Negative Catallactics

A meandering, muddy, and generally low-quality post, which mostly just recapitulates recent lines of thought on the blog, by way of some underpowered reflections on what we’re doing in social-scientific explanation.  I’m going to start by talking briefly about some themes in analytic philosophy (atomism versus holism).  Then I’m going to bring in debates within analytical Marxism over microfoundations and methodological individualism.  Then I’ll talk about the way in which Brandom’s approach to the holism-atomism debate could in principle be used to resolve the analytical Marxist debate over microfoundations.  Finally, I’ll talk about why I don’t actually want to take the Brandomiam explanatory route I’ve outlined in the post – so the post will conclude without actually settling anything, but just pointing the way towards reading tasks I’m setting myself.

~~

Start with the ‘inauguration’ of Anglophone analytic philosophy by Bertrand Russell.  In the early years of the twentieth century, Russell differentiated his new logical atomism from the holism of the then-dominant Hegelian idealist tradition.  Russell famously drew this contrast by comparing the holist picture of the universe to a bowl of jelly, and the atomistic to a bucket of shot.  (Clearly, we’re meant to think, shot is better than jelly.)  Holism sought to explain individual phenomena by seeing them as part of a larger whole, and unintelligible independently of that whole.  Atomism sought to reduce the complex whole to its simple components, and fully explain the whole as made up of those components.  For Russell, one of the core objections to holism was its mystical failure to provide any kind of explanatory clarity – in this sense, holism was anti-scientific.  By contrast, atomism was all about explanatory clarity – the goal of atomistic explanation was to break down complex wholes into their more basic parts, and this explanatory process is the core of the scientific impulse.  In this sense, an ‘analytic’ philosophy, which sought to break complex wholes down into their atomic subcomponents in a way that brought analytic clarity, was a ‘scientific’ philosophy, even if it was not scientific in the sense of being tethered to experimental observation.  (Whether Russell is right to think about science in these atomistic terms is another question which this post will not address.)

This atomism was the “fighting faith” of Russellian analytic philosophy.  One of the features that picked out this analytic tradition from other atomistic or similarly reductionist explanatory traditions was its emphasis on language.  Russell’s project was, in the first instance, a project in philosophical semantics, which proposed a specifically logical atomism.  This logical atomism was, for Russell, as for many though not all figures in the Vienna Circle, tethered to a form of Humean empiricism, in which sense-data was the only ultimate ‘content’ of atomic meaning-units.  There were all kinds of philosophical problems with this project, which the subsequent analytic tradition has litigated at huge (and intellectually valuable) length – it’s not at all obvious that atomism in anything much like Russell’s sense can even be made to make sense as either a semantics or a metaphysics.  And my general philosophical orientation is strongly anti-metaphysical, so I’m not interested in pursuing the debate between atomism and holism at the metaphysical level at all – my (conventionally post-Kantian) view is that metaphysics is basically just an ill-formed project.

With that important caveat, the distinction between atomism and holism can still be made at the level of semantics – and, separately, it can still be made at the level of ‘vibes’.  At the level of semantics, Brandom’s project is, of course, a form of holism: Brandom argues that you can’t make sense of apparently ‘atomic’ constituents of linguistic content without understanding the content of those apparently ‘atomic’ units as granted by their functional role within a much larger semantic system.  We need to explain semantic components by their place in the whole, rather than the semantic whole as made up out of the building blocks of those independently-analysable meaning-units.  Brandom, then, is reviving an explicitly Hegelian semantic holism, using resources from the ‘analytic’ tradition: from Brandom’s rationally reconstructive perspective the tradition has gone ‘full circle’ back to the Hegelian holism it began by rejecting.

Fair enough.  But Brandom’s holistic semantics is not itself explanatorily fundamental – Brandom embeds his semantics within a pragmatics, which explains the semantics as instituted by human action.  And I think one can at least arguably make the case that Brandom’s pragmatics is more ‘atomistic’ than his semantics.  This claim would need to be made with many caveats and grains of salt.  After all, at the metaphysical level Brandom argues that there is a hylomorphism between the two poles of the intentional nexus: the holism that makes individual meaning-units unintelligible outside of a larger inferential network also makes individual objects unintelligible outside of a larger network of lawlike interconnected modal dispositions.  In this sense, both the ‘normative’ and the ‘objective’ poles of Brandom’s account of (what other philosophers would call) the mind-world relation are equally holist.  Since Brandom’s pragmatics is grounded in his account of law-like dispositions, this means one can’t call his pragmatics ‘atomistic’ in any very strong sense at all, certainly not metaphysically.

Still, I don’t endorse Brandom’s metaphysics (because I don’t endorse any metaphysics).  And at the level of ‘vibes’, rather than of metaphysics, I think there’s a case to be made that Brandom’s pragmatics is towards the more ‘atomistic’ end of the ‘atomism-holism’ spectrum, in at least some sense.  That sense is the following: Brandom’s pragmatics provides ‘microfoundations’ for his semantic holism.  

My shift here from talking about metaphysics to talking about ‘vibes’ comes with an obvious risk of a complete collapse of clarity.  What am I even saying when I start discussing ‘vibes’?  In my head, it’s legitimate to think in these kinds of terms because of the “psychodynamics as first philosophy” social-psychological subject naturalism that I’m trying to work with here.  Analysing things in terms of ‘vibes’ can mean analysing the social-psychological affective structure that ultimately institutes or constitutes our other objects of analysis (or, at a minimum, our understanding of those objects).  So, in shifting here from discussion of metaphysics to discussion of ‘vibes’, I mean to differentiate two different philosophical (and social-scientific) dispositions, which relate to the category of explanation which is found (for want of a better phrase) affectively satisfying.  Recall that on the ‘default-challenge-response’ Brandomian model of authority that I’m endorsing, every community has some kind of background commitments that are locally unchallenged.  Different communities therefore have different ‘discursive locations’ where the chain of challenges comes to rest in a commitment that is accepted as presumptively legitimate.  Thinking about things in this way lets us differentiate two different (ideal-typical) categories of social-scientific explanatory community: those that are ‘satisfied’ when explanation ‘comes to rest’ at the level of a social whole, and those that are ‘satisfied’ when explanation ‘comes to rest’ at the level of the ‘atomic’ subcomponents of the social whole.  We don’t need to think of these two approaches as requiring different metaphysics, or different social ontologies (though they may) – at base they can be analysed in terms of these different kinds of affective explanatory satisfactions.  These are two different social-emotional dispositions.

So – one kind of community is typically satisfied when it can direct explanatory challenges to supra-individual ‘social facts’; another kind of community is typically satisfied when it can direct explanatory challenges to ‘atomic’ individual-level human actions or attitudes.  Very broadly, sociological theory tends to align with the former approach, and economic theory with the latter.

What I’m saying is that, although Brandom is neither a semantic nor a metaphysical atomist, Brandom’s pragmatics is sufficiently detailed in its account of the specific social practices that institute norms and meanings that it seems reasonable to me to see Brandom as offering a ‘microfounded’ account of the social practices that institute (holistically analysable) meanings.  In this specific sense, Brandom can be seen as closer to the ‘atomist’ pole of the ‘vibes-based’ spectrum of atomism/holism.  At one level, then, Brandom is maximally rejecting the Russellian project – he is rejecting semantic atomism in favour of semantic holism, and indeed a Hegelian semantic holism.  At another level, though, Brandom is engaged in the analytic decomposition of meanings into their constituent parts – he is simply doing so at the pragmatic level of actions rather than the semantic level of meaning-units.  Pragmatic microfoundations can be used to give an account of how semantic content, understood holistically at the level of semantics, is enacted.  In this sense Brandom is (as he puts it in ‘Between Saying and Doing’) extending the traditional project of analysis.

Fine.  This is all an incredibly laborious (yet simultaneously quite vague and sloppy) way of drawing a very familiar distinction: between the kinds of social science that emphasise ‘social wholes’ and the kinds of social science that emphasise ‘microfoundations’.  Still, I’ve approached things in this laborious way in part because I think some common ways of drawing that distinction are quite misleading.  I’m now going to pivot to a discussion of one such misleading way: a debate within ‘first wave analytical Marxism’ over functionalism and microfoundations.

~~

In his reconstruction of Marx’s philosophy of history, which more or less kicks off ‘analytical Marxism’ as an English-language tradition, G.A. Cohen makes the case for the legitimacy of functional explanation in the social sciences.  By contrast, in a series of texts around the same time, Jon Elster is extremely critical of what he sees as a vastly excessive use of functional explanation in the social sciences, and specifically in Marxism.  Cohen and Elster have some back-and-forth on this topic, with Elster moderating his position somewhat, but still clearly thinking that Cohen is being egregiously sloppy in his casual use of functional explanations.  The dispute, as I see it, boils down to the following.  Both theorists agree that functional explanations are sometimes legitimate.  Both theorists agree (though Elster shifts his position towards Cohen’s, on this, in response to critique) that functional explanations can be valuable even if we can’t specify the mechanism that makes them true.  So, for example, Darwin was able to give an incredibly powerful functional explanation of biological traits (evolution via natural selection) but that explanation specified mechanism only in the most abstract way – Darwin (and Alfred Russel Wallace) had no account of how evolution via natural selection actually operated in any kind of detail.  This didn’t stop Darwinism from being a scientific breakthrough – a breakthrough without (biological) ‘microfoundations’.  

But Cohen and Elster disagree over how important it is to be able to specify the mechanism that ultimately makes our functional explanation work.  How legitimate is it to give a functional explanation without detailed explanatory microfoundations?  Cohen has a pretty relaxed attitude on this score: for Cohen, sure, it would be nice if we could specify mechanism, but science progresses slowly, and mechanism can wait.  For Elster, by contrast, bogus and nonsensical functional explanations are so rampant in the social sciences that anyone who gives a functional explanation without an account of specific causal mechanisms undergirding the explanation can legitimately be presumed to be just bullshitting.  For this reason, Elster thinks that we have a strong obligation to microfound our social analysis; Cohen thinks we only have a weak obligation.

In this whole exchange, I think Cohen pretty convincingly stands his ground against Elster’s initial position (which is a too-stringent rejection of non-microfounded functional explanations).  Moreover, Elster’s unsympathetic reading of other social theorists, alongside the limitations of his own theoretical framework, can lead Elster to reject as bullshitters theorists who are actually making perfectly good sense, at least by my lights.  (I’ve argued this with respect to Elster’s critique of Bourdieu; I think Elster is also ungenerous in his too-frequent attribution of teleological functional explanations to Marx.)  Nevertheless, in terms of what ‘explanatory resting-point’ I myself find satisfying, I am, at base, with Elster.  Elster is right, it seems to me, that functional explanations which do not see themselves as obliged to give an account of actual causal mechanisms in play are a haven for poor reasoning and explanatory evasiveness.  Indeed, this goes for Cohen himself: although his arguments about functional explanation are interesting, the theory of history that Cohen reconstructs and attributes to Marx in his book is (in my view) basically silly.  This is the kind of theory that can look sort of plausible at the right level of abstraction, and as long as one doesn’t start talking specifics in too much detail, but once one starts to think concretely about mechanisms it looks extremely shaky.  At base, then, I think Elster is absolutely right to place a strong demand for detailed causal mechanisms on functionalist explanations, and the “detailed” bit of that demand means microfoundations.

Ok.  So let’s say we align with Elster against Cohen on the importance of microfoundations in social science.  Then the question becomes: what microfoundations?  (Here, by “microfoundations”, I of course mean the analytic decomposition of a large-scale phenomenon into more ‘atomic’ subcomponents – that’s the connection to the first part of this post, on holism and atomism.)

The Elster and John Roemer of this period have the same clear answer to this question of microfoundations: rational choice theory.  Both Elster and Roemer will ultimately end up breaking with traditional rational choice theory, but in this post I’m interested in the original analytical Marxist project.  Elster in particular is strongly committed to methodological individualism (a commitment he retains to this day, even though he is much more sceptical than he used to be about the “rational” bit of “rational choice theory”).

The problem with this approach to microfoundations, for analytical Marxism, as both Elster and Roemer acknowledge, is that methodological individualism stands in clear tension with one of the core substantive commitments of Marxism (and of much other social science): the idea that “social being determines consciousness” – that we cannot understand individuals just as individuals, but that even apparently individual-level intentions are in some deep way socially formed.  Here is Roemer:

Neoclassical economics, being primarily concerned with positive description, usually (though not always) postulates individual preferences as given.  I think this postulate is more than methodological: it is ideological and flows from the dictum ‘Cogito ergo sum.’  Marxism reverses the Cartesian epigram, and asserts that people’s preferences are in large part the consequence of social conditioning.  This is an important consideration, if one intends to use rational choice models for describing welfare, for making normative judgements about the consequences of rational activity.  Rational choice models should be used to develop a theory of endogenous preference formation.  A materialist psychology is necessary to derive preferences from endowments and history. (193)

And here is Elster:

rational choice theory takes the desires and preferences of the agents as given, whereas Marxists want to explain them…. the preferences of each depend on the actions of all.  Although neither Marxism nor any other social theory has gone very far towards a theory of endogenous preference formation, this is clearly a very important problem. (209)

(Both quotes are from the Roemer-edited 1986 ‘Analytical Marxism’ volume.)

So there is a problem, which both Roemer and Elster acknowledge, with using rational choice theory as our explanatory microfoundations: rational choice theory, in its traditional form, has limited resources for the analysis of endogenous preference formation.

Now, since Roemer and Elster published these pieces there has been a vast amount of work on the issue of endogenous preference formation.  I’ve read almost none of that work – this is one of the (many) reading tasks I need to spend a lot more time on, and, unsatisfyingly, I want to largely bracket the issue of endogenous preferences until I’ve done at least a little more of that reading.

Still, even bracketing the question of endogenous preference formation, the Marxist emphasis on the social determination of consciousness potentially opens up an even more ‘fundamental’ set of philosophical issues.  Granted that individual-level preferences are socially shaped, is it also the case that the very existence of preferences as such is socially instituted?  The neopragmatist approach to the problem of intentionality answers “yes” to this question.  For the neopragmatist, ‘original intentionality’ is itself a product of social practices, and therefore to stop one’s analysis at the level of individual intentionality (as one’s ‘atomic’ unit of microfounded explanation) is to make a mistake about the appropriate basic explanatory unit for our social-theoretic explanation.  If intentions in general can only be understood by reference to sociality, then an account of endogenous preference formation is still mistaking a social product for an explanatory building block. 

This neopragmatist explanatory approach suggests that Roemer and Elster are wrong about what is required to microfound Marxism.  Roemer and Elster both identify microfoundations with methodological individualism and, more specifically, rational choice theory.  But this is not the only available theoretical option.

In an important and (in my view) under-discussed contribution to the Carver and Thomas edited volume ‘Rational Choice Marxism’, Mark E. Warren challenges Elster’s version of methodological individualism.  Warren distinguishes between three categories of methodological individualism, which he calls “methodological individualism of events”, “methodological individualism of subjects” and “methodological individualism of actions or practices” (233-4).  I think this is a slightly confusing way of putting things – for my tastes, only the second of these (“of subjects”) is what people normally mean by ‘methodological individualism’ – the subject is the individual in methodological individualism.  I think it’s better to see Warren’s three categories of ‘methodological individualism’ as three different categories of microfoundations.  Rephrasing Warren’s point in my own preferred vocabulary, I take it that Warren is endorsing Elster’s explanatory emphasis on microfoundations, but rejecting what I would call Elster’s methodological individualism.  Instead, Warren is proposing that we can microfound our account of social reality on the ‘atomic’ basis of actions and practices, instead of individual subjects.

I think this is a very important point, and it opens up a research programme within analytical Marxism that has (as far as I know) been largely unpursued.  Moreover, this idea has obvious overlap with what Brandom makes of Hegel in A Spirit of Trust.  Brandom reconstructs Hegel as a ‘fundamental pragmatist’, whose semantic holism can be grounded in the microfoundations of linguistic practice.  The Hegel of A Spirit of Trust thus offers an account of meaning which is microfounded in Warren’s third sense of methodological individualism (“actions and practices”), but which rejects Warren’s second sense (“subjects”).  In other words, Brandom’s Hegel offers just the kind of apparatus that could allow us to ‘make good’ on Warren’s proposal for bringing Roemer and Elster’s commitment to microfoundations more in line with Marx’s own approach.  And, of course, this makes a lot of sense, because Marx himself is heavily influenced by Hegel.  If Brandom is right that this approach to microfoundations really can be extracted from Hegel, well, maybe that’s part of where Marx has gotten it from.  On this approach, Elster and Roemer are looking in the wrong place for microfoundations for Marx’s framework.

All of this, as I say, points the way towards a research programme.  That research programme is the following: take Brandom’s analytic reworking of Hegel as our starting point and, just as Marx ‘materialised’ Hegel, ‘materialise’ Brandom’s Hegel in a way that integrates analytic Hegelianism with analytical Marxism.  In the past, on the blog, I’ve made a number of moves in that direction.  And I’m not the only person thinking along these lines.  In 2024 there was a conference at University College London on ‘Marxism and the Pittsburgh School’ (meaning Sellars, Brandom, McDowell), where a number of speakers discussed related ideas – I still haven’t listened to all the presentations at that event, but they are available on YouTube. 

Speaking for myself, and speaking very loosely, I think there are two broad approaches one could take to bringing Brandom and Marx together.  The first approach is to offer a more ‘semantic’ reading of Marx.  There is a major figure – Habermas – who is already doing something in this space, with his reworking of Frankfurt School critical theory in a linguistic philosophical register, and his grounding of emancipatory critique in the philosophy of language as first philosophy.  Brandom himself cites Habermas as an important influence, so this is a very reasonable project to pursue.

The second approach is to offer a more ‘materialist’ reading of Brandom.  This is the project I’ve been pursuing.  The goal here would be to try to extend Brandom’s normative pragmatics beyond the domain of linguistic philosophy.

So – there’s an available project here, which would take something like the following shape.  We agree, with Elster and Roemer, that Marxism should have microfoundations.  We also agree that those microfoundations need to give an account of the social formation of preferences and opinions – intentional content.  We follow neopragmatism in adopting a ‘radical’ version of this project – adopting the view that intentionality itself, in general, is socially instituted.  We therefore follow Warren in the view that we need to reject methodological individualism about subjects as our preferred microfoundation – instead, our microfounding categories should be actions and practices.  Brandom gives us a very elaborate apparatus for making good on this approach – but Brandom is too focused on the linguistic.  So we need to extend Brandom’s practice-theoretic microfoundations of the theory of intentionality to the kinds of practices that are of interest to Marxists.

This seems to me like a project extremely worth pursuing.  Nevertheless, it’s not quite the project I want to pursue on the blog.  That’s because – as I keep saying – Brandom is an eliminativist about the mental, and I think that our social theory should not only give a central explanatory role to practices, but should also give a central explanatory role to desires.  I’m not prepared to follow Brandom in evacuating subjectivity, or in analysing desire in purely normative expressivist terms.  So the Brandomian apparatus is unacceptable to me, along this dimension.

So here is my dilemma.  I endorse the ‘microfoundations’ approach to social-theoretic explanation.  But there are two paradigms of microfoundational explanation I’ve been discussing, which have directly opposite explanatory directions.  On the one hand, there is the “desire first” explanatory direction which finds formal expression in decision theory: the individual and their desires are the basic units of analysis.  The problem with this approach is that, pretty clearly, desires cannot be adequately analysed – certainly not fully analysed – at the level of the individual.  On the other hand, there is the “practice first” explanatory direction which finds formal expression in Brandom’s apparatus: social practices are the basic unit of analysis.  The problem with this approach is that it evacuates desire – it gives desire no basic explanatory role at all, and I find this a very unappealing feature of the framework.

Is it possible to have one’s cake and eat it too, here?  Is it possible to treat desire as explanatorily basic, but not analyse desire in a methodologically individualist way?  In some sense, I think, Hegel is trying to do this – but honestly I don’t know what it would look like to reconstruct this element of Hegel in a more legible analytic idiom.  Lacan is also doing this – but Lacan is even more opaque – and much less rewardingly and justifiably so – than Hegel, by my lights.  So I feel like I have limited resources available for thinking about this question.

My current broad plan of action, then, is to attempt to advance along three fronts.  First, there’s social identity theory – that is to say, the social-scientific analysis of the way “The I that is We, the We that is I” actually works in social practice, rather than in the utopian ideal of Hegelian social totality.  Second, there’s mathematised decision theory.  Here I’m particularly, but again of course by no means exclusively, interested in social preference formation.  My basic plan is just to read more in these traditions, bashing them against each other, and see where I end up.  Third and finally, there’s the philosophical or metatheoretical level, trying to make sense of all of this philosophically.

This has been an extremely muddy post, even by my standards – there are quite a few things in here that I’m already unhappy with, even before hitting publish.  But what to do – I’m going to follow the blog’s general approach of just posting through it, and see if I end up anywhere more worthwhile, eventually.

In my previous post I distinguished three different attitudes we might take in the ‘hermeneutics of suspicion’.  These were: 1) we might take it that somebody is simply lying about their intentions; 2) we might take it that somebody is deceiving themselves about their own intentions (through suppression or denial formations, for example); 3) we might take it that somebody is not aware of the content of their own intentions, because that content “just ain’t in the head” – because, that is, we endorse some kind of externalist semantics, or externalist account of mental content.

In other posts, separately, I have distinguished different things one might mean by ‘methodological individualism’.  These include: 1) the idea that intentions can only properly be attributed to (let’s say) human individuals; any attribution of goals, etc., to other entities (states, say, or corporations, or classes) are metaphoric, and in some sense therefore methodologically improper; 2) the idea that intentions are always in some sense ‘self-interested’; 3) the idea that intentions themselves can be fully adequately specified at the level of the individual agent (at whatever scale of agent) – that is to say, intentions are not socially instituted.  I take it that mainstream decision theory can relatively easily accommodate the abandonment of methodological individualisms (1) and (2).  The real problem for mainstream decision theory is commitment (3).  And, non-coincidentally, this is the same basic commitment as internalist semantics.  The difficult problems in the ‘hermeneutics of suspicion’ and in the debates over methodological individualism therefore boil down to the same core set of issues around the ‘internalist’ versus ‘externalist’ specification of mental content, or intentionality.

When I consult summaries of the debates around externalist semantics or externalism about mental content (for example, in the Stanford Encyclopedia of Philosophy) they typically distinguish two different lines of argument for semantic externalism.  On the one hand, there are arguments deriving from Hilary Putnam’s ‘Twin Earth’ thought experiments; these arguments rely on the idea that attitude-independent facts are partly constitutive of the meaning of our claims and thoughts – for example, the idea that what ‘water’ means is partly determined by what the thing we refer to as ‘water’ actually is, whether or not we know what it is.  On the other hand, there are arguments derived from Tyler Burge’s thought experiments (around the meaning of ‘arthritis’, for example); these arguments rely on the idea that community usage or practice determines the meaning of our claims and thoughts, and the relevant dimensions of community practice may fall wholly outside the psychological life of the individual.  These are, roughly, arguments from mind-independent natural phenomena, and arguments from individual-independent community attitudes.

I endorse a social-perspectivalist pragmatist account of truth, including of scientific truth – something in the space of Peirce’s ‘ideal observer’ theory of objectivity.  From my philosophical perspective, then, Putnam-style ‘physicalist’ arguments can be reworked in a social-perspectival idiom.  In other words, I take it that Putnam-style object-dependent semantic externalism can ultimately be ‘reduced’ to Burge-style social-perspective-dependent semantic externalism.

From my point of view, then, the ‘hard cases’ in both the hermeneutics of suspicion and the debates around methodological individualism come down to the following question: how do we ‘cash out’ social-perspectival semantic externalism (or, perhaps more broadly, social-perspectival externalism around mental content)?

Burge opens his article ‘Individualism and the Mental’, in which he lays out his community-focused account of mental content externalism, by referencing Hegel:

Since Hegel’s Phenomenology of Spirit, a broad, inarticulate division of emphasis between the individual and his social environment has marked philosophical discussions of mind.  On one hand, there is the traditional concern with the individual subject of mental states and events…  On the other hand, there is the Hegelian preoccupation with the role of social institutions in shaping the individual and the content of his thought. This tradition has dominated the continent since Hegel.  But it has found echoes in English-speaking philosophy during this century in the form of a concentration on language. Much philosophical work on language and mind has been in the interests of Cartesian or behaviorist viewpoints that I shall term “individualistic.” But many of Wittgenstein’s remarks about mental representation point up a social orientation that is discernible from his flirtations with behaviorism. (73)

Burge laments the relative unclarity of the traditions focused on social environment explanations of mental content:

Philosophical discussions of social factors have tended to be obscure, evocative, metaphorical, or platitudinous, or to be bent on establishing some large thesis about the course of history and the destiny of man. There remains much room for sharp delineation. (73-4)

Burge takes his work to be a contribution to this project of clarifying the tradition that gives social-theoretic externalist accounts of mental content.

Brandom’s work is also a contribution to this project – A Spirit of Trust is surely the most impressive effort to rework or rearticulate Hegel’s externalism about mental content in a way that makes contact with the resources of the analytic tradition, bringing “sharp delineation” to that Hegelian and post-Hegelian project.

Brandom doesn’t discuss Burge much, to my knowledge – and when he does so, he is typically critical.  I haven’t read much Burge yet, and I don’t understand the points of similarity and difference between Burge and Brandom’s projects.  I do, though, at this point, have relatively well-developed views on many elements of Brandom’s project – what I endorse and what I reject.  I endorse the pragmatism, the social-perspectivalism, the inferentialism, and the semantic externalism; I reject Brandom’s rejection of ‘instrumentalism’ in the philosophy of action, as well as his linguistic exceptionalism and his eliminativism about the mental.

What this means, in effect, is that I’m casting around for ways to explore the social-perspectival externalism about mental content that Brandom articulates which do not require the additional commitments I find problematic in Brandom’s project.  In other words, I need to read more in this broader ‘externalism about mental content’ theoretical tradition.  Burge seems like a good place to start!