Connectionists: how the brain works?

Wed Mar 19 18:05:24 EDT 2014

As I have discussed (some would say ad nauseam) on this list serve before, Ockham’s razor was a terrible heuristic for planetary science for about 1500 years, not with respect to replicating the position of the planets in the night sky (big data ‘prediction’ problem as it would currently be called now) but instead as a model to actually understand planetary motion, which became the foundation for mechanics and modern physics.  thank heavens that Kepler (who was predisposed) and Newton (who was not) didn’t apply the same argument below (actually Copernicus sort of did ironically enough).

I believe that there is a very important lesson there for those of us who are actually trying to figure out how the brain works.  Simplifying for the sake of it is even less appropriate in a physical system (the brain or any biological system) that explicitly uses order and complexity to do what it does (mess with the second law).

The details matter, and the details are all we really have.  You can invent any kind of model you want to ‘predict’ (as I have said previously actually usually ‘postdict’) the data - however, history suggests and in my own experience it won’t get you very far in figuring out how things actually work.  In fact, models of this sort  distort the field and have resulting in years of running down blind alleys. 

Old expression (I didn’t make up) "in biology Ockam’s Razor will cut your throat".

For those of you interested, I just posted an essay on this general subject recently written for a publication whose editors decided not to publish:

http://jamesmbower.com/blog/?p=119

feel free to take a look - but I warn most of you - it will probably make your teeth grind.

but this:  "we throw as much data from as many levels of analysis as we can come up with into a big pot and then construct a theory”  makes my blood boil for the outright stupidity.  Why create an unstructed mess,  when you actually have structure to look at - as complex as it is

THAT SHOULD TELL YOU SOMETHING>

And from long experience in building models of the nervous system, the precise reason you include the details is because you don’t know what doesn’t matter.  And, in fact, remarkably enough, so far, they all have in completely unexpected ways.

That, of course, is the point.  Biology works because of the details - selection is precisely a mechanism that selects for the details, just as the rather remarkable system of biological coding preserves them and replicates them.  

If someone wants to declare that all parts of the brain basically work the same way, they are so far off the mark it is hardly worth responding - although, of course, I keep doing so (tilting at windmills)

Perhaps the unindoctrinated interested in these subjects will at least spend the minimal amount of time necessary to understand something about the remarkably fine and detailed structure of the brain - WHICH IS THERE FOR A REASON.

Arggg

jim

On Mar 19, 2014, at 3:07 PM, Brian J Mingus <brian.mingus at Colorado.EDU> wrote:

> Hi Jim,
> 
> Focusing too much on the details is risky in and of itself. Optimal compression requires a balance, and we can't compute what that balance is (all models are wrong). One thing we can say for sure is that we should err on the side of simplicity, and adding detail to theories before simpler explanations have failed is not Ockham's heuristic. That said it's still in the space of a Big Data fuzzy science approach, where we throw as much data from as many levels of analysis as we can come up with into a big pot and then construct a theory. The thing to keep in mind is that when we start pruning this model most of the details are going to disappear, because almost all of them are irrelevant. Indeed, the size of the description that includes all the details is almost infinite, whereas the length of the description that explains almost all the variance is extremely short, especially in comparison. This is why Ockham's razor is a good heuristic. It helps prevent us from wasting time on unnecessary details by suggesting that we only inquire as to the details once our existing simpler theory has failed to work.
> 
> Brian
> 
> 
> On Wed, Mar 19, 2014 at 12:42 PM, james bower <bower at uthscsa.edu> wrote:
> Actually, the previous statement is only true in its most abstract form -which in that form also applies to the heart, the kidney and trees too.  So not sure what use that is.  (trees used cellular based communication to react to predation by insects - and at least mine look like they are in pain when they do so).
> 
> 
> the further statement about similar developmental processes for cortical like brain structures is also only true in its most abstract sense.  In particular, the cerebellum has a quite unique form of cortical development (very different from the frontal cortical structures.  cell migration patterns, the way cellular components get connected, as well as general timing - all of which are almost certainly important to its function.  The cerebellum, for example, largely develops entirely postnatally in most mammals.  It is also important to note that cerebellar development is also considerably better understood than is the case for cerebral cortex.
> 
> Again, as I have argued many times before - in biology (perhaps unfortunately) the devil (and therefore the computation) is in the details.  Gloss over them at your risk.
> 
> Jim
> 
> 
> 
> 
> 
> On Mar 19, 2014, at 12:50 PM, Juyang Weng <weng at cse.msu.edu> wrote:
> 
> > Mike,
> >
> > Yes, they are very different in the signals they receive and process after at least several months' development prenatally, but this is
> > not a sufficiently deep causality for us to truly understand how the brain works.  Cerebral cortex, hippocampus and cerebellum are all very similar in the mechanisms that enable them to develop into what they are, prenatally and postnatally.
> >
> > An intuitive way to think of this deeper causality is: Development is cell-based.  The same set of cell properties enables cells to migrate, connect and form cerebral cortex, hippocampus and cerebellum while each cell taking signals from other cells.
> >
> > -John
> >
> > On 3/14/14 3:40 PM, Michael Arbib wrote:
> >> At 11:17 AM 3/14/2014, Juyang Weng wrote:
> >>> The brain uses a single architecture to do all brain functions we are aware of!  It uses the same architecture to do vision, audition, motor, reasoning, decision making, motivation (including pain avoidance and pleasure seeking, novelty seeking, higher emotion, etc.).
> >>
> >> Gosh -- and I thought cerebral cortex, hippocampus and cerebellum were very different from each other.
> >>
> >
> > --
> > --
> > Juyang (John) Weng, Professor
> > Department of Computer Science and Engineering
> > MSU Cognitive Science Program and MSU Neuroscience Program
> > 428 S Shaw Ln Rm 3115
> > Michigan State University
> > East Lansing, MI 48824 USA
> > Tel: 517-353-4388
> > Fax: 517-432-1061
> > Email: weng at cse.msu.edu
> > URL: http://www.cse.msu.edu/~weng/
> > ----------------------------------------------
> >
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20140319/aceaa766/attachment.html>