lundi 15 novembre 2010

Wittgenstein's dinosaurs

I moved to wordpress. Please comment this note at its new address: /

http://mlnode.wordpress.com/2010/11/15/wittgensteins-dinosaurs/

"Les gardiens du bon usage" is a critical paper written by Pierre Poirier and Nicolas Payette in response to Bennett & Hacker's Philosophical Foundations of Neuroscience.

"Le bon usage" is a phrase, which one should translate as something like "the right way, according to convention". Manuels with titles such as "Savoir-vivre et usages du monde" (from the infamous Berthe Bernage) told young women how to act as a perfect housewife in any mundane situation, and "Le bon usage" was the title of the most popular prescriptive grammar handbook. If anything, "les gardiens du bon usage" is a phrase which will strike as a rather dubious praise: one could indeed see the figure of the patriach, keeper of traditions, but also the figure of the "grammar nazi" or the authoritarian grade school teacher. One could also add that there is a generational issue in it: if the older generation remembers fondly of "Les insolences du frère Untel", which denounced the use of Québec's slang as boneless and servile, their children have not taken the carried the fight for "le bon français". In any case, the ambiguity in the phrase carries into the paper, a careful criticism of Bennett & Hacker's book which can be seen as benign or destructive depending on who you are.

B&H's argument revolves around the mereological fallacy, which is about attributing a property of the whole to one of its parts, or a property of a part to the whole. In the case of the neuroscientists, B&H think they commit the later, attributing psychological states and other properties of the mind to the brain or brain tissues. The point of their book is to expose this.

Obviously, as P&P point out, some conditions apply: if, philosophically, the theory of the brainbound mind was true (as opposed to, say, the extended mind), then there would be no problem in literally attributing psychological states to the brain, since it would be identical to the mind in a sense. But notwithstanding this, exposing a the fallacy is no easy task. As P&P note, it would be very surprising if everyone in neuroscience—including researchers who are aware of the philosophical issues concerning their field, like Damasio—had done the kind of fallacy college freshmen learn to avoid in philosophy 101 courses. Thus, some charity is expected when we read phrases such as "the brain decides" or "neurons estimate probabilities".

B&H consider four ways in which neuroscientists could be warranted in their use of these phrases: they could be using them as analogies, as metaphors, they could be modfying their meaning, or using them in a completely new sense.

Analogies have been fruitful in science: take the analogy between water flow and electricity in circuits, which has yielded successful predictions. However, they did because electricity in a circuit does indeed behave like water in pipes. Such structural correspondance be found between the brain and the mind.

Metaphors are less consequential; however, in B&H's opinion, it is hard not to get carried away. Illustrating their point with an example, they contend that using psychological metaphors will confuse neurologists and lead them to mereological fallacies. P&P seem unconvinced by B&H's evidence and point out that there is no necessity in metaphors causing this confusion. A careful usage should be a sufficient remedy.

In the third and fourth scenario, neurologists would be using such words as "think", "decide", "believe", etc. not to mean think, decide, believe, etc., but rather *think, *decide, *believe, etc., which have only vaguely similar meanings to the original concepts. This, however, would mean that we'd also have to modify the meanings of all the terminology that revolves around it, and talk about *information, *knowledge, *memory, etc. That's a bit too much in one gulp, in B&H's opinion: neuroscientists haven't done the conceptual work to warrant the use of a whole new terminology. P&P think that language doesn't evolve with definitions: people use new words (or new meanings) first, it's only when they reflect on their usage that they come up with definitions. It's the usage that gives meaning to the words not the other way round. Opposing new ways of using words would cement language and prevent its semantic evolution.

P&P invoke Patricia Churchland: inspiring herself of Quine's web of beliefs, she argues a new theory could very well change the meaning of the words it uses, such that usages which were previously unwarranted become perfectly legit. For instance, before Archimedes found a way to tell true gold from fool's gold by measuring its volumic mass, you could imagine that the definition of gold was something like "yellow metal". After Archimedes, the volumic mass became part of the definition, and as science evolved, so did the concept of gold, until we could even talk of "white gold" without contradiction. Similarly, they argue, neurological advances could modify our understanding of psychological properties, to the point that behaviour would be no longer necessary nor sufficient to recording their presence.

P&P admit that B&H are right in thinking that neurologists should be careful about the way they link psychological and neurological phenomena—if only because mistakes could lead to bad diagnostics. However, banning new usage of psychological words alltogether would probably be unwise: we should judge the tree by its fruits, and those are not ripe yet.