In the Philosophical Investigations, Ludwig Wittgenstein famously pronounced that “the meaning of a word is its use in the language”. More recently Noam Chomsky has frequently surmised that there is no semantics, but only syntax and pragmatics. Neither iconic thinker offered full-fledged theoretical underpinnings of his dictum. At the time of his pronouncement, Wittgenstein was sceptical that theorising about language is fruitful, while Chomsky, albeit a theoretician about language if there ever has been one, has always been more interested in viewing language as an abstract symbol system the fundamental structure of which is universal and known by us innately, than in questions of linguistic meaning and understanding (“semantics”) or in features of the use of language in in context (“pragmatics”).
In contrast, the Pittsburgh philosopher Robert Brandom has for the past four decades worked on a comprehensive and systematic theory of linguistic communication that enshrines these two mottos – albeit perhaps not in the way Wittgenstein or Chomsky themselves have understood them. Brandom’s results to-date depict a fascinating alternate world, compared to the one contemporary mainstream philosophers of language take themselves to inhabit. Moreover, rather like twin-earth in Hilary Putnam’s famous thought experiment – a planet superficially exactly like earth, with mountains and valleys, rivers and lakes, plants and animals, clouds, rain, and sunshine, and even people communicating in a language phonetically and syntactically exactly like English, but consisting in matter with a fundamentally different chemical constitution – Brandom’s world seems superficially exactly like the one depicted by mainstream analytic philosophers of language, while operating, as far as linguistic communication is concerned, on fundamentally different principles. Speech is propositionally contentful and has illocutionary force in Brandom’s world just as in the one of the mainstream: Brandomian speakers, just like mainstream speakers, make assertions, commands, requests, or interrogations, etc. with propositional meanings that p, that q, that r, etc. Correlatively, Brandomian speakers, like mainstream speakers, reason in the medium of propositional attitudes – beliefs, intentions, etc. that p, that q, that r, etc. Claims have names, predicates, and (often) logical connectives among their component parts. In context, names refer to mind-independent individuals and predicates represent mind-independent properties and relations. Thus, propositional contents and meaningful claims have objective truth conditions, and Brandom’s world, rather like Aristotle’s or David Armstrong’s, is, accordingly, metaphysically non-nominalist (it comprises universals – properties and relations – in addition to individuals) and metaphysically realist (the facts obtain independently of what anyone or everyone may think or claim).
For good measure, Brandom’s world, like Armstrong’s, is modally realist too: what metaphysically or by law of nature must or may be the case, is built into the structure of the world itself, again independently of what anyone or everyone may think or say. At the same time, the modal operators (“possibly”, “necessarily”) and the logical connectives behave in Brandom’s world like they behave in the one of the mainstream, in the sense that something is a tautology or a modal principle in Brandom’s world if and only if it is a tautology of classical propositional logic or a principle of classical modal logic. Epistemically, Brandomian speakers are unremarkable too. They have lots of empirical knowledge, that is, lots of true, justified beliefs (plus, perhaps, some extra that helps them not to know “accidentally” in the kinds of scenarios Edmund Gettier famously concocted half a century ago). Their beliefs are moreover justified if and only if they are formed by mechanisms that, in their given environment, are reliably truth-conducive. Thus, having knowledge, for Brandom, is a matter of being epistemically reliable, in broad accordance with Alvin Goldman’s classical externalist epistemic reliabilist theory of knowledge.
So far, all this sounds unexcitingly familiar to analytically inclined contemporary philosophers of language, mind, and knowledge. Yet beneath the surface, things are exotic in Brandom’s world. Mainstream speakers are basically updated Lockean creatures. In the order of explanation, they start out as solitary epistemic and rational subjects and agents, and they become social creatures communicating in the medium of a public language last. Antecedently, mainstream speakers are furnished with updated versions of Lockean ideas – conceptually structured, propositionally contentful mental representations (which, however, may be innate rather than explained in empiricist fashion) – in the medium of which they are initially able to reason at least to some extent. Their ability to communicate linguistically with each other is then explained as a special form of rational agency: agency through which they intentionally express or convey to each other, mediated by linguistic vocalisations, their representational thoughts. These vocalisations themselves inherit their semantic features from the representational mental contents expressed or conveyed, and these semantic features are thus themselves representational in nature. What makes concepts (originally) contentful and words (derivatively) meaningful is that they represent individuals and properties and relations, and words inherit their representational dimension (their meaning) from the representational dimension of the concepts they express or convey on particular occasions of use, which concepts the mainstream speaker is antecedently furnished with.
In contrast, in Brandom’s order of explanation, speakers start out as non-rational, norm-governed, social creatures, rather than as rational, representing, solitary ones. Instead of being antecedently furnished with explicit, propositional mental representations, they are antecedently furnished with implicit (that is, non-conceptual) practical skills or know-how to participate in norm-governed social practices. In practice, they know how to follow certain social norms and how to hold each other to them – think of social animals like elephants, wolfs, or chimpanzees. Discourse, or linguistic communication, is simply a species of such norm-governed social practices, according to Brandom, albeit one with a distinct overall structure and characterised by performances that exhibit internal syntactic, morphological, and phonetic complexity. (As far as their mastery of these syntactic, morphological, and phonetic features of these performances is concerned, Brandomian speakers may well be broadly Chomskian creatures. That is, such mastery may well develop from rich, innate abilities, as long as it is exclusively a matter of implicit grammatical skills rather than a matter of explicit knowledge of grammatical rules.) In short, in Brandom’s world, competent participation in at least the simplest genuinely discursive social practices is entirely a matter of implicit, norm-governed, practical know-how, free from any knowing that.
The most basic form of reasoning in Brandom’s world, from which all other forms derive, is discursive reasoning, which is simply an aspect of a speaker’s activity of skilfully participating in the simplest kinds of discursive practices. How so? Given the norms governing their discourse, speakers are at each stage of their exchange committed or entitled to certain performances and prohibited from certain other ones. Moreover, given the norms governing their discourse, any actual or potential linguistic performance (viz. any sentences that is part of the discourse) stands, in the given context, in multifarious inferential or incompatibility relations to certain other actual or potential performances, in the following sense: a speaker’s commitment or entitlement to a sentence in the given context consequentially commits her to certain other sentences (that is, commits her to performing the sentence in immediate response to certain triggers), entitles her to certain other sentences, and precludes her from using certain yet further sentences.
In this sense, any sentence characterising the discourse occupies, given the norms and the context, a distinct position in a web of inference and incompatibility, relating it (as premise or conclusion, or as one half of an incompatible pair) to certain other sentences characterising the discourse. Putting his project squarely into the tradition of inferential role semantics, Brandom defends the view that the propositional meaning of a linguistic performance (viz. of the corresponding sentence) deserves to be regarded as the distinct position the performance (viz. the corresponding sentence) occupies in the web of inferential relations. The meaning that p of a sentence S thus is the set of norms of inference and incompatibility relating S as premise or conclusion, or part of an incompatible pair, to certain other sentences. Given this set-up, Brandom regards propositional attitudes as complex normative attitudes discursive practitioners take towards sentences of the language in use. To gloss the idea, to believe that p, in Brandom’s world, is to acknowledge a commitment to a declarative sentence S while implicitly treating S as having a certain inferential role – implicitly treating S as inferentially related to, or as incompatible with, certain further sentences S’, S’’, etc. – to the effect that p, in accordance with what the participant implicitly takes to be the norms. Reasoning in its original, most fundamental form is inferential reasoning in the simplest genuine instances of discursive practice, that is, it is a participant’s more or less skilful and appropriate effort to trace or acknowledge implicitly in practice the norms of inference and incompatibility relating the sentences in use to each other in various ways. Thus, whereas the reasoning of mainstream speakers is originally unrelated to discourse, solitary, and representational, the original form of reasoning in Brandom’s world is fundamentally social, discursive, and inferential: is an aspect of one’s practical ability to participate in the simplest kinds of genuine discursive practices.
The norms of inference and incompatibility constituting the meaning of a sentence are thus among the norms governing the use of the sentence in discourse, in Brandom’s world. Thus, apropos Wittgenstein’s dictum, meaning is in Brandom’s world an aspect of the use of language in communication. Moreover, pragmatics concerns features of the use of language in contexts, in contrast to context-independent features of language. Thus, if we interpret Chomsky’s hunch that language has only syntactic and pragmatic features, but no semantic ones, reductively rather than eliminatively – as saying that language has no semantic features over and above syntactic and pragmatic ones, rather than as saying that it has no semantic features at all – that hunch is true in Brandom’s world. In that world, meaning reduces to norms of inference and incompatibility, which in turn are among the discursive norms governing the use of language in context. Pragmatism, in Brandom’s sense, is the attempt to ground the semantic features of language in pragmatic features of the use of language in discourse, and Brandom’s theory is thus a (neo-)pragmatist one: it attempts to reduce meaning to pragmatic norms (to wit, norms of inference) and, correlatively, speaker’s linguistic understanding to pragmatic know-how (to wit, know-how to trace the norms of inference).
How do we get from this picture of meaning and propositional reasoning as fundamentally inferential, social, and discursive to the surface of meaning, reasoning, and knowledge as empirical and representing a largely non-linguistic, mind-independent Aristotelian world of states of affairs? Obviously, the story is long and complicated (as is, by the way, the story behind what I’ve just glossed). But it is also systematic, elegant, and lush with fine technical details. Moreover, its contours are quite straightforward. In Brandom’s world, the empirical and representational dimension of language and reasoning falls out from the interplay between three dimensions of discursive practice: the fine-structure of the language in use, the fine-structure of the social dimension of discourse, and the embeddedness of discursive processes in larger processes involving the non-linguistic environment, mediated by the participants qua living, embodied organisms. First, given the syntactic fine-structure of language, and given the meaning of a perhaps relatively small, finite pool of sentences (their inferential role relative to the other members of the pool), we get, via a certain procedure of substituting terms for terms in the generation of valid inferences, a theory of term meaning – a theory of meaningful singular terms (names) and general terms (predicates). Brandom thus explains, in strict top-down fashion, the meaning of (non-logical) words in terms of the meaning of sentences in which they occur, rather than the other way around. Moreover, given the pragmatic know-how corresponding to this top-down procedure, Brandomian speakers are able to generate and understand ever new well-formed combinations of meaningful names and predicates, that is, ever new sentences, and they thus master in practice at least the beginnings of the systematic and productive features of natural languages. Second, apropos the fine structure of the social dimension of discourse, an essential aspect of speakers’ ability to participate in discourse is their ability to track in practice their interlocutors’ multifarious discursive commitments and entitlements as best as they can. With each new discursive move and each shift in context, every participant acquires some new discursive commitments and entitlements and, usually, ceases to have some old ones, and speakers track in practice these shifts in their interlocutors deontic statuses as best they can. Borrowing a term from David Lewis, Brandom calls this tracking scorekeeping. Third, qua living, embodied organisms, discursive practitioners are in ongoing, goal-directed, feedback-governed, dynamic transaction with their natural and social environments. Organisms in general, whether or not they are rational or even sentient – and even artefacts like heat-seeking missiles – engage in ongoing, feedback-governed, goal-directed processes encompassing both the organism and environment. One bit of goal-directed behaviour alters some aspect of the environment or the organism’s situatedness in its environment, yielding updated information and correlative behavioural adjustment, yielding further changes in the environment or the organism’s situatedness, yielding yet further behavioural adjustment, and so on in an on-going feedback-governed loop.
So how does the representational dimension of discursive reasoning get into the picture? It does so, Brandom proposes, if we regard discursive practitioners’ intrapersonal reasoning processes and interpersonal scorekeeping processes, with their respective fine-structures and their intimate connections, in turn as segments of the larger, goal-directed, feedback-governed, dynamic processes involving those practitioners and aspects of their non-linguistic environment. Because they are segments of such larger processes, our thought and talk turn out to be representational. Indeed, they turns out to represent our environment as part of a larger, largely non-linguistic, mind-independent Aristotelian world of states of affairs. That is, in Brandom’s world, what ensues if organisms qua systems involved in feedback-governed, dynamic processes with their environments, also keep score on each other while engaging in norm-governed inferential discursive practices with the right kind of syntactic structure, is genuine propositional, empirical reasoning, inquiry, and knowledge about an objective world of Aristotelian states of affairs. Hence the sameness in surface appearance between Brandom’s world and the one depicted by mainstream analytic philosophers of language, mind, and, which disguises fundamental, irreconcilable differences beneath the surface.
Of course, any enthusiast can conjure up a grand vision of the nature of reasoning and communication. But what I admire about Brandom’s work is the tenacity, endless ingenuity, and careful attention to detail. While much of his overall project – though not, I hasten to add, the bit about representation – is inspired by Richard Rorty, his dissertation advisor at Princeton in the 1970s, Brandom’s role model for philosophical rigour and care is David Lewis, who served on his dissertation committee back then as well. It shows. Working through Brandom’s theory one finds, among much else, a carefully developed theory of non-formal, material inference – the kind of inference that constitutes empirical content and that characterises most of our ordinary reasoning – and of how such inferences relate to laws and statistical regularities of nature. One finds, under the heading of anaphora, a theory explaining how use of a term on a particular occasions counts, in context, as the recurrences of other usages of terms, perhaps by different speakers in other contexts and perhaps with different syntactic features (e.g. pronoun vs. proper noun). This theory of anaphora is reminiscent of Kripke’s causal theory of reference. Moreover, one finds a formal expressivist theory of at least some logical vocabulary, including modal vocabulary (further developments of which are apparently under way), a carefully developed deflationary theory of truth and reference, and a reliabilist theory of empirical knowledge that differs from more mainstream versions of epistemic reliabilism by being non-naturalist and by emphasising social aspects of reliability.
Yet despite its riches and rigours, the reception of Brandom’s work has so far been more energetic in continental Europe than on the west sides of the Channel or the Atlantic. While Brandom’s name is familiar in Anglo-American philosophical circles, deep engagement with his work there has so far been largely limited to former students and people interested in neo-pragmatist approaches to reasoning and knowledge. On the other hand, interest has been sustained and vigorous on the European, ever since the appearance in the mid-90s of Brandom’s landmark book Making it Explicit. This difference in reception is, I think, no accident. While Brandom shares with the analytic mainstream the aspiration for clarity, rigour, and careful attention to detail, and while his project has deep affinities with some influential approaches to language in the analytic tradition – for example, Donald Davidson’s theory of radical interpretation – he motivates and legitimises his project not so much through direct engagement with analytic work on language, mind, and knowledge, but indirectly by developing it through a deep, albeit selective engagement with key figures in the history of modern western philosophy – first and foremost Kant, Hegel, and Wilfrid Sellars, and, to a lesser extent, Frege, Wittgenstein, Dummett, and Heidegger – figures who Brandom takes to belong to a parallel “rationalist” and “inferentialist” minority tradition in modern western philosophy, spanning from Leibniz and Spinoza to the present. Thus legitimised, Brandom is satisfied with placing his finished product alongside the ones of the analytic mainstream (belonging to a dominating “representationalist” tradition) and letting people judge for themselves. Yet while, this strategy yields, in addition to all else, engrossing interpretations of some of these historical figures – above all Hegel and Sellars – absorbing it takes time and patience, and most people working more in the mainstream are, understandably, simply too busy pursuing their own projects and responding to critics who directly engage with the fruits of their labour. On the other hand, the legacy of German idealistic philosophy, transcendental and absolute, continues to influence much contemporary work on the continent, including work by people intimately familiar with the relevant analytic literature. Moreover, Brandom’s work was enthusiastically received by towering figures on the continent, such as Jürgen Habermas. Finally, academic philosophers working on the continent perhaps also tend to be more inclined to sinking their teeth into big, monolithic, single-authored bodies of work, rather than exclusively concentrating on debates that largely unfold through writing and cross-referencing dense, elegant, short articles. If all this is right, it may explain why Brandom’s work has so far resonated more on the continent than it has in England or North America.
There are, however, some signs of a reversal of this trend, as the philosophical climate in Anglo-American countries has been morphing, for the past decade or so, into an intellectual environment more hospitable for Brandom’s project. I’m thinking of several potentially overlapping recent trends here, such as work on embodied cognition in the philosophy of mind and psychology (e.g. by Francesco Varela, Evan Thompson, and Eleanor Roche and by Shaun Gallagher); work on collective intentionality, that is, on the beliefs and joint actions of groups of socially interacting agents, in social philosophy (e.g. by Margarete Gilbert or Raimo Tuomela); work in the philosophy of psychology and biology on human nature as fundamentally social, cooperative, and norm-governed (e.g. by Kim Sterelny, Michael Tomasello, or Tadeusz Zawidzki); and more traditional work on aspects of thought and language as expressive rather than as having a descriptive function (e.g. by Huw Price, Allan Gibbard, or Simon Blackburn, among other). Without going into the details, Brandom’s theory with its focus on us as simultaneously embodied and embedded in ongoing feedback-governed processes with our environments and as fundamentally normgoverned social creatures is congenial with much of this work, and there are signs that its importance is recognised by people working in these areas (I’m thinking here of recent work on Brandom by Shaun Gallagher or Tad Zawidzki, or Huw Price’s ongoing efforts to wed a Brandom-style neo-pragmatism with a global expressivism about thought and language.)
With such an array of connections to several important contemporary trends and to important post-WWII developments in analytic philosophy – not to mention to key developments in modern Western thought generally – hopefully the times of the relative isolation of Brandom’s work in Anglo-American analytic philosophy may slowly come to an end.