Working memory localization in hippocampus

Sturla Molden stumol at
Fri Mar 13 06:49:11 EST 1998

On 13 Mar 1998 03:53:38 GMT, flefever at Frank LeFever)

>One big problem in this area is the diversity of definitions of
>"working" memory.
>Using at least some (operational as well as explicit) definitions, I
>believe oone can argue that the hippocampus is NOT very important for
>Working Memory.  Part of the difference may lie in the organisms and
>paradigms used (e.g. rats and rat tasks vs. humans and human tasks).

The claim that the hippocampus and surrounding temporal-lobe
structures are involved in episodic and/or spatial working memory is
not to say that these are the ony structures needed for this type of
memory (cf. patient NA).

However, bilateral hippocampectomy disrupts all forms of memory
lasting for more than 15 seconds and containing more than a dozen
items; i.e. the structure is needed for all forms of explicite memory
that requires plastic rather than dynamic changes in the nervous

Rawlins claim that the hippocampus was a temporary memory store was
based on such findings as the abolished partial reinforcement
extinction effect in rats with hippocampal lesions when between-trial
interwals were long but not when they were short (eg less than 10

>Soomeone (I think it may have been Joachim Fuster) said the better term
>might be "working WITH memory".  Fuster's work (monkeys again) has
>shown the importance of the prefontal cortex of selecting/activating
>specific regions of posterior cortex so as to make not just transient
>memory but long-term memory NEEDED FOR PERFORMANCE OF A TASK
>temporarily available (and, I would argue, dominant in competition with
>other memories and/or ambient stimuli).  Accordingly, "working with"
>the relevant memories can be disrupted by local cooling of either the
>relevant frontal region or the relevant posterior region (e.g. a
>particular part of parietal cortex).  (Again, easy to locate via
>In human clinical neuropsychology, performance of tasks with relatively
>little "intellectual" demands but strong "working memory" demands (e.g.
>Trails B) is impaired with frontal dysfunction (cortical or

There can be little dubt that other structures such as the prefrontal
cortex are needed for "intellectual functions" (as a biologist I try
to avoid these mentalistic terms); but hippocampal lesions
nevertheless disrupt the animals performance in working memory
dependent tasks (and this is not due to perceptual or motorical
dysfunctions). The exact function of the hippocampus remain to be
discovered, but it is clearly essential for the operation of certain
forms of memories (those that have a spatial component).

Sturla Molden

More information about the Neur-sci mailing list