Menary to Macquarie

Congrats to Richard on his taking up a post July 1, 2011 at Macquarie University in the Centre for Cognition and its Disorders and the Department of Philosophy.

Congrats, Richard!

Winner of a UoW Vice-Chancellor's Award. 

Now if we can only persuade him to channel all that work for the good guys...

Kiverstein & Farina's Example of Mental Imagery

This is nicely illustrated by Chambers and Reisberg’s (1985) finding that when individuals are shown a picture of an ambiguous image for 5 seconds, a period to short for them to discover the alternative interpretation, and then asked to find the alternative interpretation through recall, subjects fail to find the alternative image.  However when asked to draw the ambiguous figure from memory they can discover a second image in what they’ve drawn. People find it impossible to find different interpretations in their mental imagery, a difficulty they don’t have when presented with a picture on a page. (Kiverstein & Farina, forthcoming, p. ???)

This is just the kind of evidence that Pylyshyn, for example, uses to argue that mental imagery is not a matter of having little pictures in the head.  It's just the kind of thing that makes A&A not really worry about Menary's appeal to mental imagery to challenge the non-derived content condition on the cognitive.

Cognition versus Behavior According to Menary

In "Cognitive Integration and the Extended Mind," Menary writes,

This second wave of arguments also takes a more enactive approach to cognition, seeing it as constituted by our bodily activities in the world in conjunction with neural processes and vehicles (Menary, 2010, p. 227).

So, if cognition is constituted by bodily activities in the world in conjunction with neural processes and vehicles, then how is this different from behavior?  After all, think about how Skinner describes behavior in the introduction to Verbal Behavior,

MEN ACT upon the world, and change it, and are changed in turn by the consequences of their action. Certain processes, which the human organism shares with other species, alter behavior so that it achieves a safer and more useful interchange with a particular environment. When appropriate behavior has been established, its consequences work through similar processes to keep it in force. If by chance the environment changes, old forms of behavior disappear, while new consequences build new forms.
     Behavior alters the environment through mechanical action, and its properties or dimensions are often related in a simple way to the effects produced. When a man walks toward an object, he usually finds himself closer to it; if he reaches for it, physical contact is likely to follow; and if he grasps and lifts it, or pushes or pulls it, the object frequently changes position in appropriate directions. All this follows from simple geometrical and mechanical principles. (Skinner, 1957, p. 1)

Surely Menary has options here, but which one does he take? If by "cognition" Menary just means behavior, then I'm for that kind of extended cognition.  How could there no be extended hammering behavior?  Hammering behavior has to include arms and a hammer, right?

Fodor on Thinking in English/Pictures

From a book review in which Fodor spends a lot of time challenging argument for the view that we do not think in English, Menary takes the following:

 “I wouldn’t be in the least surprised, for example, if it turned out that some arithmetic thinking is carried out by executing previously memorised algorithms that are defined over public language symbols for numbers (“now carry the ‘2’ and so forth”).” (Fodor 1998, p. 72)

This looks bad for A&A, but the context matters.  So, here is more of Fodor's book review:

I want to end by returning to a point I've already made in passing. I don't think that there are decisive arguments for the theory that all thought is in Mentalese. In fact, I don't think it's even true, in any detail, that all thought is in Mentalese. I wouldn't be in the least surprised, for example, if it turned out that some arithmetic thinking is carried out by executing previously memorized algorithms that are defined over public language symbols for numbers ("Now carry the '2,'" and so forth). It's quite likely that Mentalese co-opts bits of natural language in all sorts of ways; quite likely the story about how it does so will be very complicated indeed by the time that the psychologists get finished telling it.
     But here's a bet that I'm prepared to make: For all our philosophical purposes (e.g., for purposes of understanding what thought content is, and what concept possession is, and so forth), nothing essential is lost if you assume that all thought is in Mentalese. Hilary Putnam once remarked that if you reject the analytic/synthetic distinction, you'll be right about everything except whether there is an analytic/synthetic distinction. Likewise, I imagine, in the present case: If you suppose that all thought is in Mentalese, you'll be right about everything except whether all thought is in Mentalese. More than that is maybe more than it's reasonable to ask for. (Fodor, 1998, p. 72).

Now, this is a kind of puzzling concession on Fodor's part, but he seems to be willing to make it only because he thinks nothing philosophically interesting turns on it (and because the evidence is less than decisive).  And, if I thought that I might agree.  But, something philosophically interesting does seem to me to turn on this, so I don't agree.

So, I want some empirical evidence that we do arithmetic in English or in pictures.  Fodor notes the inadequacy of the taking literally the familiar claim that we think in English.  This seems to me to be all that Menary has.

As for Dehaene, I'm not holding my breath for decisive evidence.  The issue has been around for a while without resolution to the satisfaction of all.

Adams and Aizawa's Stipulation

In the final subsection I argue that cognitive scientists don’t care about Adams and Aizawa’s stipulation. There is empirical research in cognitive neuroscience that does not respect the stipulation, therefore Adams and Aizawa ought to give it up.

Now, A&A have been pretty careful to make explicit that we are not defining cognition in terms of bearing non-derived content.  There was, however, one point in The Bounds of Cognition where we slipped a little.  (See discussion here.)  Rather than stipulating that cognitive processing involve non-derived representations, we hypothesize that cognitive processes involve non-derived representations.  In this, we think we are following the views of cognitivist cognitive psychologists.  So, Menary, it seems to me, fashions this uncharitable interpretation, then bludgeons us with it.

I take it that Dehaene and colleagues are doing empirical research on cognition and yet if we are to take Adams and Aizawa’s underived content condition seriously they cannot be. This seems to me to be decisive. Scientists often ignore the strictures and demarcations laid down a priori by philosophers. Philosophers, such as Adams and Aizawa, may think that they are providing a theory of underived content for cognitive scientists (2001, p. 50), yet cognitive scientists may not be at all interested in these theories because their work does not need them, or they may not be working only with representations with underived content. (Menary, 2010).

But, of course, scientists don't care about the stipulations of philosophers.  They often don't care about anything by philosophers.  But, that's not what we are up to.  The idea is that there is this scientific tradition of cognitivism that postulates mental representations with non-derived content.  This tradition is apparently being challenged by other scientists (and philosophers).  There are scientists on both sides of this issue, so let's cut to the chase and try to figure out which scientific theory is correct

Why can't we take the Ecumenical Turn?

I will now take it that Adams and Aizawa ought to concede that there is no difference in content between my thought that ‘the harbour Bridge looks beautiful in the sunlight this morning’ and my utterance to you that ‘the Harbour Bridge looks beautiful in the sunlight this morning.’ Indeed I might think that very sentence to myself in my head before uttering it to you. If Adams and Aizawa are happy to accept this conclusion, then we really have no disagreement, because we have an example of thinking in natural language (English in this instance). Similarly it seems obvious to me that a Venn diagram that I am imagining now has the same meaning as the Venn diagram that I am drawing on the page now.
     However, this ecumenical turn contradicts some of their earlier claims, where, for example, they “presuppose that thoughts have non-derived content, but that natural language has merely derived content.” (Adams and Aizawa 2008 p. 35) Given that commitment in conjunction with their stipulation, they cannot really help themselves to my example above. That is why I have always found Adams and Aizawa’s position so strange, it leads to clear problems and even contradictions.

I don't think the ecumenical turn contradicts some of our earlier claims; it is one of our earlier claims.  It really looks like Menary's view is that thinking in English is a matter of having a token of English that bears derived content occuring in the brain or a picture with derived content occuring in the brain.

Now, later, Menary writes this,

I suspect that it just turns out that Adams and Aizawa are making a merely trivial claim (which no one I take it denies): a pencil drawing of a Venn diagram on a piece of paper cannot literally be found in the brain—you wouldn’t find pieces of paper in my head if you opened it up and had a look. It doesn’t follow from this claim that the meanings of Venn diagrams aren’t the same whether they are represented as images neurally or as circles on a page.

Yes, we are making the trivial claim that no one denies, namely, that mental imagery is not a matter of little pictures bearing derived content being found in the brain.  Thinking in English is not a matter of sound streams bearing derived content being found in the brain.  So, Menary has to work out the idea of mental image and thinking in English in such a way that there are some other sorts of mental representations that have derived content in the mind.  That's the tricky part.

And, yes, it doesn't follow from this claim that the meanings of Venn diagrams aren’t the same whether they are represented as images neurally or as circles on a page.  But, A&A already said that mental representations and physical representations can have the same content.  That's what we are noting here:

The first thing to observe about the derived/non-derived distinction is that it concerns the conditions in virtue of which an object bears a particular content. A thought might bear the content that the cat is hungry in virtue of satisfying some conditions on non-derived content, whereas a particular inscription on a piece of paper might bear that same content by satisfying some other conditions on derived content. To put the matter another way, there are two questions one might ask of a representation. The first is what content that representation bears; the second is what conditions make it the case that it bears that content.  (Adams & Aizawa, 2010, p. 582).

Menary on the Harbour Bridge

I will now take it that Adams and Aizawa ought to concede that there is no difference in content between my thought that ‘the harbour Bridge looks beautiful in the sunlight this morning’ and my utterance to you that ‘the Harbour Bridge looks beautiful in the sunlight this morning.’ Indeed I might think that very sentence to myself in my head before uttering it to you. If Adams and Aizawa are happy to accept this conclusion, then we really have no disagreement, because we have an example of thinking in natural language (English in this instance). Similarly it seems obvious to me that a Venn diagram that I am imagining now has the same meaning as the Venn diagram that I am drawing on the page now.

Yes, A&A agree that there is no difference in content as outlined in the first sentence here.  We note this here:

The first thing to observe about the derived/non-derived distinction is that it concerns the conditions in virtue of which an object bears a particular content. A thought might bear the content that the cat is hungry in virtue of satisfying some conditions on non-derived content, whereas a particular inscription on a piece of paper might bear that same content by satisfying some other conditions on derived content.

Menary thinks this concession, however, cooks our goose.  He observes that the first thought is an instance of thinking in English and that we have mental imagery.  And, of course, there is thinking in English and there is mental imagery.

But, this apparently depends on a particular interpretation of "thinking in English" and there being mental imagery.  To a first approximation, this only shows that there are instances of derived representations occurring in thought if there is a token of an English sentence in the head or a picture in the head, rather than merely a mentalese representation of an English sentence in the head or a mentalese representation of an image.  It looks as though this is a problem for A&A only if thinking in English and mental imagery are cases of non-derived representations occurring in thought.  But, Menary gives no reason to think this.  He takes it as obvious.  But, of course, there is a vast literature on these issues, a literature that A&A think does not bode well for Menary.

Why doesn't the derived/non-derived distinction matter?

Menary's discussion here is confusing to me.  I am not sure that I understand his objection.  So, the section begins,

In their joint paper for this issue, Adams and Aizawa respond to some of my
concerns about their usage of the intrinsic/extrinsic underived/derived distinctions as applied to representations and content (Menary 2006). Their concern is how something gets its content, the pertinent question is: how does the way that a representation get its content matter to whether or not a representation can count as a cognitive representation? It turns out not to matter much, because Adams and Aizawa really think that there is no difference in kind between underived and derived content. Both underived and derived representations have the same content, they just get those contents determined differently (2010b).

In the last sentence, Menary seems to get our point that there are two questionsone might ask of a representation: 1) What is its content? and 2) In virtue of what conditions does it get that content.  That's what we are on about here:

The first thing to observe about the derived/non-derived distinction is that it concerns the conditions in virtue of which an object bears a particular content. A thought might bear the content that the cat is hungry in virtue of satisfying some conditions on non-derived content, whereas a particular inscription on a piece of paper might bear that same content by satisfying some other conditions on derived content. To put the matter another way, there are two questions one might ask of a representation. The first is what content that representation bears; the second is what conditions make it the case that it bears that content.  (Adams & Aizawa, 2010, p. 582).

Now, what's puzzling in this light is Menary's claim that "Adams and Aizawa really think that there is no difference in kind between underived and derived content."  Well, no.  There is a difference in kind between underived and derived content, namely, the conditions in virtue of what something bears a given content.

Thinking in English and Thinking in Pictures

I will now take it that Adams and Aizawa ought to concede that there is no difference in content between my thought that ‘the harbour Bridge looks beautiful in the sunlight this morning’ and my utterance to you that ‘the Harbour Bridge looks beautiful in the sunlight this morning.’ Indeed I might think that very sentence to myself in my head before uttering it to you. If Adams and Aizawa are happy to accept this conclusion, then we really have no disagreement, because we have an example of thinking in natural language (English in this instance). Similarly it seems obvious to me that a Venn diagram that I am imagining now has the same meaning as the Venn diagram that I am drawing on the page now.

Menary seems to me not to respect the difference between having a thought that P and have an English sentence with the content that P in one's head.  Sure, one can have a thought with the content that the harbour Bridge looks beautiful in the sunlight this morning, and one can even "think this in English", but that's not to admit that there is an English sentence in the head that has the content "that the harbour Bridge looks beautiful in the sunlight this morning".

So, I now think that maybe A&A indulged the Clark and Menary discussion of these cases a bit too much.  Spending a lot of time explaining their view may have obscured our view, which I think is pretty simple. 

1) All thought is in mentalese (a system of mental representation that differs from all natural languages) which gets its semantic content by way of satisfying conditions on non-derived representations. 

2) Thinking in English is forming mentalese representations of English sentences; it is not a matter of having tokens of English sentences in the head.

3) Thinking in images is forming mentalese representations of pictures; it is not a matter of having tokens of pictures in the head.

These are just statements of our view.  There is a huge literature on mental imagery and a large literature on thinking in a natural language, so I would think it would take a lot of work on Menary's part to show that we are wrong on any of 1)-3).  And, he is trying to show, or maybe is just assuming, that we are wrong on 1)-3).

Menary on the A&A appeal to Searle

it is odd to use the demarcation as a principle of defining cognition, since the demarcation was introduced by John Searle (1980) precisely to attack classical cognitivism and to show that the symbolic representations beloved of classical cognitivism, including cognitive psychology and AI, were examples of derived representations.

*

A&A have a somewhat different view of Searle's (1980) argument.  It's conclusion is that nothing gets to have semantic content simply in virtue of running a formal computer program.  So, the naturalized semantics stuff tries to provide an account of how something gets semantic content other than by simply running a formal computer program.  Not a bad fit, it seems to me.

I don't want to go so far as to say that the latter program arose as a response to the Chinese room argument, but there is a confluence between the Searle argument and the naturalized semantics stuff.

Menary (2006) on Integration and Constitution

Controversially, on this account, the process of remembering is constituted by the integration of internal and external vehicles and processes which complement one another in the completion of a cognitive task (Menary, 2006, p. 330).

So, here Menary is invoking a notion of constitution (that he thinks A&A need to explicate for him) and "integration" seems to be another causal coupling kind of notion (only under a new term).  So, as I read things, there are causal and constitutive notions in play here, ripe for the C-C fallacy.

The picture looks to be like this.  There is a task, say, getting to the MoMA.  Then there are these things brainy representations in Otto's brain and these inscriptions in this book, among other things.  It is the integration--the causal coupling--between all these things that enable Otto to get to the MoMA.  So, all those integrated things--all those causally coupled things--constitute the process of remembering.  The fallacy looks to be that simple.

How do you get to HEC from here?

I have repeatedly stated my position that cognitive integration starts from the following positions:

• That we are actively embodied in a socially constructed cognitive niche and
• That phylogenetically and ontogenetically there is good evidence to suppose that we acquire cognitive capacities to create, maintain and manipulate the shared cognitive niche and
• That this has led to the development of hybrid cognitive systems where the bodily manipulation of vehicles (some of them representational) in the niche involves the coordination of neural, bodily and environmental vehicles.
• Cognitive processing sometimes involves these online bodily manipulations of the cognitive niche, but also collaborative thinking and offline private thinking.

Consequently, I propose that I am not committed to the view that cognition is first in the head and then gets extended into tools. Nor does it follow that I am committed to the idea that pencils can think for themselves. (Menary, 2010)

So, as I read these claims, I don't see HEC on the list.  I don't see anything like the claim that cognitive processes are realized by (or supervene on) brain, body, and world.  But, maybe that's just my shortsightedness.  Doesn't Menary believe that cognitive processes are realized by or supervene on brain, body, and world?

Or, maybe there's an argument that links these theses to HEC.  If so, what is it?

Wherein Menary is Right

I have repeatedly stated my position that cognitive integration starts from the following positions:

• That we are actively embodied in a socially constructed cognitive niche and
• That phylogenetically and ontogenetically there is good evidence to suppose that we acquire cognitive capacities to create, maintain and manipulate the shared cognitive niche and
• That this has led to the development of hybrid cognitive systems where the bodily manipulation of vehicles (some of them representational) in the niche involves the coordination of neural, bodily and environmental vehicles.
• Cognitive processing sometimes involves these online bodily manipulations of the cognitive niche, but also collaborative thinking and offline private thinking.

Consequently, I propose that I am not committed to the view that cognition is first in the head and then gets extended into tools. Nor does it follow that I am committed to the idea that pencils can think for themselves. (Menary, 2010)

A&A agree with all of these claims.  We think Menary is right on this stuff.  (Incidentally, this is why we largely ignore Menary's four claims about integration: "Adams and Aizawa, strangely, fail to notice the discussion of four different theses that are supposed to motivate integration: the manipulation thesis, the hybrid mind thesis, the transformation thesis and the cognitive norms thesis." (Menary, 2010).)

But, notice that this list of positions does not contain HEC: "According to this view ... human cognitive processing literally extends into the environment and surrounding organism, and human cognitive states literally comprise--as wholes do their proper parts--elements in that environment" (Rupert, 2004, p. 393).  Menary does not include the claim that "the material vehicles of cognition can be spread out across brain, body and certain aspects of the physical environment itself’"

So, maybe he is not defending EC after all, or at least the version of EC that Clark defends.  (Of course, Menary does claim that "cognitive integration starts from the following positions", so maybe he still does want to defend HEC.  Not absolutely clearly.)

But, why would A&A have thought that Menary defends the kind of EC that Clark does?  His 2006 paper begins with this:

Recently internalists (Adams & Aizawa, 2001, 2006; Rupert, 2004) have mounted a counter-attack on the attempt to redefine the bounds of cognition. Their counterarguments are aimed at the extended mind hypothesis, which, as Andy Clark has recently put it, is the view that ‘‘the material vehicles of cognition can be spread out across brain, body and certain aspects of the physical environment itself’’ (Clark, 2005, p. 1).
     However, I think that the extended mind hypothesis is part of a more radical
project which I call ‘‘cognitive integration,’’ which is the view that internal and
external vehicles and processes are integrated into a whole. It is this more radical
project that Clark and others are really engaged in.  (Menary, 2006, p. 329).

So, insofar as Menary's complaint is that we have not challenged the items in his bulleted lists, he's right.  We have mostly set aside discussion of those things, since we think they are correct.  What we have strived to challenge is HEC (and mostly the arguments typicaly given for HEC), or the idea that cognitive processes are typically realized by processes in brain, body, and world.

Sterelny as Fan of EC?

I have not read Sterelny's paper in Phenomenology and the Cognitive Sciences, but I am skeptical that he is the fan of EC that Menary suggests he is.  Here is Sterelny from a paper back in 2004.

"to depict much of advanced cognition as rooted in the operation of the same basic kinds of capacity used for on-line, adaptive response, but tuned and applied to the special domain of external and/or artificial cognitive aids-the domain, as I shall say, of wideware or cognitive technology."((Clark forthcoming) chapter 8)

To a significant extent we are serial processing inference engines. But we are so only because of our reliable, designed, adjusted coupling with a staggering array of cognitive artefacts, including linguistic and quantitative systems of serial representation. Our extended mind uses and processes linguiform representations even though (quite likely) naked human brains do not.
     Furthermore, some of these artefacts can literally be parts of an agent's cognitive system. In The Extended Mind, Andy Clark and David Chalmers develop a thought-experiment about an Alzheimers sufferer (Mr T, as I shall call him). Mr T cannot unaided remember the location of an exhibit he wants to visit. But he manages such problems by writing down in a notebook crucial information for his daily plans notebook. He then acts by consulting this book. Clark and Chalmers argue that the information in the notebook plays the same functional role for Mr T that an ordinary (non-occurrent) belief plays for ordinary human agents. They conclude we should count the notebook as part of the patient's mind, and the location of the exhibit as one of Mr T's beliefs (Clark and Chambers 1998). Clark is careful not to trivialise this extension of the boundaries of the mind. He insists that agents' minds include only those exter¬nal tools to which they have regular, unfettered access: "the props and aids which can count as part of my mental machinery ... are at the very least, reliably available when needed and used or accessed pretty much as automatically as biological processing and memory" (2001, p139).
     While agreeing with Clark on the fundamental role of epistemic agency in explaining human rationality, I have reservations about this picture. Even when there is a reliable link between user and tool, there are important differences between internal and external cognitive resources. The external storage of information is very important, but the psychological and evolutionary dynamics of mind /filofax relations are critically different from those of mind/memory interactions. So I do not think it is helpful to think of epistemic artefacts as literally parts of the minds of the agents that regularly use them. Moreover and more importantly, Clark underplays the importance of non-exclusive use of epistemic artefacts. Many of our most important cognitive tools are common-use tools, not parts of coupled systems. (Sterelny, 2004, p. 245).

Far from supporting EC, it seems to me that Sterelny comes very close to the Adams and Aizawa view.  Instead, he seems to be supporting something like HEMC.  

The way the Sutton and Menary papers are shaping up, it sounds to me as though they are unhappy that I am challenging, say, Clark's version of EC, but not their version of EC. 

Menary and Sterelny are for HEMC?

Indeed, Sterelny argues that our cognitive capacities are deeply dependent upon the cognitive niche such that the cognitive capacities of the agent are transformed by interactions with the niche. It would be closer to the mark to talk of the niche ‘extending’ into the agent, in the sense that the agent acquires new  representations and new capacities by being ‘coupled’ to or scaffolded by the niche. For Sterelny, there is no cognitive agent without embedding in the cognitive niche.

This is consistent with HEMC:

Cognitive processes depend very heavily, in hitherto unexpected ways, on organismally external props and devices and on the structure of the external environment in which cognition takes place (Rupert, 2004, p. 393).

And, of course, interacting with the environment causes changes in the mind and brain.  That's what learning is all about.  Moreover, we talk about the impacts of environmental interaction in the development of visual capacities in Chapter 9 of The Bounds of Cognition.  It's common ground that agents are transformed by their interactions with their niche.

Menary on the Point of the Ecological Turn

The point of the ecological turn in cognitive science is to show that there are deep consequences for our conception of cognitive agents once we consider their embodiment and embedding in an environment. The whole purpose of Adams and Aizawa’s arguments is to deflate those consequences, but you can’t do so just by assuming that agents are cognitive independently of their environmental embedding.  (Menary, 2010)

This comment I think confuses what is at issue.  A&A think it is fine to consider the embodiment and embeddedness of cognitive agents in environments.  That's the whole point of our concession to Sutton, et al., for example, and our point in the early pages of The Bounds of Cognition, that no one (short of Leibniz) doubts that cognitive agents are causally embodied and embedded in an environment.  This is the deflationary side of our view.

What we resist is not looking at brain-body-world interactions.  What we resist, among other things, is the claim that the same theory of cognition that applies to intracranial cognitive processes also applies to transcranial causal processes.  We resist what I have sometimes called a "fractal scaling" approach to brain-body-world interactions.  This is the idea that we have one and the same processes occurring at different scales of nature, in just the brain alone and in brain-body-world.  This is to resist the more familiar version of Clark's EC according to which what is going on in Inga's brain is roughly functionally equivalent to what is going on in Otto+notebook.

And, yes, there are versions of EC that abandon the fractal scaling idea and say that the intracranial processes do not have to be like the transcranial processes.  But, then what is the basis for saying that both are cognitive?  What is the advantage to cognitive science of doing this?  It would appear not to be some deep theoretical unity.  What is the advantage or theoretical insight in claiming that intracranial processes constitute cognition of one sort and transcranial processes constitute cognition of another sort?  Maybe, instead, it is just loose metaphorical talk.

Note that we do not just assume that agents are cognitive independently of their environmental embedding (where this claim is interpreted synchronically).  We have drawn attention to certain discoveries in psychology that provide a principled basis for saying that what takes place in the brain is plausibly construed as cognitive and is qualitatively distinct from what goes on in brain-body-world systems.  See, for example, section 4.2.1 of The Bounds of Cognition.  See much of the discussion in Chapter 9.  In chapter 9, for example, we discuss how it appears that agents can perceive perfectly well despite being completely immobilized by neuromuscular blockade.  As I've mentioned before, Gangopadhyay has written a very detailed empirically informed critique of Chapter 9.  She sees that there are empirical arguments to be met.

I Think Menary Missed an Argument

In the light of these remarks let’s look a little closer at Adams and Aizawa’s
account of causal-coupling to see where they have gone wrong. “If a cognitive agent causally interacts with some object in the external world in some “important” way—if that agent is coupled to an object—then that agent’s cognitive processing is constituted by processes extending into that object.” (2010b) This is precisely what I was objecting to in 2006. It assumes an already formed cognitive agent with, presumably, internal representations manipulated by computational processes, who just happens to interact with the environment. Adams and Aizawa are here leveraging their argument on a premise that I think is false—that we can consider cognitive agents independently of their environments (apart from the inputs from and outputs to the environment).

Menary thinks that we cannot consider cognitive agents independently of their environment.

But, A&A replied to this.  We think this move does not help.  Here is what we said in The Bounds of Cognition, pp. 102-3:

The suggestion appears to be that we should never think of a lone human being as a discrete cognitive system.  Humans are, so this line goes, always cognitive systems integrated into a network of interacting components.  Humans in their mere biological being are never cognitive systems.  Put more boldly, perhaps, insofar as humans are cognitive beings, they are essentially users of external vehicles.
     Suppose, just for the sake of argument, that it is true that, insofar as humans are cognitive agents, they are never entirely bereft of external vehicles that they manipulate.  That is, suppose that every human cognitive agent always engages some external vehicle or another in her cognitive processing.  Even this concession is not adequate to circumvent the coupling-constitution fallacy.  We can simply reformulate the problem to incorporate Menary’s idea.  So, suppose, simply for the sake of argument, that Otto’s biological mass never in itself suffices to form a cognitive system.  Otto’s cognitive being is always enmeshed in a network of tools.  Still, think of “young Otto” before the onset of Alzheimer’s disease.  Young Otto was embedded in one network of tools.  Presumably this network of tools will not include the notebook that will one day, say, 30 years later, be manufactured in some factory and subsequently purchased by “Old Otto” who has come to suffer from Alzheimer’s Disease.  That is, assume that ones cognition does not extend into currently non-existent tools that one will use in the future.  Now consider “Old Otto” following the onset of Alzheimer’s, but prior to the purchase of the notebook.  Still, the notebook lying on a store shelf never seen by Old Otto is not part of Old Otto’s cognitive apparatus.  How, then, does the notebook become part of Old Otto’s cognitive apparatus on a coupling argument?  One might suspect it begins with Otto’s coming to regularly use the apparatus.  It begins when Old Otto begins to manipulate his notebook.  But, it is right here that the coupling-constitution fallacy is committed.  It is committed when one makes the move to include new cognitive processing mechanisms, such as the notebook.  So, even Menary’s strong hypothesis that cognitive agents are never without their cognitive processes extending into tools is not enough to avoid the coupling-constitution fallacy.

Why isn't this a correct statement of Menary's view and also an adequate reply?

Now, later Menary comments, "I am not committed to the view that cognition is first
in the head and then gets extended into tools."  Right.  That's why we wrote,

"The suggestion appears to be that we should never think of a lone human being as a discrete cognitive system.  Humans are, so this line goes, always cognitive systems integrated into a network of interacting components.  Humans in their mere biological being are never cognitive systems.  Put more boldly, perhaps, insofar as humans are cognitive beings, they are essentially users of external vehicles."

Here is how our argument works though.  At t0 Otto is committed to one set of tools, hence not bereft of tools.  But, then at t1 he acquires a new tool, hence is still not bereft of tools.  At t1, Otto is now coupled to something else, but that new coupling does not make the processes that take place in that new tool cognitive processes.  So, we tweaked the C-C fallacy point to deal with change of tools used.  We are trying to give Menary this point, but show that it is of no help to him.  I don't see why this is otiose.

I think there are a couple of slips here ...

I take the following to be the commitments of those working in the extended mind framework: cognitive processes are causal processes (cognitive processes cause certain effects, presumably cognitive states and behaviour), the extended mind holds that certain kinds of coupled processes (processes that span brain, body and local environment) are cognitive processes. So it is certainly true that coupled processes are taken to be cognitive processes but this is a statement of identity.  (Menary, 2010).

I think there are a couple of slips here.

First, it's not a statement of identity that coupled processes are taken to be cognitive processes.  Coupled are processes are cognitive processes is probably what Menary means.

Second, if that's what he means, then while it is an identity statement, it is not a true identity statement, even according to EC.  Couple two pendula together. The joint swinging isn't a cognitive process.  Couple evaporation and condensation in a distillation apparatus.  The distillation doesn't make for a cognitive process.  So, it's clearly not true that coupled processes are cognitive processes.

This brings us to the claim that coupled processes of a certain kind are cognitive processes. That is, the "certain kinds" qualifier obviously disappeared from the conclusion.  Now we have something that appears to be correct, since that is a view one would have if one thought that the coupled process of the right sort is one in which a system manipulates non-derived representations in specific sorts of ways.  Menary and others see that this weak identity claims helps them, insofar as it enables them to avoid the objections that have been leveled against specific kinds of coupling.  But, it's not much in the way of a theoretical advance; instead it is a mere weakening.

And, of course, there is a lack of consensus on how to cash out the qualifier "of a certain kind".

Menary on Manipulation and Cognitive Processes Extending

It is very clear, to me at least, that the manipulation thesis does not depend upon
any kind of causal coupling (in Adams and Aizawa’s sense) and does not make any claims about cognitive processes extending from brains into bodies and tools.

And, I agree that the manipulation of structures in the environment does not make any claim about cognitive processes extending from brains into bodies and tools.  But, then what is the link between manipulating environmenal structures and extended cognition?  Or, if you don't like the term "extended", what is the link between manipulating structures in the environment and cognitive processes being realized by those manipulation processes?  I've been assuming that observations regarding the manipulation of environmental structures is somehow supposed to bolster the case for EC.  But, that looks to put Menary on the road to committing a version of the C-C fallacy.