Sunday, April 13, 2008

To be, or not to be.... a proponent of the multi-system view?

Super corny title, yes.

For anyone that has iTunes, and that's probably everyone, I found a Stanford podcast that relates directly to the topic we're discussing, though I haven't had a chance to read it yet. Get on iTunes, select "iTunes Store" from the toolbar on the left > search "Stanford U" > select "Stanford U"> go to "Health and Medicine" > choose "Mental Health > the podcast is "Perception, Decision and Reward: Toward a Neurobiology of Decision-making" by William T. Newsome. Hope you enjoy!

I'd like to start the body of my post with a quick summary of the Somatic Marker Hypothesis, which I feel that the Bechara et al., article did not quite explain. (This is a mix my my own limited knowledge supplemented by Wikipedia.)Essentially, the SMH proposes the existence of a mechanism through which emotional processes may either guide or bias behavior, particularly in the realm of decision-making. This proposal indicates that the view held by the authors is one of a multi-system process in decision-making. Oftentimes, one has to make a decision between conflicting alternatives, at which point cognitive processes may become overloaded and are unable to provide an informed option. It is here that somatic markers come into play; somatic markers are psychologically affective states that have been induced by reinforcing stimuli from the environment.
On a superficial level, I was highly entertained by the way Bechara et al. went about defending their hypothesis, and given what little I do know about the matter, I adamantly support them in their defense. Though I am not certain that I wholeheartedly agree with their hypothesis, they present incredibly valid points. For instance, Bechara et al. studied patients with VMPC damage, whereas Maia et al. studied regular Moreover, Bechara et al. point out that Maia et al.'s study, "undermines traditional methods for identifying implicit knowledge" (159). The reason for this accusation is that Maia et al. simply questioned their participants about what they know, undercutting the idea of implicit knowledge, which may be unconscious. In the end, Maia et al. show that even normal participants without damage to the VMPC, with adequate knowledge, are not guaranteed to make the correct decision.
Moving on, while reading the article by DeMartino et al., I found myself thinking, "Aha! This sounds like the SMH". However, I feel that while DeMartino et al. present two systems through which information is processed in decision-making, that they put forth the idea that the overriding system is one of "simple heuristics," or trial and error, while Bechara et al. (though I may be mistaken), propose that emotions, of affective states, are the deciding factor. To simplify, both believe that decisions are not made in the brain by only one system; when the fast and dirty system falls through, another one comes up to take the slack and makes the final decision. The difference between the two papers is that the authors differ on which of the two systems is the overriding one.

When I read the title for the Sanfey et al. article, I was a bit taken-aback to see the term "Neuroeconomics". However, once I began to read the article, it made perfect sense to me. An idea that caught my attention right off was the idea that behavior can be interpreted as choosing alternatives with the goal of maximizing utility. To me, this seems intuitively true, and as such, I thought it might be interesting to discuss this idea in class and see how other people feel. Later, Sanfey et al. provide two processes for decision-making (as do Bechara et al. and DeMartino et al.): automatic processes and controlled processes. To my understanding, automatic processing, true to its name, is quick, or "fast and dirty" and can be compared to the low road in the brain system, while controlled processing, as it is flexible and can support many goals, resembles the high road. Included in the functions of controlled processing are introspection, reasoning, etc., and so it would be reasonable to fit emotions into this category rather than that of automatic processing. This model for the interaction between the two systems, at least to me, resembles that presented by Bechara et al., though I do not wish to simplify the complexities of each of the different models. So my question is simple: Have I gotten it all wrong?
I would really love to hear others voice in on the similarities and differences of all these models of multi-system processing proposed by the different authors we read.

2 comments:

Lily Thom said...

It seems to be that emotions would be part of the automatic system if you think of them as visceral, low road responses. However, feelings if defined as conscious appraisals of the emotions, would be part of the controlled processing. It then follows that people with "better and more refined representation" of their emotions are able to use the controlled processing to their advantage (De Martino et al). Thus and make rational decisions that better recognition of the automatic processing would help you make more rational decisions.
I read this week's readings right before I saw the movie '21' in which blackjack players count cards in order to gamble only on "good decks". It is interesting that this method of playing requires ignoring the emotional aspects or somatic markers of decision making. I wonder how the deliberate non-emotional approach taxes the exploration-exploitation system. For some of the players in the film it proved impossible to gamble without eventually taking risks based on emotional desires, curiosity or, perhaps, some type of somatic marker.

A. J. Marr said...

Limitations of the Iowa Gambling Task and the Somatic Marker Hypothesis

The somatic marker hypothesis imputes that an affective ‘gut level’ somatic response (i.e., tension mediated autonomic arousal) results from the appraisal of surprising counterfactual outcomes or discrepancies that are implicit in choice. However, the immediate perception of these outcomes is also affective, and this affect is mediated by the neuro-modulator activity (I.e., activity of midbrain dopamine systems) that modulates choice. Thus not one but two ‘gut level’ affective responses occur in tandem during the IGT. Specifically, neurally based affect as elicited by the perception of decision-outcome discrepancy is not addressed in the experimental measures of the somatic states that are induced by the IGT and its various experimental iterations. But this is not a limitation of the IGT but of the observational tools that may be applied concurrently with the experiment. As originally implemented, the IGT represents an experimental ‘snapshot’ of the covert and overt behavior concomitant with decision making or choice. However, in studies that replicate the IGT, only a subset use psycho-physiological data, and these data generally use indirect measures of muscular tension as indicated by the SCR. The unique psycho-physiological affective responses due to the perception of positive or negative discrepancy as a subject chooses cards are not measured. This represents a serious omission in the experimental investigation of the somatic marker, since all the relevant affective events occurring due to choice are not considered. Specifically, in decision making peripheral and neurologically based affective responses often occur in tandem, yet no experiments have measured their correlation with the experimental contingencies that underlay choice. Thus the covert responses that mediate choice between alternative contingencies are measured separately through measures the source of peripheral (e.g., muscular tension) based affect (e.g. EKG, SCR), or in-vivo brain imaging that measures the source of neurologically based affect (e.g., activity of dopamine neurons). However, none have measured them together. The question is how covert and overt behavior under the IGT or in a more general sense alternative response contingencies or ‘decision making’ can be explained through an integration of the data provided by both of these experimental perspectives as reflected in the comprehensive discipline of affective neuroscience.