Brain Maps, Icons & Indices, continued

A representation of any process seems, paradoxically, more abstract and removed from reality until time is visually accounted for--nevermind that time is probably one of the most abstract concepts we've devised for ourselves to deal with. Very briefly, the history of brain-imaging techniques shows a progression from direct observation (sawing into the skull and prodding), to observation via static x-ray images, to static MRI-imaging, to SPECT images (also static, and revelatory of only appearance and not function at the time of the image), to PET images (reconstruction of imagistic "slices"), to functional imaging (depictions of the brain as it's working), to various still-experimental techniques of analyzing the brain from within (such as tiny "devices" like bacteria that could sense the localization of blood flow to a region of the brain and follow it there to record). Overall, the progression is from image-as-representation (icon) to image-as-diagnostic tool (index). The important change from the static to the functional representation is the addition of time--and the corresponding but implicit assumption that time implies causality.

It's difficult to look at a time-series progression of images of the brain without importing some causality from one image to the next. (Really, this is the simple principle that accounts for the magic of those early-animation flip-books.) Especially when a series of brain images depicts a "real-time" event like a movement of the arm, for example, it's almost irresistibly easy to regard the image as the event's equivalent. However, it might be more conceptually sound to regard it as no more than an analogy. ("No more than" should not lessen the usefulness of the analogy, though.)

An analogous rather than a causal link is useful for both practical and metaphysical reasons. Practically, it maintains the distinction between brain and brain image, between object and exploratory method. If the interest of science is objective observation, it should at least focus some of that observation on its own methods, and any initial conflation or blurring of important distinctions might make science blind to itself and as such a rather unwieldy beast. Metaphysically, to accord an analogous relationship to the time-series image of the brain and the brain itself at least leaves room for a distinction between mind and brain, which (it has been argued*) is conflated by increasingly precise time-series images of the brain without a corresponding increase in accuracy.

My summary of the argument is: scientists are much more precise than they are accurate. Overly precise images of the brain might give the mistaken impression that we know more than we do. For instance, a time-series of a brain showing an arm movement might give us very precise information about the brain during segments of this event, and, along with the illusion of causality, this precision might suggest that the cause of the event--the impetus for movement--is discernable. When really, we might be looking at the wrong thing entirely--or at the thing in the wrong way entirely.

*This was Prof. Alan Gross' argument in a lecture of his I recently attended. This post owes much to his lecture and to the audience's questions during the lecture.

Tbc, again...

Brain Maps, Icons & Indices

There are certain things we take for granted about science, that paradigm of objectivity. For one, there's its objectivity; that science is not objective seems a vague and fundamental contention whose proof seems as obvious and elusive as the horizon. There's another thing we take for granted: that models mirror the causal structure they depict. It's easy to forget that images generated by apparatuses such as MRI and electron microscopy are still essentially maps, and it's easy to construe the image with the thing generating the image--a tool that's the product of another tool (the scientific method) that's the product of The Human Brain, a popular subject of so many models and images.

First a few concessions: these images are for teaching and learning; the people who drew the amygdala in early issues of Brain knew very well that it wasn't perfectly congruent with the original. It was The Amygdala, a representation, not an amygdala functioning as part of a system filled with idiosyncracies and undocumentable anomalies. Really someone just wanted to convey the idea of the amygdala, enough said.



But these icons can become a little more connected to the real thing: they can become indeces. Comparing a representation of the brain to an actual brain, certain variations become noticeable. A representation that records this variation is an index: basically, a repository for information storage and retrieval. Now the diagram is a little more implicated in the system it represents.

There are other examples of images that are both icon and index: for instance, a map drawn by someone who, nevermind the circumstances, is quite without significant parts of his left brain. For this person, the map is an icon--it's a familiar, sorta gestalt representation suggestive of the represented (the actual content of which is often not experienced, or not able to be experienced, entirely and simultaneously). So, working from a map, this person draws, say, the U.S's West Coast, sees that he's copied what the map shows, and calls it complete. For a psychologist, this map isn't an icon, but an index of the guy's brain injury. The guy was his own diagnostic tool. The diagram is an important part of the system at work in this guy's brain.

To be continued...

"Prototype for new understanding"

Totemic and athletic shoe sculptures. I've been liking this lately.