The Social Construction of the Self

May 28, 2011

{Post three of four}

So, continuing backwards through the errors I listed earlier, consider the idea that the aspects of self that should be used to push with (against other aspects of self, or against circumstance) should be truly internal to the self – innate – core.

I don’t want to be unfair to my earlier self here [and, to be clear, the ‘earlier self’ I’m engaging with is one from prior to my online presence – I don’t take myself with be arguing with the author of ‘Praxis’.]  On the one hand I don’t find it at all implausible that there are important aspects of self that are ‘core’, in the sense of being biologically coded in such a way that they are not vulnerable to transformation via experience.  On the other hand, my earlier self’s position was not as naive as might appear but was (rather) uncertain, or self-consciously contradictory.  My earlier self was interested in tracing the origins and guiding principles of normative demands through the vicissitudes of the instincts that produced them, in order to locate simpler demands that underlay these – a discourse on desire in the psychoanalytic sense.  At the same time, my earlier self was aware of the facts that a) the transformations that instincts undergo as biologically simple impulses interact with a complex social field (itself made out of such impulses) and are impacted by that interaction, are real transformations – the fact that a normative demand can (perhaps) be analysed into the complex interaction of simple ‘instinctual’ components does not in itself debunk the normative demand.  (Nor did Freud think this, I should add.)  At the same time b) there is some sense (and what sense, exactly, my earlier self always had a hard time articulating) in which these instinctual demands, when taken as demands, are themselves socially constituted.  How to understand this ‘social constitution’ in a way that was not metaphysically idealist, and yet, also, not biologically essentialist, was a problem that my earlier self had some difficulty navigating.  (And, to put it mildly, the metaphysical inclinations of psychoanalytic theory – which was the principle intellectual resource I was making us of to try to understand these questions at that time – didn’t help.)

I believe I can now provide a much better and more detailed account of all of this.  In particular, there are two ways in which an apparently ‘simple’ impulse can be understood as socially constituted.  In the first place, the nature of the disposition itself can simply be influenced by social interaction — this is true for many dispositions that are classically (and falsely) thematised as ‘natural’.  [I’m not, for example, unsympathetic to the idea that there are strongly socially uninfluenced aspects to individuals’ sexualities – but even if this is the case, the socialisation of the human organism is sufficiently impactful that surely any given individual sexual behaviour beyond at least the very youngest infancy is going to be partly socially produced in some sense.]  This is obvious, though.  The more important point is that putting all this aside, and taking no stance at all on the origins of any given biological disposition, Brandom’s apparatus (which I of course endorse) makes clear to us that to the extent that any practice is taken as normative, this must be because it is taken as normative – in other words, the property of being-normative is not innate, but is only ever socially granted.  A discourse on desire in the psychoanalytic sense is intrinsically a discourse about normative practices.  Desires, drives, investments (cathexes), tell us what we should do (even if that ‘should’ is contradicted – and, potentially, suppressed – by other components of the psychic economy).  And this ‘should’ can only ever come from the drive in question being part of a larger social set of practices.  In principle (as I’ve said before) this larger set of practices could be restricted to the internal dynamics of a single organism – but in practice, if we are dealing with human desire and human dispositions, the larger set of practices will be social in the regular sense – that is, composed of the interacting practices of a number of different human organisms.  In this sense, even if in no other, desire is always already social.

Now if we connect this idea up to the understanding of the content of commitments that I’ve recently articulated, we can see that to the extent that a desire is a desire – to the extent that a disposition is normative (again, I take these to be different ways of talking about the same set of social-psychic phenomena), I take it that a desire also involves a normative commitment (because desire is inherently (even if implicitly) intentional, and intentionality is incomprehensible without the idea of commitment – even if that commitment is one that we simultaneously disavow or sublate).  And, as we have seen, the implicit normative content of a normative practice can only be understood from another social-perspectival location — and there is always the possibility of multiple competing social-perspectival locations providing rival accounts of the implicit content of any given normative disposition.

So to take a desire as possessing content means treating of not just the simple biological disposition, but also a much larger and more complex social system, from a location or locations within that system from which this attribution can be made.  And the ‘simplicity’ of the simple desire (the base impulse that – at least in some versions of psychoanalytic theory – grounds the psychic edifice) is really no such simplicity, but also involves the many implicitly normative practices that enable the attribution of this content (and the same is of course true of all of these practices in turn).

How we understand the ‘core’ of the self, then – and, indeed (since there is – on Brandom’s account – nothing to normative content beyond the attribution of normative content) what that ‘core’ of the self even is, cannot be understood except from within a larger social space.  In this quite strong sense, the self is socially constituted, even if (as is unlikely) the actual ‘core’ dispositions we’re dealing with are unaffected by lifelong social experience.

All we need now, however, is for the self’s dispositions to be affected by what the self understands the content of its dispositions to be [note that all of the above would still be the case even without this additional assumption – but the assumption is, I believe, correct], and this process of attribution of normative content can be formative of the self in an still stronger sense – in the original, obvious sense given above.  I am talking about, in the first place, empathy – the understanding and formation of the self by the self seeing its self through the minds of others – and I think that this is, indeed, a powerfully formative aspect of socialisation.  It should be noted, as well, however, that although I just wrote of ‘seeing’ through the ‘minds’ of others, this kind of formative process need not be conscious – and need not be empathic.  The consciously psychological processes I’m talking about here are only one kind of normative social practice – we know for a fact that there are other kinds and, as I will eventually argue, these can be formative of the self in structurally parallel ways.

Sticking with the more overtly psychological aspects of this account for now, this account gives us several distinct ways in which the content of the self is formed by what the self is taken to be.  This is a mirrored taking.  The self ‘sees’ itself through the normative attitudes of others.  And how it ‘sees’ itself is formative (in the two distinct senses given above) not just of what it takes itself to be, but of what it is.

My earlier self aimed to understand this dynamic through psychoanalytic theory (which theory has obviously significantly influenced how I’ve articulated the position above).  He understood this process as, fundamentally, about desire, and the social constitution of desire.  I no longer give desire this foundational role.  Desiring-dispositions are only one set of dispositions among others associated with the human organism, although they are important.  But my earlier self was not wrong to think about desire – and love – in these broad terms, I think. Two aspects of the standard thematisation of love – the desiring-disposition and the inhabiting of another heart and mind – are connected in this account, and shown as mutually implicating.  And, on the Brandomian account, of course, the inhabiting of others’ perspectives, in whatever sense, is a practice without which no other normative practice would be possible.

In addition to all this, this account explains, I think, why my earlier self was wrong to see circumstances as determining in quite the strong way he did.  For the self is formed by the social sphere.  But this social sphere is not just the social sphere the self finds itself in and happens to experience, but also, within that sphere, the social-perspectival locations that the self chooses to identify with, and how.  The self is made by whose attitudes it experiences itself through – or, more properly, what practices it interacts with itself via.  The self can – to some greater or lesser extent, depending, of course, on objective circumstances – choose what other social-perspectival locations it chooses predominantly to inhabit, and to endorse as sources of legitimate judgement – and this choice is formative in the ways I’ve just articulated. We choose what we love, and who.  We choose what communities we identify ourself with, and how.  We choose whose practices become part of our own.  This is, as I now see it, the rational core of my earlier self’s wrestling with the psychoanalytic concept of libidinal investment.  And in this way the self is made, my love – belonging not just to itself, but to you.


12 Responses to “The Social Construction of the Self”

  1. Jed Harris Says:

    This post certainly makes clear you don’t reject “built in” dispositions, as I was concerned you might in my first comment on “Self; Errors”.

    A few thoughts, which I’m afraid don’t add up to a clear “point”:

    Your emphasis on “simple” dispositions may be misleading. For example the human disposition to acquire language may be analyzable into a set of “simple” dispositions (though we don’t yet know comprehensively what they are) but certainly they interlock to constitute a complex abstract disposition. Also, certainly, this disposition is pretty much universally “built in” to humans. This disposition is so powerful that humans growing up together with no language to learn will invent one (the strongest examples are congenitally deaf children growing up with no sign language “speakers” in their environment).

    I agree with your general point about even built in dispositions being socially constructed through being taken normatively in one way or another.

    But I think it is important to recognize that there’s a “strange loop” here. We agree that taking anything as normative requires a social context, and also requires inhabiting others’ perspectives. But I think in humans our social context and our ability to inhabit others’ perspectives depend on fairly complex built in dispositions — perhaps as abstract and complex as our disposition to acquire language. These dispositions give our social context and perspective taking a “grain” that makes some sorts of normative takings easy and others very hard.

    I believe I can come up with thought experiments that would explore this further, postulating credible populations that have a completely different dispositional basis for social context and perspective taking. But I haven’t done that yet, and even if I do it would take longer to explain them than would be appropriate in a comment.

    I’d like to be able to point to science fictional examples, but as far as I know they don’t exist, precisely because strong examples would involve beings that would be very hard to identify with or empathize with. The potential audience would not want to read stories that depended on these examples.

    This leads to a potential problem. I think we will tend to implicitly take the grain of our social / perspectival dispositions as analytically universal, because they are universal in our experience, and because our own dispositions make it hard to think outside of them. This is probably OK up to a point, but without a stronger grasp on the problem it is hard to know when we’ve crossed the line and are illegitimately basing conclusions on these implicit biases.

    A different, perhaps minor point about perspective taking: Perspective taking is distinct from and perhaps prior to social behavior. For example, it is useful to a predator to be able to take the perspective of prey — this enables the predator to hide more effectively, to better predict how the prey will attempt to escape, and so forth. Similarly it is useful for prey to be able to take the perspective of potential predators, to detect hiding places, choose unexpected escapes, etc. Even among social beings (e.g. human hunting bands) taking the perspective of other beings that aren’t members of the social context is important to survival.

    The salience of this sort of perspective taking among young animals (including humans) is very high. Many popular childhood games involve hiding and finding — peekaboo, hide and seek, etc. “Horror” films involve a lot of strong emotions evoked by playing with this theme. So I’d guess there are strong built in dispositions that drive the development of perspective taking skills somewhat independently of social context. Unfortunately I don’t know the research in this area very well.

  2. Jed Harris Says:

    I just saw an interesting post by Will Wilkinson that directly addresses the metaphysics of the self, and of normative judgements, in the context of some “experimental philosophy” data. (I guess experimental philosophy uses the methods of cognitive psychology to investigate questions with philosophical implications.)

    The post is at the Economist blog and you “just have to know” that WW is Will Wilkinson.

    Perhaps the most relevant (summary) paragraph:

    My own view is that the sense of a stable self is an evolutionary construction with a certain social function, which our intuitions about authenticity reflect. The primary human means of survival is social cooperation. But cooperation is fragile. We need to trust one another to follow through, to not take advantage. Coordinating on a common moral ideology facilitates cooperation, but only if we all stick to it. We cannot make others trust that we will stick to it if we cannot trust ourselves not to opportunistically change our stripes. So we build a sense of self upon the shared moral ideology of our local culture. We come to feel that to betray these values would be to betray the essential self. To prize integrity is to fear disintegration. To violate our constitutive values is to risk falling apart. This fear of falling apart—of losing one’s self, of standing for nothing—prods us to keep our oaths, to pull our weight, and thus to be truly trustworthy, even when it would be to our advantage, in some sense, to cheat. So the sense of self enables social cooperation. But what matters most is not so much the content of our moral ideology, but simply that we all stay pretty much the same over time, so that we can continue to trust ourselves and one another. This is not to say that the values upon which we build stable, cooperation-enabling senses of self can be anything at all. But anything that works works, and probably there are many moral ideologies that work reasonably well.

    I don’t know if you’d agree or disagree with this, but it certainly bears on your discussion and Brandom’s.

    I’ll also note that one interpretation of Wilkinson is that people have an evolved (and thus mostly built-in) disposition to “prize the integrity of their self”, to “fear dissolution of the self”, etc. However this “self” is constructed by the individual in response to their social context, can of course change (gradually or abruptly), etc.

    It seems to me that these feelings of fear, pride, and so on, associated with challenges to our personal identity, are pretty well attested in literature and our personal experience. Strong feelings pretty much imply to me a tie to some underlying built-in disposition(s). This sort of feeling related to identity doesn’t seem to me to be any more unlikely than the feelings related to territorial challenges, which we know are built in in many species.

    Of course territories are socially constructed (more or less by definition). However of course their definition is also non-verbal in non-humans, and has a big non-verbal component even in humans. Similarly, I would guess that personal identity doesn’t have to be tied very strongly to language.

  3. duncan Says:

    Hi Jed – sorry as usual that it takes me an age to respond. Just busy.

    I take your point (in the other thread) about complex and abstract dispositions. Language acquisition is a good example – I guess birds building nests and so on would be another. So yes – we shouldn’t rule out the idea that there are biologically untransformable dispositions at both an individual and a species level that are extremely complex in their structure and their consequences. The human organism in general is a hugely complicated thing, that performs countless incredibly rococo activities that are minimally impacted by socialisation – and some of those activities (like those that enable language acquisition, say) are strongly social in their consequences and nature. So I think we’re fundamentally on the same page. If I seem to downplay the importance of biologically invariant dispositions, I think it’s because –

    a) I’m ultimately interested in socially transformable dispositions and their consequences – things like structures of political and economic organisation, which are highly culturally variable. So although there’s a huge amount of interesting research to be read and done in the area of what the invariant dispositions are that enable social phenomena, the things I’m ultimately interested in analysing are a long way ‘downstream’ from these questions, and don’t I think require any firm conclusions on the biological-invariants front in order to be discussed. And then also;

    b) I perceive excessive ‘naturalisation’ as a more active conceptual problem than excessive ‘social constructivism’ in the research areas that most interest me. That may be wrong, depending on the direction specific people are coming from – but I’m in any case interested in elaborating a relatively strong social constructivism that’s actually compatible with empirically supported biological predisposition stuff. There shouldn’t be a conflict between these theoretical orientations, imo. My worry, of course, is with the intellectual inclination that takes as its premise that there are some biologically invariant dispositions of the human organism (true), and then immediately reaches the conclusion that, say, the institution of the New York Stock Exchange is an everlasting natural necessity. I’m likewise concerned to refute the philosophical suggestion that biological (or transcendental) invariance is a prerequisite of normative or conceptual legitimacy – and one reason to push hard on the social constructivism side of things is to make the point that even if we were maximally socially plastic (in some sci-fi scenario in which there were no species- or individual-invariant biological dispositions, for example) this would still not have any negative consequences for normative and conceptual content. There’s several things squashed together in this paragraph – but my point is that I think what’s going on here is just a difference in emphasis – I don’t detect a substantive disagreement between us on these issues.

    If you haven’t seen it you might find that my conversation with Pete Wolfendale here might address some of the issues that you raise in slightly more detail – for example predator-prey relationships.

    I’ll try to speak to your other remarks in a different comment…

  4. duncan Says:

    The Will Wilkinson post you link to is interesting – it prompted thoughts in a few different directions. The experimental data he’s referring to is interesting and fun. I don’t know where (if anywhere) it’s written up properly – I don’t know Knobe, Newman and Bloom’s work. My initial thought from a research-design point of view was that I hope they included questions about the research subjects’ assessments of the ethical value of the changes described, independent of the research subjects’ assessments of ‘true self’, etc. It’s possible in a questionnaire like this for research subjects to want to express ethical approval or disapproval, and to find the best proxy for that expression, even if it doesn’t precisely tally with their views. In other words, it’s possible that some research subjects are using the ‘true self’ language simply to express ethical approval, in a questionnaire format that doesn’t allow any other way to express that view. Plus it’d be interesting to get a direct view of the relationship between ethical judgement and judgment of true self, rather than simply using expressed political alignment as a proxy for the former (as seems to be the case? As I say I can’t find – on a rather cursory search – a full write-up of the research.) Even if all that’s the case, though, the results are still fun, and confirm my own intuitions, which of course I like 😛

    That said, I think there’s something problematic in the passage from Wilkinson’s post you cite. (Obviously I speak for myself here, not Brandom.) On the one hand, Wilkinson’s making noises I’m inclined (as you know) to agree with, about the social construction of self, and the idea that the aspects of self we treat as if metaphysically core are apparently so because we so treat them. I obviously agree with the basic orientation there. On the other hand, I believe Wilkinson tacitly reproduces the structure of argument he’s criticising in his own framing of the issue. The key line here is

    This fear of falling apart—of losing one’s self, of standing for nothing—prods us to keep our oaths, to pull our weight, and thus to be truly trustworthy, even when it would be to our advantage, in some sense, to cheat.

    Wilkinson is here, I think, tacitly positing a self capable of being advantaged and disadvantaged – and, probably, capable of making assessments in this regard – that can be understood distinct from the socially specific way in which that self has been constructed to respond to socially specific normative frameworks. His underlying picture, I suspect, is of a core self-interested homo economicus, with culturally derived ethical demands overlaid on top of that core self for reasons of broader evolutionary advantage. I suspect Wilkinson, in other words, of participating in the same structural move he criticises in proponents of ‘false consciousness’ theory. After all, the idea of core interests overlaid with and counteracted by culturally specific normative demands is the thing he’s complaining about – but he doesn’t seem to regard it as problematic to regard norms as such as a solution to the problem caused by an intrinsically self-interested individuals needing to engage in co-operative activity. If we’re serious about the social construction of self, of course, individual self-interest is similarly so constructed, and should not be contrasted with ‘external’ normative frameworks. There’s a double standard operating in Wilkinson’s framing of the problem, I think – an ideologically charged one. (Wilkinson’s own sense of the core self in homo economicus terms is part of the reason why he writes for The Economist and not, say, New Left Review, I’d guess.)

    But these are just some thoughts – apologies again for the usual delay in responding.

  5. Jed Harris Says:

    I guess I’ll reply in multiple pieces because I don’t have time to write all my thoughts about your (very interesting) comments in one shot.

    So replying just to your morning comment:

    – We appear to agree on the basic issue of built in dispositions. Your bird nest example is better than any of mine at illustrating this point. Good we have converged!

    – Birds’ nests, etc. are mostly independent of social context — assuming that birds aren’t imitating each others’ nest construction styles, which as far as I know they don’t. Territories, languages, etc. can’t on the other hand even be defined independent of social context. I don’t know that this makes any difference but perhaps it is worth keeping in mind.

    – Excessive naturalization is a valid concern. We have certainly had this problem from some people waving the banner of evolutionary psychology. Whatever else we do in this space, we need to clearly push back against that sort of “preformationism”. Something similar has been a problem in developmental biology, where people tend to interpret genes as “blueprints” rather than as developmental dispositions that can play out in a wide variety of complex ways, with the same genes perhaps generating very different phenotypes. And that is even independent of social context.

    I guess I have two concerns that pull me in a somewhat different direction from your and probably Brandom’s programs as they are playing out here. One of them is maybe just a difference is our priorities, but the other I think is more fundamental and is hard to avoid.

    – My priority is to move toward a good theory or language to describe these sorts of (human) dynamically produced and maintained structures. Such a language would of course have to systematically avoid preformationism, but it would also have to help direct our attention to sufficiently useful, general, perhaps even necessary features of the substrate.

    In contrast, as I understand your project, your priority is to find the most general analytic characterization of normative or meaning producing systems, even if (as I would assert) most of the space of systems thus described couldn’t be occupied by human social formations. Apologies if I have mis-described your intent.

    I couldn’t of course disagree with this choice, and would just hope that our respective projects would be useful to each other.

    – The other concern is a bit harder to articulate because it depends on a lot of background from game theory and related analytic modeling work. We already know a lot of constraints on the possible form of stable or self-reproducing social constructions — both very local, e.g. two players, and very large and statistical. Of course these depend on assumptions, which are very abstract and general, but also can vary in ways that radically change the conclusions.

    Without rehearsing all this background, one conclusion is that the maintenance of (what Brandom calls) norms requires very special background assumptions about the motivations or possible choices of the “players”. These assumptions are very abstract and general but also exclude many more possible modes of interaction than they permit.

    This modeling work is also very far from delivering a complete map of this territory, and the partial map constructed so far doesn’t include most of the phenomena of interest to you — for example mostly these theoreticians aren’t addressing anything like a persistent “self”, though some of their ideas could perhaps be configured to bear on issues related to the self (such as pre-commitment to some general choices).

    My own response to this situation is to (1) follow this literature as far as it seems useful in my own project, and (2) try to work from both ends — the very general but inadequate analytic framework and the relatively concrete and rich but perhaps too limiting materials describing human sociality.

    I guess my concern about your project is that the modeling that you (and I guess Brandom) are doing seems to generate results that are a lot more specific and closely mapped to human sociality than can be built using any formal models I am aware of. My guess is that you are implicitly relying on a sense of what is and is not possible / relevant for human beings to constrain your models. But an implicit background like that, while probably necessary in this sort of project, is also dangerous and needs to be constantly questioned and re-explored.

    More later.

  6. duncan Says:

    Thanks for this. With apologies I’ll reply to your first comment before your second one arrives, just because I’m worried I’ll be swept away by other obligations if I wait. Not meaning to pre-empt, though.

    I’m glad we agree about built in dispositions – and I like this comment on naturalisation a lot:

    Something similar has been a problem in developmental biology, where people tend to interpret genes as “blueprints” rather than as developmental dispositions that can play out in a wide variety of complex ways, with the same genes perhaps generating very different phenotypes.

    This strikes me as exactly right, and I’m sufficiently ignorant in this area that I wouldn’t be able to formulate points like this except haltingly – thank you.

    On the two points of divergence you mention, in order…

    as I understand your project, your priority is to find the most general analytic characterization of normative or meaning producing systems, even if (as I would assert) most of the space of systems thus described couldn’t be occupied by human social formations.

    This is Brandom’s project, and I’ve been inhabiting Brandom’s system so much on the blog recently that it’s reasonable to have concluded that it’s my project too – but I’m actually quite critical of this project as stated. Brandom frames Making It Explicit by asking, essentially, what features of an entity allow us legitimately to characterise it as sapient? He directly discusses the question of correctly identifying extra-terrestrial intelligence, and he is also interested in AI. Brandom appears to believe that his system gives the necessary and sufficient features of sapience – if and only if an entity exhibits the practices described in Making It Explicit can it be called sapient, in Brandom’s opinion.

    For myself, I have no interest in this question. In fact, more strongly, I don’t think it’s a question that can be conclusively answered (although we can formulate hypotheses about it that would be open to empirical refutation). This is one of the points of disagreement between myself and Pete Wolfendale, who (unlike me) endorses this aspect of Brandom’s project, and believes that this question can be answered ‘philosophically’. I think that the question of what behavioural features are necessary and sufficient for sapience is principally an empirical question (although there’s an analytic side to it too, since we can define sapience as coterminous with certain behaviours – but if we do so we’re not thereby discovering anything substantive about sapience). I therefore regard this question as essentially outside my purview, since I’m not doing empirical research in this area – and I also regard it as something that it would be very difficult for any empirical research to answer, since I think we humans are generally quite bad at anticipating unexpected structures, and have a tendency to mistake imaginative limits for impossibility. I think this is an area where epistemic humility is very much appropriate. We can know quite a lot about the forms of sapience we regularly experience, but in my opinion it’s foolhardy to generalise from these to features of sapience as such.

    So I see where the impression that I’m interested in the general characterisation of normative systems has come from, but as it happens it’s not a motivating interest. What I’m really interested in is, quite literally, human socio-economic structures. It’s extremely eccentric of me to have backtracked to this level of philosophical generality in order to get a handle on that topic (the fact that my academic training is in analytic philosophy no doubt has something to do with it) but my own conclusions ultimately aspire to be pitched at a much lower level of abstraction than Brandom’s.

    That said, I’ve also been pushing back against Brandom in a direction of still greater generality than his system proposes, which I realise no doubt also contributes to the impression that I’m interested in general features of sapience. That’s because I’m vexed by what I regard as Brandom’s unwarranted linguistic exceptionalism. Brandom thinks that language is a necessary condition of sapience. I think he’s almost certainly wrong about that, and I’ve been trying (a little half-heartedly, since as I say this isn’t my own area of research interest) to offer counterexamples and counterarguments. Speaking more incautiously than I tend to on the blog, I’d be astonished if, for example, elephants and gorillas aren’t sapient. This is, in the first place, a gut-level thing – it strikes me as obvious that these animals are sapient, just watching them, and I don’t understand why it doesn’t strike everyone else as obvious. I can provide more careful and properly sourced arguments in this area as well – although again, it’s not my own area of research-interest – but I’m at root just sort of offended by the human-exceptionalism and linguistic-exceptionalism that are common in philosophical spaces, and that Brandom’s work also exhibits. You may disagree with this, in which case no worries, it’s not (I don’t think) of huge relevance to most other topics under discussion. But I’ve been pushing back against Brandom in this area out of vexation, and this probably looks as if I’m interested in maximising the generality of my account.

    So in other words I think I’m probably much closer than it may seem to the project you describe here:

    My priority is to move toward a good theory or language to describe these sorts of (human) dynamically produced and maintained structures. Such a language would of course have to systematically avoid preformationism, but it would also have to help direct our attention to sufficiently useful, general, perhaps even necessary features of the substrate.

    With the proviso that I’m probably ultimately interested in a narrower field of human practice than even this suggests.

    On your second point – I know almost nothing of the game-theoretic literature; I’ll obviously have to fix that at some point, but it means, I’m afraid, that I don’t have much to say on this set of issues. But I think you’re right that both Brandom and myself are implicitly relying on a set of ideas about what is and is not possible / relevant for human beings. I don’t myself find that intrinsically problematic – I’m happy for the theory of practice I end up with to be restricted to humans, so long as this theory of practice isn’t falsely presented as applying to the practices of all sapient creatures. One of the concerns here is the practical consequences of a theory either assuming that certain behaviours are impossible when in fact they’re perfectly tenable; or assuming that certain behaviours are possible – or easily institutionalisable – when in fact they’re very hard or impossible for humans to adopt. If these are our concerns (and this may very well not be where you’re coming from, in which case apologies) my feeling is that this is where historical and sociological/ethnographic work comes into play. We have a limited but still extensive sample of the range of possible human behaviours – this gives us material to draw on in our analytic descriptions of possible human behaviour (it’s surprisingly common for theorists to claim the impossibility of a documented human behaviour), and in our proposals of desirable alternative institutions. If some social group somewhere has systematically reproduced a specific behaviour or social form, this is good evidence that such a behaviour or social form could be systematically reproduced elsewhere, and again. Such “mining” of the historical and ethnographic record, for the seeds of possible alternative social structures, and the proposal of institutional forms that could enable the reproduction of these desirable behaviours, is, imo, the most pressing task of the politically engaged intellectual. This is the relevance, such as it is, of the kind of work I’m doing now to the issues we discussed back in September of last year – all this Brandomian stuff is intended to provide a meta-theoretical framework that permits an adequate practice-theoretic account of (among other things, but centrally) what’s going on the the reproduction of social forms.

    But I’m afraid I’m veering drastically away from anything we were actually discussing. So I’ll call a halt for now.

  7. Jed Harris Says:

    I’m enjoying this discussion a lot, but it is hard to keep up! No need to apologize for anything.

    It is great that we’re in such good sync and I’m happy my comments are useful to you. It would be fun to follow up some of those threads but I’ll restrain myself so I have energy for the rest.

    Thanks for clarifying your position relative to Brandom. I did pick up on your earlier rejection of Brandom’s linguistic exceptionalism and agree wholeheartedly that it is not needed for norms, a sense of self, etc. I’m not sure I have an opinion regarding what “sapience” is or need to our purposes.

    You are right that beyond that I did identify your interests more or less with Brandom’s and am happy that our interests are actually closer than I would have guessed.

    It seems to me we are also closer than perhaps you think. When I say I want to work from (among other things) “the relatively concrete and rich but perhaps too limiting materials describing human sociality” I mean almost exactly that I want to “[mine] the historical and ethnographic record, for the seeds of possible alternative social structures, and the proposal of institutional forms that could enable the reproduction of these desirable behaviours”.

    In this context perhaps our methods are somewhat divergent but I think more due to history and skills than goals or values.

    The problem, of course, with anything as complex as our records of human sociality is that we have to filter them down to relevant aspects and summaries that are small and simple enough to think with. And that filtering is always conditioned by our practices of construal, the language we have to characterize aspects and to summarize patterns, and so forth.

    Sometimes we are lucky and just happen to have practices and language that fit our needs. But we both have found that our received practices and language aren’t adequate to this project, and we both feel a strong need to respect the material rather than force it into a framework that doesn’t seem to fit very well.

    So we have found we need to develop a better / more suitable conceptual framework, within which we can evaluate and revise our practices for working with descriptions of sociality. I guess you are working mostly with meta-theoretical perspectives derived from Brandom and more broadly from the analytic tradition. I’m working with a variety of perspectives, including population biology, game theory, of course analytic philosophy (why I’m here!). I’m not a practitioner in any of these fields; I extract what seem to be relevant ideas, try to figure out how to integrate them together across disciplinary boundaries, and also try to figure out what the underlying problems are when they clash.

  8. Jed Harris Says:

    Apologies for screwing up the link on the reference to Wilkinson’s piece. The original post by Knobe (one of the researchers) is here.

    I just looked for an actual paper, but I guess it isn’t available yet. However Joshua Knobe’s home page links to interesting material on philosophy, conventional and experimental (he’s on the philosophy faculty at Yale).

    Your questions about the interpretation of the results are appropriate and maybe Knobe and/ or his coworkers have considered those — certainly they are valid methodological issues. If you find Knobe’s work interesting enough you might want to ask him. I suspect the answer would throw light on other issues as well.

    I like your point about Wilkinson’s implicit dependence on some immutable homo economicus in each person. He is an economist so I’m sure has some pretty deep commitments to that as a construct. However I think this quote and others like it also illustrate a difficulty with the project we’re pursuing here — certainly not a fatal one, but a fairly persistent one.

    On the one hand we’d like to open up a conceptual space for as wide a range of social practices regarding personal identity as humanly possible. On the other hand we have bodies, life histories, various ways of anchoring claims, expectations and sanctions in attributions of identity, etc. and we need language that lets us talk coherently about how all these elements fit together, and what the consequences are of different ways of fitting them together.

    Unfortunately I don’t have a satisfying way of talking about the aggregate of dynamic, typically somewhat inconsistent, partially imputed, partially internalized, partially biologically grounded patterns that are interwoven in our day to day references to “self”, and the practices that depend on those references.

    So interpreting Wilkinson charitably, he might be groping for such language from his own perspective, and might agree that he is illegitimately postulating an innate homo economicus. But also he might challenge us, asking “How should I talk about the interplay between social context and individual motivations — especially when there’s a conflict between short term and long term consequences?” I don’t have an answer that I find adequate in general, and in some cases I have to admit that homo economicus is a useful way to talk about some aspects of the causal structure that stabilizes the “self”.

    There are bigger points here but right now they are hovering out of reach, so I’ll stop for the time being.

  9. Jed Harris Says:

    Trying to grab one of those points hovering out there — I think one that is sufficiently central that it may drag many of the others along with it:

    Acemoglu and Jackson have a paper on the social emergence of norms (!) that is pretty much in the evolutionary game theoretic tradition, which more or less corresponds to Wilkinson’s perspective. An page with the abstract and an ungated PDF is here.

    The paper is worth reading if only to get a fairly comprehensive view of the relevant current views in evolutionary game theory. The introduction requires no math and very little background in game theory. The authors provide lots of intuitive guidance in the rest and most of the math is fairly easy. The hard stuff can probably be skipped with little loss.

    The paper is asking and to some extent answering questions that are, I think, very useful in the project being pursued here. However that isn’t the primary reason I mention it.

    The paper catalyzed this comment because I realized that, given what you have said of Brandom, he has to be reasoning according to some schema along these lines — and therefore, I guess, you must also depend on something like this. But it seems to be implicit (at least in what you have said or alluded to) and that is limiting and potentially dangerous.

    To be a bit more explicit: let us assume a broadly naturalistic context, in which higher level patterns (such as norms) have to arise as stable causal patterns from lower level patterns (such as reproducing interacting populations). If norms arise through sanctions, then beings who create / learn / are imbued with those norms must be in some kind of persistent relationship with those who could sanction them, and whom they could sanction. They have to be able to learn patterns in that relationship, they have to be motivated to stay in the relationship, and (by hypothesis) they have to be motivated to adapt to the sanctions by avoiding the sanctioned behavior. There are probably other background requirements, but these are enough.

    I don’t see how at least these assumptions can be avoided in any account of the emergence of sanctions (except simply by being vague). But what Acemoglu and Jackson in their paper, and for that matter Wilkinson in the lines you quote, are trying to do is very little more than think clearly about what these assumptions imply — to be sure, in ways that are limited by their familiar perspectives, but that is true of all of us.

    To bring this down to a relatively specific point about the self: I don’t see how any account of the emergence of sanctions can avoid positing that each member of the population has a “persistent self” with some dispositions — the dispositions to maintain relationships with conspecifics, to react to certain behavior aversively, but not so aversively that they leave the relationship(s), to sanction behavior among conspecifics that they don’t like, to learn certain sorts of patterns in the behavior of conspecifics, etc. To a large extent these are the sort of things we talk about when we use self-talk. Fred likes, believes, did, desires, etc.

    So it seems to me that we don’t know a way to avoid some sort of talk about the “self” that’s more or less equivalent to the talk Wilkinson was using.

    We may or may not agree on all this, and rather than elaborating I’ll just see if / how you respond. However I do want to make one point that ties back to our earlier comments. I am not (I hope obviously) advocating this framework instead of attention to the rich and surprising details of human sociality. That would lead to a narrowing of vision and sad neglect of many of the most interesting issues that need to be understood. But conversely, to get what I want out of working with these rich materials I also need to work on the theoretical analysis of their causal structure, the emergence of higher level patterns, etc. Given how much effort you’ve devoted to Brandom, and other frameworks before that, I guess you agree.

    I’d be very appreciative if you can help identify where / how we disagree (if we do) because I think it would help me see the limitations of my own perspective.

  10. Jed Harris Says:

    Just occurred to me: The social dispositions necessary for norms are probably fairly specific to the evolutionary context of the primate band. Happily, a lot of work has been done on primate bands — deep evolutionary history, ethology, population biology studies of existing bands, theoretical studies of the dynamics of evolution in that context, etc. Unfortunately I don’t know this body of work very well, but it is relatively accessible and I should be able to plunder it fairly easily.

    In a quick scan, I didn’t see work that directly addresses the questions we’re discussing here. But that may be because I don’t know the right terms of art to use.

    I think this could form a useful intermediate level, much richer and more concrete than game theoretic models, but much less complex and daunting than the full complexity of human culture.

  11. duncan Says:

    Hi Jed – as I say in my latest post above, I’m putting the blog into hiatus for a time. I don’t want to drop this conversation, though, which I’m finding very valuable. It may take me a little while for me to get round to responding properly – sorry about that – but that’s not from lack of interest.


  12. Jed Harris Says:

    Thanks for the comment, Duncan. Darn, I was enjoying your blog before our conversation, but the exchange made it even better. I’ll miss having your thoughts available.

    You have my email address if you want to continue outside the comments. And I’ll get email notification if you do post.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: