In article <nRZy7.65160$My2.35990924 at news1.mntp1.il.home.com>, "Lance Sherman"
<lancesherman at home.com> writes:
>However, IMHO, he doesn't give enough importance to temporal issues - I
>think Hebb built on processing of sequences. Not just cell assemblies but
>"phase sequences". And I still think this is the best bet for modeling the
>operation of the brain.
>>I'll confess: tho i think Moby is somewhat in his own little world (like,
>I'm not?) I admire and agree with his commitment to processing temporal
>information. And Ron Blue, he is closer to my thinking, at least in his
>commitment to balance - as he calls it, "correlation opponent processing".
>Too bad these guys don't write as clearly as Calvin. Too bad I don't,
>either.
>>So, for those who think there is something to these time based ideas, try
>Peter Cariani's site, www.cariani.com. Neural timing nets - there's a big
>idea.
>
I would agree that temporal coding is often overlooked, but in my opinion, the
most inovative aspect about Cariani's ideas is that he's beginning to look at
all the ways neurons can code information in a parallel way. There's a figure
on his web site entitled "Three complementary types of neural codes" that
speaks to this. I differ with Cariani in that I believe that such coding is
overlapping and simultaneous. Much of the sensory integration that I study is
composed of at least a combination of two of the types listed (time-of-arrival,
spatio-temporal, channel, and latency). I'm currently working on tools to
evaluate temporal pattern coding independent of common stimulus input, so I
can't yet speak to that end.
Mike