Monday, 15 March 2010

The only non-representational cognitive psychologist in the village

Hi. I’m a cognitive psychologist, but I’m not that kind of cognitive psychologist. Specifically, I don’t believe in representations, and I reject the computational model of cognition. Yes, this makes me very unpopular. I this post I want to quickly review the dominant cognitive approach and then briefly raise several potential problems with this framework. I will go through these issues in detail in later posts, but I want to go ahead and present the big picture here.

Cognitive scientists tend to view cognition as computation. In this model, representations are data structures and cognitive processes are algorithms acting on these data structures. Input => transformation of input via manipulation of discrete symbols (representations) => output. The clear analogy is to information processing in a computer. Another way to think about representations is as internal mediating states (cf. Dietrich & Markman, 2003). For instance, when I see a cup, the stuff happening in the visual system will probably be a better match to my “cup” representation than to my “glass” representation. So, I correctly identify the object as a cup. In other words, my ability to identify an object depends on consulting a discrete, internal representation of that object.

There are a number of unresolved issues with this representational stance: First, there is no theory of what representations actually are or of what information they contain. Second, many cognitive phenomena seem to defy a computational explanation. For instance, attempts to use a computational framework to model cognitive behaviours have often failed to produce anything as flexible or interesting as what we humans get up to. Third, alternative stances (e.g., that there are no discrete representations or that they are not processed algorithmically) have not been thoroughly explored. Cognitive psychologists usually take representations for granted; their existence is assumed, rarely defined or tested. This just isn’t good science. I’m just raising these points here; in future posts I’ll lay out the evidence.

My goal is to spend some time discussing these issues and to think about alternatives to representation. As a cognitive psychologist, I could get away with not understanding or caring about perception. Honestly, it just doesn’t come up much. When it does come up (e.g., Barsalou) it’s in terms of the “sensation based theory of perception” (see previous post), which we know is outdated. In the long run, I want to discuss how it might be possible to ground the study of cognition in Gibsonian perception/action. This is risky since, at the moment, I have no idea what such a framework would look like. But, cognitive psychology needs to evolve, and this is currently my best bet on how that might happen.


  1. You can smack down all the sciences you want, I care not for categorization, just plain evidence.
    I don't agree much with the Computational models of cognition either.

    Are you against all types of current data acquisition methods other than behavioral(EEG, fMRI)? (Should I picture you as a grumpy old man in a bath robe hitting an MR scanner with a cane?)
    And if so, what are you Ecological approaches on, say for example, pain research?

  2. I didn't post this one :)

    But I will be down on fMRI for missing the point at some point.

  3. I just saw that.

    But what about my questions? I'm interested in the answers, they're not rhetorical.

  4. Yes, by all means picture me as a grumpy old man in a bath robe assaulting neuroimaging equipment. I am not a fan of the way most psychologists use neuroimaging.

    My primary concern is that we mistake an effect (oohh, that part lights up) with an explanation. If a participant engages in two distinct tasks, then it would be extraordinary if that didn't correspond to some change in brain activity. However, this doesn't really tell us anything about the mechanisms underlying a particular effect. Neither our understanding of neuro - anatomy/physiology nor the reliability of effects across studies is sufficient. My second gripe is that people use expensive imaging techniques when a good old fashioned behavioural measure would do the job. My third concern is that effects that turn up in imaging data are given special status as somehow more real, and, sometimes, more "innate", than effects found in behavioural data. Again, considering the totally unsurprising connection between changes in brain activity across tasks, there doesn't seem to be anything particularly revelatory about seeing something happen in the brain. Psychologists frequently interpret results from fMRI, etc., as representing some hardwired preference (e.g., mens' neural response to certain waist-to-hip ratios).

    I say this as someone who started grad school studying neuroscience and who won a pretty sweet neuroimaging fellowship to continue along that path. In the end, I decided to abandon ship for behavioural measures. I think neuroimaging is unbelievably cool, but I don't think the science is there (in psychology, at least) to really support the using these techniques effectively.

    I don't know anything about ecological approaches to pain research, but it doesn't seem like the model you mention, especially in the contribution of "top-down" processes would require representations.

  5. Thanks for the answer.
    I actually thought the kiwi posted that in the first place, hence the "old man" rant.

    I'm not a psychologist, and I agree that behavioral tests can give more insight than a head scan. It's just a matter of knowing how to use the tools and not over interpret the results or blobs.

    What I can't stand, are people with no understanding of the tool, who just want to see pretty pictures in their study.

    That said, I do use neuroimaging tools like EEG to support specific hypotheses backed up with behavioral studies.

  6. That said, I do use neuroimaging tools like EEG to support specific hypotheses backed up with behavioral studies.
    Can you provide something specific, that is an experiment you couldn't have done behaviourally?

    My hunch is that the only experiments that require neuroimaging are ones that actually ask 'what lights up in the brain?', and I think that question is not as interesting as people think.

  7. I think that sentence came out wrong.
    What I meant was, I use neuroimaging tools supported by behavioral studies.
    I wouldn't do an only Neuroimaging study. Thought there you have the Morphometry studies where you don't actually need behavioral, if your question is about the differences in gray matter size between certain groups.
    But not for a psychological study. For that, you must rely on behavior as your main tool.

    My last study was a pain study, entirely behavioral with physiological measures to support claims about the subject's pain reports.
    That study alone answered all the questions, and now I'm adapting it to fMRI.
    Not because I care where the pretty lights are, but because I'm interested in the underlying network that influences the pain perception, specifically the brainstem.

    Hope that answers the question.

    I love this debates. Keep it up.

  8. Two things:

    if your question is about the differences in gray matter size between certain groups
    Obviously you need to image here: but the question is always, why is this interesting/how do we interpret the result? It's here that imaging studies tend to half-ass their way through, with no real justification other than 'we wanted to know the brain network underpinning performance' which is only interesting if either a) the brain network is the only thing underpinning performance (never true) or b) if you explicitly ask the question with the fact this isn't true forefront in your mind (which never happens).

    entirely behavioral with physiological measures to support claims about the subject's pain reports
    This is the other thing: if it's in the brain, it is more real than if it's just in your behavioural measure.

    These are just some of the assumptions built into neuro-imaging work that are almost never questioned.

  9. I'm not a psychologist, I believe that verbal or written reports given by the subjects are subjective enough to not trust them without the aid of at least physiological measures (such as heart rate).

  10. Here is a question for you; What about the plight of Non-Imagers? I am not talking about weak imagers, I am referring to people like me who have NEVER had a mental image nor have I ever remembered (or had fleeting glimpses of) any 'dreams' I may have had? Do either of you have an opinion on whether or not an fmri would help at all in discovering if this is a neurological or psychological issue?
    I only ask because I thought that would be a perfect starting point for a researcher assuming that anyone will ever research this...
    I know the APA does not even consider the idea and almost noone has even given thought to such an idea like this being possible!
    Check out if you do not understand what I am referring to, also I would like to point out that this was concept was first (to my knowledge) reported in The Journal of Philosophy (Vol. III, No. 26) on Dec 20, 1906!!

  11. I do not think that MRI would be a useful tool for addressing that question. One reason for this is that there is not a clear distinction between neurological and psychological. Differences in brain structure can reflect genetics or experience (and, most likely, both). Furthermore, our understanding of the relationship between brain structure and behaviour is not nearly sophisticated enough to shed much light on the issue. I don't think fMRI would fare much better for similar reasons (e.g. our understanding of the relationship between brain activity and behaviour is also pretty poor).

    I think the most important direction for future research in this area is to figure out precisely what role the sensation of mental imagery plays in behaviour. Current arguments about the importance of mental imagery clearly overstate the role it plays in cognition, as evidenced people who have been commenting on this post. There appear to be some real differences, but if we were to believe some of the claims made in cognitive psychology, those who do not experience imagery would be very dysfunctional. This is clearly not the case.

    So, we need to conduct a systematic comparison of performance across a wide variety of behavioural tasks to assess exactly where and how performance differs between those who report experiencing mental imagery and those who do not. This good old fashioned behavioural analysis is the essential first step. Jumping right into neuroimaging will likely lead to some very impressive claims about neurological differences, and speculation about what these differences mean. However, without a solid empirical underpinning, these results will be useless.