Transcranialism

I am finally reading Clark's Being There, but couldn't help look at what he is doing right now. He actually has some extremely relevant stuff about "material symbols" here. He describes EM (The Extended Mind Hypothesis) as it applied to symbols:

In the present treatment I explore a third option, which I shall call the ‘‘complementarity’’ view of language. According to this third view the actual symbol structures of a given language add cognitive value by complementing (without being replicated by) the more basic modes of operation and representation endemic to the biological brain. The ‘‘cognitive bonus’’ that language brings is, on this model, not to be cashed out either via the ultimately mysterious notion of ‘‘thinking in a given natural language’’ or via some process of exhaustive translation into another inner code. Instead, we should try to think in terms of a kind of coordination dynamics in which the forms and structures of a language qua material symbol system play a key and irreducible role. Understanding language as a complementary cognitive resource is, I argue, an important part of the much larger project (sometimes glossed in terms of the ‘‘extended mind’’) of understanding human cognition as essentially and multiply hybrid: as involving a complex interplay between internal biological resources and external non-biological resources.

As we know, the main proponents of transcranialism (this is apparently Adams and Aizawa's polemical term for them) are Clark and Chalmers (1998), Merlin Donald (1991), Daniel Dennett (1996) and Ed Hutchins (1995).

Link

New knowledge on genes warrants reinterpretation of Evolutionary Psychology

In a couple of recent articles in Trends In Cognitive Sciences the relationship between genes and psychological traits, so important for evolutionary psychology, is treated. First of all “generalist genes” have been found that are used in building many different modules

Abstract: ”In the ‘generalist genes’ hypothesis, it is suggested that the same genes affect most cognitive abilities and disabilities. This recently proposed hypothesis is based on considerable multivariate genetic research showing that there is substantial genetic overlap between such broad areas of cognition as language, reading, mathematics and general cognitive ability. We assume that the hypothesis is correct and consider here its implications for cognitive neuroscience. In our opinion, the two key genetic concepts of pleiotropy (in which one gene affects many traits) and polygenicity (in which many genes affect a trait) that underlie the generalist genes hypothesis imply a ‘generalist brain’. That is, the genetic input into brain structure and function is general not specific.“ [TICS 10.5. 2006:198-203]

This does not, however preclude a domain specific and nativist theory, since these genes may just be used to manufacture recurrent features common to different domains [TICS 10.9 2006: 397-398].

It also seems that the nature-nurture debate has finally been called of, since as good as all psychological traits have genetic as well as environmental causes. Although single gene disorders exist they seem to be the exception [TICS 10.9 2006: 399-400].

Thus the mapping from one gene to many different traits under the continual interaction with the environment seems to be the emerging consensus. This minimises the role of single genes, on which mutation works. That does not necessarily falsify the central claims of EP, but I do think that it has the following consequences: 1) It is not possible to stipulate very well-formed and detailed modules for very specific purposes, but instead more general (within a specific domain) biases. 2) Much of the natural selection lies not in phylogeny, but in ontogeny and cultural history.

God Exists (a recent conversation)

Have you seen this?
http://www.dr.dk/Nyheder/Indland/2006/12/19/214556.htm

Peter
__

The really revealing thing is that these guys apparently
don't believe in Santa!!! It'll make his work a whole lot easier,
of course....

Jesper

__

And the full version:
http://www.information.dk/InfWebsite/FremvisningPHP/Common/Information.php?p

Don't these results pose a problem to the idea that religious
representations are decoupled? If they were, wouldn't
Santa and God beprocessed in the same way?

/P

__

Dear Peter
I'm not sure I understand why that would be the case. Decoupled cognition only means the ability to represent and to simulate agents, actions etc. not present. Representing Santa as coming down the chimney and representing God as listening are both examples of decoupled cognition, even if their epistemic status is different.

I didn't find the result surprising (but its still nice with some evidence). I mean, what is the alternative to understanding conversations with a god as with another person (I actually do have one suggestion presented below)? My guess is that it is very much like talking to a phone answering machine. You expect someone hears you, but you get no immediate feedback. So, the metarepresentations decides the status of the 'receiver', or said differently there are metarepresentations whose content decides whether there is a 'receiver' at all. (I guess the guys from Indre Mission must have felt pretty strange asking Santa for stuff--I know I would)
My question would be how or whether it differs from the 'internal dialogue', i.e. the conversations have with ourselves? As I understand it, this is pretty much a universal phenomena and might be the template for 'addressing an agent not answering' (unless of course there is an answer in the internal dialogue, but my unscientific intuition is that this is usually not the case and that if this appears, we are approaching split personality).


Does this make any sense?


Jesper

_


Dear Jesper

No, I don’t think it makes sense. The internal dialogue has nothing to do with these experiments as far as I could gather. They have to do with brain activation used when you engage in social interaction. I agree with Peter that this does go against the traditional meta-representational account. I also agree that both situations are decoupled, so it is not the decoupling per se. According to the traditional account the brain activation we would expect is the same, with the addition of a marker of whether the content is held to be true or false. Areas associated with correct or incorrect assumptions exist, and have been used by Kevin Dunbar (as far as I remember). This is apparently no the pattern we see. The part about social interaction is left out in the Santa condition.

Best regards

Anders

__

So Durkheim was right!!!
Lars

__

Don't these results pose a problem to the idea that religiousrepresentations are decoupled? If they were, wouldn't Santa and God beprocessed in the same way?

/P

__

Hey guys

I may be getting this all wrong, but as far as I can see, there are a number of "problems" here.
Decoupled representations are usually said to be "relevantly similar" (cf. Goldman), in that they involve SOME of the same functional areas. As a typical functionalist metaphor the idea of tagging something as either real or pretend has some merit, I guess, but none of us probably buy the idea that for every semantic representation of something, there is a specific place in the brain which has the function of marking it "real" or "imaginary" by setting that bit to 1 or 0 respectively. Reality status evaluation as a cross-domain task MAY be evaluated in specific regions, but I would suspect that multiple, and different, areas would be involved as well: Motor imagery uses motor areas, visual imagery visual areas. Also, imagining something is an act, but there are many different acts in that grab-bag, and they are NOT functionally alike, for an example take pragmatic motor imagery vs. visual imagery. Apart from the fact, that I would have liked to see some non-believers perform the exact same tasks, I also think that even if you focus on Indre mission, you need a bit more:

The question is what the most relevant contrast condition(s) is/are in relation to the prayer performed by a believer-task. Is it whether God is felt as real, or is it whether God is present or nonpresent? Real people can be both present and non-present as actual embodied individuals, but God cannot.
The article does not mention other tasks than a nursery rhyme and adressing Santa (or did I miss something?). The first goes for the linguistic (sematic-syntactic) output similarity only, as it does not have the pragmatic structure of adressing someone or something - so that would be used as a contrast to adressing or not adressing someone with the same kind of output. But in contrast to God cognized and adressed by believers, The Santa task obviously involves talking to someone cognized as "actually fictional" and non-present, not just someone not present. This remark of mine is, of course, begging the question since it is seems to be the difference between two fictional agents that he wants to show, but I think that by focusing on that, he misses the mark. Do we really believe that those tasks are relevantly similar for religious people: Does it show that they treat god as a social agent, but not Santa? To me, the relevant contrast is not between Santa and God, but between real people and God.
As far as I can see, a relevant comparison would be keep the nursery rhyme and the Santa data but in addition ask the same subjects to adress, in their mind, someone, and ask them to do something, and contrast that with actually asking them to do something, both while in the scanner. And preferably another task too: Asking someone to do something that they know can be done easily, and something that demands a lot more effort from the person in question.

And so on. But good to see both Santa and God making the news.


Andreas

__

Dear Jesper
No, I don't think it makes sense. The internal dialogue has nothing to do with these experiments as far as I could gather. They have to do with brain activation used when you engage in social interaction. I agree with Peter that this does go against the traditional meta-representational account. I also agree that both situations are decoupled, so it is not the decoupling per se. According to the traditional account the brain activation we would expect is the same, with the addition of a marker of whether the content is held to be true or false. Areas associated with correct or incorrect assumptions exist, and have been used by Kevin Dunbar (as far as I remember). This is apparently no the pattern we see. The part about social interaction is left out in the Santa condition.


Best regards
Anders

__

No, Tylor was right!
a

__

And I am not even finished here: I just thought of the Danish book "Det er et bånd der taler", where they look at schizophrenics and their statements. I seem to remember that they frequently adress both authority figures such as God, the King and the Chief of Police, so just for fun I'd like to see those believers adress someone they believe are both powerful and real. Maybe the Minister for the Church. Maybe religion rides piggy-back on evolved dispositions to represent and believe in powerful, socially relevant agents? Oh well, lunch...

/Andreas

__

It seems that at least theories are marked as being plausible or implausible. This implies reference to dedicated areas. The article by Kevin Dunbar showing this is called: “Brain-based mechanisms underlying complex causal thinking”: The abstract is:

We use functional magnetic resonance imaging (fMRI) and behavioral analyses to study the neural roots of biases in causal reasoning.
Fourteen participants were given a task requiring them to interpret data relative to plausible and implausible causal theories. Encountering
covariation-based data during the evaluation of a plausible theory as opposed to an implausible theory selectively recruited neural tissue in the
prefrontal and occipital cortices. In addition, the plausibility of a causal theory modulated the recruitment of distinct neural tissue depending
on the extent to which the data were consistent versus inconsistent with the theory provided. Specifically, evaluation of data consistent with
a plausible causal theory recruited neural tissue in the parahippocampal gyrus, whereas evaluating data inconsistent with a plausible theory
recruited neural tissue in the anterior cingulate, left dorsolateral prefrontal cortex, and precuneus. We suggest that these findings provide a
neural instantiation of the mechanisms by which working hypotheses and evidence are integrated in the brain.
© 2004 Elsevier Ltd. All rights reserved.

Link to pdf: http://www.dartmouth.edu/~kndunbar/pubpdfs/FDneuropsych2005.pdf

There does therefore seem to exist something tagging assumptions about the world, if not as “real” or “imaginary”, then at least as probable or improbable,

a

__

My Danish isn't quite good enough to read it, but a small comment anyway. My understanding of decoupling is that it wouldnt necessarily have anything directly to do with SHA at all, but concerns instead propositions. Decoupling is our ability to remove propositions from their 'normal' contexts the classical example for which is the ability to embed propositions within one another, so you can say: He is crazyHe said "He is crazy"and He said that he is crazyand these all embody different semantics. The last one is especially important because the truth conditions change when it is embedded. This is where the term de - coupled comes from, I believe, in the sense that it is no longer attached to literal use. And of course "entertaining" a proposition fits into this category.not just any representation can be true or false -- only a proposition -- Anyway, on another subject, dont religious people think God is answering them all the time? I mean if someone prayed to God and thought he didnt respond at all that wouldnt be very satisfying.and anyway arent our prayers carried up to God by angels and other emissaries? On a further note I was talking about this with my wife a while back and concluded that especially in the states Santa serves as a kind of early encoding of the protestant God. There are so many movies about "just believing" in santa as the essence of Christmas -- it teaches kids about the doctrine of faith alone. thanks for indulging - maybe someone can explain to me what they actually conclude in the study.

Gabriel

__

Andreas point about schizophrenia indicates to me that the idea of prayer being related to internal dialogue is not as far fetched as Anders makes it sound. As prayer in some form seems to be a human universal, we need to find a relevant origin to the ease in which humans engage in dialogues with non-present agents. In internal dialogue we often address a more or less specified other agent, and I therefore thought this might tbe a relevant origin of this funtional capacity (lacking answering machines in the pleistocene). So I would still like to see this contrated with internal dialogue. Anyway, this would entail a more thorough knowledge of the features underlying internal dialoque that I don't have.About the metarepresentational account, I still think this is the best explanation so far. I cannot se how these experiments in any way questions it. As Andreas points out, metarepresentation is a functional description of out ability to give some kind of epistemic 'tag' to representations. This does not entail that there is a special area of the brain, but more likely that epistemic status is an intergral part of the representation itself that will limit the extensions and inferences a concept are likely to produce. So, when rebuilding my chimney, Santa is not relevant (as I do not 'believe' in him--otherwise he would be), but an Indre Mission Christian might have strong inferences about God in relations to many procedures. So far this description does not entail any strong constraints on how this is physically instantiated.This brings me to Uffe's results. The real interesting point, as I see it, is that the Indre Mission guys activates brain areas associated with social cognition, wheres ordinary danish Christians don't. So one might discuss how metarepresentations of the 'reality' of an agent are created by using specific areas of the brain (this, however, does involve a shift of explanatory level that is not unproblematic). I think there are many means. 'Traces in the sand' is one: I believe that the author Tom Wolfe exists because I just read a book written in his name. I would thus not be surprised if someone referred to Tom Wolfe doing this or that, even if I never met the guy and thus really do not know that he exists. Having some kind of indication that an agent exists, e.g. books in his name, seeing a photograph, other people interacting with him etc. tags him as having 'real existence' and this will work as long as there is not overwhelming evidence to the constrary (e.g. children believing in the toothfairy until other indications pointing to tha fact that mum placed the gift). Such representions might make me interact with the guy (long distance) and I might indulge in decoupled practices such as imagining what Tom Wolfe would have said about a particular case. All in all, the 'signs' from our environment that a given agent is real, even though not present, result in possible inferences as to the ability to communicate with him or her.Another way that might be more relvant in this case is through specific practices involving the relevant connention between the activation of a concept (God) and a practice involving social interaction (asking someone for help). The rehearsal of the practice might be necessary to connect a particular concept to a reliable activation of social cognitive processes. It would thus be interesting if new converts who have never prayed before would have stron activation immediately, or whether the thing have to be 'build' through repeated practices

./Jesper

PS i side with Anders. This is more in line with Tylor than with Durkheim....

__

First of all the background to Gabe: The study showed that fundamentalist christians have a different brain activation pattern, when addressing god with the lord's prayer than when they adressed Santa with a similar prayer, and when they recited a nursery rhyme. In the ordinary christian griup there was no difference in activation. The brain activation pattern was the same as when you engage in social relations with other people,I just want to point out a few things:1) I do not have any opinion about whether or not internal dialogue is involved in prayer, I just pointed out that Uffe's study does not tell us anything at all about that hypothesis2) I take Uffe's word that the study does not address de-coupled cognition at all, because he simply knows more about it. 3) Consider these quotes: "This does not entail that there is a special area of the brain, but more likely that epistemic status is an intergral part of the representation itself that will limit the extensions and inferences a concept are likely to produce. (..) So far this description does not entail any strong constraints on how this is physically instantiated." I think this is symptomatic for the use of neurological evidence in Evolutionary Psychology. I do not herby intend to imply that Jesper is an evolutionary psychologist, although he is influenced by their way of arguing. Then I find a study that actually demonstrates such an area in the brain which is associated with correct and incorrect assumptions about the world, and it is completely ignored. Instead an account of how we should not expect any neurological evidence is offered. I have elsewhere argued this (see my article here: http://www.csr-arc.com/view.php?arc=4 ). Either the neurological studies count and are able to falsify theories at the cognitive level or you must give up the claim that you have a materialistic theory. 3b) This calls for some principles for the use of functional neuroimaging, because oftentimes you will also hear neurologists say that you cannot use this or that result because it is much more complex. The question is, do we need neuroscience at all ro should we just ignore it altogether? I might write something more on this shortly

All the best,Anders

__

I think that this is an extremely relevant discussion. Does thesimulating parts of the brain, actually thinkt that thecommunication/decoupled experience is going on, or are the "scopetacs", Theory or whatever seperates simulation/decoupling frokm othercogniton, only added in the conscious/higher order experience?The way I see it, most of the brain just does what it usually does inprimary/naive representation, nonewithstanding whether the'representation' is decoupled or not. Derived activations, such as -for instance - movement, is however inhibited. Semantics, such asUffe's 'orders' can activate any and all systems needed for, say, aconversation with God or Santa, but it woukd appear that only theGod-conversation is activated "properly". The interesting part reallybeing, that only God seems to register as a 'true person', and thusthe whole array of ToM, etc. along with him. Edelman's basic argumentwould be, that inhibition from different sub-cortical nuclei preventsthe "less believed"conversation to reach its full distributedprocessing-potential. Finally, fMRI is recorded through a lot ofnoise, so there might still be slight activation - from semantics, forinstance - which isn't seen too well. I am looking forwards to hearingUffe's own conclusions.But yeah, It would seem that Tylor was right in his basic assumptions.Boyer just did a rebottleing. I seem to recall Jesper giving me an Afor that conclusion some time back ;-).

/Andreas L.

__

Count me in the same group as Gabe here, i.e. those needing some background- I did not know that Schjødt also had a group of ordinary christians inthat study, which makes most of my initial remarks somewhat off mark - toput it mildly. I guess he is also making use of other studies documentingsocial cognition areas in the brain. Just goes to show that one should neverbelieve what one reads in the newspaper, if that is all one reads...Shame onme.Anders: I do not discount "raw" neurological research, but I find myselfmore comfortable with things like articles from TICS, if you know what Imean. That being said, is the importance of neuroscience really a questionof proof? Isn't it more reasonable to say FMRI and other imaging techniques(and academic work in general, for that matter) can make certain hypothesesmore or less plausible? As for the Dunbar reference, the abstract statesthat the primary differences relate to evaluating consistency orinconsistency in complex causal relations. I see no problem in this beingtrue in combination with the fact that social relations may recruit verydifferent areas of the brain. This doesnt rule out something like a generalconsistency evaluation structure; Following DUnbar or similar results,something like a central evaluation structure may (or may not) be involvedin evaluating complexes of differing kinds, which could explain the somewhatsimilar cognitive-emotional dissonance effects we seem to have in severaldomains. I have heard a few references to pre-frontal areas in connection tosuch functions.But, I am not quite sure whether that has any bearing on how people relateto God. Prayers do not involve evaluation of consistency - and the questionof how you believe in God probably doesn't either. I mean, these two groupsbelieve in different ways, and they seem to pray in different ways, whichmay be a product of certain evaluation procedures at some other point intime, but surely they are not constantly evaluating complex causalstructures in their everyday religious practices...or...?My reference to de-coupling was from Goldman(2006), his recent book aboutsimulation and mindreading, which I have only just started leafing through;very good so far. Leslie uses the term a lot in his pretend theory, but Imust admit that I got re-confused when I looked at his 1987 paper - itseemed a lot more "understandable" in Nichols and Stich (2003) and inGoldmans paraphrases. Whether its plausible, I dont know, but it has alwaysseemed just a little too boxy-functionalist and perhaps a little too"Language-of-thought"-ian to me. On the other hand, what I've read ofLeslies recent papers (in e.g. TICS) I found extremely thorough and veryconvincing, but please don't ask me to summarise it...Goldman argues thatthe Leslie papers up thorugh the years show some inconsistencies, so I guessits not obvious how to understand the term.

Andreas

__

Sorry, Anders, I might be stupid, but I still don't get it. I'm not saying that neurostudies cannot be used (neither are most EP, I would guess). I'm merely pointing out that they operate on two distinct explanatory levels and that you cannot go directly from one to the other. I have not talked to Uffe about this, but, no his study does not address the question of decoupled cognition--not was it meant to! This is not something that can be addressed on a brain scanner unless you have a very precise neurocongitive model of the structures underlying it (its probably not an 'it' anyhow). You are right that we need to think about how neuroscietific results can be used in cognitive descriptions. The only thing I said was that the result does neither confirm nor disconfirm notions of metarepresentations. I fail to see how this question can be addressed in a scanner unless you have a precise model.There is a more fundamental matter in your reply, however. If you argue that it is only through neuroimaging etc that we can defend a materialistic theory, we'll never agree. That would throw experimental psychology out the window!! Arguing that information processing is depended upon and constrained by a material medium is not the same as being able to explicate the exact neural mechanisms underlying it--as well as understanding brain mechanisms does not necessitate an explanation on the level of quantum physics, but do believe that atoms underlie it. These are distinct explanatory levels. Claiming that eating is necessary for an organism to stay alive is a materialistic theory (arguing a causal interaction between material matter) even if I have no exact model as to how digestion works. Of course, knowledge from one level should constrain knowledge on the other level, but results from one level cannot falsify theory on the other level, unless you 'translate' the theory to that level (it might, however, through reduction make the level irrelevant). Such 'translation' is exactly what is happening in the excellent Dunbar article you send (I now had time to scim it). They go from a cognitive level (Gingerenzer etc.) to a very specific neurocognitive level involving modelling used to generate specific hypothesis. These, in turn give rise to expectations about differential activation of specific brain areas in specific cognitive task. If they had not found what they were looking for, would that have falsified for instance Gingerenzer's findings and his explanations of these? NOT at all, as these are well documented in experimental psychology. It would, however, question the translation of these findings into neurocognitive modelling. Does that mean that Gigerenzers explanations have nothing to do with neurofindings. NO! If someone comes up with an alternative explanation to Gigenrenzers findings and this model can be translated into neurcognitive modelling that fit experiments on this level better, then the results point to the new theory being more true than the old one (it has an overall better coherence between levels).What I object against is the conflation of explanatory levels. If we want to address the question of decopuled cognition or metarepresentations (or any other cognitive level concepts or processes) we need to create neurocognitive models that can give rise to specific hypothesis at the level of the brain. Thus, because there might be a relation between two things on a cognitive level (say concepts and decoupling) one cannot use studies of one of these to falsicify or verify theories of the other (but one can be inspired to model the other area and then conduct studies...)

/Jesper

__

Now I have posted some further thoughts on the subject on our blog: http://cognitioncircle.blogspot.com/2006/12/whither-neuroscience-hail-gospel-of.html you are welcome to leave coments there,

a

Whither neuroscience? hail the gospel of cognitive science?

Cognitive scientists claim that they are constructing materialist theories. This makes the connection between cognitive function and neural substrate critical, especially for evolutionary psychologists (cf. my article ”What’s HIDD’n in the HADD” http://www.csr-arc.com/view.php?arc=4). Yet very few ever worry about this (with the noteable exception of (Boyer and Lienard 2006) ). On the other hand when finally you try to use the studies conducted, chances are that in the audience will be a neurologist, who will say that you cannot argue from firing patterns in an fMRI study to cognitive systems. If cognitive scientists don’t care about the neural basis and neurophysiologists think that the neural basis is too complicated to be used for arguments on cognitive function, should we just make the de facto divorce official then?
First some cynical observations:
Cognitive scientists are really not interested in neurolophysiology at all. They only need it because of the superiority it gives them to be able to claim that their theories are materialist, and therefore not mystical dualist ones. They think that if no neurological evidence can be found it is just due to the inferior current state of experimental machinery. If we had better machinery we would see the theories suported, but neurological function is more complicated.
Neurophysiologists are most interested in finding out more about how the brain works. They have a different background and they are the ones who know how to operate the scanners. They would really like to keep it that way. If the results were immediately understandable they would be reduced to mere technicians. This is why they also have an interest in limiting the connection between the neural substrate and cognitive function. Again the reason adduced is that neural function is more complicated.
(maybe I should mention the philosophers who also want a piece of the cake. Their claim to fame is that the scientists have not really thought through the basic assumptions behind their work, guess what: It is more copmlicated!)
To be completely blunt: If cognitive scientists cannot use neurophysiological evidence even with their limited educational background and necessary simplifications, you might as well turn off the scanners and divert the energy into spa baths so someone will at least benefit from it. Conversely if cognitive scientists are not constrained by the neurological evidence, their theories are not materialist, and they have to compete on equal terms with other dualist theories such as the SSSM (Barkow, Cosmides et al. 1995) and Buddhism.



References

Barkow, J., L. Cosmides, et al. (1995). The Adapted Mind. New York, Oxford University Press.
Boyer, P. and P. Lienard (2006). "Why Ritualized Behavior? Precaution Systems and Action Parsing in Developmental, Pathological and Cultural Rituals." Behavioral and Brain Sciences Forthcoming.

New Post from Sperber

Here is an interesting article on Theory of Mind that Sperber just posted on his website. They apparently found children as young as 13 months possess 'an incipient metrepresentational ability'.

Abstract:
In two experiments, we investigated whether 13-month-old infants expect agents to behave in a way consistent with information to which they have been exposed. Infants watched animations in which an animal was either provided information or prevented from gathering information about the actual location of an object. The animal then searched successfully or failed to retrieve it. Infants’ looking times suggest that they expected searches to be effective when—and only when—the agent had had access to the relevant information. This result supports the view that infants’ possess an incipient metarepresentational ability that permits them to attribute beliefs to agents. We discuss the viability of more conservative explanations and the relationship between this early ability and later forms of ‘theory of mind’ that appear only after children have become experienced verbal communicators.

From www.dan.sperber.com