IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

[Neuroscience] Re: Philosophical Foundations of Neuroscience

jonesmat via neur-sci%40net.bio.net (by jonesmat At physiology.wisc.edu)
Sat Dec 2 03:04:07 EST 2006

Matthew Kirkcaldie wrote:
> Gentlemen, please!

Matthew, I'm quite sure I don't care for your tone!
Just what are you insinuating?

> Surely you are arguing different points.  Isn't the "mereological
> fallacy" a catchy term for the fact that the boundary of "brain" as the
> source of behaviour is far blurrier than is typically acknowledged?

Ok, I think that's a very optimistic way of rephrasing that term.

You'll get no argument from me that the boundary between brain and
behavior is fuzzy (actually, I don't think there's a boundary at all -
that's sorta my point). And you give very nice, specific and well
thought out, examples of how fuzzy the boundary is.

But please remember that this discussion got started by specific
reference to this recent book by Bennet & Hacker, "The Philosophical
Foundations of Neuroscience".
(by the way, has anyone seen the original poster, John H, anywhere?
Wonder what he makes of all this.)

Now that's a pretty ambitious title, wouldn't you say? Without knowing
anything else about it, I would be tempted to run out and by a book
with a title like that, hoping to learn something useful about the
history and philosophy behind the field of which I am so fond.

Ay, there's the rub. I do know more about it now. There are several
reviews available. In those reviews, I learn that this book appears to
be an attack on the *conceptual* foundation of neuroscience, and that
this attack revolves around the term "mereological fallacy".

So, now I'm thinking that the *literal* meaning of this phrase actually
might matter to understanding the "philisophical foundations of
neuroscience." Yeah?

The NDPR review tells me this:
"In Chapter 3 of Part I - "The Mereological Fallacy in
Neuroscience" - Bennett and Hacker set out a critical framework that
is the pivot of the book. They argue that for some neuroscientists, the
brain does all manner of things <snip list of all manner of things
neuroscientists have said about the brain - and with which I largely
agree>...Implicit in these assertions is a PHILOSOPHICAL MISTAKE (my
caps), insofar as it unreasonably inflates the conception of the
'brain' by assigning to it powers and activities that are normally
reserved for sentient beings....To attribute such capacities to brains
is to commit what Bennett and Hacker identify as "the mereological
fallacy", that is, the fallacy of attributing to parts of an animal
attributes that are properties of the whole being."

and this:

"Neuroscientific research that proceeds from CONCEPTUALLY FLAWED (my
caps) premises is likely to yield INCOHERENT (my caps) empirical
questions and answers that ESCAPE ILLUMINATION OF ANY KIND (my caps -
couldn't resist - that last bit was sooooo dramatic)."

This is serious stuff, right? What's at stake here is whether
neuroscience is incoherent, and will never go anywhere, because we are
pursuing a line of thinking that is utterly flawed.


Right off the bat, I object to these criticisms for two reasons:

Objection #1)
It turns out that "mereological" is already a widely used word in a
completely different context: that of formal logic, philosophy and
artificial intelligence. It has a very precise technical meaning
already, and there's a large literature on it. And nowhere in that
existing meaning do the words "animal" or "brain" occur at all.

So that seems a bit fishy to me. Are Bennett and Hacker paying any
attention to the existing meaning of the word, or are they just
appropriating it for their own purposes, the way a new-age mystic might
talk about "spiritual energy", when "energy" is already a very
precisely defined term that essentially excludes the concept of
"spirit"? I still don't know the answer to this, but I'm gradually
resigning myself to buying and reading the book, if for no other reason
than to study the opposition, so to speak. I recently slogged my way
through a book called "The Mysterious Flame: Conscious Minds in a
Material World" (another very ambitious sounding title) by famous
philosopher Colin McGinn. I can now heartily recommend that any serious
neuroscientist or psychologist or philosopher NOT read that book. I
found it a complete waste of time, filled with loose logic and sloppy
reasoning (but lots of catchy phrases). At some point I'll probably
have a more informed opinion of the B&H book as well.

Ok, so to summarize my first objection, I have the very strong feeling
from reading the reviews that the authors are NOT paying much attention
to the existing discipline of mereology itself, and are just
appropriating high-fallutin' terms to sell books. That sort of pisses
me off in general, and moreso if it's being used as a tactic to
undermine confidence in a discipline that has actually had a lot of
practical success in helping the human condition over the last 1.5
centuries or so. You see where I'm coming from?

Objection #2)
Let's forget my anal-retentive desire for literal definitions for a
while, and just go with the common language understanding of
"mereological fallacy" as described in the reviews: "the fallacy of
attributing to parts of an animal attributes that are properties of the
whole being." This is the damning conceptual mistake that
neuroscientists make, apparently.

Is this, in fact, a fallacy at all? Is it unacceptable to do this? Does
doing it invalidate any interpretations of data, or cause experiments
to be guided down the wrong road such that they will never lead
anywhere? Certainly this seems to be the premise of the authors (as
described in the reviews). Would you agree that that is the point of
this whole discussion, ignoring issues of precise definitions?

I assert that is NOT NECESSARILY a mistake or fallacy to "attribute to
parts of animal attributes that are properties of the whole being".

It surely CAN be a mistake. The following show cases in which such
attribution is nonsensical:

Attribute = Reaching:
Reaching is an attribute of a whole person.
I reached for the whisky glass. (valid - the whole person reaches)
My liver reached for the whisky glass. (nonsense - the part of the
person that is the liver can't reach in that sense)

Attribute = Running:
Running is an attribute of a whole person.
I ran to catch the bus. (valid - the whole person runs)
My brain ran to catch the bus. (nonsense - the part of the person that
is the brain can't run in that same sense)

However, sometimes it makes perfect sense to attribute the same
function to the part as to the whole, because that part IS the part of
the whole that mainly performs the function:

Attribute = Reaching:
Reaching is an attribute of a whole person.
I reached for the whisky glass. (valid - the whole person reaches)
My hand reached for the whisky glass (VALID!)

The latter is valid because the fact that reaching is an attribute of
the whole person DOES NOT detract from the fact that reaching is ALSO
an attribute of the hand, and in fact, when the whole person reaches,
it is actually MAINLY the hand that is doing the reaching.

Attribute = Running:
Running is an attribute of a whole person.
I ran to catch the bus. (valid - the whole person runs)
My legs ran as fast as they could carry me (VALID!)

The latter is valid for the same reason described above. Running is an
attribute BOTH of the whole person, AND of the appropriate part of the
person. There is NO FALLACY in making the attribution either way.

Now, consider these:

Attribute = Feeling
Feeling is an attribute of a whole person.
I feel bad. (valid - the whole person feels bad)
My stomach feels funny (valid or invalid?)

Is "feeling" an attribute of a stomach? Aren't we committing a
"mereological fallacy" by saying our stomach feels a certain way, when
everybody from Wittgenstein to Bennet & Hacker tell us that only a
"whole person" can display the attribute of "feelings"? I think most
educated adults would say that the stomach is not really doing the
feeling, rather their "mind" is doing the feeling. Most neuroscientists
would probably say that the feeling is occuring "in the brain", and is
being associated with the stomach via topological mapping of the
stomach afferents onto the somatosensory cortex or whatever. On the
other hand, my 6 year old son would probably be perfectly comfortable
with saying that it was the stomach *itself* that had the feeling. As
far as he's concerned, his stomach is telling "him" that he ate too
much of Dad's lamb vindaloo.

More to the point, suppose that, like my son, we decide that saying "my
stomach feels funny" LITERALLY means that the stomach itself is
"experiencing a funny feeling", and that it is trying to tell us
something about what we ate, and that we shouldn't eat anymore of that
until the feeling goes away. Is this "mereological fallacy" going to
lead us down the wrong path? Are we going to lose anything by
attributing this property (feeling) of the whole animal to just one
part of the animal? To paraphrase from the quote above, are we going to
proceed from conceptually flawed premises (e.g., the stomach feels
something) to yield incoherent empirical questions (e.g., if I eat any
more vindaloo, will I get the runs?) and answers (e.g., yes, I'll
probably get the runs, so I better not eat any more vindaloo) that will
escape illumination of any kind (I can only assume this would be the
state of not knowing whether we had the runs or not)?

No! That's utterly absurd. If your stomach "feels funny" after you eat
vindaloo, it's probably the best possible idea to stop eating the damn
curry. Mereology aside, this "fallacious" attribution is NOT going to
lead you astray. Quite the opposite.

[Let me now make it perfectly clear that I personally do not believe
there is any such thing as eating too much curry - curry is a gift from
God - my son has not yet had this epiphany, but I'm working on it]

And how about this one:

Attribute = Thinking
Thinking is an attribute of a whole person.
I think I'll have another whisky. (valid - the whole person thinks)
My striatum thinks I should have another whisky, but my frontal cortex
thinks I should go talk to that girl. My medial amygdala thinks I
should hurry up and go talk to that girl as soon as humanly possible!
Unfortunately, my cerebellum thinks I should just stay in my seat to
avoid falling over and making a fool of myself. My frontal cortex sees
the logic in that, but amygdala is, as usual, being very persistent.

Valid or invalid? You tell me. Frontal cortex is usually worth
listening to, but rarely any fun to go drinking with. Maybe brain truly
is irrelevant, because experience demonstrates that bladder will
inevitably get the last word.

Anyway, I think I've illustrated various cases in which mixing
attributions among the part and the whole can either be a totally
stupid thing to do, or a totally fine thing to do, or an ambiguous
thing to do.

I conclude that simply the fact that attributions are sometimes mixed
like this in neuroscience IS NOT necessarily a damning fallacy. And it
is most definitely not a "mereological" fallacy unless one can derive
the appropriate mereological formal proof, as I argued previously. I
conclude, as I said at first (just before I was accused of having my
head and both legs up my ass), that no "mereological fallacy" has in
fact been committed by n euroscientists.

> Now I expect you will both attack me instead!  Still, I'd like to know
> your reactions and whether this is a reasonable paraphrase of your views.
>       Cheers, MK.

Oh, damn. I almost forgot that I'm supposed to be "insufferably
arrogant", and use a lot of "ad hominem" arguments and gratuitous
invective. How's this:

Piss off, you wanker!

(did that sound at all convincing?)



More information about the Neur-sci mailing list

Send comments to us at biosci-help [At] net.bio.net