Loading ...
Sorry, an error occurred while loading the content.

[MURG] why precise neuron models?

Expand Messages
  • R. Paul McCarty
    I ve seen frequent mention here of biologically realistic models of neuron activity (sophisticated compartmental models, etc.). It seems to me once we ve made
    Message 1 of 10 , Sep 1, 1999
    • 0 Attachment
      I've seen frequent mention here of biologically realistic models of
      neuron activity (sophisticated compartmental models, etc.). It seems to
      me once we've made the assumption that the underlying computations which
      result in our conscious selves need not be tied to biological neurons
      we're just making our life more difficult be attempting to model neuron
      activity precisely. The only reason to do it would appear to be to
      capture the current behavior of the human brain at the lowest necessary
      level and hope all the other stuff is there. This seems good, but
      inherently difficult and uncertain.. we really wont know if it works
      until we copy a whole individual and see how they do.. I suppose we have
      to do this anyway, but it would be quite disappointing to spend your
      whole life working on this and discover it doesn't work until you model
      x,y and z.

      A far easier method, would be to merge with an already conscious
      electronic system. Or to slowly augment your brain electronically,
      until a large portion of what makes you *you* is electronic, and you can
      just shake off your flesh when it expires. I paint a gruesome picture
      don't I? ^_^

      This has always seemed the shortest path to the uploaded state to me.
      But obviously, you have lost some of /you/ in the process.. but it can
      be minimized by making the electronic side proportionately larger then
      the biological.

      What other reasons are there to concentrate on realistic models of
      neurons?

      -Paul
      --
      R. Paul McCarty / rpmc@... / x52059
      317 Lattimore Hall, University of Rochester, Rochester, NY 14627
      Life is nothing if you're not obsessed. -Pecker
    • Joseph J. Strout
      It s simply a matter of personal identity (sigh... here we go again). Suffice to say that I believe my identity lies in mental structure, which in turn arises
      Message 2 of 10 , Sep 1, 1999
      • 0 Attachment
        It's simply a matter of personal identity (sigh... here we go again).
        Suffice to say that I believe my identity lies in mental structure,
        which in turn arises from the workings of the brain. No amount of
        electronic augmentation (if such is even possible, except for very
        limited sensory augmentation -- anything more than that seems
        unlikely) will change that. If my brain dies, and it has not been
        copied into a functionally equivalent duplicate, then *I* die.

        Now, it's possible you could make a functionally equivalent model of
        the brain at a slightly higher level than the compartmental model --
        an elaborate kernel-based model might do, for example. OTOH, even
        compartmental models might not be enough; Eugene advocates starting
        with molecular modeling and building your way up from there. I guess
        you have to pick the level of modeling you think is most likely to
        work, and yet is efficient enough to work with near-term technology.
        Then, you have to gather lots and lots of data to see how far off
        your model is.

        Cheers,
        -- Joe
        ,------------------------------------------------------------------.
        | Joseph J. Strout Biocomputing -- The Salk Institute |
        | joe@... http://www.strout.net |
        `------------------------------------------------------------------'
      • R. Paul McCarty
        ... Sorry, that wasn t what I was poking at. ^_^ ... But your mental structure, your neurons and synapses, are constantly changing. Let s not even talk about
        Message 3 of 10 , Sep 1, 1999
        • 0 Attachment
          Joseph J. Strout wrote:
          >
          > It's simply a matter of personal identity (sigh... here we go again).

          Sorry, that wasn't what I was poking at. ^_^

          > Suffice to say that I believe my identity lies in mental structure,

          But your mental structure, your neurons and synapses, are constantly
          changing. Let's not even talk about electronics, what if I just added a
          few more folds to your cortex? The new cortical surface is free to use
          however you want.. it has no inherent function yet. Eventually you will
          grow into it. Conversely, what if you suffer a stroke and lose 10% of
          your brain. The human mind would take these things in stride. I don't
          see why it would fall apart as soon as you increase your brain size 100,
          200, 1000%. You at least follow my line of thinking, correct? even if
          you don't like this approach.

          > which in turn arises from the workings of the brain. No amount of
          > electronic augmentation (if such is even possible, except for very
          > limited sensory augmentation -- anything more than that seems
          > unlikely) will change that. If my brain dies, and it has not been
          > copied into a functionally equivalent duplicate, then *I* die.

          But you assume if you increase the substrate of your mind (through
          additional neural "stuff") that it stays put.. I don't think your
          identity is trapped in your 100billion neurons. It can grow and fill
          whatever computational system it can, that's why adding new neurons, or
          old neurons dying doesn't change you significantly.

          > Now, it's possible you could make a functionally equivalent model of
          > the brain at a slightly higher level than the compartmental model --
          > an elaborate kernel-based model might do, for example.

          sure, but regardless of the level you choose your model has to be close
          to 100% accurate at predicting the activity of efferent/afferent signals
          in and out of the brain.. my point is I don't think you need to *know*
          how the brain works.. you just need a means for the mind to grow and
          fill new computational structures beyond the human brain.

          I'm sorry if we're coming back to the personal identity debate again..
          I'm not trying to.

          -Paul
          --
          R. Paul McCarty / rpmc@... / x52059
          317 Lattimore Hall, University of Rochester, Rochester, NY 14627
          Life is nothing if you're not obsessed. -Pecker
        • Joseph J. Strout
          ... Yes, but slowly. My identity changes slowly too -- slowly enough that I m still mostly the same person I was many years ago. ... In the latter case,
          Message 4 of 10 , Sep 1, 1999
          • 0 Attachment
            At 3:59 PM +0000 09/01/99, R. Paul McCarty wrote:

            >But your mental structure, your neurons and synapses, are constantly
            >changing.

            Yes, but slowly. My identity changes slowly too -- slowly enough
            that I'm still mostly the same person I was many years ago.

            > Let's not even talk about electronics, what if I just added a
            >few more folds to your cortex? The new cortical surface is free to use
            >however you want.. it has no inherent function yet. Eventually you will
            >grow into it. Conversely, what if you suffer a stroke and lose 10% of
            >your brain. The human mind would take these things in stride.

            In the latter case, depending on which 10% it was, I could lose my
            identity to a significant degree. I wouldn't consider that "in
            stride". As for hypothetically adding more cortex: there's no
            reason, in particular, to suppose that this would ever get used for
            anything. Cortex isn't just "stuff" that your brain uses to think;
            it's divided into dozens of specialized processing centers, each
            doing specialized things by virtue of its inputs, outputs, and
            internal structure. That's why I doubt we'll be able to augment a
            biological brain in any significant way, whether by electronics or by
            somehow cramming more neurons into the brain.

            >But you assume if you increase the substrate of your mind (through
            >additional neural "stuff") that it stays put..

            See, actually I'm assuming (asserting?) that there is no "substrate
            of the mind" or "neural stuff". What your proposing is vaguely like
            throwing opening your radio, cramming in a random jumble of
            electronic components, and expecting it to play music better or lose
            its reliance on the original components.

            > I don't think your
            >identity is trapped in your 100billion neurons. It can grow and fill
            >whatever computational system it can, that's why adding new neurons, or
            >old neurons dying doesn't change you significantly.

            Nope, I just don't think so. When new neurons are added (which
            doesn't happen very much, BTW), they grow in specific places and make
            specific connections with other parts of the brain. When neurons
            die, functionality is lost -- this is very gradual, but undeniable.

            My identity isn't just in any old 100 billion neurons -- it's in many
            many specific ones. Habits in my cerebellum, object recognition in
            temporal cortex, recurring dreams maybe in the thalamocortical loops,
            plans in the frontal cortex -- we don't actually know most of how
            this works, but it's clear that there's functional specialization,
            and it's definately not true that more cortex here makes up for less
            cortex over there.


            >sure, but regardless of the level you choose your model has to be close
            >to 100% accurate at predicting the activity of efferent/afferent signals
            >in and out of the brain..

            Yes, it must reproduce them exactly.

            > my point is I don't think you need to *know*
            >how the brain works.. you just need a means for the mind to grow and
            >fill new computational structures beyond the human brain.

            Well, suit yourself. But I want to live as me. :)

            >I'm sorry if we're coming back to the personal identity debate again..
            >I'm not trying to.

            I guess it's unavoidable. But if we try, maybe we can stick to the
            neuroscience, as it does seem your theory is based on some
            questionable assumptions about how the brain works.

            Cheers,
            -- Joe

            ,------------------------------------------------------------------.
            | Joseph J. Strout Biocomputing -- The Salk Institute |
            | joe@... http://www.strout.net |
            `------------------------------------------------------------------'
          • R. Paul McCarty
            ... That is in contrast to a large volume of research showing the contrary, that brain functions can be adopted by new structures after damage to the brain(
            Message 5 of 10 , Sep 1, 1999
            • 0 Attachment
              Joseph J. Strout wrote:
              >
              > At 3:59 PM +0000 09/01/99, R. Paul McCarty wrote:
              >
              > See, actually I'm assuming (asserting?) that there is no "substrate
              > of the mind" or "neural stuff". What your proposing is vaguely like
              > throwing opening your radio, cramming in a random jumble of
              > electronic components, and expecting it to play music better or lose
              > its reliance on the original components.

              That is in contrast to a large volume of research showing the contrary,
              that brain functions can be adopted by new structures after damage to
              the brain( such as after stroke), and that (at least during development)
              the cortex is easilly reorganized to accomodate the environment. For
              example there are several papers on how motor cortext is reorganized if
              you tape two fingers together, etc. Development work has been done, for
              example if you destroy the cortical regions associated with visual
              processing they will reform elsewhere. There seems to be alot of
              research pointing towards the possibility of integrating new neural
              tissue into our brains in a useful way.

              Unlike my transistor radio, my brain is constantly adapting to my
              environment. A better analogy would be grafting a tree branch into an
              existing tree.

              > > I don't think your
              > >identity is trapped in your 100billion neurons. It can grow and fill
              > >whatever computational system it can, that's why adding new neurons, or
              > >old neurons dying doesn't change you significantly.
              >
              > Nope, I just don't think so. When new neurons are added (which
              > doesn't happen very much, BTW), they grow in specific places and make
              > specific connections with other parts of the brain. When neurons
              > die, functionality is lost -- this is very gradual, but undeniable.

              yes, it's very gradual, but that doesn't mean you can't turn over
              neurons or add new structures no faster then this.

              > My identity isn't just in any old 100 billion neurons -- it's in many
              > many specific ones. Habits in my cerebellum, object recognition in
              > temporal cortex, recurring dreams maybe in the thalamocortical loops,
              > plans in the frontal cortex -- we don't actually know most of how
              > this works, but it's clear that there's functional specialization,
              > and it's definately not true that more cortex here makes up for less
              > cortex over there.

              Not to start a new thread here, but once uploaded does everyone imagine
              they will remain simple 100 billion neuron individuals? I think it's
              inevitable that once uploaded we will wish to add to ourselves and
              improve on our brains/minds to make ourselves smarter and to evolve as a
              new form of life. Is this such a radical idea? o.o

              > > my point is I don't think you need to *know*
              > >how the brain works.. you just need a means for the mind to grow and
              > >fill new computational structures beyond the human brain.
              >
              > Well, suit yourself. But I want to live as me. :)

              I just want to live. :)

              > But if we try, maybe we can stick to the
              > neuroscience, as it does seem your theory is based on some
              > questionable assumptions about how the brain works.
              >

              ouch, that hurts.. you know this M.S. degree I have in neuroscience
              didn't come in a cracker jack box. :)

              -Paul
              --
              R. Paul McCarty / rpmc@... / x52059
              317 Lattimore Hall, University of Rochester, Rochester, NY 14627
              Life is nothing if you're not obsessed. -Pecker
            • Joseph J. Strout
              ... Is it? Seems to me the theory of equipotentiality was debunked decades ago. If you want to argue this point, I guess we ll have to start digging out
              Message 6 of 10 , Sep 1, 1999
              • 0 Attachment
                At 6:10 PM +0000 09/01/99, R. Paul McCarty wrote:

                > > See, actually I'm assuming (asserting?) that there is no "substrate
                > > of the mind" or "neural stuff".

                >That is in contrast to a large volume of research showing the contrary

                Is it? Seems to me the theory of equipotentiality was debunked
                decades ago. If you want to argue this point, I guess we'll have to
                start digging out references.

                >that brain functions can be adopted by new structures after damage to
                >the brain( such as after stroke), and that (at least during development)
                >the cortex is easilly reorganized to accomodate the environment.

                Do you know anyone who's had a stroke? Did they recover fully? If
                so, they were lucky, and this was probably not due to other neurons
                taking over the lost functions, but due to temporary deactivation &
                restoration of a small area (yes, that happens). What stroke victims
                and other brain-injury patients do is learn coping mechanisms. It's
                just like a blind person learning to get around by hearing and touch;
                they learn to cope, but it doesn't mean they can see.

                In early development, things are more plastic in certain regions, but
                not much. If you have damage to your speech centers, you'll never
                have natural speech, no matter how healthy the rest of your brain is.
                If damage to visual cortex, you'll never learn normal vision. If
                damage to your inferotemporal cortex, you'll form proper long-term
                memories -- just ask H.M. (and ask him again the next day; he won't
                mind, as he'll have forgotten you by then).

                > For
                >example there are several papers on how motor cortext is reorganized if
                >you tape two fingers together, etc.

                Ah yes, that's true -- the amount of sensorimotor cortex *apparently*
                used for control or activated by stimulation can be changed by
                various manipulations, and on an extremely fast time scale (seconds)
                -- much faster than neurons could possibly be reorganizing. Probably
                what's going on here is that the connections from, say, your left
                ring finger are actually distributed much more widely than is usually
                apparent, but the fringe activity is suppressed through competition.
                When the competition is removed, the activity grows to encompass a
                wider area, and we think the cortex has reorganized. But you won't
                find the activity expanding into non-sensorimotor cortex.

                > Development work has been done, for
                >example if you destroy the cortical regions associated with visual
                >processing they will reform elsewhere.

                Um, no, they won't. If you destroy visual cortex the subject will be
                blind (except for some limited unconscious visual abilities due to
                the noncortical targets of the retina).

                > There seems to be alot of
                >research pointing towards the possibility of integrating new neural
                >tissue into our brains in a useful way.

                Well, there are people hoping to add neurons (derived from fetal or
                stem-cell tissue) to produce dopamine in Parkinson's patients. But
                these neurons aren't expected to do any information processing;
                they're basically just secretory organs.

                You could maybe also add neurons usefully to the hippocampus, which
                is continually growing and changing anyway.

                But that's about it, AFAIK.

                >Unlike my transistor radio, my brain is constantly adapting to my
                >environment. A better analogy would be grafting a tree branch into an
                >existing tree.

                True, the radio analogy was weak. But I'm not sure a tree is much
                better; unlike a tree, a brain is fundamentally an information
                processor. It's not enough that the graft live; that's easy. But to
                be useful (with the above exceptions), it must also connect up in
                very complex and specific ways with other parts of the brain.

                >Not to start a new thread here, but once uploaded does everyone imagine
                >they will remain simple 100 billion neuron individuals? I think it's
                >inevitable that once uploaded we will wish to add to ourselves and
                >improve on our brains/minds to make ourselves smarter and to evolve as a
                >new form of life. Is this such a radical idea? o.o

                No, I agree many people will want to do that and will proceed to do
                so. Once uploaded it'd be fairly straightforward. Of course, many
                of those experiments will end tragically. Personally, I'm content to
                just stay myself (sans mortality), at least for the next few
                centuries. But that's just me. :)

                >ouch, that hurts.. you know this M.S. degree I have in neuroscience
                >didn't come in a cracker jack box. :)

                Mine too. :) Sorry, didn't mean to offend. Well, we seem to have
                some real scientific issues to chew on here. Maybe we should start
                digging out specific papers and haggling over interpretations?

                (Don't expect me to have too much time for in it the next month or
                two, though; I have a three deadlines in the next several weeks.)

                Cheers,
                -- Joe

                ,------------------------------------------------------------------.
                | Joseph J. Strout Biocomputing -- The Salk Institute |
                | joe@... http://www.strout.net |
                `------------------------------------------------------------------'
              • Eugene Leitl
                ... We re making our lives easier, because we already have a clue how to model biologically realistic neurons. We don t at all know how to achieve a more
                Message 7 of 10 , Sep 1, 1999
                • 0 Attachment
                  R. Paul McCarty writes:
                  > I've seen frequent mention here of biologically realistic models of
                  > neuron activity (sophisticated compartmental models, etc.). It seems to
                  > me once we've made the assumption that the underlying computations which
                  > result in our conscious selves need not be tied to biological neurons
                  > we're just making our life more difficult be attempting to model neuron
                  > activity precisely. The only reason to do it would appear to be to

                  We're making our lives easier, because we already have a clue how to
                  model biologically realistic neurons. We don't at all know how to
                  achieve a more compact encoding while preserving the identity.

                  > capture the current behavior of the human brain at the lowest necessary
                  > level and hope all the other stuff is there. This seems good, but
                  > inherently difficult and uncertain.. we really wont know if it works
                  > until we copy a whole individual and see how they do.. I suppose we have
                  > to do this anyway, but it would be quite disappointing to spend your
                  > whole life working on this and discover it doesn't work until you model
                  > x,y and z.
                • Keith Wiley
                  ... The trend in this line of conversation stemmed from whether or not new neural material can be integrated into the brain. Are you saying you don t believe
                  Message 8 of 10 , Sep 1, 1999
                  • 0 Attachment
                    > >that brain functions can be adopted by new structures after damage to
                    > >the brain( such as after stroke), and that (at least during development)
                    > >the cortex is easilly reorganized to accomodate the environment.
                    >
                    > Do you know anyone who's had a stroke? Did they recover fully? If
                    > so, they were lucky, and this was probably not due to other neurons
                    > taking over the lost functions, but due to temporary deactivation &
                    > restoration of a small area (yes, that happens). What stroke victims
                    > and other brain-injury patients do is learn coping mechanisms. It's
                    > just like a blind person learning to get around by hearing and touch;
                    > they learn to cope, but it doesn't mean they can see.
                    >
                    > In early development, things are more plastic in certain regions, but
                    > not much. If you have damage to your speech centers, you'll never
                    > have natural speech, no matter how healthy the rest of your brain is.
                    > If damage to visual cortex, you'll never learn normal vision. If
                    > damage to your inferotemporal cortex, you'll form proper long-term
                    > memories -- just ask H.M. (and ask him again the next day; he won't
                    > mind, as he'll have forgotten you by then).

                    The trend in this line of conversation stemmed from whether or not new
                    neural material can be integrated into the brain. Are you saying you
                    don't believe this is possible at all? The human body's ability to
                    repair the brain is limited, but that isn't directly related to the
                    ability of a neural-net type of data structure to slowly absorb new
                    neural material that can then be accessed or used for neural-net type
                    functions. For example, like you said, the human brain doesn't
                    instantly fix stroke lesions. However, it seems possible that by
                    growing neurons and inserting them into the brain we could increase our
                    brain mass. After all evolution has been doing this with dramatic
                    success and the example of the primate brains and hominid -> human
                    evolutionary lines are positively explosive in the raw size of their
                    brains. This suggests to me that there is not only feasibility but also
                    an evolutionary tendency for our brains to keep growing, probably far
                    beyond present levels once we can get uploading going. In 100 years I
                    just want to be alive at all, but in 1000 years I hope my neural net has
                    trillions of nodes.

                    . . .. ... ..... ........ ............. .....................
                    .. ... ..... ....... ........... ............. .................
                    . .. .... ........ ................ ................................
                    Keith Wiley *
                    keithw@... * * * * * *
                    http://www.unm.edu/~keithw *** ** * * ** *
                    http://www.amp3.net/keithw (mp3 music) * ** ** ***
                  • Joseph J. Strout
                    ... At all is too strong -- see comments about dopamine-releasing neurons, possible use of new neurons by the hippocampus, etc. But, barring those
                    Message 9 of 10 , Sep 1, 1999
                    • 0 Attachment
                      At 3:38 PM -0600 09/01/99, Keith Wiley wrote:

                      >The trend in this line of conversation stemmed from whether or not new
                      >neural material can be integrated into the brain. Are you saying you
                      >don't believe this is possible at all?

                      "At all" is too strong -- see comments about dopamine-releasing
                      neurons, possible use of new neurons by the hippocampus, etc.

                      But, barring those exceptions, yes.

                      Of course, it's possible you could revert the brain (or parts of it)
                      to a developmental stage, with rapid growth & remodeling. But the
                      result would be (a correlate of) infant amnesia -- who can remember
                      being born? I'll pass, thanks.

                      > However, it seems possible that by
                      >growing neurons and inserting them into the brain we could increase our
                      >brain mass. After all evolution has been doing this with dramatic
                      >success and the example of the primate brains and hominid -> human
                      >evolutionary lines are positively explosive in the raw size of their
                      >brains.

                      Evolution doesn't conserve individuals. But there is a point there,
                      in that it could be that a few minor genetic changes basically tell
                      the brain when to stop growing, and as that time was delayed for our
                      evolutionary line, we just had extra cortex lying around, which
                      self-organized into doing useful things like math and writing COBOL.
                      The argument against this notion is that, if our brains just
                      self-organized, each brain would end up different -- and they don't,
                      not in gross details. Your language areas are in roughly the same
                      place as mine (well, maybe not mine since I'm a lefty, but you get
                      the idea).

                      > This suggests to me that there is not only feasibility but also
                      >an evolutionary tendency for our brains to keep growing, probably far
                      >beyond present levels once we can get uploading going. In 100 years I
                      >just want to be alive at all, but in 1000 years I hope my neural net has
                      >trillions of nodes.

                      Well (seems like I'm saying this a lot today), suit yourself. In
                      1000 years I just want to be still alive. I'm happy with the way I
                      am -- I just prefer not to die if I can help it.

                      Cheers,
                      -- Joe

                      ,------------------------------------------------------------------.
                      | Joseph J. Strout Biocomputing -- The Salk Institute |
                      | joe@... http://www.strout.net |
                      `------------------------------------------------------------------'
                    • R. Paul McCarty
                      ... This is getting away from what I meant to say. No, the cortex isn t equipotential in the developed brain, but it has the ability to reorganize itself to
                      Message 10 of 10 , Sep 1, 1999
                      • 0 Attachment
                        Joseph J. Strout wrote:
                        >
                        > At 6:10 PM +0000 09/01/99, R. Paul McCarty wrote:
                        >
                        > > > See, actually I'm assuming (asserting?) that there is no "substrate
                        > > > of the mind" or "neural stuff".
                        >
                        > >That is in contrast to a large volume of research showing the contrary
                        >
                        > Is it? Seems to me the theory of equipotentiality was debunked
                        > decades ago. If you want to argue this point, I guess we'll have to
                        > start digging out references.

                        This is getting away from what I meant to say. No, the cortex isn't
                        equipotential in the developed brain, but it has the ability to
                        reorganize itself to some extent to compensate for changes in peripheral
                        innervation (limb loss) and damage. How adaptive varies and I think is
                        largely unknown and may be altered with neurotophic factors. The motor
                        cortex seems to be the most flexible allowing complete reorganiztion to
                        process changes in the skin and muscles. And there during development
                        the cortex appears to be /almost/ equipotential allowing even visual
                        cortex to develop occular dominance columns in other areas of cortex
                        (I'm trying to track down this reference). And the ability(or minimal
                        ability) to reorganize may not indicate that *new* structures and input
                        can not be adapted to and incorporated.

                        a quick link on somatosensory reorganization:
                        http://www.psychologie.hu-berlin.de/klieng/klifo51.htm

                        Reorganizing finger maps: Michael M. Merzenich et al, Neuroscience
                        10:639-665 (1983); Annual Reviews of Neuroscience 6:325-356 (1983);
                        Journal of Comparative Neurology 224:591-605 (1984); Journal of
                        Neuroscience 6:218-233 (1986); Nature 332:444-445 (31 March 1988). For a
                        recent review, see J. T. Wall, "Variable organization in cortical maps
                        of the skin as an indication of lifelong adaptive capacities of circuits
                        in the mammalian brain," Trends in Neurosciences 11(12):549-557
                        (December 1988).

                        >
                        > >that brain functions can be adopted by new structures after damage to
                        > >the brain( such as after stroke), and that (at least during development)
                        > >the cortex is easilly reorganized to accomodate the environment.
                        >
                        > Do you know anyone who's had a stroke? Did they recover fully?

                        I've only heard anecdotal evidence about stroke recovery, like people
                        who couldn't walk, relearning how to walk, speak, etc. Whether complete
                        recover is possible is irrelevant, it is just evidence to suggest the
                        human brain has the ability to adapt to changing structure.

                        > Ah yes, that's true -- the amount of sensorimotor cortex *apparently*
                        > used for control or activated by stimulation can be changed by
                        > various manipulations, and on an extremely fast time scale (seconds)
                        > -- much faster than neurons could possibly be reorganizing. Probably
                        > what's going on here is that the connections from, say, your left
                        > ring finger are actually distributed much more widely than is usually
                        > apparent, but the fringe activity is suppressed through competition.
                        > When the competition is removed, the activity grows to encompass a
                        > wider area, and we think the cortex has reorganized. But you won't
                        > find the activity expanding into non-sensorimotor cortex.

                        Well there wouldn't be any point would there? from the brain's
                        perspective the missing or changed limb sensations are needed downstream
                        for motor activity and sensation, if things change, it needs the new
                        information. There would be no point in taking over other areas of
                        cortex in the developed brain. However what happens with *new* cortical
                        structres?

                        >
                        > > Development work has been done, for
                        > >example if you destroy the cortical regions associated with visual
                        > >processing they will reform elsewhere.
                        >
                        > Um, no, they won't. If you destroy visual cortex the subject will be
                        > blind (except for some limited unconscious visual abilities due to
                        > the noncortical targets of the retina).

                        In developed individuals yes, since the optic nerve isn't plastic enough
                        to reinervate a new piece of cortex, but, during development I believe
                        they have shown it can change.

                        > Well, there are people hoping to add neurons (derived from fetal or
                        > stem-cell tissue) to produce dopamine in Parkinson's patients. But
                        > these neurons aren't expected to do any information processing;
                        > they're basically just secretory organs.

                        They are already experimenting with adding new neurons to stroke
                        patients to recover damaged brain function, so I'm not the only one
                        crazy enough to think this will work. ^_^

                        -Paul
                        --
                        R. Paul McCarty / rpmc@... / x52059
                        317 Lattimore Hall, University of Rochester, Rochester, NY 14627
                        Life is nothing if you're not obsessed. -Pecker
                      Your message has been successfully submitted and would be delivered to recipients shortly.