Alhassanain(p) Network for Heritage and Islamic Thought

1. Introduction: Language and Structure

We are tempted to think that the action of language consists of two parts; an inorganic part, the handling of signs, and an organic part, which we may call understanding these signs, meaning them, interpreting them, thinking. These latter activities seem to take place in a queer kind of medium, the mind; and the mechanism of the mind, the nature of which, it seems, we don’t quite understand, can bring about effects which no material mechanism could …

Frege ridiculed the formalist conception of mathematics by saying that the formalists confused the unimportant thing, the sign, with the important, the meaning. Surely, one wishes to say, mathematics does not treat of dashes on a bit of paper. Frege’s idea could be expressed thus: the propositions of mathematics, if they were just complexes of dashes, would be dead and utterly uninteresting, whereas they obviously have a kind of life. And the same, of course, could be said of any proposition: Without a sense, or without the thought, a proposition would be an utterly dead and trivial thing. And further it seems clear that no adding of inorganic signs can make the proposition live. And the conclusion which one draws from this is that what must be added to the dead signs in order to make a live proposition is something immaterial, with properties different from all mere signs.

But if we had to name anything which is the life of the sign, we should have to say that it was itsuse .[1]

If language grants the possibility of sense to a human life, then the systematic inquiry into its structure consigns this life to an ambiguous basis in the relation of signs to their meanings. For as soon as it becomes the object of systematic analysis, the totality of language both demands and refuses completion by a principle of meaning exterior to its own economy. Wittgenstein’s text, written in 1933 or 1934 as part of a series of notes intended for his students at Cambridge, identifies the desire for such a completion in the thought of his great philosophical progenitor, Frege. The anxiety to which this desire responds is one of death, specifically a death of sense in the materiality of the sign. The characteristic resource it marshals against this anxiety is the life of the human being who speaks, understands, intends and thinks.

One of the most significant projects of the analytic tradition in the twentieth century has been its attempt to envision and comprehend the structure of language. From the first moments of the development of its characteristic modes of analysis, reflection, and inquiry, the analytic tradition has attempted to grasp language as a regular structure of signs accessible to rational elucidation. Wittgenstein’s analysis of the tendencies operative in Frege’s thought displays, particularly clearly, some of the constitutive tensions to which this attempt is prone. As we shall see over the next several chapters, the analytic tradition’s search for a comprehensive description of the structure of language has also involved a complex consideration of thelife of the human user of language. It has pictured this life, alternatively, as the self-consciousness of a subject of experience or as the shared life of a community of speakers, the mutuality that is seen as the foundation of any possibility of communication. The implications of the analytic tradition’s joint envisioning of language and life are varied and far ranging; a historically based exploration can help to elucidate the broader legacy of the analytic tradition itself for contemporary critical thought and action. In its diagnostic modality, this exploration looks toward the clarity of a life that no longer seeks its significance in the problematic attempt to master the relationship of signs to their meanings, but might find in the withdrawal of this relationship into abeyance the vanishing of the problem it represents.[2]

I.

Analytic philosophy’s engagement with language has been, on any account, longstanding, sustained, and determinative for both the tradition’s main methods and its most significant results. It is probably impossible to identify asingle conception of the nature of language that underlies all of the tradition’s varied analytic, reflective, and critical projects. Nevertheless, there is a distinctive set of interrelated theoretical and methodological commitments that have repeatedly made reference to “language” itselfpossible for many, if not most, of the projects of analytic philosophy throughout the twentieth century that have discussed it. Introducing these commitments briefly may serve to facilitate reflection on the nature of this reference and focus some of the questions that it raises.

As I shall discuss it in this work, thestructuralist picture of language consists in four interrelated central commitments and a fifth, less central one that often (though not always) goes along with the first four:

1. Language as a whole can be understood as asystem orstructure of signs, words, propositions, sentences or other significant terms.

2. The logical, grammatical, or structural interrelations among these terms, as well as their ordinary use in speaking or writing, are wholly or partially constrained by a corpus of intelligiblerules orregularities.

3. These rules or regularities are describable and their description can account for thecorrect ornormal use of terms in everyday interlocution.

4. On the basis of such a description, it is possible to determine themeaning ormeaningfulness of terms or combinations of terms used on particular occasions.[3]

5. The rules or regularities that thus constrain the use of language are essentiallypublic ,intersubjective , andsocial in character.[4] [5]

These interrelated theoretical commitments, naturally linked to one another in the vision of language that they determine, have had far-ranging methodological consequences from an early moment in the development of the tradition. At its beginning, they provided the methodological basis for the projects of conceptual or logicalanalysis that characterized the tradition in its early stages, and indeed originally gave it its name. For the practitioners of these early projects, the solution or dissolution of philosophical problems depends on the clarification and description of logicalstructure .[6] Analysis of propositions, facts, or concepts into their structurally simpler elements serves to reveal the real or genuine form of these individual items or their systematic interrelationships, over against our ordinary tendencies to mistake or misconstrue these forms or relations. The demonstration is, in particular cases, to be guided by an overarching elucidation of the structure of a set of systematically interrelated terms, whether these terms are conceived as elements of language (at first they were not), as objects of knowledge, or as individual concepts, thoughts, senses or meanings.

An early and influential expression of one such project can be found in the manifesto “The Scientific Conception of the World: The Vienna Circle,” released in 1929 to summarize the project of the circle of philosophers and scientists that had been, since the early 1920s, meeting in Vienna around Moritz Schlick. The manifesto describes the “scientific world-conception” of the Circle as consisting in two main features: first an “empiricist” and “positivist” orientation demanding that “there is knowledge only from experience;” and second, the application of “a certain method, namelylogical analysis .”[7] The authors (chiefly Hans Hahn, Otto Neurath and Rudolf Carnap) explain the implications of this method:

It is themethod of logical analysis that essentially distinguishes recent empiricism and positivism from the earlier version that was more biological-psychological in its orientation. If someone asserts “there is a God”, “the primary basis of the world is the unconscious”, “there is an entelechy which is the leading principle in the living organism”, we do not say to him: “what you say is false”; but we ask him: “what do you mean by these statements?” Then it appears that there is a sharp boundary between two kinds of statements. To one belong statements as they are made by empirical science; their meaning can be determined by logical analysis or, more precisely, through reduction to the simplest statements about the empirically given. The other statements, to which belong those cited above, reveal themselves as empty of meaning if one takes them in the way that metaphysicians intend. … Analysis … shows that these statements say nothing but merely express a certain mood or spirit.[8]

According to the Vienna Circle authors, then, the analysis of propositions decides the meaningfulness of statements of ordinary and philosophical language by elucidating the extent of their logical connection to propositions already known to be meaningful (in this case, propositions describing the “empirically given.”) This determination provides the basis for drawing a line, within language as a whole, between statements that are meaningful (those of empirical science) and those that lack meaning (typically those of “metaphysics”). As the authors of the manifesto make clear, this demarcation of language into meaningful and meaningless regions itself depends on a systematic elucidation of the logical interrelationships of the concepts of science.[9] In the spirit of the “scientific world-conception” that they saw as gaining ground in Europe, England, and the U.S., the authors of the manifesto thus looked forward to the complete dissolution of metaphysics through the clarification of the logical structure of meaningful language. The project, despite the similarities they noted to earlier versions of empiricism and positivism, had its methodological basis in the new apparatus of logic that the Circle philosophers had available to them, and in the conception of language as a total structure of signs that it suggested.[10]

When the manifesto authors wrote in 1929, there were already significant precedents for the practice of logical analysis that they espoused. Perhaps the most decisive early influence was Frege’s conception of a systematic notation for the clear logical expression of thoughts, the so-calledBegriffsschrift or “concept writing.” The new syntax was to bring out in symbolic form the underlying structure of thought. Frege compared the improvement over ordinary thinking that such a notation would afford to the advantage of a microscope over the eye. Like a specialized visual instrument most useful for special investigative purposes, the more precise symbolism would, without replacing ordinary language, facilitate the special work of an analysis of the logical structure of concepts actually underlying ordinary claims and judgments.[11] The analysis would be particularly important, Frege thought, in consolidating the rigor of mathematical proof and placing mathematics on a firmer logical basis. This hope to find a rigorous logical basis for mathematics led him to pursue the project later calledlogicism : the reduction of the claims of mathematics to a basis in a small set of axioms and their logical, deductive consequences.

This logicist program culminated in Russell and Whitehead’sPrincipia Mathematica of 1910-1913. The three-volume work, which provided a logical analysis of the basic notions of set theory and number theory, represented the most detailed and rigorous development of the conception of logical analysis that Russell had originally reached, at least in embryonic form, around the turn of the century, and which the Vienna Circle authors themselves cited as a definitive inspiration for their own project. As early as 1900, Russell, rejecting the holism and monism that characterized the then-dominant Hegelian Idealism, had declared the utility of an analytic approach: “That all sound philosophy should begin with an analysis of propositions is a truth too evident, perhaps, to demand a proof.”[12] Russell’s colleague Moore had used the term “analysis” inPrincipia Ethica (1903) to characterize his own investigation of the concepts of ethics. Russell’s theory of descriptions, expounded in the 1905 article “On Denoting,” demonstrated the utility of the method by offering a powerful early example of a successful logical analysis, showing that the actual meaning of a large class of sentences could be exhibited, against the obfuscating effects of ordinary language, by clarifying their underlying logical form.[13]

 Subsequently, Russell articulated and defended the “logical-analytic” method again in 1914 in the lectures that becameOur Knowledge of the External World .[14] Here, it formed the methodological basis for a wholesale project of epistemological clarification of the nature and basis of empirical knowledge. Russell’s work in the foundations of set theory had led him to suggest that the referents of a large variety of ordinary-language terms might be treated as “logical constructions” of simpler elements. In the particular case of ordinary spatial objects of perception, for instance, analysis could decompose them into the simple sensible particulars that made them up and exhibit the logical relations among these particulars. Analysis, he suggested, could resolve the “inferred entities” of ordinary experience into the logical constructions that they in fact were; in this way, their ultimate constituents would be revealed and the possibility of our knowledge of them explained.[15]

For all of the early analytic philosophers who appealed to logical structures in practicing the new methods of analysis, a primary motivation for the appeal was their desire to safeguard theobjectivity of contents of thought, over against the threat posed by subjectivist theories of them. Onlylogically structured contents, they thought, could genuinely be objective in the sense of existing wholly independently of anyone’s acts of thinking of them, grasping them, considering them or entertaining them. Frege’s conception of the objectivity of logically articulated contents of thought, and his resultant polemics against psychologistic and historicist theories of content, figured prominently in his writings on logic and the foundations of mathematics from nearly the beginning of his career.[16] A similar motivation, directed initially against the then-dominant post-Hegelian Idealism, underwrote Russell’s initial realism about “propositions” and “meanings” and his resultant conviction that logical analysis could demonstrate, through decomposition, the actual constituents of the world.[17] For Schlick, Carnap and other members of the Vienna Circle as well, an analysis of the logical structure of the propositions of science was essential to demonstrating their objectivity and distinguishing them from the claims of pseudo-science or metaphysics. Only by displaying their underlying logical structure, Schlick and Carnap thought, could the propositions of empirical science be purged of any essential dependence on ostensive or demonstrative elements, and so portrayed as genuinely independent of the acts or occasions of their discovery or verification.[18]

Logical analysis, as it was first conceived within the tradition, therefore sought to demonstrate the actual logical relations that determine the contents of thoughts or propositions, and so also the meaning of the linguistic terms that express them. But the first analytic philosophers (in particular, Frege, Moore, and Russell) did not yet see their methods as grounded primarily or specifically in the analysis of language. Their attention to ordinary language most often had the aim of exhibiting its tendency to obfuscate and conceal rather than any analysis of language for its own sake. For Russell (at this time) and Frege as well as for Moore, the object of investigation was “thoughts,” “concepts,” “meanings,” or “propositions” (conceived as non-linguistic but structured constituents of the world) rather than words or sentences.[19] InThe Philosophy of Logical Atomism , culled from lectures he delivered in 1918, Russell described his method of analysis as dividing the world into mutually independent “facts,” each of which were further decomposable into more basic simples or particulars. Most basically, Russell thought, a fact was what made a proposition true or false; its further logical decomposition would resolve the particular simple objects to which the simplest terms of such a proposition, if fully analyzed, refer.

The young Wittgenstein, in theTractatus Logico-Philosophicus , drew out the semantic and ontological consequences of this “logical atomism” for the relationship of language to the world. With its invocation of “logical form” as responsible not only for the laws of logic but thepossibility of any meaningful language, theTractatus was the first to suggest that the solution to the problems caused by the forms of ordinary language could lie in analysis of these forms themselves. They were to be elucidated by means of reflection on the varieties of sense or nonsense they permit; by drawing a line between sense and nonsense, the analyst could hope to clarify the logical structure shared by language, thought, and the world. The suggestion, and its quick reception (and partial misconception) in the logical empiricism of the Vienna Circle, brought the structuralist conception of language in its fully developed form to bear on the problems of philosophy and the meaning of everyday language alike.

Although the structuralism of early analytic philosophy was not, initially, explicitly linguistic, the structuralist conception of language as a totality of signs governed by logical rules thus appeared relatively quickly to be the most natural setting for its distinctive methods of analysis. It captured, as well, much of what analytic philosophers explicitly carried forward from earlier philosophical projects, and hence came to define the relationship of early analytic philosophy to the larger philosophical tradition which it sought to update. One important precedent for the structuralist picture of language as a system governed or determined by rules for the intercombination of signs was the logical system of Leibniz. In some of his earliest writings, Leibniz had suggested the idea of amathesis universalis or “universal character,” a symbolic language of logic that would, like Frege’s own “concept-writing,” clarify human reasoning by giving it a unified, systematic mathematical calculus for the evaluation of the validity of arguments and conclusions.[20] Russell’s first substantial philosophical work, written in 1900, took up the question of the relationship of Leibniz’s conception of logic to his metaphysics against the changed backdrop of the new forms of logic derived from Frege and his recent forebears, and Rudolf Carnap cited Leibniz’s project approvingly as a precedent for his own logical analysis project in his first masterpiece,The Logical Structure of the World .[21]

To this determinative rationalist influence deriving ultimately from Leibniz, the philosophers of the Vienna Circle added a picture of experience drawing on the empiricism of Hume, Locke, and Berkeley, as well as the positivism of Mach, Poincaré, and Duhem. Logical analysis was, among other things, to clarify the inherent structure of the given contents of experience, clarifying its simple elements and describing their structural interrelationships. But even more methodologically decisive for the Vienna Circle’s project of linguistic and logical analysis was the legacy of Kant’s critical project of tracing the boundaries of reason’s legitimate employment in relation to our knowledge of the world. Several of the Vienna Circle philosophers had themselves been deeply influenced by the neo-Kantianism of philosophers like Cassirer, Cohen, Natorp, and Rickert.[22] These philosophers had already undertaken to update Kant’s critical project by reflecting on the way the formal and symbolic structures of language condition the possibility of human knowledge. Now, the availability of the new methods of logical analysis suggested that formal logic itself could be the basis for a critical delimitation of the boundaries of language or of sense, clarifying the scope of possible experiential meaning and thus carrying forth the Kantian limit-fixing project in an updated logical-linguistic mode.

From near the beginning of its itinerary, the structuralist conception of language existed in an uncertain relationship with the empiricist conception of subjective experience that also regularly accompanied it. Beginning in the late 1920s, the philosophers of the Vienna Circle envisioned the analysis of scientific propositions as elucidating the total logical structure of science, as well as the possibility for some of its claims - the so-called “protocol sentences” - to be directlyverified by experience.[23] The ensuing debate about the form of protocol sentences their relationship to the other propositions of science touched on a large number of interrelated philosophical issues within epistemology, metaphysics, and the nascent field of “philosophy of mind.” But at the core of many of these issues was the question of the relationship of language, conceived as a total structure of rule-bound symbols whose meaning is defined only relationally, to immediate, first-person experience.[24] The structuralist conception of language and objectivity demanded that the concepts of science be defined by their structural role in the system of science as a whole, independently of their relationship to experience. But in order to distinguish genuinely empirical propositions from non-empirical ones, it was necessary to give an account of the role of experience in grounding or verifying them. The question of the status of protocol sentences - in particular, whether they should be pictured as verified by experiential events outside the realm of logically structured, objective science or as ordinary elements within this structure - was never resolved by the participants of the Vienna Circle themselves. Subsequently, Schlick’s murder in 1936, along with the political events of the 1930s, led to the Circle’s breakup and the indefinite suspension of its project.

During and after World War II, the new forms of analysis and projects of what was only then first widely called “analytic” or “analytical” philosophy proposed new methods and means of linguistic clarification, reflection, and critical demarcation.[25] These methods ranged from those of the “ordinary language” school that developed at Cambridge and Oxford to Quine’s logical “regimentation” of ordinary language using the formal apparatus he had learned from Carnap’s original syntactical project. Essential to many of these new projects was their repudiation of what they took to have been the Circle’s “reductionism” and “verificationism.”[26] In the widely influential “Two Dogmas of Empiricism” of 1950, Quine sought to replace the “dogmas” of analyticity and reductionism with aholistic picture of confirmation that drew explicitly from Neurath’s earlier anti-foundationalist conception of the structure of language.[27]

Subsequent historical retellings of the development of the analytic tradition have often seen Quine’s argument, and the more general repudiation of verificationism that quickly became established wisdom, as essential to the abandonment of the original methods of analytic philosophy, in particular thedecompositional method of analysis that had been suggested by Frege, Russell and Carnap. But as we shall see over the next several chapters, attention to the ongoing role of structuralist assumptions about language demonstrate the actual continuance of the most significant methodological threads of the original project even in those postwar projects that claimed most directly to repudiate it.[28] For the ambiguities of the relationship of logical structure to its elements that had proved fatally problematic for the Vienna Circle’s project of analysis remained in place, and continued to produce characteristic difficulties for structuralist reflection on language and its meaning, even when “reductionism” was replaced with “holism” and verificationism was replaced with a more nuanced conception of empirical confirmation.[29]

My aim in identifying the influence of structuralism upon the analytic tradition is not to impose a false unity upon a tradition that has certainly been marked, at least since the 1950s, by an extremely diverse and heterogeneous set of philosophical methods and practices.[30] But I do hope to show the deep and pervasive way in which the methods and results ofmany of these practices, even including the ones that have officially rejected structuralism or sought explicitly to limit its influence, can be seen as deeply influenced by the problems and ambiguities that arise from it. These problems and ambiguities have continued to exert a decisive influence on the methods and results of analytic philosophy, I shall argue, long past the widespread midcentury rejection of reductive and atomistic forms of “conceptual analysis.”[31] Indeed, even many of those contemporary projects that reject the entire idea of a specific relevancy of linguistic reflection to the problems of philosophy, preferring to define themselves as pursuing “metaphysics” (in some non-pejorative sense) or empirical contributions to psychology, sociology, or biology, nevertheless inherit styles of argumentation, methods of reasoning, and writing practices that originated within the earlier project of structuralist analysis and remain subject to its specific instabilities. Historical reflection on the origin and persistence of these instabilities within the methods of analytic philosophy helps to demonstrate the contemporary relevance of these projects to the (still very much open) question of the nature and basis of linguistic meaning, and of the continuing possibility of philosophy’s recourse to it.

In choosing the term “structuralism” to characterize the particular set of commitments underlying the picture of language that has been most widely influential within the analytic tradition, I intend also to gesture toward the close conceptual and methodological connections between this set and the tradition of European (chiefly French) thought that has been called by the same name. Although my chief concern here is to identify and trace the role played by the structuralist picture of language in the analytic tradition, the texts of philosophers, linguists, and anthropologists such as Saussure, Jackobson, Levi-Strauss, and Benveniste show the influence of a similar picture just as pervasively.[32] The sustained inquiry into the systematic character of language and linguistic meaning begun by Saussure has, over the course of the twentieth century, situated and given essential shape to the political, social, and philosophical contributions of phenomenology, psychoanalysis, and critical thought. Part of my hope in characterizing this inquiry as parallel in many important respects to the analytic one is that the unfortunate and divisive legacy of disregard and mutual misunderstanding that has existed, throughout the second half of the twentieth century, between representatives of the two structuralist traditions, European and Anglo-American, can be recognized as baseless and finally overcome, presumably to the mutual benefit of both.

II

From the beginning of the tradition, the structuralist picture of language as a totality of rule-governed signs directed the attention of analytic philosophers toward the analysis of the structure of propositions, facts, or terms and the systematic clarification of their logic. This structuralist clarification, at first (for instance in Frege) cited logical rules that were conceived as underlying the objectivity of thought in an ideal sense, quite independently of the actualpractice of language. But with the explicit turn to language, philosophers quickly began to see reflection on linguistic practice as the most natural home for analysis of logical or grammatical structure. At this point, it became natural to consider linguistic terms and sentences as, among other things, objects ofuse . Such use is, besides being explicable in terms of rules or regularities, to be understood as essentially publicly and intersubjectively learned and controlled. The fifth commitment that has often held by analytic philosophers who have held a structuralist conception of language, accordingly, concerns its essentiallypublic nature:

5. The rules or regularities that constrain the use of language are essentiallypublic ,intersubjective , andsocial in character.

This commitment is logically independent of the other four. As we have seen, the first analytic philosophers, Frege and Russell, did not hold it. But it is already at least implicit in Schlick and Carnap’s descriptions of the project of analysis as grounded in the elucidation of “rules of use” governing the application of terms in the practice of a language as a whole. It would soon thereafter come to play an increasingly explicit role in many philosophers’ statements of their own projects. It follows, in any case, naturally enough from structuralism’s picture of language as a system or structure of rule-governed signs. Once language itself is seen as a regular structure of signs governed by rules determinative of meaning or meaningfulness, it is natural to suppose that the relevant rules are primarily applicable to, and evident in, the establishment and maintenance of social practices, especially communicative practices of judgment, assertion, rational evaluation, and criticism.

In many of the texts of immediately postwar analytic philosophy, in particular, commitment to the “public character” of language was held to be essential to repudiating theindividualistic or methodologicallysolipsistic assumptions of earlier philosophical projects.[33] Here, it was supposed, only a fundamental insistence on the essential publicity of linguistic concepts and their basis in intersubjective practices of using and learning language could remedy the (now widely repudiated) reductionist and foundationalist assumptions of an earlier phase of analytic reflection. Insistence on the essential publicity of language and concepts seemed to offer new and pervasive grounds, as well, for continuing the critique ofpsychologism that had figured centrally in the analytic tradition’s methods of logical reflection. For if our very access to the concepts in terms of which we describe our immediate experience is dependent on our understanding of a language, learned in public and controlled by public criteria of applicability, then there is no hope for psychologistic theories of meaning that base it instead in the experiences or phenomena of an individual mind, consciousness, or subject of experience.

A socially based theory of the learning and communication of linguistic terms and their regular interrelationships thus came to seem requisite for a comprehensive understanding of the structure of language itself.[34] The hope for such a theory indeed became almost ubiquitous in the projects of midcentury analytic philosophy of language. The continued complicity of these projects with structuralist assumptions about language was hardly noted. But as we shall see, the fundamental problems and inherent ambiguities of the structuralist picture of language in fact remained determinative in producing the theoretical tensions to which these projects of analytic reflection were repeatedly prone.

Significantly for the continuing reception of the analytic tradition, the fifth assumption of structuralism about language continues to play a pervasive role, as well, in prominent projects of analytic philosophy today. The thought that an accounting for linguistic meaning and meaningfulness depends on a description or analysis of social practices of assertion, communication, and judgment, plays a foundational role, for instance, in the projects of Davidson, Dummett, Brandom, Kripke, and Rorty, among others.[35] For these philosophers, the rules, regularities, or norms that determine the actual and correct usage of terms and locutions in a language are to be discovered, at least in part, in theinstitutions of social practice that govern the intersubjective behavior of the language’s speakers in discourse and communication. Such social practices are inextricably connected with non-linguisticpraxis as well, and normally include our ordinary ways of interacting with and shaping our environments. What they involve, on any particular occasion, is thought largely to be learned along with, or as an essential part of, the learning of a first language. After they are learned, they are maintained, and enforced, through essentially intersubjective and social mechanisms for the evaluation, critique, endorsement or censure of particular linguistic performances, insofar as these performances comport, or fail to comport, with them.

In some recent projects, social practices are seen as providing an explanation not only for the actual facts of language use, but also for thenormativity of language or concepts and therationality of their users. That is, our social practices are seen as providing a basis not only what we in fact do say, in a variety of contexts and situations, but for determinations of what weshould say if we want to draw rational inferences, or respond appropriately to the utterances of our peers, or cooperate with them in making claims that lead us reliably to the truth. The various practices of deliberation, correction, consideration and evaluation that normally accompany the venturing and verification of claims in everyday discourse are thus seen as embodying, through the rules or norms of usage ordinarily governing them, the claims ofreason orrationality to which traditional philosophy devoted a complex and self-critical reflection. The real object of this traditional reflection can then be seen as, in Sellars’ memorable phrase, the socially inculcated and maintained “game of giving and asking for reasons,” the set of communicative practices ultimately responsible both for the meaning of propositions and the validity of the claims they formulate.

Some of the contemporary and recent philosophers who assume a basis for the meaningfulness of language in publicly learned and socially maintained practices claim to draw inspiration from the late Wittgenstein, and in particular from his considerations of “rule-following” and the idea of a “private language” in thePhilosophical Investigations . On a hasty reading, it can indeed seem as if Wittgenstein’s scattered references to the forms of human understanding, thought, and perception as grounded in “language-games” articulate a (perhaps largely “implicit” or suggestive) theory of these forms, or indeed of the “practice of language” itself, as grounded in regular, describable, public social practices or institutions. Such an account is often seen, moreover, as including a “use-theory” of meaning that accounts for the significance of the various terms of language by reference to the facts or norms of their ordinary application. Partisans of such an interpretation often hold that Wittgenstein himself did not work out such a theory in detail, but that one could be developed, consistently with Wittgenstein’s underlying intentions, either through empirical research into sociology, psychology, linguistics, biology or some combination thereof or through philosophical description of the underlying structure of our practices.[36]

Such interpretations, as I shall argue, ignore not only Wittgenstein’s lifelong and methodologically essential animadversions against mistaking positive theory (especially of an empirical type) for philosophical work, but indeed miss, as I shall argue, one of the most significant critical points of theInvestigations . This point is not at all to confirm or consolidate accounts that place rule-bound practices at the basis of the ordinary meaningfulness of language. It is, rather, almost the direct opposite: to criticize the structuralist conception of rules and rule-following that provides the ordinary setting for such accounts on the level of their picture of language as a whole. Taking Wittgenstein to be supporting a “practice”-based account of language, commentators and subsequent philosophers have largely missed the deep and pervasive way in which his consideration of rule-following actually undermines any such account.[37] They have thus persisted uncritically in a structuralism about language and practices that a fuller reading of Wittgenstein’s internal critique of structuralism otherwise might have, long ago, taught them to doubt.

Indeed, the assumption of a social basis in intersubjective practice for linguistic meaning and meaningfulness has repeatedly obscured the far-reachingcritical implications of the analytic tradition’s sustained consideration of the structuralist picture of language. Even when, as with Wittgenstein’s critique of rule-following or Quine’s thesis of the indeterminacy of radical translation, these critical consequences have appeared with a fair degree of explicitness, they have seemed, within the ambit of social-practice accounts of language, simply to bear against one or another morerestricted picture of the structure of language, and their more general significance as internal critiques of structuralism has been missed. The usual result has been the continuance of an underlying structuralism about language, despite relatively superficial changes in the form it takes, and a recurrence of the problems to which it leads.