Bence Nanay --- Naturalizing Action Theory

[New Waves in Philosophy of Mind]

Naturalizing Action Theory

Bence Nanay

The aim of this paper is to give a new argument for naturalized action theory. The sketch of the argument is the following: the immediate mental antecedents of actions, that is, the mental states that makes actions actions, are not normally accessible to introspection. But then we have no other option but to turn to the empirical sciences if we want to characterize and analyze them.

Video of talk: http://www.youtube.com/watch?v=a7JojUpS-DU

A PDF of the paper is ready to view and download in the attachment below. 

A direct link to the PDF: http://goo.gl/AQNqW

[Ian Phillips]

Hi Bence

I really enjoyed this, not least aspects of the opening polemic. A couple of initial concerns to get the ball rolling:

Why should we assume that the pragmatic representations (which you argue are inaccessible) are the antecedents which make actions actions, as opposed to mere bodily movements? I didn't immediately see why an opponent couldn't accept what you say about these guiding representations but resist the conclusion that action-constituting representations are inaccessible. Clearly you raise a challenge which needs meeting. So here's one v tentative try: the action-constituting representations come from consciously accessible experience in that these representations provide the action-goal -- to grasp *that* disc, to shot *that* hoop. However, once the goal is set then non-conscious processes play a causal but non-constitutive role in guiding the action. These are inaccessible but they're also not what makes the action an action. (Perhaps they are even sometimes involved in mere bodily movements.)

I suppose a more general concern was why we should think that there actually are any antecedents of actions which ground the difference between actions and mere bodily movements. In your brief remarks about mental actions (and perhaps non-goal directed actions too?) you suggest that such antecedents aren't necessary in those cases. Yet someone might wonder whether the reason you give for this (namely: "one could argue that this would be duplicating mental processes") didn't actually also apply to actions themselves. After all, aren't actions mental events/processes (as well as sometimes also being bodily)? I realize your argument is basically a conditional claim, but I wondered what you would say to an opponent who took your argument to be an argument against a certain kind of reductive approach to action theory as opposed to an argument for naturalism.

Be great to hear your thoughts. As I say, really enjoyed the paper.

Cheers

-- Ian

[Olle Blomberg]

(Note: This is just a re-posting of my post from 4 hours ago, but with the many typos corrected…)

Hi Bence,

I really enjoyed this paper! I agree with its overall gist, but here are a couple of worries concerning details in your argument.

You formulate the main question that you suggest should be approached in a naturalized empirically-informed way like this: 

"What is the difference between performing the action of raising my hand and having the bodily movement of my hand going up (maybe as a result of a neuroscientist manipulating my motor cortex, see Penfield 1975))?" (p. 4)

But wouldn't the bodily movements of Penfield's patients be guided by pragmatic representations? Arguably, these movements do not belong to the category of "reflexes and other mere bodily movements". So maybe the reference to Penfield here is a bit misleading, since it suggests that either a bodily movement is an intentional action, or it is merely a bodily movement---a dichotomy that I take it you want to avoid.

I also wonder what exactly the nature of your 'pragmatic representations' are, and whether you are right to identify them with the 'immediate mental antecedents of actions'. You write:

"Leaving the specifics of these accounts behind, the general insight is that the immediate mental antecedent of action has two distinct components: one that represents the world, or the immediate goal of the action, in a certain way, and one that moves us to act. These two components can come apart but the immediate mental antecedent of actions consists of both (at least in most cases)." (p. 5)

But it seems to me that your pragmatic representations don't represent the "immediate goal of the action", but rather represent aspects of the environment that are action-relevant (affordances, perhaps). The pragmatic representation represents the actual size of the central poker-chip in the Ebbinghaus illusion for example, not the goal of picking up the chip. So the pragmatic representations have a mind-to-world direction of fit (you say they are "correct or incorrect" later on p. 5). But this is not what Brand talks about when he refers to a cognitive component of the mental antecedent of an action--as far as I can tell anyway--but rather, he refers to the representation of a future potential outcome, a goal (world-to-mind direction of fit). I take it that the same could be said about Bach, Searle and Millikan as well. 

All this ties in to Ian's proposal above. Arguably, the mental antecedents that make bodily movements into actions are goal-representations which have a world-to-mind direction of fit. Thus your pragmatic representations are not action-constituting representations. (Perhaps this is only true of intentional actions though, and not actions in general…?)

Looking forward to responses and further comments on your paper!

Best

Olle

[Emanuele Podio]

I found the topic of the paper incredibly interesting and productive for the discussion. 

More and more are now proposing an naturalist approach to philosophy of action and I do believe this is the way.
I have a couple of points to make:
1. you refer in the text to "empirical evidence": do you think there is one discipline or method (behavioural neuroscience, brainimaging, single case studies in neuropsychology, etc,) that is more suitable than another for the philosophical debate? 
2. do you think do we still need the concept of representation in philosophy of action and perception? you briefly cited the visual dichotomy and the issues within this framework (interaciton between the streams, etc.). It seems to me that recent developments are pushing for a terminological revision that could involve the same idea of "pragmatic representation". What do you think about that?

Manu

[Declan Smithies]

Hi Bence, thanks for the interesting paper. What do you think about the phenomenon of anarchic hand syndrome? On the one hand, this doesn't look like a case of mere bodily movement, since there are certainly immediate mental antecedents - unconscious representational and motivational states - that explain the complex properties of the behavior. On the other hand, it doesn't look like a case of agency - after all, many of the distinctive features of agency are missing, including a feeling of doing, first-person knowledge of what one's doing, a sense that one can be held responsible for what one's doing, and so on. If that's right, then we can't answer the fundamental question of action theory - what makes something an action? - just by appealing to unconscious representational and motivational states that are the immediate mental antecedents of behavior. It seems to me that some kind of consciousness is needed for genuine action, e.g. a feeling of doing, that might contribute to an explanation of the agent's knowledge of what she's doing and our sense that she can be held responsible for what she's doing.

Perhaps you'd disagree with that, e.g. you might say that anarchic hand syndrome is a kind of agency, although it's not a case of autonomous agency (or something like that). I'm sure there's lots more to be said about this issue. But doesn't it seem like we need to sort out these kinds of conceptual issues about the nature of agency before we hand over to the cognitive neuroscientists? I've no doubt that cognitive neuroscience has all sorts of important implications for action theory, but my worry would be that we won't know what we're looking for, or how to understand the significance of various scientific findings, without a fairly sophisticated conceptual framework for understanding action in the first place. Is this congenial to the kind of naturalization project that you have in mind?

[Bence]

Thanks a lot Declan – I was secretly hoping someone would bring up AHS. So I think these cases where the representational component of the immediate mental antecedent of action is present, but the ‘moves us to act’ component of the immediate mental antecedent of action is missing are really interesting. Besides AHS, there are other examples – like Darwin’s where in the zoo he firmly decides not to jump back when the snake strikes, but when it does strike, he of course jumps back and one I have been using about someone on the other side of a thick plexyglass wall throws a ball at me and I react (I don’t claim that these examples are all the same – there are also important differences). I came up with the fascinating label of ‘semi-actions’ for bodily movements like this.

 

The short story is that I consider the movements of the AHS patient semi-actions. So in the case of AHS patient, we have pragmatic representations – they perform incredibly nuanced and sophisticated actions that can’t be explained without positing representations that guide these actions. But whatever moves them to act is not something they can control and there’s no sense of agency. Are they bona fide actions? I’m not sure – and I’m not sure it’s really worth arguing about. Remember, what is supposed to ‘make actions actions’ is the immediate mental antecedent. And pragmatic representations are one of the two components of this.

 

If we make this distinction between the representational and the ‘moving us to act’ components of the immediate mental antecedent of action, we get a fourfold distinction (both, this one only, that one only, neither). If we have neither, we have mere bodily movements. If we have both, we have fully fledged actions. If we have only one, we are at the grey zone of action attribution. As we have some empirical evidence that there is in fact a grey zone when we categorize bodily movements as actions and non-actions, and the examples I mentioned belong to this grey zone, I take this to be a desirable feature of an account of what makes actions actions. I have a paper on semi-actions (and a section in my book) with a lot of discussion of the AHS case – I’d be happy to send it to you if you’re interested.

 

The other point about the naturalizing project: I don’t think we disagree. I certainly do not want to hand everything over to the cognitive neuroscientists. I do think that naturalized philosophy is still philosophy. And if we talk about the two-way interaction between philosophy and the sciences, there needs to be something different from the sciences that can interact with them… So yes, you’re right – there is need for genuine philosophizing – my point is merely that this philosophizing needs to be informed by and needs to be able to (at least potentially) inform empirical research. 

[Bence]

Thanks a lot for this, Emanuele.

Re. 1: I’d take any empirical evidence there is that would help us philosophers. For the more specific questions of this paper, I think the cognitive neuroscience of action is the most relevant.

Re. 2: I myself think we have very strong reasons (both empirical and conceptual) to talk about representations when we talk about the mind. I am very much against the recent anti-representationalists. 

But one very big picture way of locating where I’m going with this pragmatic representation framework is this. The classic view about the mind is a version of computationalism (or maybe propositionalism, under some understanding of ‘propositional’): what mediates between perception and action are these computational/propositional states that connect up with each other. The more recent anti-representationalists push the idea that nothing mediates between perception and action – it’s all just a dynamic interaction, etc. My approach is supposed to carve out an intermediary position – we do have representations mediating between perception and action – the computationalist is right about this.  But they are very different from beliefs and desires: they are very much action-oriented – the anti-representationalist is right about that. So the hope is that both those with computationalist leanings who are worried about overintellectualizing and those with anti-representational leanings who want more than just metaphors (ouch...) can hop on board.

[Bence]

Thanks a lot Olle – very helpful.

First, you’re completely right about Penfield – it is a confusing contrast in so many ways. I should get rid of that.

Second, again, you’re completely right – the ‘representation of goals’ somehow remained there from a much earlier version. I do have a longish spiel how my view is different from the (very respectable) view that what makes actions actions is the representation of goals in a longer version of this (by the way, I think Brand oscillates between these two proposals, as does Jeannerod).

 

The short story is that while the representation of goals is not necessary for action, the representation of action-properties (the representation of the objects as having the weight that would allow me to lift it with a certain force, etc) is necessary. Now, I don’t want to deny that very often we do have these representations of goals (besides introspection, we have a lot of empirical evidence that we do). But it is not a necessary feature. I also think that what may sound like a harmlessly simple representation of goals is in fact not so harmless – it involves the representation of a counterfactual situation – I see the window closed and I represent it opened (by me). So if we took this route, it would make it very difficult to talk about the action of animals that are incapable of representing counterfactual situations. I consider this to be a major drawback.

[Bence]

Thanks a lot, Philip, very spot on as always.

 

About the first point, the worry was that there is another mental state kind, M, that ‘makes actions actions’: pragmatic representations are not constitutive of action.

 

First, a clarification: my claim is what makes actions actions is the immediate mental antecedent. And pragmatic representations are only one of the two components of this immediate mental antecedent. So there are two versions of your objection:

1. The ‘moving to act’ component of the immediate mental antecedent of action is the one that is doing the real work – that is what makes actions actions – the representational component (ie, pragmatic representation) doesn’t.
2. What is constitutive of actions is something that has nothing to do with the immediate mental antecedent of action as I described it – maybe it is the representation of the goal of the action.

 

My response in both cases is twofold. This other mental state ‘M’ is either not necessary for action or it is itself also normally unconscious (or sometimes both).

So, is the ‘moving to act’ component of the immediate mental antecedent of action necessary for action? I would say it’s not – see the examples in my response to Declan’s question. Importantly, and this is something I should elaborate on in this paper, I don’t this the ‘moving to act’ component is normally conscious. Take the following example (borrowed from Robert Musil’s The Man without Qualities): I am lying in bed, really tired and I know I have to get up. I make a decision to get up, I form an intention to get up, I represent everything around me in such a way that would facilitate my getting up, but I just go on lying there. And all of a sudden, I’m up. Whatever pushed me to in fact move is not something that is available to introspection.

How about the representation of goals? As I say in response to Olle’s question, I doubt that it is necessary for action. And we also have reason to doubt that this representation would be necessarily conscious. People who argue that it is necessary for action (Butterfill and Siniogaglia, Millikan, Jeannerod in some of his odd moments) all assume that it is not normally conscious.

 

I’ll have to think more about your other (more general) worry – I am a bit busy with my own conference at the moment.

[Olle Blomberg]

Thanks Bence, your responses to the questions and comments have clarified your position a lot to me. 

Do you think there are good reasons to think that pragmatic representations are less accessible to conscious awareness than goal-representations? (Seems plausible to me.) If so, then I think this would clarify the dialectic of your argument. As a reader one might wonder: why does Bence appeal to the unorthodox view that only pragmatic representations are necessary for action, if he could reach the same conclusion--that action theory must be naturalized--with appeal to the more widely accepted view that goal-representations are necessary for action too? If you could argue that the content of pragmatic representations are *in principle* inaccessible to conscious awareness (a very strong version of Milner & Goodale's two visual systems hypothesis), that would of course be great for your argument.

I'm inclined to think that goal-representations are necessary for action, but I agree, of course, that such representations are often not conscious. A big question: What's at stake here? Why isn't this just a terminological question about what we should call 'action'? We all agree--I think--that there are interesting differences between actions that are driven by goal-representations and actions (or bodily movements) that are not, between autonomous actions and anarchic hand-type actions that are disowned by the agent herself, etc. Once we divorce action theory from ethics and questions about responsibility, then this becomes a lot less clear to me. I'm not denying that it is valuable to make distinctions, but perhaps the term 'action' itself is not that important…

Olle

[Bence]

Hi again, Ian. So, about your second worry:

This may be just a terminological disagreement – correct me if I’m wrong. I’m inheriting the term ‘immediate mental antecedent of action’ from Brand, but maybe it is misleading here. Maybe I should say the ‘immediate mental antecedent of the bodily movement’. So I’m very open to a view where the mental state I’m interested in is a proper part of action. So action would be bodily movement plus this ‘immediate mental antecedent’. If you go this way, I certainly don’t want to posit yet another mental state that is the antecedent of this complex of a mental state and the bodily movement – this would indeed amount to ‘duplicating mental processes’.

Does this clarify things? Thanks again!

[Bence]

Olle, you are absolutely right that for the purpose of the paper, one could go that way and argue that the representation of goals is normally inaccessible to introspection. The reason why I didn't do this is that (i) I have independent worries about the claim that the representation of goals is necessary for acting (see above) and (ii) I don't think that the argument I gave for the claim that pragmatic representations are normally inaccessible to introspection works for the representation of goals (some other argument might though).

Thanks again!

[Declan Smithies]

Thanks, Bence - I'd be interested to read your paper on semi-actions. From your description here, it wasn't entirely clear to me why anarchic hand syndrome doesn't count as fully fledged action by your lights, since I take it that there are both representational and motivational (i.e. 'moving us to act') states involved in explaining the behavior. Maybe you think the motivational component has to satisfy certain further conditions, e.g. phenomenological conditions, in order to count as a fully fledged action? Anyway, I'd be interested to read your paper on this if it's available.

[Bence]

Declan, sure, I'll send it to you in a pm.