I wonder if there are books/articles that would provide evidence on this question: Do working and semantic memory affect speech representation in a different way?
Semantic memory requires only a network in which spreading activation
can take place. For an association between two items only a connection
is needed to establish a relationship. Such networks can support word choice
and represent context when listening to language.
A visualization can be seen here.
http://www.newscientist.com/article/mg21728984.400-take-a-peek-inside-the-brains-filing-cabinet.html
http://gallantlab.org/semanticmovies/
http://www.plosbiology.org/article/info:doi/10.1371/journal.pbio.0060159
http://www.eurekalert.org/pub_releases/2007-10/plos-nip102407.php
http://www.eurekalert.org/pub_releases/2009-08/sicp-wts081109.php
Working memory, on the other hand, must be able to store tuples
of symbols. A, B, B, A is different from A, B, A, B. This can't be
explained by exciting the symbols A and B, because there must be
order and repetition. One way of explaining working memory is: WM is
based on a more general ability to store patterns. E.g. feedback
networks like Hopfield nets can do this.
http://www.eurekalert.org/pub_releases/2004-05/asu-aac051004.php
http://arxiv.org/abs/1007.3122
Newer research points to synchronized brainwaves as
the underlying glue of working memory.
http://www.eurekalert.org/pub_releases/2012-11/niom-ibw110212.php
http://www.nature.com/neuro/journal/v15/n3/full/nn.3038.html
Regards,
Joachim
Semantic memory and working memory as terms originate from two different historical streams in learning and memory research.
Semantic memory was contrasted to episodic memory in Tulving's seminal work in the early 70s at the University of Toronto. Semantic memory content is that which is deeply embedded within the memory association networks and can be accessed through numerous associations. Episodic is information organized in a temporal fashion and associated with all the things that happened during that event, but is not yet fully integrated with an individual's knowledge schema and therefore not yet part of semantic memory. As can be guessed, most of Tulving's memory research dealt with students learning lists of words.
Semantic and episodic were one of memory two types of memory theories that became prevalent, and grew out of the earlier distinction of short term and long term memories.
Two types of memory theories were the easiest to evolve due to simple experiments that would focus upon the nature of one type of memory and its qualities, versus memories that were not this type.
Working memory is contrasted to Reference memory (again evolving out of the short term versus long term memory distinction). Working memory versus reference memory grew out of a specific rodent learning task known as the radial-arm maze developed by David Olton at John Hopkins in the mid seventies, In which animals learning to run down arms of a maze that radiate out from a center point to forage for food at the ends of the arms. If the maze has 8 arms and 4 arms are always baited and 4 are always not. Then a reference memory error is one where the rat would go down an always unbiated arm. A working memory error would occur if the rat obtained the bait from an arm and then later during the same trial ran down that arm again looking for food it had already obtained.
Once these memory theories were in place regardless, of their origins researchers would further abstract their concepts and begin to contrast them among different theories to develop a better overall schema of memory.
People, who recently (within the last decade) have tried to focus on overall encompassing memory theories have been Larry Squire of UCSD and Howard Eichenbaum of Boston University.
So it is fine to take older terms from originally separate memory theories and contrast them, as a new mental exercise that may bring new facets of memory to light, but it should also be understood that our theorizing and testing of memory may or may not have any bearing of how memory is encoded or represented or organized in the Brian. Thus, memory research of the mind can be totally independent of memory mechanisms of the brain. All depends upon the goals of the research.
For more details and references on learning and memory, particularly David Olton's radial- maze and reference working memory distinctions can visit my website and view the article on the hippocampus.
WWW.PageoNeuroPlasticity.com
- Badges
- Science topic