Igor Rubets wrote:
>> Alexander R Terrill wrote:
>> > Does anyone know what the role of the hippocampus in human memory? As
> > far as I can tell it somehow seems to make the conversion form STM to
> > LTM. Do we know how it does this, or why?
>> In 1962 Milner showed that STM was impaired in patients with hippocampal
> lesions.
> Lesions of Hip cause:
> - loss of passive evasive behavior though do not affect active evasive
> behavior (in some experiments it becomes even better);
> - also, re-education is almost impossible;
> - animals could not locate objects in space correctly;
> - long-term conditioning is entirely impossible.
> Thus Hip is really vital to the STM ---> LTM conversion.
Yes HM... but, it is =important= to be clear about the fact that the
hippocampus is not where the LTMs are stored... the hippocampi store
supersystem =configuration= information... externally-relevant
information is correlated with such configuration information, but only
via an indirect route... basically, the hippocampi function as TD
E/I-minimization "tools"... they blindly strive to achieve TD E/I(min),
and the supersystem configurations are produced as by-products of the
degrees to which TD E/I(min) is, in fact, achieved... it's the
configurations that the hippocampus imposes upon the supersystem that
"hold the sypersystem steady" which enable LTM consolidation... of
course, this "cascade" can be interfered with at many points... it's
been the case that the hippocampi have been viewed as the "seat of
memory" =only= because the hippocampi are so much more
compactly-localized than is the rest of the neural architecture that's
involved in the formation of LTM... if the hippocampi are lost, the
effect is more dramatic re. memory than is, say, if an equivalent volume
of non-primary neocortex is lost... but this doesn't mean that "memory"
is "in" the hippocampi...
...it's important to get these things straight because, if one doesn't
one is "thrown off the scent" with respect to the whole nervous system's
information-processing dynamics...
...I've worked out such matters for the whole CNS... I have an hypertext
doc that briefly outlines such... runs on MSDOS, Windows, Windows95
machines... gratis... ken collins
