Self-Consciousness, Interaction, and Understanding Others

Presenter: Katja Crone, University of Mannheim

Commentator 1: Joel Krueger, University of Durham

Commentator 2: Kristina Musholt, London School of Economics


  1. Kristina, many thanks for your great commentary on my paper! You point to some important issues, and I am very glad to have the opportunity to briefly respond to them and hopefully to become more clear about the argument at issue.

    I will comment on your concern relating to the concept of interaction that you think to be too demanding. Interaction, understood as a personal stance, which in turn requires persons to acknowledge each other as self-conscious and free agents, seems to require strong conceptual and cognitive abilities, and IT-approaches seem to take interaction/social cognition to go without, at least very often.
    Fichte’s theory spells out what it means to relate to another person as opposed to other things. This relation is further described as a mutual acknowledgement as self-conscious and free agents. What I think is relevant here is the distinction of different but equally relevant components or layers of instances of social cognition. However, I am not sure whether relating to the conceptual – nonconceptual distinction is really helpful here. As said in the paper, I think it would be highly misleading to think that each and every time one deals with other persons (either from remote or through direct contact) one has to consciously perform what Fichte calls “Aufforderung” (demand) and “Anerkennung” (recognition). What we get from Fichte’s argument is above all a structural description of a specific attitude, that is, of a disposition to act or behave – in a given situation – in a certain way rather than in another. The view explains that persons who adopt a personal stance are prepared to treat the thing they face as a free agent. It is especially important to note that this stance functions as a background of concrete performances of social cognition. It is (only) one part of a much bigger picture. My suggestion is to distinguish between the structure of this personal stance and concrete performances of social cognition based on it. It may be useful to reevaluate the conceptual-nonconceptual distinction with respect to these components or layers of social cognition. Do we need certain conceptual abilities in order to be prepared to treat someone as a self-conscious and free agent? As enabling conditions: yes (and this is compatible with findings in developmental psychology according to which rudimentary conceptual abilities of infants develop into fully-fledged abilities). Does it follow from this that conceptual abilities are always required for “directly perceiving” what another person thinks or feels? I am not sure about this. Moreover, I am not sure whether the ability to directly perceive is, according to IT, really supposed to go without conceptual abilities for Gallagher is dedicated to show to what extent the perception of others is “smart”, and I would assume that this refers at least in part to conceptual abilities.
    Overall, I think the structure of the personal stance elucidates in a very basic way the particular I-you relationship that IT takes to be the core of social cognition.

  2. Hi Katja,

    Thank your for your response! I think I agree with you that we should distinguish between the structure of the personal stance and concrete performances of social cognition based on it, as you put it. I would also agree that if we re-evaluate the conceptual/nonconceptual distinction in light of this difference, we will probably end up saying that while we need conceptual abilities as enabling conditions for the taking the personal stance, it is not necessarily the case that we exercise conceptual abilities each and every time we “directly perceive” another person’s mental states. (Btw., am I right in taking you to say here that we do not form explicit/conceptual judgments each time we are engaged in social interaction?) Still, with regard to the types of examples that Gallagher refers to, such as early-infant imitation, detection of sensori-motor contingencies, or shared attention, I would say that these do not even suggest the presence of conceptual abilities understood as enabling conditions. So I would say that these early forms of social interaction are not (yet) instances of taking a personal stance – even though they might be (necessary) precursors for the development of the latter. It might well be that the personal stance is what characterizes the I-you relationship at a more mature level (and that it is a useful way of elucidating IT when it comes to this level), once the relevant conceptual abilities have been acquired, even if this does not imply that these abilities be exercised in the form of explicit judgments in all interactions. I just doubt that very early forms of infant social interaction are best understood in this way. (And you might just want to say that this doesn’t matter too much, as you are not concerned with an examination of the ontogenetic development of intersubjectivity. Though it might still be interesting to say more about the relation between these early forms of social interaction and later, more mature forms of social interaction.)

    I also wonder whether you could say a bit more about the way in which you see your account in relation to De Jaegher’s. In your response you said that you were not sure that direct perception is to go without conceptual abilities, given that Gallagher puts so much emphasis on the “smartness” of perception. Interestingly, De Jaegher seems to criticize his account at precisely this point, because she worries that this pushes IT back into the “cognitivist” camp. This is why she emphasises what she takes to be irreducibly intersubjective forms of interaction, understood in terms of dynamical systems. However, this makes me wonder both to what extent her account is compatible with yours (as it doesn’t seem to imply anything close to the recognition of another as a self-conscious and free agent, at least not when it comes to the examples she cites), and whether you think there is a different way of establishing an irreducibly intersubjective I-you relationship (namely in terms of a mutual recognition of free and self-conscious agents). If so, it might help to spell out in a bit more detail how your account differs from Gallaghers with respect to the problem that De Jaegher raises.

  3. These are interesting questions. Let me first turn to De Jaegher’s example of someone pointing at something for someone else to see. My suspicion is that De Jaegher takes for granted what I (and Fichte) would call the realm of self-conscious subjects and free agents (which I believe is explanatorily relevant). De Jaegher mentions the “partners” who create meaning through coordinate movement of pointing and following the gesture. But how do we account for the fact that a “partner” would never even hope to achieve shared “pointing” in the ‘face’ of a stone, a table or a fly? Partners (in De Jaegher’s sense, as I take it) address each other in a person-specific way when they enter into the process of meaning creation. And the person-specific way to address someone else can thus not be an outcome of collective meaning creation and action – but rather its condition. One has to be at least implicitely aware of the fact that the thing one enters into the process of meaning creation with is a subject capable of certain things. (As you correctly say, no explicit conceptual judgement is needed here). This view may be compatible with De Jaegher’s, it may just add a further aspect. Would you agree?
    I will come back to the issue of early social cognition and differences to Gallagher’s account later, if it is okay.

  4. Hi Katja

    [This is related to Kristina’s worry about infants, but raises slightly different issues, I think] I wonder what you think about the attribution of psychological states to non-humans? In many respects, this seems similar to the human-human case. Yet it’s not so plausible, it seems to me, that we do, or are even prepared to, think of many such animals as free agents.

    An unrelated point (and not really a question, just a mild rant!). I personally think that it’s unfortunate that perceptual accounts of mindreading tend to get associated with interactionist views. The latter are, I think, rather doubtful. Certainly the face-to-face version that you argue against is clearly false (we attribute psychological states to people we see on the TV). However, it’s far from clear why that would shed any doubt on the claim that perception is a source of knowledge of others – since it’s not obvious how interaction is supposed to relate to such perceptual knowledge (at least, not obvious to me anyhow).

  5. I think it’s very plausible that forms of early social cognition like shared attention and social referencing are precursors of social cognition in a more “sophisticated” sense, which entails that one has an implicit awareness of the other as self-conscious and free agent. According to findings in developmental psychychology, young infants not only distinguish between self and world, but also between self and other subjects. Only at the age of around four children start to conceive of others as having their own first-person perspective, from which they may represent parts of the world (possibly) differently from oneself – depending on the information available (or unavailable) to them. They start to understand that others are self-conscious beings that have represenations on their own and that are relevant for their behavior. And this aquired awareness has an impact on how one is able to make sense of others’ behavior. Coming back to De Jaegher’s example: pointing at something for someone else to see surely works – to a certain degree – with infants. But is the pointing at something and following the gesture really “the same” once individual subjects have aquired an understanding of others as self-conscious subjects? This seems to be what some proponents of IT hold. I think it is to some degree a different thing, for parts of the underlying structure has changed. For instance, both partners assume the possibility of misunderstandings due to different perspectives etc.
    I think we can make a similiar case for (some) non-humans (thanks for the reference John, I will definitely take a look at it). Some non-human primates, for instance, apparently do have the ability to represent other subjects having their own perspective, although in a rudimentary way. And their ability to understand others’ mental states seems to be related to it. There seems to be evidence (according to Tomasello, for instance) that they are able to perceive goals and intentions of others. I would think that these are precursors of more complex abilities adult human subjects possess.

  6. Hi Katya

    I meant to raise the issue of our attributions of mentality to animals (not of animals’ attributions). If mindreading requires us to be willing to treat the other as self-conscious free agents, how can we explain the fact that I attribute mental states to my guinea pigs? I don’t know whether they are self-conscious free agents, but I rather doubt it.

  7. Hi Katia, Joel, and John,

    On a related point (to the one of our attribution of mental states to animals): you raise the question as to how we can make sense of the fact that we wouldn’t treat a stone or a table in the same way as we do another human being. Indeed, empirical studies seem to suggest that infants at a very early age show a disposition (perhaps innate) to treat conspecifics (and animate beings in general) very differently from inanimate ones. However, these studies also seem to suggest that what infants are sensitive to in showing these different reactions are things like differences in movement. (In fact when you consider the famous example of the triangles “chasing” each other, you find that even as adults we are inclined to ascribe intentionality on the basis of rather simple features.) However, it is not clear at all that what underwrites these implicit ascriptions of intentionality has much to do with a conception of the other as a free and self-conscious other. Now, the infant case could be seen as providing insights into the precursors of genuine mindreading. (I agree with you that this a plausible way to look at things.) But the fact that even adults ascribe intentionality to animals (or even triangles) suggests that perhaps there are aspects of social cognition and mindreading that are different from the ability to take a personal stance in the sense of treating the other as a free and self-conscious agent (which is not to say that the latter does not also enter into our intersubjective relationships). (Though Helen Stewards claim that animals are free-agents potentially adds an interesting angle to this.)

  8. Hi Joel (this concerns post #7),
    In my view these ascriptions are derivative and require as-if assumptions. The readyness to make such aspcriptions probably also very much depends on the context (closeness to one’s beloved pet).

  9. Hi Katja,

    I like the idea of “distinguishing between the structure of this personal stance and concrete performances of social cognition based on it”. As I understand it, your claim is that in order to avoid pretty decisive counter-examples, those who support an interactionist account of social understanding will need to retreat to the notion of a ‘personal stance’, which has the structure of the concrete performances, without requiring the performances themselves.

    I’m sympathetic to this, but I wonder whether the ‘personal stance’ as you describe it does have the structure of interactionist theories (or indeed the structure of Fichte’s theory as you lay it out).
    IT requires both participants to be involved. Similarly, you describe Fichte as saying that participants have “a mutual acknowledgement as self-conscious and free agents”.

    When it comes to the personal stance you outline, though, it seems that there is something less than this: it’s a matter of treating someone else as a self-conscious and free agent. But, doesn’t this miss out half of the IT theorist’s (and Fichte’s) claim that understanding others requires mutual acknowledgement? That is, there is no mention, within the personal stance as you outlined it, that – by treating someone else as a self-conscious and free agent – you are also thereby considering yourself as potentially such an object of their stance too. To put it another way, I can treat another person as a self-conscious and free agent, without thinking that they have the capacity to do so with regards to me; but doesn’t the structure and the spirit of the initial interactionist proposals require that much too?

  10. A different way of building on Kristina’s proposal in #8 above would be to say that all the relevant phenomena — self-consciousness, free agency, the attribution of mental states, etc. (incl. having concepts?) — come in degrees. There are lots of views in the literature that develop positions along these lines, e.g. Bermudez and various Gibsonians on self-consciousness, Steward on free agency, Susan Hurley on concepts, any number of authors on infant and animal minds, …

  11. John: Indeed that’s the route I choose in the development of my own position (e.g., in the paper of mine that Katja cited). I think there are indeed different degrees (or levels) of self-consciousness and intersubjectivity. I’d be interested to know if Katja thinks that this is compatible with her view. However, I would also want to say that there are minimality conditions for self-consciousnes, intersubjectivity, etc. So on my view not all the phenomena under consideration would count as instances of genuine self-consciousness or mindreading (for instance because they do not imply the ability to distinguish self from other). This depends, of course, in part on how one defines the relevant notions.

  12. Hi Michael,
    Thanks for your comment (#10).
    What I wanted to stress in my sketchy reconstruction of Fichte’s argument is this: the interpersonal stance has, as you correctly say, a specific structure. And this structure has to be distinguished from concrete performances of social cognition. Now adopting this stance basically means to be part of an intersubjective setting in which subjects are principally prepared to treat each other as self-conscious and free agents. So the mutuality you rightly point to is indeed part of this structure. Otherwise, Fichte’s account would be just another example of real-time engagement, which is precisely what I criticise in IT accounts.
    I can see that my account so far is really a bit misleading in this respect.

  13. Hi Katja,

    Thanks for that. Actually, your account was very clear and not at all misleading – I just jumped the gun in responding, and should have read the discussion much more carefully than I did.

    Since I’m here, I’ll ask some questions that (I hope!) are a bit more to the point:

    When you talk about the structure of the interpersonal stance, would you understand this as (i) something that is internal to each participant (i.e. private, for example, an internal cognitive state or state of consciousness), or (ii) would you understand it, as you say above, as ‘part of the setting’ – an environment, say, composed of subjects disposed to treat each other in certain visible ways (i.e. as self-conscious and free agents)?

    With (i), it seems that you would part company with the IT theorist to the extent that the interpersonal stance would be part of the causal conditions required for interaction – it would perhaps share less of a structure with interactionism, and more of a structure with something like the we-mode that John Searle talks about.

    With(ii), it seems that you would be more in line with the IT theorist to the extent that interaction would be a concrete filling-in of the abstract structure composed of the interpersonal stance. But, I suppose, I would have more trouble understanding what such an environment or setting amounted to.

    Perhaps what I’m saying is this: you describe a stance which has certain epistemological implications – by virtue of taking the stance, you come to be aware of other people as conscious beings who treat you in the same manner. But those epistemological implications needn’t get a metaphysical result: I might be mistaken in thinking that the person I’m interacting with is capable of relating to others as free, conscious agents (due, say, to a neurological disorder on their part). Here I might be taking an intersubjective stance, but there is no scope within the environment – between me and that disordered other – for actual intersubjectivity.

    The kind of intersubjectivity that De Jaegher endorses is (I think – I might be wrong), of a more metaphysical kind. It involves the concrete interaction between individuals such that the dynamics of the interaction can’t be reduced to the actions of each individual.

  14. Hi Kristina and John,

    Yes, I completely agree on that. I also think that there are different degrees of self-conscsciousness as well as of intersubjectivity. I think that the existing philosophical accounts (including yours, Kristina) as well as empirical findings are very compelling in this respect. Kristina, as you rightly say, not all phenomena under consideration would count as genuine self-consciousness or mind-reading. I have the impression that the latter (the question of what would count as cases of mind-reading) remains rather unclear in the debates on social cognition, and it would be interesting to try to find common grounds. For instance, I have difficulties to accept that seeing somebody minimally altering the position of his eyeball should count in any meaningful sense as such a case.

    Michael, I will come back to your comment a bit later.

  15. Hi Katja, Kristina, Joel, et al,

    Thank you very much for your paper and the commentaries. I am learning a lot about intersubjectivity and Fichte.

    The personal stance, as I understand it given Katja’s argument, might play an important role in social cognition. Whether that is true and to what extent I hope to set aside for a moment. What interests me is what the specific contribution an attribution of free agency (will?) plays in Fichte’s (and Katja’s!) argument. It seems to me that much if not all of the work that the personal stance is expected to do in social cognition can be done by something less sophisticated, such as the intentional stance, as Daniel Dennett characterizes it. Or, perhaps, there is no difference between the intentional stance? I understand the latter as merely a strategy for explaining and predicting behavior that depends on attributions of beliefs and desires. And surely we can successfully do that even with rocks! I would be happy to elaborate if it is not clear what I have in mind.

    My question is, I guess, mostly for Katja, since she explicitly invokes Fichte’s work in articulating the structure of the interpersonal stance, but I would be grateful for clarification from anyone.



  16. Just a quick thought in response to Michal’s great question (Hi, Michal!): Above I mentioned Helen Steward’s recent work on freedom, and there she suggests a close tie between being a free agent and being a proper subject of the intentional stance. According to Steward, we should think of organisms as free just in case adopting the intentional stance is the *only* way we have of making sense of their behavior. (So rocks wouldn’t qualify, even if we can adopt the stance “successfully” w/r/t them.) Perhaps this position is unworkable, but it does provide a way of bringing together these two ideas.

  17. John, thanks – I will also respond to Michal’s question shortly, and this may fit to what you just said (I really have to get this book!).

    First, I want to reply to Michael (#15):

    It’s surely something along the lines of number 2 that I have in mind. I like your description instances of concrete social cognition “would be a concrete filling-in of the abstract structure composed of the interpersonal stance”. I understand this “environment” as a disposition to engage in concrete social cognition.
    I think your example nicely shows in what sense the abstract structure implies a certain readiness with respect to the fill-ins when encountering a subject. It accounts for implicit expectations concerning the behavior of the other as well as the way oneself is being viewed. And if they are not met – as in the case of a disordered person – one will have to make certain adjustments and include further information beyond the scope of the interpersonal stance.

    As for what you call the “metaphysical implication” (however, I am not sure whether I understood you correctly) I would think that because the interpersonal stance is a necessary condition for (normal) cases of concrete social cognition, the latter cannot – at least not in one relevant sense – be reducible to the cognitive make-up of one individual nor to one-sided action.

  18. Hi John! Glad to see you here!

    Steward’s book sounds great. I have some worries about what you said and specifically about how we could individuate proper subjects of the intentional stance from those that are not so proper without having an independent grip on free agency and, alas, we don’t have that as far as I know. I hope that is not what Fichte’s and Katja’s argument depends on!

    Maybe my question was not so innocent and clarificatory.


  19. Hi Michal,

    Let me try to be more clear about the relation between being a subject of the personal stance and ‘free agency’. The personal stance I talk about differs from Dennett’s notion of an intentional stance in some important respects. The first difference is that the personal stance is by definition based on mutual acknowledgment. Again, this is not to say that each time one encounters a person one has to consciously perform this acknowledgement or renew it. It is the structure of the stance. (Maybe you want to go back to some previous posts in which this was at issue). The second difference is that Fichte’s argument sets out the idea that one’s own awareness of being an individual subject depends on being part of what I have called an interpersonal setting.
    In order to be more specific about the aspect of free agency one has to go back to Fichte’s notion of self-consciousness. As pointed out in my paper, Fichte’s notion is practical at its very roots. So even being conscious in the sense of both creature and state consciousness means being mentally active (formally). Another practically related feature of self-consciousness is the ability to actively direct mental activity to an object or redirect it (figuratively speaking). And crucially, one can use certain results of mental acts (representations) to set goals for one’s actions. This may not entirely answer your question but it hopefully explains the interrelation between being a subject of the personal stance and agency.

  20. Hi Katja,

    This was very helpful as far as drawing a distinction between the personal and intentional stances. Many thanks for taking the time to respond.

    The first difference is being able to engage in mutual acknowledgment, which seems equivalent to two persons taking the intentional stance towards each other and recognizing that they are doing so. Is that correct?

    The second difference is that being aware of oneself as a subject depends on an interpersonal setting, which is, roughly, a setting where subjects are principally prepared to treat each other as self-conscious and free agents. So, we extrapolate from such interpersonal setting to ourselves being subjects. All of this makes sense to me, especially if I take a Sellarsian gloss on it, but there remains a question of how it is that treating someone as self-conscious and a free agent contributes to social cognition in a way that goes beyond treating someone as merely having beliefs and desires. Or so it seems to me.

    I hope you now see the source of my original question. Neither of the differences you mentioned between the personal and intentional stance seems to be doing any work in social cognition.

    Another way to see my point is to consider conditions associated with a diminished capacity for social cognition, such as autism, which are not typically characterized as deficits in one’s ability to treat someone as self-conscious or as a free agent. They are characterized as deficits in one’s ability to extrapolate from behavior to beliefs and desires–whether by inference or simulations or whatnot.



  21. Hi Katya,

    Thanks again for your response – it’s very useful and very interesting. I hope you don’t mind me asking more questions, because I’m really interested in this area, and you’re taking an approach which I hadn’t properly considered before.

    1) I think when I saw the term ‘stance’, my mind immediately leapt to Dennett and his ‘Intentional stance’, and I thought that you were presenting something similar, but perhaps just more full-blooded. But, if I understand you right, the notion of a ‘stance’ you have in mind is more like a practice – or perhaps better, the scaffolding for a practice? – than Dennett’s notion is.

    So, it seems to me that your notion of a stance will involve a permanent, standing readiness to treat others, and to be treated as, a conscious, free being. So, it differs from Dennett (as I understand him) not just in the respect that it adds both a conscious and a reciprocal aspect, but also in the respect that it is standing disposition, rather than a strategy of interpretation which you might exercise from time-to-time in the face of appropriate beings. [I take it that you require it to be a standing disposition, rather than an occasional strategy, to allow for the fact that one might be being watched at any moment, as with the spy example you used].

    2) I think the worry that I had (and perhaps Michal’s worry is similar), is that I understood the ‘personal stance’ as a stance taken by an individual. That would then seem to leave it open to the objections that (i) it is individualistic, rather than genuinely intersubjective; (ii) it is, after all, something along the lines of Dennett’s stance, just more full-blooded; and (iii) that it doesn’t after all have quite the structure of IT, which would seem to require at least two people in order to occur. (i think those three are just the same point, taken from different directions). But, the spirit of your proposal and your responses is that there is some genuine intersubjective, intertwining going on in the personal stance. Which makes me think that it would not be right to think of it (the stance) as something that is taken by an individual, but rather that it is something taken by people in plural; i.e. it is a stance that we humans take. That is, in order to be a member of the world of conscious, free beings, you need to enjoin with the others in taking the stance – the nearest touch-stone I can think of here (in my ignorance of continental philosophy) would be something like McDowell’s ‘Second nature’.

    3) My feeling, following on from this, but maybe at a tangent, is that any intersubjective approach to mentality is going to come up against something like the following problem eventually: a robust notion of intersubjectivity seems to call for a non-reducible intertwining of subjects; but a robust notion of conscious experience seems to require a single subject as the locus of that conscious experience. So, the former allows in a weakened form of mentality (a very bodily form, as in IT, or a nonconceptual form perhpas), but not the richer form. While the latter allows in the richer form of conscious experience, but only a weakened form of intersubjectivity (where the presence of the other might be necessary for the experience in some way, but the other is only an object of the experience rather than the subject).

    I’m not sure if I’ve got your views right or not, but I would be very interested to hear whether I have, and also whether your view might have something to say about the third point above.

  22. Many thanks for your remarks. I’m sorry I couldn’t come back to you any sooner.
    Michael, I think I see what you mean. You are pointing to a difference between a robust notion of intersubjectivity that requires a non-reducible intertwining of subjects, and a robust notion of conscious experience, that seems to be a question of one subject only. I wonder whether there is really a clear-cut line to be drawn between the two forms of awareness related to the two notions (bodily/non-conceptual and conceptual or weaker and stronger)? Furthermore, according to my argument the stance in any form of its manifestation necessarily implies non-reducible intertwining of subjects at its very root. So it way well be the case that the aspect of intersubjectivity remains below conscious experience (I think this plausibly to be the case in the situation of mere readiness to treat someone as a self-conscious and free agent). My claim is, that it is yet functionally related to social cognition.

  23. Michal, thanks very much to you, too. I think there is an important difference between merely ascribing beliefs and desires to someone (as a strategy to interpret someone else’s behavior) and being prepared to treat someone as a self-conscious and free agent. The difference is that in the latter and not in the former case we conceive of the other as having his/her own first-person perspective. This implies to see the other as a being that may or may not use own represenations about the world for actions. The simple having of beliefs and desires lacks, for instance, the capacity to metarepresent, which I take to be necessary for self-determined actions.
    You mention the autism case. I think autistic persons precisely lack the ability to represent others as having their own perspective, the fact that others may represent the world differently from their own perspective. This, I believe, is not reducible to simple belief-desire ascriptions.

Comments are closed.