience and why her initial model of experience failed to account for it. Further, the learner often examines consciously her experiential model, that is, engages in ’reflected abstraction,’ in order to understand why her initial action produced an unexpected or undesired result. Regardless, while synthesizing a viable solution the learner utilizes reflected abstraction to reorganize her model of experience and the activity that model guides. Once a viable solution is constructed, the perturbation is neutralized and cognitive equilibrium is re-established.
The reference to “re-presenting” experiences in this quotation introduces another key element in GRC’s account of cognition. Traditional epistemology, it asserts, conceives of ideas as “representations” of reality. But to speak of representation, GRC insists, implies the existence of the thing represented, thus such a conception is objectivist and implicates the correspondence notion of truth. But jettisoning it leaves the GRCist with a problem. How is one to account for the individual’s ability to relate a current experience to a previous one? The very notion of action-schema, not to mention concepts and cognitive operations, presupposes such an ability; concepts and conceptual structures are built up through comparison of experiences and new experiences have a cognitive impact by being referred to previous ones. Some account must be given of this process if the notions of assimilation and accommodation are to be plausible.
By introducing a hyphen into the word representation, GRC seeks to distance itself from the traditional view of ideas as “pictures” while retaining the notion that the process of relating one experience to another involves some kind “retention”: the storage of a previous experience such that it can be imaginatively replayed and compared against a current one. But to posit such a storage device would seem to run counter to GRC’s developmentalism; what kind of storage device can it be that precedes and makes possible the fundamentals of cognitive actions? Clearly it cannot be innate. Von Glasersfeld finesses this problem by arguing that the storage device is itself an adaptive product. He coins the terms “proto-space” and “proto-time” to refer to the initial spatial and temporal features of the device, features that become more systematically realized as organized cognition emerges. However, he acknowledges (1995a: 95) that beyond this GRC has little to offer concerning this crucial element of the theory:
I do not know how re-presentation works. In fact, no-one, today, knows how it works. We have not even the beginnings of a plausible functional model of human memory, let alone a model of human consciousness. Yet, something we want to call memory, as well as something we want to call consciousness are involved in the kind of replay of past experiences that I was describing.
The final element of GRC I will discuss is its conception of the social dimension of experience. What does GRC have to say about the interpersonal nature of knowledge? After all, knowledge is something shared, both at the everyday level and at the level of science and mathematics. The concepts we employ and the beliefs we have are ones we hold in common with others, and the activities in and through which knowledge is put to work in our lives are social activities. The belief in objective reality that GRC aims to explain is a belief in a world that is, in some sense at least, the same for everyone. At some point, then, GRC needs to explain how knowledge becomes interpersonal and collective. In line with its basic approach, GRC’s analysis of the interpersonal dimension is developmental, in two related respects. First, it argues that the capacity to cognize other persons is a sophisticated one that emerges over time, issuing out of more basic forms of cognition. In the early stages of development, the individual experiences reality in a wholly “egocentric” fashion, making no distinction between her- or himself and other objects, let alone distinguishing her or his own perceptions from those of others. Only when egocentric thought has been overcome and “decentred” cognitive structures are in place is the individual capable of cognizing others and regarding her or his environment as an interpersonal one. The mutually constituted phenomena of “self” and “other” begin as models, but subsequently become assimilated as a reality. The second developmental significance of interpersonal awareness concerns its role in the emergence of objectivity. The progressive decentring of cognition that makes it possible to cognize others also engenders a grasp of reality’s independence from any one knower. All vestiges of solipsism drop out of the individual’s thinking as her or his understandings are “corroborated” in and through interaction with other people. On the basis of such corroboration, knowledge begins to extend beyond the immediate experience of the individual. She or he takes it that she or he “knows” things never experienced personally but encountered only indirectly through communication with others. At this intersubjective level (von Glasersfeld, 1995a: 120),
one is led to believe that concepts, schemes of action, goals and ultimately feelings and emotions are shared by others, and, therefore, more ’real’ than anything experienced only by oneself. It is the level on which one feels justified in speaking of ’confirmed facts,’ of ’ society,“social interaction’ and ’common knowledge.’
Other persons, then, function as sources of “corroboration” of the subjective belief in a determinate external world. By means of agreement with others the world as we are familiar with it is made possible. As will become clear later in the paper, this radical constructivist view corresponds to a key assumption of social constructionism.
Some Conceptual Problems with Radical Constructivism
It is an understatement to say that there is much in the above that is unclear and/or puzzling. I will focus on three related issues that are especially problematic and for which Wittgenstein’s writings have particular relevance. They are: (a) GRC’s notion of knowledge as “internally” constructed, (b) the distinction between “experience” and “knowledge” that is central to GRC’s account of concept formation and (c) the concept of “re-presentation.” There are numerous other difficulties that will have to be left for another time. Among these are the misleading use of biological and cybernetic analogies in GRC’s talk of “adaptation” and “equilibration,”[Note 9] the ill-defined character of the concept of “viability,” GRC’s conception of the mind as some kind of mechanism, and the systematic ambiguity between physicalist and intentionalist description in GRC theorizing (e.g. referring to persons as “cognizing organisms”).[Note 10] I will also make no comment on GRC’s account of the subjective construction of mathematical knowledge, nor on the strange coincidence that the active, structure-creating mind constructs precisely and only those logico-mathematical structures already formulated by logicians and mathematicians
There is a strong rhetorical flourish to GRC’s alleged shift away from “traditional epistemology” towards a radically different conception of knowledge. Yet what sort of change does GRC actually involve? Despite the rhetoric, the claimed radical break from conventional philosophical thought seems anything but; it is thoroughly imbued with assumptions drawn from the tradition of philosophical skepticism deriving from Descartes.[Note 11] The notion of the mind as subjective and the dichotomy between objectivity and subjectivity that drives GRC’s conception of knowledge are cornerstones of that tradition. At the heart of skepticism is the puzzle of how it is possible for the subjective mind to contain within itself knowledge of an external, objective world. In the Cartesian tradition the internal and external realms stand in opposition to one another in every way. Crucially, they differ in the kind of access to them the individual can have. The contents of the mind can be known directly in an unmediated way: the individual has only to “look inwards” to see her or his ideas displayed there. Access to the external world, in contrast, is indirect and deeply problematic. The only contact the individual has with the world is via her or his senses, and these are utterly unreliable as a source of knowledge. The problem is not simply that our senses may sometimes deceive us. After all, recognizing certain perceptions as mistaken or delusive presupposes the possibility of others that are valid. The more serious difficulty is that the information received by the mind from the senses is conditioned by the contingent nature and capacities of our sensory equipment. Therefore, asks the skeptic, how can one be sure that such information corresponds to that which it seems to inform one about? How can a contingent phenomenon guarantee the validity of one’s knowledge?
GRC is correct in its assertion that none of the proposed philosophical solutions to this problem have convincingly resolved it. The question is whether GRC’s attempt fares any better. The answer, I suggest, is no; its account of the internal processes of knowledge construction is arbitrary and conceptually confused. Take, for example, the explanation of concept construction. As noted above, in von Glaserfeld’s account the foundation of knowledge consists in the way in which concepts are constructed through a process by which “experiences” are compared over time. There are several difficulties with this account. The first concerns the distinction between “knowledge” and “experience,” in which the former is conceptual and the latter not. Now it is perfectly true that in everyday language these concepts are quite distinct, in that they have different uses (for example, we talk of having “good” and “bad” experiences, but not of “good” or “bad” knowledge). However, GRC conceives of them in a specific theoretical fashion, as standing in a causal relationship. Therefore one can ask whether the theoretical distinction it makes is a coherent one.
What, then, is an “experience”? For experiences to explain concept formation presupposes that whatever an experience is, it is extra-conceptual in nature. But, as W. A. Suchting (1992) points out, following Wittgenstein (1967), experience is both a modal concept (one does not simply “experience” per se – one hears, smells, sees, feels, imagines, dreams, etc.) and an object-defined one (it is “intentional” in phenomenological language). In other words, one has an experience of something, whether this involves a sensation (a feeling of hunger, a sense of excitement), a perception (a bright morning, a loud bang) or neither (a bad dream). Furthermore, the thing one has experience “of” defines the experience, or to put it differently, the object of the experience constitutes its nature. Thus, for example, if one experiences a bad dream, then “bad dream” is both what one has had an experience of and the nature of the experience one has had. This basic conceptual point renders incoherent the claim that concept formation can be causally explained in terms of experiences, since any experience – to be an experience – must already be conceptually identifiable. This does not mean, of course, that one can always describe the experience clearly and precisely. Sometimes the best one can do is to characterize it or describe it by analogy (“It was a bit like. …”).
It is important to be clear that the sense in which the concept, experience, is “object defined” is a strictly grammatical one, since otherwise one could be mislead into thinking that having an experience involves the existence of some mental thing that the person possesses. Thus GRC might respond to the points above by asserting that what it is referring to is not experience in any “linguistically described” sense, but rather the existence of some (pre- linguistic) “subjective feel” that the individual possesses as a result of a given action. This notion of the phenomenal quality of an experience, conceived as some sort of mental thing, goes under the heading “qualia” in contemporary cognitive philosophy (Seager, 1999). By invoking the notion of qualia GRC might seek to avoid the difficulties raised above, claiming that “what it feels like” to the individual as she or he experiences something is a phenomenon wholly subjective and pre-conceptual. Thus in the case of the young infant who has yet to acquire language, “subjective feels” are all there is to experience. Language develops as a result of “inner comparisons” made between these subjective phenomena, through the way in which the individual assigns labels to types of experience that have the same or similar “subjective feel.”
However, the notion of “subjective feels” is no more coherent than that of concept-free experience. The concept of qualia requires one to accept that there is a “way it feels” to have any particular experience, where that phenomenal quality is something distinct from what it is that is being experienced. It is important to remember that this concept is not referring to the individual’s attitude to what she or he is experiencing; the point is not whether she or he is relieved, sad or angry at perceiving this or that. Rather, it asks one to believe that every experience is accompanied by a subjective state known only to the individual. Thus, allegedly, there is a subjective feel that accompanies seeing a book on the desk in front of one, or hearing the telephone ring. As Peter Hacker (2002) notes, Wittgenstein (1980: para.896) anticipated this view in his writings on the philosophy of psychology. He wrote:
The content of experience. One would like to say ’I see red thus’, ’I hear the note that you strike thus’, ’I feel sorrow thus’, or even ’This is what one feels when one is sad, this when one is glad, etc. One would like to people a world, analogous to the physical one, with these thuses and thises. But this makes sense only where there is a picture of what is experienced, to which one can point as one makes these statements.
Wittgenstein’s point is that the idea of the “subjective experience of experience” is deeply obscure. It makes perfect sense in respect of certain kinds of experiences to ask “how does it feel?” For example, if a patient complains of a pain, the doctor may well ask, “What sort of pain is it?” “How bad is it?” and so on. But what is being asked about here is the pain itself, there is no relevant distinction here between the object of experience – the pain – and how it feels. The claim that such a distinction is a general feature of all experience, like the idea that one must always have an attitude to the things one experiences, makes no sense. The concept, “qualia,” cannot come to GRC’s aid for the simple reason that it is a non-concept.
Against Linguistic Subjectivity
Despite their force, these points might still be insufficient to convince the proponent of GRC that she or he is mistaken in attempting to detach experience from concepts. Surely one would not deny that children have experiences prior to the development of conceptual understanding? Such a denial would seem absurd; the child hears, sees, feels all kinds of things before possessing the concepts by which such experiences might be described. The infant cries because it is hungry or in pain. Experience – of these kinds, at least – comes first, concepts later; how can one argue with that? Of course, one would not deny this obvious point but GRC is wrong in assuming that this truism provides any support for its theory of knowledge. It would only do so if its conception ofconcepts as subjective constructions was valid.
To see why this conception does not follow from these trite observations I will turn again to Wittgenstein, to his famous discussion of the notion of “private language.”[Note 12] The general aim of this discussion is to establish the social character of language. Wittgenstein’s strategy for demonstrating this is to expose the incoherence of the notion that the meaning of an experience can be fixed subjectively through introspection. A private language, in Wittgenstein’s terms, is one which an individual creates by assigning labels her or his experiences through a process of inward inspection of those experiences. The key point is that she or he does this not by employing the known concepts of a shared language, but by creating concepts de novo through reflection upon her or his inner experiences. Since the experiences are private only the individual having them can know “just what” they consist in; consequently, the meanings of the words used to identify them can only be known only by the individual herself or himself. Only she or he will know what is being referred to by a particular expression; all others will have to guess or infer its meaning. This is essentially the view held by GRC. The entire GRC account of concept formation presupposes that individuals assign meaning to their experiences by inward inspection, concepts thereby being formed on the basis of inner judgments of sameness and difference. Von Glasersfeld (1989: 133) asserts “the essential and inescapable subjectivity of linguistic meaning,” while Hardy and Taylor (1997: 141) say that
Language is comprised of symbols which have no inherent meaning and must, therefore, be interpreted. Accordingly, neither symbols nor linguistic expressions acquire meaning prior to being associated with one or more concepts, and since concepts are internal to the knower so are linguistic and symbolic meanings.
The privacy of language, then, consists in the way an individual gives names to her or his private experiences. These names become concepts as further experiences are recognized as “the same” as ones already named. But what does it mean to say that one can give names, for example, to this or that sensation, not by employing a public language of shared concepts but by creating concepts de novo that refer only to the private sensations one is having? Wittgenstein points out that under the skeptical assumptions that ground this view, it is not just the meaning of the same sensation word for two different speakers that is problematic. If meaning is nothing more than a word – private object correspondence, there is nothing to guarantee that the same word will have the same meaning on any two occasions of its use, no matter who is using it. Thus, even in the case of a given individual using the word on two separate occasions, the problem recurs. How can she or he be sure that what she or he means by “backache” now is what she or he meant by it last week? One might say “Well, it has the same meaning because she or he can tell that the pain (the object) is the same.” But what is meant by “the same” here? The very notion, sameness, only makes sense in terms of publicly accountable criteria. In another famous passage (1967; para 258), Wittgenstein proposes
Let us imagine the following case. I want to keep a diary about the recurrence of a certain sensation. To this end I associate it with the sign ’S’ and write this sign in a calendar for every day on which I have the sensation. – I will remark first of all that a definition cannot be formulated. – But still I can give myself a sort of ostensive definition. – How? Can I point to the sensation? Not in the ordinary sense. But I speak, or write the sign down, and at the same time I concentrate my attention on the sensation – and so, as it were, point to it inwardly. – But what is this ceremony for? for that is all it seems to be! A definition surely serves to establish the meaning of a sign. – Well, that is done precisely by the concentration of my attention; for in this way I impress upon myself the connection between the sign and the sensation. – But ’I impress it on myself’ can only mean: this process brings it about that I remember the connection right in the future. But in the present case I have no criterion of correctness. One would like to say: whatever is going to seem right to me is right. And that only means that here we can’t talk about ’right.’
Wittgenstein’s point is that the idea of an individual naming a sensation through a process of “inward pointing” makes no sense, since to name something one must be able to distinguish and recognize instances of the thing named. This requires, in effect, that the individual gives herself or himself a definition of the name by indicating to herself or himself the private object to which it refers. Thus to keep a diary in which each experience of a given sensation is marked by the name “S” presupposes that the sensation is independently identifiable as such each time it occurs. But how does the individual know that what she or he is feeling now is the same as what was felt before?
GRC tells us that individuals can “re-present” their experiences – a sort of internal re-running or re-imagining of what was seen, heard, felt, etc. on a previous occasion. Such experiences are “stored” somewhere in the mind and when retrieved can be compared against a current experience to judge whether they are the same. But this elaboration of the private language story simply generates more difficulties. How does the individual know where to look to find a relevant previous experience; what is she or he looking for? How does she or he decide that “this” memory rather than “that” is the best one to select for comparison with what is being experienced right now? As Rush Rhees (1970: 60) notes:
This is not a question of whether I can trust my memory. It is a question of whether it makes sense to speak of remembering; either of a good memory or a faulty one. If I thought I could not trust my memory, then of course I might look for confirmation. But there cannot be any question of confirmation here, nor any question of doubting either. There is just no rule for what is the same and what is not the same; there is no distinction between correct and incorrect; and it is for that reason that it does not make any difference what I say. Which means, of course, that I say nothing.
Ironically, given its espoused opposition to the “copy theory of knowledge,” GRC’s difficulties arise out of the assumption that there are two separate things: the experience of the sensation on the one hand and the naming of that experience on the other, such that the meaning of the name involves a relationship of correspondence between the two. But if there can be correspondence there can also be mis-correspondence. A definition is a rule of use and, as Wittgenstein notes, where there are rules it must be possible to make mistakes, to mis-apply the rule. In other words, it must be possible, under this account, for the individual to experience, say, a pain, to mis-describe it, or be mistaken about it. However, the notion of “believing that one is in pain but being wrong,” like the notion of “believing the pain was in my leg when it was actually in my chest,” makes no sense. If sincerely stated, such assertions indicate that the speaker is delusional.
GRC’s conception of concept formation as a process of “inner naming” runs up against an even more troublesome Wittgensteinian argument. The early sections of Philosophical Investigations comprise a systematic critique of the view that language consists in names and that the foundation of linguistic meaning is the act of naming an object. This view, a cornerstone of empiricist thought at least since Locke, is rejected on two main grounds. First, Wittgenstein points out that naming objects is just one kind of use that is made of language and that there are no theory-independent reasons for regarding it as any more “basic” that any other uses. Secondly, he argues that naming is itself a sophisticated linguistic activity, one that presupposes a mastery of a range of other linguistic and social competencies: “Only someone who already knows how to do something with it can significantly ask a name” (1967: para 31).
With these points in mind we can return to the case of the infant crying with hunger or pain. The parent hears the child’s cries and can infer that it is, say, hungry, but surely only the child itself “knows” what the cry means; despite all the arguments above, the temptation remains to say that there is something essentially subjective about sensations and therefore about the meaning of the ways we give expression to them. The GRC view must be right about this, surely? Wittgenstein makes two important points here. The first concerns the position of the parent (or other observer) who witnesses the crying and the kind of knowledge she or he has of the child’s state. He notes that talk of “inferring” another’s sensation state is highly misleading. One does not “infer” such things as pain from “outward signs”; this is not the correct way to understand the relationship between pain and pain behaviour. Pain behaviour is not an indicator of pain, in the sense in which, say, a red evening sky indicates good weather on the morrow. Rather, crying or screaming are, in the appropriate circumstances, criteria of pain. That is to say, in those circumstances one knows that others are in pain because one can see that they are. Another way of putting this is to say that the relationship between pain and its expression is a “grammatical” or normative one rather than contingent and inferential. Given the pain behaviour, one is entitled to say that the other is in pain, just as one is entitled or required to react in the appropriate way, e.g. by giving help. The crucial point is that these are facts about us, parents and other competent users of the language, and about the language we share. They are not facts about the child as a subjective experiencer.
So what, then, of the infant itself? Is it correct to say that the child “knows” it is in pain, and what kind of pain it is in, and that its behaviour is an outward expression of this knowledge? No, it is not. First, one does not “know” one is in pain, since knowledge, in its verb form at least, is an “achievement” word. It simply makes no sense to ask of someone “But how do you know you’re in pain?” “How did you find out?” “What makes you so sure?” etc. Secondly, what can “kind of pain” possibly mean here? How can the child know the sort of pain it has before it possess the relevant concepts? We teach the child what pain is by teaching it the proper ways to express sensations in language. Thus it learns what the difference is between an ache and a sting, or a headache and a stomachache, by learning the language we use and the ways we use it. Therefore sensations are not to be thought of as “private objects,” but as part of a linguistically realized “way of life” shared with others: “Grammar tells what kind of object anything is” (Wittgenstein, 1967: para. 373). For Wittgenstein, then, an adequate appreciation of sensations and sensation terms begins with the recognition that having sensations and expressing and describing them are intimately related. Their relationship constitutes a social practice, one which we learn by virtue of being members of a human community.
Escaping Constructivist Dualism
The above arguments demonstrate that GRC presupposes a dualistic perspective in which the possibility of knowledge is held to be accountable only in one or other of two opposing ways. On the one hand knowledge can be viewed as a relationship between the mind and an “external,” mind-independent reality, such that ideas that properly count as knowledge are those that provide an inner mental “copy” of the outer objects and properties to which they refer. On the other hand, knowledge is seen as a wholly internal, the objectivity of reality being explained as the product of an emergent relationship among ideas themselves. On the assumption that these two theories define and exhaust the range of epistemological options, GRC offers what it takes to be convincing arguments for the rejection of the former, leaving the latter to command the field. No serious consideration is given to the possibility that its fundamentally 17th century dichotomy between an “inner” subjective realm and an “outer” objective one may be conceptually misconceived. Such a possibility is confirmed in the later writings of Wittgenstein. Wittgenstein’s arguments demolish the thesis of epistemic privacy and the subject/object dualism from which it derives. They present a powerful case for a social view of knowledge in which ideas are not the components of individual minds but the constituents of shared human practices. Thus Wittgenstein’s thought marks a genuinely radical break with the Cartesian tradition within which radical constructivism remains enmeshed. Little wonder, then, that his writings have been a major influence in the formulation of a very different constructivist approach known as “social constructionism.” It is to this brand of constructivist thought that I now turn.
Bloor’s Social Constructionism: Knowledge as Consensus
Like von Glasersfeld’s radical constructivism, Bloor’s social constructionism is committed to the goal of a causal theory of knowledge. Unlike GRC, however, BSC conceives the causes of knowledge to reside in the social structures of collective life rather than the working of the individual mind. Bloor draws heavily upon a reading of Wittgenstein to support his view that knowledge is essentially social and that it is by means of agreement among the members of a community that the objectivity of phenomena is constructed (Bloor, 1992). Thus scientific knowledge, like all other forms of human knowledge, is made possible by the institutional organization of society and the ways in which social factors constrain and shape thought. To say that knowledge is socially constructed is to assert the dominance of such factors over the processes of knowledge production. The research agenda of BSC, known as the “strong programme,” thus consists in empirical demonstrations of the role of social factors in shaping the epistemic content of science, technology, mathematics and so forth.[Note 13]
The fact that BSC seeks to explain such fields of formal knowledge in terms of social factors gives rise to a contrast between “internal” and “external” explanations (see Barnes, 1974, ch.5). The characteristic (indeed, defining) feature of formal knowledge is that the processes of knowledge production are themselves socially organized, such that knowledge claims in any such field are subject to institutionalized practices of assessment and decision. Another way of putting this is to say that science and other kinds of formal knowledge are epistemically self-defining. Within science, justifications of knowledge claims are required to be cast in terms that invoke science’s institutionalised practices and the evidential standards and criteria those involve. This fact has enormous implications for the kind of sociological explanation of knowledge that BSC aims to provide. It means that in a field such as science “explanations” are already available for why a particular theory or finding has that status of accepted knowledge. There is a sense in which it is part of science’s essential business to make available such explanations. Thus if one were to ask why, say, relativity theory replaced Newtonian dynamics as a foundational theory in physics, any physicist can provide an answer, where that answer will be cast in terms of the scientific reasons for this transformation.[Note 14]
In contrast to such “internal” explanations, BSC aims to explain scientific knowledge in ways that step beyond the self-referring accounting practices of science itself It distances itself from such practices by adopting a number of methodological assumptions that directly contrast, so it claims, with those of internal accounts. These methodological assumptions (Bloor 1976) are of causality, impartiality, symmetry and reflexivity.[Note 15] Briefly, these amount to a commitment to explanations of scientific knowledge (a) that are cast in causal rather than normative, reason-giving terms, (b) that make no distinction for explanatory purposes between accepted and rejected scientific ideas, (c) that display the same form for true as for false beliefs, and (d) whose form applies equally to the practices and findings of SC as to any other branch of science. Thus, while SC seeks “external” explanations relative to the accounts of scientific knowledge generated by the internal accounting processes of any given scientific field, it regards itself as consistent with (indeed, as a continuation of) the overall project of empirical science to establish systematic and general empirical knowledge.
To pursue its goal, then, BSC must redefine the epistemic status of internal accounts of knowledge. If such internal explanations are deemed adequate there would seem to be no need – indeed, no “epistemic space” – for an explanation in terms of extra-scientific factors. Therefore, in order to create such a space, BSC has to argue that internal accounts of knowledge are necessarily inadequate and incomplete. Bloor invokes Durkheim (1976) in asserting that the reason-giving practices integral to science are to be seen as ceremonial activities whose function is to enhance science’s “sacredness,” rather than bona fide explanations of why a particular theory or finding was found compelling. Additionally, and more importantly, BSC treats such practices as theories and marshals epistemological arguments in an attempt to show that the explanations deriving from such theories are logically incapable of accounting for scientific knowledge. These epistemological arguments come in various forms and draw upon a number of different sources, but what they have in common is the claim that knowledge is under-determined by the kinds of explanatory factors appealed to in internalist theories. In other words, however conventionally reasonable, plausible or compelling they may appear, such accounts do not explain how it is that a given piece of knowledge comes to have that status. I will mention just two of the epistemological arguments BSC recruits to its cause: the “Duhem-Quine” thesis and Wittgenstein’s remarks on rule following.
One argument that Bloor leans upon heavily concerns the alleged under- determination of theory by evidence. This argument, known as the DuhemQuine thesis, states that the choice between alternative theories in a field cannot be made by appealing to a given body of empirical evidence, however elaborate this body of evidence may be. There are two reasons for this. First, any theory is a complex network of concepts and assumptions which only “attaches” to empirical evidence at certain points, such as specific empirical predictions derived from the theory. Consequently, any mismatch between a theory-derived prediction and empirical data leaves open a number of possibilities apart from wholesale rejection of the theory, such as the revising of a particular aspect of the theory network to accommodate the problematic observations. In W. V. O. Quine’s famous words “our statements about the external world face the tribunal of sense experience not individually but only as a corporate body” (1953: 41). The second reason involves the “theory-laden” nature of observation. Quine’s rejection of the distinction between analytic and synthetic truths undermines the positivist view that scientific theory is grounded in a theory-independent “observation language.” Any observation statement involves concepts whose meaning is shaped by the theoretical framework that defines the phenomena to which the concept intendedly refers. This means that the notion of what a theory is has to be expanded. Rather than observation standing over against theory, the former is shaped by the latter such that what counts as relevant evidence is effectively part of the theory itself.
The impact of these arguments, at least as far as BSC is concerned, is to render internalist explanations of scientific knowledge incapable of accounting for theory choice. By appealing to such things as “the experimental evidence,” such explanations generate a relativist situation in which any one of a number of competing and irreconcilable theories could legitimately claim to be justified. That theory conflicts in science do get resolved, with one theory emerging as scientifically accepted while its competitors are consigned to the history books, thus is a fact that requires a different kind of explanation. Scientific knowledge cannot be accounted for simply in terms of factors internal to scientific practice, and the “boundary” dividing such practice from the wider society in which it is embedded must be seen as weak and permeable.
While the Duhem-Quine thesis has been a major buttress for Bloor’s project, an even more powerful resource has been found in Wittgenstein’s writings on the problem of rules and rule following. Lengthy sections of both Philosophical Investigations (Wittgenstein, 1967) and Remarks on the Foundations of Mathematics (Wittgenstein, 1978) are devoted to this topic. Human behaviour, Wittgenstein argues, is normatively ordered, that is, involves the following of rules. Thus in relation to both ordinary language and mathematics, one can ask what such rule following amounts to, i.e. what the correctness of a rule- following action consists in and how the ability to follow a rule correctly is to be accounted for. In line with his general philosophical position, Wittgenstein does not regard such ability as residing “inside the head” of the rule follower as some kind of mental structure; neither, for the reasons discussed earlier, does he believe that following a rule can be conceived as any kind of essentially “private” individual act. He insists that rules are social phenomena and rule following is a learned, social ability. But, and this is the key question, just what kind of social ability is it?
In his interpretation of Wittgenstein’s views on this question, Bloor broadly follows the position taken by Saul Kripke. According to Kripke (1982: 60), “Wittgenstein has invented a new form of skepticism.” Wittgenstein’s analysis of rule following involves a “sceptical” view in which the rule as stated underdetermines the actions that constitute its correct application. Kripke refers to this as “Wittgenstein’s paradox,” taking as his key text the opening sentence of para. 201 of Philosophical Investigations:
This was our paradox: no course of action could be determined by a rule, because every course of action can be made out to accord with the rule.
Wittgenstein illustrates this problem by asking one to imagine a teacher instructing a child how to add in twos. The teacher demonstrates adding in twos by showing the child examples of the correct series up to 1,000 and by setting tests which the child passes successfully. She or he then asks the child to continue the series beyond 1,000, whereupon the child writes 1,004, 1,008, 1012, and so on. The teacher objects that the child has not followed the rule “Add in twos,” whereupon the child replies “I thought that was how I was meant to do it.” Not only that, but further attempts to show the child that she or he has misunderstood the rule are unsuccessful: she or he continues to insist that her or his actions conform to the rule as stated. What is the basis of the teacher’s certainty that her or his view – that the child has deviated from the rule – is the correct one? What is it that makes her view right and the child’s wrong? After all, the teacher’s instructions involved showing the child in a limited set of examples what correct following of the rule involves. Nothing, one might say, has been demonstrated beyond those examples. Relating the rule to new cases would seem to require a judgment; therefore who is to say that the child is wrong when it applies the rule as it does to numbers beyond 1,000?
Wittgenstein’s “paradox” remark seems to be suggesting that whatever determines this, it cannot be an appeal to “what the rule says,” since it is precisely that which is at issue between the teacher and the child. But if the correct application of a rule does not rest upon what the rule says, what does it rest upon? How does any rule user actually know what the correct way of following a rule is, as opposed to what she or he might think it is? Kripke finds the answer to this question in the paragraph immediately following 201, where Wittgenstein says
And hence ’obeying a rule’ is a practice. And to think one is obeying a rule is not to obey a rule. Hence it is not possible to obey a rule ’privately’: otherwise thinking one was obeying a rule would be the same thing as obeying it.
Kripke takes this remark to indicate that the solution to the skeptical problem is to be found outside the rule itself and beyond any single rule user, in the existence of a “community” of rule-users (1982: 89ff). Since the skeptical problem cannot be resolvedfrom within, its solution is to be found in the shared understanding of the rule that is enforced within the community. The teacher, or any other rule user, can appeal to such shared understandings to justify her or his application of the rule. Only given such shared understandings, Kripke argues, can Wittgenstein’s distinction between “thinking one is obeying the rule” and actually obeying it stand fast. In conforming to the “community view,” the individual rule user knows what correctly applying the rule involves.
Ever since its publication, Kripke’s interpretation of Wittgenstein’s remarks on rule following have been repeatedly and powerfully criticized by Wittgenstein scholars (Baker and Hacker, 1984; Hanfling, 1985; Diamond, 1989; Putnam, 2001; most of the contributors to Miller and Wright, 2002). Despite this, Kripke’s views have been adopted pretty much wholesale by Bloor, who has consistently championed a Kripkean Wittgenstein (or “Kripkenstein” as Putnam (2001) prefers to call him) as a source of theoretical support for BSC. The reason is not hard to see: Kripke opens the door to a causal, sociological explanation of rule following. Thus under the heading of “rule finitism” (Bloor, 1983; 1996; Barnes et al., 1996) Bloor presents a version of Wittgenstein’s views that differs from Kripke’s only at the margins. Bloor’s version of the argument emphasizes the role of consensus within a community of rule users. Such consensus, he asserts, is what transforms mere “belief” into knowledge; it is the foundation of the sense of “objectivity” that attaches to the knowledge the members of the community share. Without such consensus there is nothing to limit the interpretive possibilities linked to a rule. With it, the meaning of the rule is “fixed” and has the status of both objective fact and normative requirement.
For BSC, then, objective knowledge is made possible by the institutionalization of consensus in society. It is the fact that members of society agree in speaking and acting in particular ways that explains the “sense of objectivity” their knowledge has for them. The rules, facts, confirmed theories, and so forth, that comprise science and mathematics, no less than the ideas and beliefs that make up common sense knowledge, are “constructed” out of such consensus. In this respect they are no different than other institutionally produced social objects, such as money. Thus what makes a piece of metal a coin, says Bloor (1996: 29), is the agreement to treat it as such within the community:
Metal discs are coins because they are called coins.. .. On this usage, speaking of a thing as a coin isn’t meant to refer to a purely verbal act, but to the whole pattern of behaviour into which such explicit verbalizations are woven. Bearing in mind the compression involved, the important point is that the group practice of calling a certain type of object a coin, makes that object into a coin.
Of course, as a good Durkheimian, Bloor does not view this consensus as the product of any kind of “contract” between individuals. Rather, it is a phenomenon sui generis, a “social fact” of an irreducible kind. As Durkheim (1976: 434) proposed, concepts are “the work of the community,” not simply in the sense that they play a part in communal life but in the causal sense that they are “collective representations” made possible by and reflecting the organized nature of social life. He further insisted that it is by virtue of their social basis that our concepts have for us the character of reflecting an objective world (1976: 437): “A collective representation presents guarantees of objectivity by the fact that it is collective.” Finally, like Durkheim, BSC emphasises that the social determination of knowledge extends all the way up to the most formal aspects of thought. Durkheim (1976: 444) states that “attributing social origins to logical thought is not debasing it or diminishing its value or reducing it to nothing more than a system of artificial combinations; on the contrary, it is relating it to a cause that which implies it naturally.” For both Durkheim and Bloor, the correct explanation of the very possibility of science, mathematics and logic is to be found in the scientific study of society.
Wittgenstein against Rule Scepticism
There are grounds for questioning whether either the Duhem-Quine thesis or the “rule finitist” interpretation of Wittgenstein provide the foundation for BSC that it claims they do. To discuss the problems with the former would take a paper at least as long as the present one. Suffice it to say that both Quine’s rejection of the analytic/synthetic distinction and his holistic conception of knowledge are highly contentious and have met with strong resistance, especially from Wittgensteinian scholars (see, for example, Arrington and Glock, 1996; also Hunter, 1995). However, even were these problems put to one side, the Duhem-Quine thesis would not provide BSC with support for rejecting internal scientific accounts, since it has nothing to say about such accounts. What the Duhem-Quine thesis denies is a certain philosophical account of scientific reasoning, one that claims a tight deductive relationship between theory statements and empirical statements in science. The assumption of such a relationship is a cornerstone of rationalist idealizations of science, in both their positivist and fallibilist varieties. However, unless one is naïve enough to assume that such philosophical idealizations closely resemble the way science actually works, the Duhem-Quine thesis has nothing whatsoever to say about actual practices of scientific reasoning.[Note 16]
The problems with the latter are more easily stated. There are grounds to argue that Bloor’s account of rule following is both incoherent and unWittgensteinian. Let us look again at what Wittgenstein says about rules. One place to begin to see that the Kripke/Bloor reading of Wittgenstein on rule following is dubious is with para. 201 itself Kripke quotes only the first sentence, but the entire paragraph reads as follows:
This was our paradox: no course of action could be determined by a rule, because every course of action can be made out to accord with the rule. The answer was: if everything can be made out to accord with the rule, then it can also be made out to conflict with it. And so there would be neither accord nor conflict here.It can be seen that there is a misunderstanding here from the mere fact that in the course of our argument we give one interpretation after another; as if each one contented us at least for a moment, until we thought of yet another standing behind it. What this shows is that there is a way of grasping a rule which is not an interpretation, but which is exhibited in what we call `obeying the rule’ and ’going against it’ in actual cases.Hence there is an inclination to say: every action according to the rule is an interpretation. But we ought to restrict the term ’interpretation’ to the substitution of one expression of the rule for another.
Do these remarks justify the claim that Wittgenstein believes there is a “sceptical paradox” associated with rule following? They seem to suggest the very opposite of this. The under-determination of rule following actions comes about because there is no way of telling, from the rule as stated, which among a number of possible interpretations of its meaning and thus its application to new cases is the correct one. But notice that this is precisely the view that Wittgenstein refers to as a “misunderstanding.” To conceive of rule following as involving interpretations of the rule is where we go wrong, he seems to be saying. One requires an interpretation, presumably, because the rule in and of itself is insufficient to guide one as to what is required. However, this assumes a questionable distinction between “the rule” on the one hand and “actions that implement the rule” on the other. It is as though these are distinct phenomena that are only contingently related. Because this relationship is causally weak, something else is required to mediate it, i.e. an interpretation. And because there is a plurality of possible interpretations, something else again is required to determine which one will prevail: the socialisation of the rule follower into a “community consensus.” But this seems to be the view that Wittgenstein is criticizing, not the one he is endorsing! The idea that rule following is interpretive is what he is rejecting in para. 219 when he says: “When I obey a rule I do not choose. I obey the rule blindly.”
Kripke’s “paradox” arises in part because of a misconception about what is being referred to when one speaks of “the rule.” The skeptical reading perceives a gap between the rule and its application because it takes the former to be referring to a statement, such as “Add in twos.” This statement makes no mention of the situations in which it is to be applied, therefore how can it determine what one is to do in such situations? But this is false problem, since it is simply a mistake to equate a rule with a statement of that rule. The rule is not the statement, it is what the statement (if it is correct) states. Thus the rule may be capable of being stated in a number of different ways. Clearly, what makes these various statements expressions of the rule is that what they are stating is the same, not that they are verbally identical.
This error feeds into a further one: the notion that a rule stands towards a rule following action as an explanation of it. Here as elsewhere in the Philosophical Investigations, Wittgenstein’s method is to state as powerfully as possible a view that might be taken about some aspect of language and human action, such that the reader is provoked into agreement and the grounds of the plausibility of that view are clearly exposed. He then picks away at those grounds with argument after argument until that plausibility is revealed as a chimera. So it is in the discussion of rule following. As Warren Goldfarb (2002: 105) notes, the idea that a rule explains rule following actions is identified by Wittgenstein as just such a plausible but misconceived view:
This elaboration can seem ineluctable; it is certainly familiar. In saying that Wittgenstein means to provoke it, however, I do not mean that he finds it acceptable. In fact, I believe he wishes to depict it as the first, and fatal, misstep, the decisive move in the conjuring trick.
Bloor’s “rule finitism” poses the problem of how a rule whose meaning has been learned with reference to a finite set of examples can explain what one should do when presented with a new situation. But, as Gordon Baker and Peter Hacker (1984), Michael Lynch (1992b) and Wes Sharrock and Graham Button (1999), among others, point out, to understand a rule is to know what is involved in applying it, whether to situations one has encountered before or those one has not. To say that the novice has learned the rule, but only in respect of the cases in which it has been exemplified, such that nothing can be said about how she or he might respond to new cases, makes no sense. What is she or he supposed to have “learned” if she or he cannot see how it is that the cases she or he has been shown as exemplifying the rule actually exemplify it? If all she or he has learned to do is to imitate the teacher, what she or he says may be “correct” but one cannot say that she or he has learned the rule. She or he is doing imitating, not rule following. What, then, is the difference between following a rule in familiar and in unfamiliar cases? The point is this: if one does not know how to correctly follow the rule “Add in twos” when one gets past 1,000, one does not understand the rule – not simply in this “new situation” but at all. The rule and its applications are not related contingently but grammatically (in Wittgenstein’s use of this term). Thus the rule does not “explain” the rule following action, nor can it “fail to explain” it, because rule and action do not stand in that sort of relationship.[Note 17]
This does not mean, of course, that we do not sometimes cite a rule to explain a particular action. We do so, for example, when someone who does not know the practice of which the rule constitutes a part asks “why has she done that?” (e.g. someone watching a chess match who does not know chess.) We may answer this question by explaining the rule that is being followed by the actor. But notice what has been explained here. The questioner’s problem is not that she or he knows the rule but cannot see how it relates to what is being done; no, the puzzle arises for her or him precisely because she or he does not know the rule, and therefore cannot recognize what she or he is seeing as a rule following action. Once she or he knows the rule the puzzle disappears, since now she or he can see what is being done.
This last point brings me to something crucially important. As just indicated, rules do not exist in isolation but form part of social practices, or what Wittgenstein refers to as “language games.” Persons learn rules in and as they learn how to participate in this or that practice. The important word here is participate. Learning rules is not to be thought of as a cognitive process or achievement, but as an interactional one; one learns how to take part successfully in certain kinds of activities involving others. To say that an activity is rule governed is to say that there are ways things have to be done if they are to count as part of the activity in the first place. Furthermore, practices in their turn are not isolated phenomena; they are located in our social lives with reference to other practices, together with which they comprise a complex way of life. One does not have to endorse Quinian holism to recognize the internal connections between the rules and activities that make up a practice and the inter-relatedness of the various rule governed practices that make up our life together.
This is where Bloor’s causal approach breaks down. He rightly insists on the social nature of rule following, but understands this to mean that “community agreement” is a causal factor which explains rule following’s normative character. Such agreement, in other words, explains the possibility of social practices (such as scientific inquiry). As Button and Sharrock (1993) have noted, Bloor’s concept of consensus is a cognitive one. The members of a community construct a shared, and therefore “objective,” world by thinking and speaking in the same ways. Because of such agreement they are able to follow the same rules. But this cognitivist notion of consensus cannot explain what Bloor claims it can. It is only with reference to an existing practice and the place of that practice in our lives that it makes sense to speak about collective agreement in his sense. What counts as such agreement is a normative matter and as such is constituted by the practice; thus it is the practice that makes such agreement intelligible, not the other way round.
The simplest way of illustrating this point is with reference to Bloor’s own example of coins. He says that “calling a certain type of object a coin, makes that object into a coin.” In other words, it is the consensus within the community of coin users that transforms what would otherwise be simply metal discs into coins. The reality of coinage is made possible by a shared definition of what a coin is. But several questions need to be asked here, such as: How many members of the community need to agree before a coin is a coin? and What kinds of metal discs are members of the community agreeing to call coins? The first question is really raising the issue of what Bloor means by a “community.” It would appear to be essentially a statistical notion (as his reference to “patterns of behaviour” confirms), thus something akin to a population. But a population is a very different sort of thing to a community (or “society”); since the latter is a normative phenomenon whereas the former is not. To be a member of a community involves much more than simply thinking and speaking in ways that match other people. One’s relations with others have a moral status and this means, among other things, that one is obliged to respond to certain things in the same way as them. The normative character of coins, then, is not the product of a “consensus” that transforms non-normative metal discs into normative coinage, but of a bunch of other normative stuff, such as the nature of the State as the body with monopoly authority over the society’s currency. One is tempted to mis-quote Clifford Geertz (1973) here and say that social life is normative all the way down.
The point can be taken further by noting that coins are not just any metal discs, but rather ones with a particular range of size and weight. Think what daily life would be like if pennies were twelve inches across and weighed twenty pounds. With the exception of the ruin it would bring to the personal fitness industry, it is hard to think of any positive consequences. The serious point is that coins are as they are because of their place in our way of life. One does not have to research the social history of coinage to be able to say that it is the conditions of that way of life that have constrained the properties coins must have if they are to be of use to us. In other words, the practice of using coins is shaped by other practices.
To all of this Bloor might respond that Wittgenstein himself emphasizes the role of agreement in social practices. He makes reference to it at various points in the Philosophical Investigations, most notably at para. 241, where he says:
‘So you are saying that human agreement decides what is true and what is false?’ – It is what human beings say that is true and false; and they agree in the language they use. That is not agreement in opinions but in form of life.
Notice once again Wittgenstein’s dialogic method. He begins by giving expression to an interlocutor’s view, in order to clarify how his own view differs. His notion of “agreement in form of life,” I suggest, is several light years away from Bloor’s “consensus.” The best way to appreciate this is by asking: What would a form of life be like in which its members did not agree? Perhaps it would be something like a community in which some people use pennies for doing weight-training exercises instead of as currency, while others read newspapers and books from the bottom up (line-by-line) rather than the top down, still others believe that wine is for washing one’s hair with and shampoo is for drinking with friends and – most interesting of all – some regard defecation as a public act to be done only in the full view of others. These Pythonesque thoughts make it clear, I hope, that what Wittgenstein is saying here has nothing whatsoever to do with conceiving of “agreement” as an explanation for anything.
Bloorian Social Constructionism: No Place to Stand and Not Social Enough
Where does this leave us with respect to BSC’s project? The above discussion of rules and practices and what a correct Wittgensteinian view of them involves might seem somewhat esoteric, but it raises serious difficulties for the “externalist” explanation of knowledge that BSC proposes. I noted earlier that this conception runs foul of the fact that practices such as science and mathematics are epistemically self-defining. What makes any assertion recognizably a knowledge claim within science or mathematics is its internal relations to the network of normative practices that define what can count as doing, say, physics or calculus. Any attempt to formulate an external account of such claims simply presupposes such relations, since it is only by virtue of them that any phenomenon to be “explained” exists in the first place. In other words, the relations between a knowledge claim and the normative context which is its disciplinary field are “grammatical” and mutually defining. And this holds not just for accepted knowledge claims, but also for non-accepted, controversial ones. As Lynch (1992b) notes, this has profound consequences for BSC; it vitiates the methodological assumptions of causality, symmetry and impartiality that define its project. For example, the idea that BSC can treat all knowledge claims equivalently regardless of their epistemic status is incoherent, since what can count as a knowledge claim is an internally defined matter (Lynch, 1992b: 294):
A controversial theory’s very standing as a theory about which there is controversy is ’internally related’ to the equipment, techniques, literary practices, observation language and accepted concepts in a field. This public relationship holds even when historians, or the scientists who originally promulgated the theory, later characterize the theory as a ’misunderstanding’ or mistake.’ Consequently, not every imaginable alternative to accepted theories in a discipline counts as a controversial theory, nor can an outside analyst presume to apply a policy of symmetry to every conceivable claim about disciplinary matters of interest. There is no room in the world for such a (non)judgmental standpoint.
Lynch’s point has equally profound significance for BSC’s basic assumption: the need for “external” causal explanation. This assumption is defended on the basis that there is something “missing” from internal disciplinary accounts, but the argument rests on the assumption that such accounts are explanations and, as such, are only contingently related to the knowledge claims they seek to explain. Just like the Kripkian rule follower who thinks she or he knows why she or he follows a rule in the way she or he does (“because that’s what it says”), but whose explanation is inadequate, BSC’s scientist or mathematician is one who can give an account of why something is correct but who fails to see the limitations of that explanation. However, as Lynch’s remarks suggest, to conceive of accounts in such terms is to miss the complex interplay of elements that constitute a disciplinary field as a normative context and the role that accounting practices have within that context. Only by removing disciplinary accounts from their context and treating them as free-standing attempts at explanation does their “validity” become questionable. But the language game of account giving is not separate from the practices of science; it is a central and constitutive part of those practices. The production of such accounts is wholly misconceived if it is treated as science attempting (and failing) to explain itself. What it actually is is science simply getting on with its business.
I began this paper by suggesting that it is characteristic of some contemporary constructivist thought in social science to conflate conceptual with empirical questions. The consequence is that although such theories seek to claim the status of empirical science, what they are actually engaged in is philosophical argument. What is more, the philosophical theorizing they engage in is of a particular kind, the sort that is standardly referred to as “foundationalism.” Both GRC and BSC seek to reveal the foundations of human knowledge; their aim is nothing less than to explain how it is that empirical knowledge, whether of the “disorganised” everyday kind or the organized, formal knowledge of science, is possible. In the current intellectual climate in the human sciences, such attempts at foundationalist theorizing are courageous if not foolhardy. But the mere fact that something is unfashionable should not be an argument against it, any more than reckless bravery justifies exemption from critical comment. If GRC and BSC are to be judged misconceived projects, this assessment has to be grounded in a close examination of the assumptions they make and the arguments they offer.
In presenting such an examination I have argued that GRC and BSC involve assumptions for which no convincing justifications can be given. More seriously, they lead to claims about the nature of knowledge and its acquisition that are essentially incoherent. It is on the basis of these arguments, and not out of any postmodernist prejudice, that I am led to conclude that the foundationalist goals of GRC and BSC are unrealizable. Both approaches conceive the question of the possibility of knowledge to be an intelligible one, capable of being given a general answer. Furthermore, both pursue this answer by putting on one side the practices in and through which knowledge actually functions in human life, in order to focus on the realities that lie behind these practices. But in their attempt to locate these realities, GRC and BSC are looking in the wrong place. Knowledge is not constructed by hidden processes – whether located in the recesses of the human mind or in society-at-large. Issues concerning what is “physic-ally real” (i.e. real given the concerns and techniques of physics) or what can or cannot be mathematically proven are internal to the practices that constitute these fields of inquiry, as are the judgments that resolve them and the arguments establishing that no satisfactory resolution is possible. This fact has a notable consequence. It follows that the construction of knowledge takes place where scientists and mathematicians (not to mention historians, novelists and poets) think it does, in the activities that comprise their intellectual work.[Note 18] If one wishes to describe what that construction consists in, to show how the practices of intellectual inquiry and creation relate to their outcomes, it is to these activities one must direct one’s attention.
Epilogue: A Note on Realism
It perhaps should be made clear before finishing that the Wittgensteinian arguments presented in this paper give no comfort to so-called “realist” critics of constructivism. Especially so since I have accused constructivists of “conflation,” a charge which is also made against them by realists. According to such critics, constructivism fails because it conflates two fundamentally different questions: (1) in what ways is knowledge the expression and product of human, social endeavors? and (2) how is it that such knowledge has successful application to the world beyond those endeavors? The realist insists that these questions demand answers of a different order such that any answer to the former has no bearing whatsoever on answers to the latter. Constructivism is accused of treating the second question as reducible to the first, thereby conflating representations of natural phenomena with the phenomena themselves. For the realist, explanations of “real” or “natural” facts cannot be reduced to or identical with explanations of “social” or “cultural” facts, since natural phenomena and social phenomena have existentially different properties.
If the point being made here was simply that natural science and social science are separate enterprises concerned with quite different phenomena, such that explanations in the one have little or nothing to do with those given in the other, then in the spirit of Winch one would sit back and applaud. Unfortunately realists argue something more controversial and incoherent. An example is John Searle’s (1995) The Construction of Social Reality. Searle presents a realist theory of knowledge in explicit opposition to constructionism. The core of his argument is a distinction between two kinds of facts: “institutional facts,” which depend for their very existence upon representations, and “brute facts,” which do not. He argues that the truth of these two sorts of facts is differently constituted. Searle is happy to accept a constructivist, Bloorian view about the truth of institutional facts; this consists in agreement among the members of the relevant culture, or what Searle calls “collective intentionality.” But he insists that the truth of brute facts cannot be accounted for in this way. For something to be a brute fact means that it stands in an extra-social relationship to that to which it refers. Realism, for Searle, tells us that “the world is independent of our representations of it” (1995: 153), and that the truth of at least some of our statements consists in their correspondence to the constituents of such a world. Furthermore, brute facts are ontologically prior to institutional facts and underpin them. Searle’s argument here is that the representational character of the latter involves the assigning of a meaning to some pre-existing entity, the latter providing the “raw materials” for the construction of the former. While for most institutional facts such entities are themselves symbolically defined, at a certain point the meaning assigning process must touch rock bottom, with an institutional fact being created by assigning meaning to a brute fact, the nature of which is not determined by any practices of human agreement. In Searle’s words, “The ontological subjectivity of the socially constructed world requires an ontologically objective reality out of which it is constructed” (1995: 191).
The problems with Searle’s analysis are manifold.[Note 19] It assumes an ontological dualism of “social” versus “natural” phenomena that is as difficult to sustain as the 17th century dichotomy between “primary” and “secondary” qualities, of which it is the contemporary descendant (Hacker, 1987). The difficulties with the correspondence theory of truth are too well known to need rehearsing here (Austin, 1961; White, 1970; Strawson, 1971). Likewise the representationalist view of knowledge (Toulmin, 1976; Rorty, 1979). Suffice it to say that one does not have to subscribe to constructivist theory to find Searle’s distinction between institutional and brute facts unconvincing. The flaw in the argument is clear as soon as one considers some examples of “brute facts” (e.g. daffodils are yellow; the moon is a sphere). Searle is correct, of course, that neither the colour of daffodils nor the shape of the moon is created by human beings. (Is this what he thinks constructivists believe?) That is what we mean when we speak of such things as daffodils and moons as “natural objects.” But insofar as these features are facts it makes no sense to say they are “independent of representations,” since like any other facts they are constituted by concepts. It is not simply that concepts are necessary in order to state what the facts are. Without concepts and the practices of which they form part, there would be no facts to state in the first place. How can there be a fact of a daffodil’s colour or the moon’s shape without our practices of distinguishing colours and shapes? As Wittgenstein has taught us so effectively, concepts are human institutions. Speaking of concepts as “representations” is too simplistic (and too cognitivist). It entices one to forget the intimate relationship between concepts and the human practices they make possible and within which they play their various roles.
Wittgenstein provides us with reminders of just this relationship. What realists such as Searle specifically need reminding of is that “fact” itself is a concept and, as such, has a grammar of use that we ignore at our philosophical peril. This grammar is elaborate yet quite precise. For example, we can “uncover” facts but not “open” them; “hide” them but not “mislay” them; “chase them down” but not “follow” them; “digest” them but not “eat” them; “face” them, even “turn one’s back” on them, but not “greet” them; “assemble” them but not “make” them, still less “make them up.” Do we do these things differently for “institutional” than for “brute” facts? Simply posing this question should suffice to make it clear that Searle’s distinction has no purchase here. Facts are facts.
Searle’s realism, like von Glasersfeld’s constructivism and Bloor’s constructionsism, exemplifies Winch’s (1958/1990) point that many theories in social science amount to “misbegotten epistemology.” For Winch, as for Wittgenstein, the origin of such misconceived theorizing lies in the tendency to treat certain words as though they hold the key to a special realm of mysteries. The correct response to this tendency is not to seek new theoretical solutions to the pu771es so formulated, but rather to show that the puzzles are dissolved as soon as the terminology used to express them is returned to its contexts of ordinary use. As Lyas (1999: 202) notes in his discussion of Winch, this applies as much to the concept of the real as to any other:
When I reflect on the way in which the term ’real’ is used I may see that, in one use or in some cluster of its uses, when I say that something is real, I can only do so if I am prepared to say that the something to which I am referring exists independently of me. Our practices commit us to the independent existence of things.
Like constructivist theoreticism, realist theoreticism loses sight of the mundane grammatical realities that give shape to language and the world. In so doing it also fails to recognise the impossibility of its quest, to see that it is seeking an answer to a question that cannot intelligibly be stated. Both theoreticisms wish to account for the possibility of knowledge, conceived as a relationship between “representations” and “the world.” But there is no answer to the question “how is knowledge possible” – not because we lack an adequate theory but because the question itself is misconceived. Constructivists and realists alike would do well to heed Wittgenstein’s (1967, para 97) remark that
if the words ’language’, ’experience’ and ’world’ have a use, it must be as humble a one as that of the words ’table’, ’lamp’, ’door.’
They might then appreciate that the abstractions they assume under such headings as “knowledge,” “reality,” “fact,” and “representation” have no solid connection with the actual uses of these concepts in the practices of speech and action from which they derive and that comprise their natural home.[Note 20] Wittgenstein’s and Winch’s arguments provide telling grounds for the view that the debate at the level of general theory between constructivists and realists is nugatory, since both attempts to explain knowledge amount to little more than empty exercises in theoreticism. The conclusion to which those arguments inexorably lead is that there is no “problem of how knowledge is possible,” therefore nothing to have a theory (still less a theoretical debate) about.
Altheide D. (2002) Creating Fear: News and the Construction of Crisis. New York: Aldine de Gruyter.
Arrington R. & Glock H.-J. (eds.) (1996) Wittgenstein and Quine. London: Routledge.
Austin J. L. (1961) Truth. In: Philosophical Papers. Oxford: Clarendon Press.
Baker G. P. & Hacker P. M. S. (1984) Scepticism, Rules and Language. Oxford: Blackwell.
Barnes B. (1974) Scientific Knowledge and Sociological Theory. London: Routledge and Kegan Paul.
Barnes B., Bloor D. & Henry J. (1996) Scientific Knowledge: A Sociological Analysis. London: Athlone Press.
Barrett R. (1988) Clinical Writing and the Documentary Construction of Schizophrenia. Culture, Medicine and Psychiatry 12: 265–299.
Berger P. & Luckmann T. (1966) The Social Construction of Reality. New York: Doubleday/Anchor.
Bijker W. (1992) The Social Construction of Fluorescent Lighting, or How an Artifact Was Invented in Its Diffusion Stage. In: W Bijker and L. Law (eds.) Shaping Technology/Building Society. Cambridge: MIT Press.
Bijker W. (1993) Do Not Despair: There is Life after Constructivism. Science, Technology and Human Values 18: 113–138.
Bijker W., Hughes T. & Pinch T. (eds.) (1987) The Social Construction of Technological Systems. Cambridge: MIT Press.
Bloor D. (1976) Knowledge and Social Imagery. London: Routledge and Kegan Paul.
Bloor D. (1982) Polyhedra and the Abominations of Leviticus: Cognitive Styles in Mathematics. In (ed.) L Essays in the Sociology of Perception. London: Routledge and Kegan Paul.
Bloor D. (1983) Wittgenstein: A Social Theory of Knowledge. London: Macmillan.
Bloor D. (1992) Left and Right Wittgensteinians. In: A. Pickering (ed.) Science as Practice and Culture. Chicago: University of Chicago Press.
Bloor D. (1996) Wittgenstein, Rules and Institutions. London: Routledge.
Brown J. R. (ed.) (1984) Scientific Rationality: The Sociological Turn. Dordrecht: Reidel.
Button G. (ed.) (1993) Technology in Working Order. London: Routledge.
Button G. & Sharrock. W. W. (1993) A Disagreement over Agreement and Consensus in Constructionist Sociology. Journal for the Theory of Social Behavior 23: 1–25.
Chalmers A. F. (1990) Science and its Fabrication. Milton Keynes: Open University Press.
Coleman W. & Sharrock W. (1998) Unconstructive. In: I. Velody and R. Williams (eds.) The Politics of Constructionism. London: Sage.
Collins H. (1983) An Empirical Relativist Programme in the Sociology of Scientific Knowledge. In: K. Knorr-Cetina and M. Mulkay (eds.) Science Observed: Perspectives on the Social Study of Science. London: Sage.
Coulter J. (1979) The Social Construction of Mind. London: Macmillan.
Coulter J. (1989) The Socio-logic of Knowledge and Belief. In: Mind and Action. Cambridge: Polity Press.
Crews F. (1986) Skeptical Engagements. Oxford: Oxford University Press.
Diamond C. (1989) Rules: Looking in the Right Place. In: D. Z. Phillips and P. Winch (eds.) Wittgenstein: Attention to Particulars. London: Macmillan.
Durkhem E. (1976) The Elementary Forms of the Religious Life, 2nd ed. London: George Allen & Unwin.
Feyerabend P. (1975) Against Method. London: Verso Books.
Fosnet C. T. (ed.) (1996) Constructivism: Theory, Perspectives and Practice. New York: Teachers College Press.
Francis D. (1994) The Golden Dreams of the Social Constructionist. Journal of Anthropological Research 50: 1–22.
Frankenburg R. (1993) White Women, Race Matters: The Social Construction of Whiteness. Minneapolis: University of Minnesota Press.
Garfinkel H. (1967) Studies in Ethnomethodology. Englewood Cliffs: Prentice Hall.
Garrison J. (2000) A Reply to Dayson-Galle, Science and Education 9: 615–620.
Geelan D. (1997) Epistemological Anarchy and the Many Forms of Constructivism. Science and Education 6: 15–27.
Geertz C. (1973) The Interpretation of Cultures. New York: Basic Books.
Gelles R. (1975) The Social Construction of Child Abuse. American Journal of Orthopsychiatry 45: 363–371.
Gergen K. (2001) Social Construction in Context. London: Sage.
Gergen M. & Gergen K. (eds.) (2003) Social Construction: A Reader. London: Sage.
Gilman S. (1988) Constructing Schizophrenia as a Category of Mental Illness. In: Disease and Representation. Ithaca: Cornell University Press.
Goggin G. & Newell C. (2003) Digital Disability: The Social Construction of Disability in New Media. Lanham: Rowman and Littlefield.
Goldfarb W. (2002) Kripke on Wittgenstein on Rules. In: A. Miller and C. Wright (eds.) Rule-Following and Meaning. Chesham: Acumen Publishing.
Gross P. & Levitt N. (eds.) (1994) Higher Superstition: The Academic Left and Its Quarrels with Science. Baltimore: Johns Hopkins University Press.
Hacker P. M. S. (1987) Appearance and Reality. Oxford: Blackwell.
Hacker P. M. S. (2002) Is There Anything It is Like to Be a Bat? Philosophy 77: 157–174.
Hacking I. (1991) The Making and Moulding of Child Abuse. Critical Inquiry 17: 253–288.
Hacking I. (1999) The Social Construction of What? Cambridge: Harvard University Press.
Hanfling O. (1985) Was Wittgenstein a Sceptic? Philosophical Investigations 8: 1–16.
Hardy M. & Taylor P. (1997) Von Glasersfeld’s Radical Constructivism: A Critical Review. Science and Education 6: 135–150.
Harre R. (1986) The Social Construction of Emotions. Oxford: Blackwell.
Hunter G. (1995) Quine’s “Two Dogma’s of Empiricism”: Or the Power of Bad Logic. Philosophical Investigations 18: 305–328.
Kenny A. (1991) The Homunculus Problem. In: J. Hyman (ed.) Investigating Psychology: Sciences of the Mind after Wittgenstein. London: Routledge.
Kitzinger C. (1987) The Social Construction of Lesbianism. London: Sage.
Knorr K., Krohn R. & Whiteley R. (eds.) (1980) The Social Process of Scientific Investigation. Dordecht,: Reidel.
Knorr-Cetina K. (1981) The Manufacture of Knowledge: An Essay on the Constructivist and Contextual Nature of Science. Oxford: Pergamon.
Kripke S. (1982) Wittgenstein on Rules and Private Language. Oxford: Blackwell.
Kuhn T. (1970) The Structure of Scientific Revolutions. Chicago: University of Chicago Press.
Kusch M. (2004a) Rule Skepticism and the Sociology of Scientific Knowledge: The Bloor Lynch Debate Revisited. Social Studies of Science 34: 571–591.
Kusch M. (2004b) Reply to my Critics. Social Studies of Science 34: 615–620.
Latour B. & Woolgar S. (1979) Laboratory Life: The Construction of Scientific Fact. London: Sage.
Livingston E. (1986) Ethnomethodological Foundations of Mathematics. London: Routledge and Kegan Paul.
Lyas C. (1999) Peter Winch. London: Acumen.
Lynch M. (1985) Art and Artefact in Laboratory Science. London: Routledge and Kegan Paul.
Lynch M. (1992a) Extending Wittgenstein: The Pivotal Move from Epistemology to the Sociology of Science. In: A. Pickering (ed.) Science as Practice and Culture. Chicago: University of Chicago Press.
Lynch M. (1992b) From the “Will to Theory” to the Discursive Collage: A Reply to Bloor’s “Left and Right Wittgensteinians.” In A. Pickering (ed.) Science as Practice and Culture. Chicago: University of Chicago Press.
Lynch M. (1993) Scientific Practice and Ordinary Action: Ethnomethodology and Social Studies of Science. Cambridge: Cambridge University Press.
MacAndrew C. (1969) On the Notion that Certain People Who are Given to Frequent Drunkenness Suffer from a Disease Called Alcoholism. In: S. Plog and R. Edgerton (eds.) Changing Perspectives in Mental Illness. New York: Holt, Rinehart and Winston.
Mackenzie D. (1989) From Kwajalein to Armageddon?: Testing and the Social Construction of Missile Accuracy. In: D. Gooding T. Pinch and S. Scaffer (eds.) The Uses of Experiment. Cambridge: Cambridge University Press.
Michael M. (1996) Constructing Identities: The Social, Non-Human and Change. London: Sage.
Miller G. & Holstein J. (eds.) (1993) Constructionist Controversies: Issues in Social Problems Theory. New York: Aldine de Gruyter.
Miller A. & Wright C. (eds.) (2002) Rule-Following and Meaning. Chesham: Acumen Publishing.
Modgil S. & Modgil C. (eds.) (1982) Jean Piaget: Consensus and Controversy. New York: Holt, Rinehart & Winston.
Ngaire N. (1987) Female Crime: The Construction of Women in Criminology. London: Allen and Unwin.
Piaget J. (1929/1997) The Child’s Conception of the World. London: Routledge.
Pickering A. (1985) Constructing Quarks: A Sociological History of Particle Physics. Chicago: University of Chicago Press.
Posner T. (1977) Magical Elements in Orthodox Medicine. In: R. Dingwall (ed.) Health Care and Health Knowledge. London: Croom Helm.
Putnam H. (2001) Was Wittgenstein Really an Anti-realist about Mathematics. In: T. McCarthy and S. Stidd (eds.) Wittgenstein in America. Oxford: Clarendon Press.
Quine W. V. O. (1953) Two Dogmas of Empiricism. In: From a Logical Point of View. Cambridge: Harvard University Press.
Quine W. V. O. (1960) Word and Object. Cambridge: MIT Press.
Rhees R. (1970) Can There Be a Private Language. In: Discussions of Wittgenstein. London: Routledge and Kegan Paul.
Rorty R. (1979) Philosophy and the Mirror of Nature. Princeton: Princeton University Press.
Ross A. (ed.) (1996) Science Wars. Durham: Duke University Press.
Rotman B. (1977) Jean Piaget: Psychologist of the Real. London: Harvester Press.
Seager W. (1999) Theories of Consciousness: An Introduction and Assessment. London: Routledge.
Searle J. (1995) The Construction of Social Reality. Harmondsworth: Penguin.
Sharrock W. W. (2004) No Case to Answer: A Response to Martin Kusch’s “Rule Skepticism and the Sociology of Scientific Knowledge.” Social Studies of Science 43: 603–614.
Sharrock W. W. & Anderson R. J. (1984) The Wittgenstein Connection. Human Studies 7: 375–386.
Sharrock W. W. & Button G. (1999) Do the Right Thing! Rule Finitism, Rule Skepticism and Rule Following. Human Studies 22: 193–210.
Sharrock W. W. & Read R. (2002) Kuhn: Philosopher of Scientific Revolution. Cambridge: Polity.
Shotter J. (1995) In Dialogue: Social Constructionism and Radical Constructivism. In: L. Steffe and J. Gale (eds.) Constructivism in Education. Hillsdale: Lawrence Erlbaum.
Spector M. & Kitsuse J. (1977) Constructing Social Problems. London: Cummings.
Steffe L. & Gale J. (eds.) (1995) Constructivism in Education. Hillsdale: Lawrence Erlbaum.
Strawson P. F. (1971) Truth. In: Logico-Linguistic Papers. London: Methuen.
Suchting W. A. (1992) Constructivism Deconstructed. Science and Education 1: 223–254.
Sutherland P. (ed.) (1992) Cognitive Development Today: Piaget and His Critics. London: Paul Chapman.
Toulmin S. (1976) Knowing and Acting. London: Macmillan.
Tuchman G. (1978) Making News: A Study in the Construction of Reality. New York: The Free Press.
Turner S. (2002) Brains, Practices, Relativism: Social Theory after Cognitive Science. Chicago: University of Chicago Press.
Velody I. & Williams R. (eds.) (1998) The Politics of Constructionism. London: Sage.
Von Glasersfeld E. (1989) Cognition, Construction of Knowledge and Teaching. Synthese 80: 121–140.
Von Glasersfeld E. (1992) Constructivism Reconstructed: A Reply to Suchting. Science and Education 1: 379–384. http://cepa.info/1437
Von Glasersfeld E. (1995a) Radical Constructivism: A Way of Knowing and Learning. London: Falmer Press. http://cepa.info/1462
Von Glasersfeld E. (1995b) A Constructivist Approach to Teaching. In: L. Steffe and J. Gale (eds.) Constructivism in Education. Hillsdale: Lawrence Erlbaum. http://cepa.info/1460
Waxler N. (1981) Learning to be a Leper: A Case Study in the Social Construction of Illness. In: E. Mishler et al. (eds.) Social Contexts of Health, Illness and Patient Care. Cambridge: Cambridge University Press.
White A. R. (1970) Truth. London: Macmillan.
Williams M. (1999) Groundless Belief: An Essay on the Possibility of Epistemology, 2nd ed. Princeton: Princeton University Press.
Winch P. (1958/1990) The Idea of a Social Science and Its Relation to Philosophy, 2nd ed. London: Routledge.
Wittgenstein L. (1967) Philosophical Investigations. Oxford: Blackwell.
Wittgenstein L. (1978) Remarks on the Foundations of Mathematics. Oxford: Blackwell.
Wittgenstein L. (1980) Remarks on the Philosophy of Psychology, Volume 1. Oxford: Blackwell.
Woodmansee M. & Jaszi P. (eds.) (1994) The Construction of Authorship: Textual Appropriation in Law and Literature. Durham: Duke University Press.
An earlier version of this paper was presented as part of the panel on ’Rethinking Constructivism’ at the American Education Research Association (AERA) conference, Chicago, April 2003. My thanks are due to Doug Macbeth for the invitation to take part in the panel. In revising this paper I was aided enormously by the extensive, detailed and supportive comments of an anonymous Human Studies reviewer.
Winch’s views are thus at odds with others, such as Quine (1960), who argue for a “naturalised epistemology.” On the relationship between Quine’s thought and Wittgenstein’s, see the contributions to Arrington and Glock, 1996. For a critical discussion of Quine from a Wittgensteinian standpoint, see Hunter, 1995.
The term “constructivism” as used in this article has a broad reference, to include “constructionism,” “social constructionism,” “the strong programme” (Bloor, 1976) and “the empirical programme of relativism” (Collins, 1983), not to mention ethnomethodological studies of formal and everyday reasoning (Garfinkel, 1967). The conceptual and methodological diversity of constructivist studies has lead one commentator to refer to “epistemological anarchy” (Geelan, 1997) and another to instruct us, “Do not despair, there is life after constructivism” (Bijker, 1993). The range of constructivist approaches is usefully surveyed in Miller and Holstein, 1993; Velody and Williams, 1998; Hacking, 1999; Gergen, 2001; and Gergen and Gergen, 2003. For an insightful discussion of the relationship between radical constructivism and social constructionism that comes to some of the same conclusions as are presented here, see John Shotter, 1995. As it will become clear, I am in full agreement with Shotter’s conclusion that “Both, it seems to me, still retain a desire… for the foundational epistemology they both claim to repudiate” (1995: 54).
While all constructivist studies have attracted controversy from those who interpret them as denying the possibility of objective knowledge in this or that field of inquiry, the constructivist approach to science has raised particular ire. The debate between constructivists and their critics has been characterised as “the science wars” (Ross, 1996). For the main objectivist attack on constructivist sciences studies, see Gross and Levitt, 1994.
In Lynch’s view “the starting point for such investigations is the “epistopics” – the discursive themes that so often come up in discussions of scientific and practical reasoning: observation, description, replication, measurement, rationality, representation and explanation. The epistopics provide foci for classic epistemological and methodological discussions, but they are no less relevant to vernacular inquiries” (1993: 299).
The term “theoreticism” is taken from Frederick Crews’ (1986) essay ’The Grand Academy of Theory.”
Social constructionism, especially, has generated a wide range of empirical studies into the social correlates of specific scientific controversies and developments. For a critical review, see Lynch, 1993. This is less true in the case of radical constuctionism; however, useful collections on constructivism in educational research are Steffe and Gale, 1995 and Fosnet, 1996. For critical reviews of Piagetian research, see Modgil and Modgil, 1982 and Sutherland, 1992.
Few if any empiricist philosophers have regarded knowledge as a straightforward “copy” of reality. From Locke to Russell, they hold the view that the mind shapes, organizes and transforms sensory information to create conceptual understanding. The distinctive assumption made by empiricists is that the mental operations involved in such shaping, organizing and transforming are constrained by the material they have to work with. The mind does not have a free hand to construct “just anything.” On the problems raised by this conception, see Williams, 1999.
For an insightful discussion of these misuses in Piaget’s writings, see Rotman, 1977.
The tendency to elide the difference between intentionalist and physicalist vocabularies, which is widespread in cognitive philosophy, gives rise to what is known as the “homunculus problem.” On this, see Kenny, 1991.
Radical constructivism’s indebtedness to Descartes and the Cartesian epistemological tradition has been remarked on by Garrison (2000). Referring to GRC’s conception of knowledge as an inner mental construction that enables the individual to conceive of a reality external to the mind, he notes that “this is the subject versus object dualism; overcoming it established the epistemology industry that has dominated Western thought since Descartes.”
The critical relevance to radical constructionism of Wittgenstein’s remarks on “private language” is noted by Suchting (1992). However, his brief discussion of their implications for GRC’s claims does not explicate Wittgenstein’s arguments in any detail. Even allowing for this, von Glasersfeld’s (1992) response to Suchting fails to address the main thrust of his objections.
Bloor’s attempt to recruit Wittgensten in support of a causal sociology of knowledge has been critiqued before, most notably by Sharrock and Anderson, 1984 and Coulter, 1989. The critical literature on the “strong programme” is voluminous. Especially useful discussions can be found in Knorr, Krohn and Whiteley, 1980; Brown, 1984; Chalmers, 1990 and Lynch, 1993.
An objection might be raised that such an account amounts to little more than “textbook history,” the product of an ideology of rational progress built into the institutional organization of science. Such an objection misses the point, which is that for any account to qualify as a candidate explanation it must be cast in relevant scientific terms. If this is an “ideology,” it is one that is constitutive of the phenomenon of science.
Bloor’s criteria have been the focus of much critical discussion over the years. For references to such discussions, see the sources in Footnote 13 above.
“Naïve enough” since, as the historical accounts of actual cases of scientific reasoning provided by Kuhn (1970) and Feyerabend (1975) make clear, rationalist idealizations bear little relationship to scientific reality. It is ironic that such writers as these are cited by BSC in support of its claim that internal scientific accounts are inadequate, since the point about the Duhem-Quine thesis applies to their works also. These are fundamentally misunderstood if they are taken as critiques of actual processes of reasoning in science. For example, Feyerabend’s famous mottos, “Against method” and “Anything goes,” frequently are taken within SC as commenting on the failure of methods to account for change in science, but they are nothing of the kind. The “scientific method” that Feyerabend is “against” is the rationalist reconstruction of science and the idea that change can be accounted for by general rules. The assertion that “anything goes” has to be understood in this light: Feyerabend’s point is that, if one is to suppose that a general rule needs to be formulated that can account for such change, “anything goes” is the rule that works best in this regard. In other words, Feyerabend’s critique is properly understood as an “internal” one, but internal to the philosophical project of reconstructing scientific reasoning in general terms, not in the sense that he is critiquing the actual reasoning practices of scientists. A similar point can be made about the writings of Kuhn. See, for example, Sharrock and Read (2002).
In a recent discussion of the “Bloor-Lynch debate” Martin Kusch (2004a) defends Bloor’s position on rule finitism against Lynch’s (1992a, b) criticisms. His argument is complex and opaque in places but the crux would seem to be that Bloor’s views are to be preferred since they present a “direct” answer to Kripkian rule scepticism whereas Lynch and other critics of Bloor present a “diagnostic” answer. In other words, Bloor takes Kripke’s formulation of the sceptical problem as requiring a conceptual solution, which he provides in terms of “finitism.” Lynch, on the other hand, following Baker and Hacker (1984), argues essentially that there is no problem to solve, since the assumptions on which rule scepticism rests are misconceived. In Kusch’s view, despite the fact that Bloor’s solution is “unsatisfactory” it at least has the merit of recognising the genuine nature of the problem. Unfortunately, Kusch provides little more reason for taking rule scepticism seriously than does Bloor. Against Sharrock (2004) for example, he argues (Kusch, 2004b) that Kripke’s position on the relation between a rule and its applications is “internalist” not “externalist.” But this only undermines the Lynch/Sharrock stance if what (Kusch’s) Kripke understands by internalism is what the critics mean by it. This is difficult to assess, since Kusch’s explanation of the internal/external distinction is notably underdeveloped. In any case, the key “diagnostic” point, as Sharrock makes clear, is that the alleged problem of rule scepticism has not been, and possibly cannot be, coherently stated.
My own view is that the studies which attend most closely to the endogenous order of intellectual activities and therefore exemplify constructivism at its best are those carried out under the auspices of ethnomethodology. For examples see Livingston, 1986; Lynch, 1985 and Button, 1993.
A number of them are identified by Stephen Turner in his critical discussion of Searle (Turner, 2002). For example, Turner notes (2002: 56) that “The best evidence we have of the inner workings of the physical world according to advanced theory is precisely squiggles and remnants on detection devices; these are evidence for anything only by virtue of extremely complex sets linked of experimental practices. This makes the idea that there is some radical difference between the objects of institutionalised scientific knowledge and those of institutionalised financial dealings somewhat suspect.” Turner’s point, of course, is precisely what constructivist students of natural science have been arguing for thirty years.
The tendency among some constructivists to treat words like “reality” as though they pose deep epistemological puzzles requiring theoretical solutions is one I have critiqued before. See Francis, 1994.
Found a mistake? Contact corrections/at/cepa.infoDownloaded from http://cepa.info/2932 on 2016-07-26 · Publication curated by Alexander Riegler