Thanks to Buddy, I came up with a way of reconstructing Bechtel and Mundale’s argument that does not rely on the tacit premise that univocal localization entails univocal realization. Recall that B&M state:
Nevertheless, it is important to note that in interpreting
these deficits, researchers implicitly reject multiple realization among human
brains and assume that damage to a brain area in anyone will result in a
deficit to a particular cognitive function that is performed by that area in
undamaged brains. (Bechtel & Mundale, 1999, p. 184).
Here is the rough reconstruction:
P1:
Researchers assume that damage to a brain area in anyone will result in a
deficit to a particular cognitive function that is performed by that area in
undamaged brains.
P1’:
Researchers assume that cognitive function F is realized in area A in both lesioned
individuals and normal individuals
P1’’:
Researchers assume that cognitive function F is always realized in area A.
P1’’’:
Researchers assume that cognitive function F is univocally realized in area A.
P1’’’’:
Researchers assume that cognitive function F is univocally realized.
C:
Researchers implicitly reject multiple realization (of F) among human brains.
This gives the argument the form of just a strict series of entailments. I don’t think this argument works either, but how about this time I throw it open to others to defend or reject this.
That seems fine, but I can already see where we will disagree. You are going to say the move from P1” to P1”’ is supposed to be deductive, and it based on the faulty lemma that UL implies UR. I agree that if the move is meant to be deductive, it is faulty. But I will say that the move from P1” to P1”’ is meant to be abductive, and might very well be justified.
If the move from P1” to P1”’ is meant to be abductive, it can certainly still be criticized, but not simply by pointing out that UL doesnt logically (deductively) imply UR.
Ok. Let me work up to your last point. This will help explain why I have reconstructed B&M as I did earlier and why the defeasible inference idea doesn’t work (as I see it). In the passage quoted, it might look as though B&M are only interested in reporting on what neuroscientists, or brain scientists, think. So, it might look as though they are only interested in science reporting or science journalism. But, at other points, they seem to want to draw conclusions about MR. The major premise of one of their central arguments “For multiple realization to be a serious option, brain taxonomy would have to be carried out both indpendently of psychological function, and without comparative evaluation across species”. So, it seems to me that they have some sort of defeasible inference rule such as “If neuroscientists believe/assume that p, then p” (One defeater for this rule would be if we knew that neuroscientists believe p on the basis of faulty evidence or inference.) So, let me apply this defeasible inference rule to the preceding argument.
P*1: Damage to a brain area in anyone will result in a
deficit to a particular cognitive function that is performed by that area in
undamaged brains.
P1*’: Cognitive function F is realized in area A in both
lesioned individuals and normal individuals
P1*’’: Cognitive function F is always realized in area A.
P1*’’’: Cognitive function F is univocally realized in area A.
P1*’’’’: Cognitive function F is univocally realized.
C*: F is not multiply realized among human brains.
This enables us to talk about what the neuroscientific evidence shows (or does not show), independently of what the neuroscientists think the evidence shows. I am assuming that we want to know what the evidence shows, rather than merely what neuroscientists thinkg the evidence shows.
So, this enables us to evaluate the inference from P1*” to P1*”’. In particular, it enables us to evaluate the inference as abductive, rather than deductive. Right?
So far, so good?
First of all, of course the point about science journalism is well taken. But still, I dont think there is an inference from P1*” to P1*”’. There is an inference from P1 all the way to C, and then from C to C* using the rule you mention (R1 below).
But look, your main point is well taken. Even if its true, as B&M argue, that scientists tacitly assume UR, if its ALSO true that they are unwarranted in tacitly assuming UR (because UL would do), then it might defeat the inference from belief among scientists to truth.
But I’m not sure why we should accept anything like (call it R1) “If neuroscientists believe/assume that p, then p” in the first place (though I agree B&M expect us to.)
I think its especially weird to invoke R1 if you admit, as they do, that belief in UR is expedient–“brain taxonomy would [be too hard]” without it. In other words, even if I accept R1, the fact that belief in UR (or in UL for that matter) is expedient might also be a defeasor on R1.
Ok. It seems to me that we are making progress toward mutual understanding here.
I have not been so much interested in what scientists tacitly assume, e.g. if they do tacitly UR. I have been trying to argue that this UL to UR inference is not good, so that perhaps scientist are making a bad inference. That is, I think we agree that “Even if its true, as B&M argue, that scientists tacitly assume UR,
if its ALSO true that they are unwarranted in tacitly assuming UR
(because UL would do), then it might defeat the inference from belief
among scientists to truth”
Regarding (R1), we seem to agree that B&M expect us to accept it. And, in addition, I think we both believe that (R1) is problematic, at best. I’m invoking (R1) not because I believe it or endorse it, but because I think it is a principle that is at work in B&M’s thinking. I’m using it to help articulate why I earlier proposed to “distill the essence” of B&M’s argument as:
P1
Lots of Fs are UL.
P2
F is UL entails F is UR.
C
So, lots of Fs are UR
You asked me why I wanted to read their premise as P1. I’m still working on that. So far, so good?
OK, so you think R1 is best applied to A1: “Researchers assume that lots of Fs are UL” in order to get to “lots of Fs are UL”. If so, then yes, P2 would be needed to get their desired conclusion, and P2 is problematic.
But you also seem to be admitting that it is also true that A2:”researches assume that lots of F are UR.”
So, the question is: is there better grounds for applying r1 to a1 than to a2?
Here’s where things get a bit strange: M&L seem to think A2 is the case because the assumption of UR is needed to make brain research possible (somthing like that). You seem to be saying: “UR isnt needed, only UL is needed. Hence R1 should only be applied to A1 not to A2.”
So, you seem to hold a principle along these lines: “If scientists hold a claim C for expedient reasons, but a weaker claim C* would be just as expedient, then we should only infer to the truth of C*, not to the truth of C”.
I don’t know what I think about this. I’m not sure we should infer to the truth of ANY claim just because scientists hold it for expedient reasons.
In my philosophizing, I don’t like to rely on (R1). I wish B&M wouldn’t, but they do. Here is why I don’t like (R1).
I think (R1) is mostly plausible when neuroscientists have good reason to believe that p and least plausible when neuroscientists do not have good reason to believe that p. I think, “Why not just cut to the chase.” What matters is whether or not there is good reason to believe that p. Leave aside the sociology of neuroscience; leave aside the science journalism.
So, I do think that it’s basically ok to apply R1 to A1: “Researchers assume that lots of F are UL” because I have read journal articles that explain the results of brain lesion studies, fMRI, etc. (I hope we can save for another time the matter of how good this evidence really is. For the sake of the present debate, I have been conceding what B&M want.)
I am hesitant to apply R1 to A2, “researches assume that lots of F are UR”, if all their evidence is just Fs are UL, because I’m thinking that that evidence for A2 is not so good.
Here again, we seem to be on the same page, right?
B&M’s first premise –
“P1: Researchers assume that damage to a brain area in anyone will result in a deficit to a particular cognitive function that is performed by that area in undamaged brains.”
is just false. This is a bigger problem here. Some see localization as fruitful, plenty don’t.
Over the course of the discussion here I have worried about that, because I know that there are researchers who deny the localization of cognitive function, e.g. Stephen Small at UC.
There is a kind of irony here. Where many philosophers, such as apparently B&M, abhor “armchair philosophy” they are tacitly quite happy to do “armchair sociology”. They read a several articles where several neuroscientists believe that P, then infer that neuroscientists believe that P. But, if you want to know what neuroscientists believe, then reading neuroscience articles might not be the way to go. Maybe you have to do certain kinds of surveys to get a representative sample. (If I were a sociologist, I might know.) A sociologist friend of mine thinks “armchair sociology” is worse than “armchair philosophy”, since the former at least purports to be an empirical science.
Just another reason to abandon the pseudo-scientific sociology and go with claims about what the evidence shows.
So, I would agree with the point you raise.
Funny you should say. I’m actually working in Steve Small’s lab at the moment. There are a (growing) number of other anti-localizationists as well, even in lesion researchers. In a recent chapter Hanna and Antonio Damasio go so far as speaking of a ‘new’ lesion method. They say “The NEW lesion method is not concerned with ‘localizing functions,’ nor is it a contest for ‘localizing lesions’. It is a means to test, at a systems level, hypotheses regarding BOTH neural structure AND cognitive processes.” (‘Structural Imaging of Patients in Cognitive Neuroscience’, In: Patient-Based Approaches to Cognitive Neuroscience, Farah and Feinberg, eds., 2nd Ed., 2006, MIT, p. 22)
I completely agree with your point about “armchair sociology.” Difficult, though, to conduct such surveys, both in finding all the people you should be asking and getting them to respond. Maybe as hard as reviewing all the relevant articles. Certainly doesn’t mean it shouldn’t be done, though. B&M could just as easily have made their point by saying ‘some research makes this assumption, and our claims apply to such research…’ If published today and not in 1999, perhaps they would have, post Uttal’s ‘The New Phrenology’ and there having been in general a significant move away from modularity and localization (by some).
I was going to ask about some of the other anti-localizationists. I knew of Uttal’s book. The case there is, I think, pretty unconvincing though. I guess I have found the neuroimaging literature so vast as to be a deterrant to trying to do any philosophical work on it. What I have seen seems to me to be in good hands.