Looking through cats' eyes

Matt Jones jonesmat at ohsu.edu
Wed Oct 20 14:25:49 EST 1999


In article <13682-380D1B94-23 at storefull-174.iap.bryant.webtv.net> ,
Y-chat at webtv.net writes:
>Since the ARTICLE refers to monkey research, I will uncharacteristically
>ASK Matt Jones to tell me the difference in these studies, other than
>the methodology of using cats and the extraction of reproducible signals
>from the data.

OK. 

Rather than try to summarize all the monkey research that's ever been
done, or to go through the Logothetis article point by point, I'll tell
you why the basic -subject- of the data reviewed by Logothetis concerns a
fundamentally different -subject- than that addressed in the cat article
by Dan and colleagues.

Logothetis article (and a lot of his own research) is about how we can
try to understand how neurons function during conscious behaviors, and
concerns questions like "Is there any information contained in the
activity of the neuron that directly relates to the conscious decisions
or actions being undertaken by the animal?" An alternative hypothesis
would be something like "Consciousness is either not contained in the
firing of neurons, or is contained in such a way that no individual
neuron can signal anything about the 'higher order' conscious state of
the animal."  This is a pretty abstract question to be asking about, and
obviously requires that the experimenter be able to a) record the
activity of neurons, and b) simultaneously be able to "interrogate" the
animal to determine exactly what it is conscious  of (i.e., aware of and
able to report about) and what it is not. The monkeys can't just tell us
these things, so people have developed all sorts of clever tests that
reveal what the monkey "knows" and what it doesn't. 

Anyway, the scientist then tries to see if there's any correlation at all
between the firing of a neuron (or group pf neurons) and whether the
monkey is consciously aware of something that happened. A very important
series of these experiments (largely carried out by Logothetis) involves
using optical illusions. When you see an optical illusion, your eyes are
literally seeing the "real" thing that is being shown, but you are not
conscious of what you are "really" seeing. You are being fooled by the
illusion. If someone asks you what you see, you'll give the wrong answer.
It turns out that there are some neurons that also give the wrong answer.
They are being fooled too, which means that there is some infomation in
their activity about what you -think- you see, not what you are really
looking at. This is why he calls the paper "Vision: a window on
consciousness". By studying the reponses of neurons in relation to
conscious behaviours, you can get a glimpse of how they are active during
consciousness.  

Now not all of the neurons in the brain act like this. Specifically, the
LGN neurons studied in the Dan paper do not act like this. They instead
report very faithfully what you are -actually- seeing, not what you
-think- you see (indeed, they continue to encode this information even
when you are anesthetised and not thinking anything at all). In contrast,
the neurons in visual cortex that respond in relation to conscious
behaviour are very bad at reporting what you actually see when you are
shown an optical illusion. The LGN neurons ENCODE THE LIGHT INFORMATION
COMING INTO YOUR EYEBALLS, whereas (oh, hell, let's just take the
plunge...) the CONSCIOUSNESS neurons in cortex ENCODE INFORMATION ABOUT
WHAT YOU THINK YOU SEE.

So, the  two sets of research are not only in different animals, but they
are in two completely different parts of the brain, involving neurons
that encode two completely different sorts of information,  and address
two completely different issues about how the brain works. They are just
plain not the same thing at all. Let me repeat that. They are just plain
not the same thing at all. One more time. They are just pla... oh enough
already. They're different.

Cheers,

Matt Jones



More information about the Neur-sci mailing list