Thursday, April 24, 2008
The search for the phonological store: From loop to convolution
Former TB West grad student, Brad Buchsbaum has found the phonological store. But don't try to punch the talairach coordinates into your NPS (Neural Positioning System), because it's not in any one place. Here's the background...
As early as our 2000 TICS paper (Hickok & Poeppel, 2000, TICS, 4:131-138), we have been saying that our concept of a sensory-motor based auditory dorsal stream is THE neural circuit that supports “phonological” working memory. Specifically, in 2000 we said, “We hypothesize that activations attributed to the functions of the phonological store reflect the operations of the proposed auditory-motor interface system... [this] is not the site of storage of phonemic representations per se ... but rather serves to interface sound-based representations of speech in auditory cortex with articulatory-based representations of speech in frontal cortex.” p. 136.
The idea is that the storage of phonological information is not contained in a dedicated buffer that lives in the parietal lobe (see any Smith/Jonides paper from the late 1990s), but rather that “storage” is nothing more than the active maintenance of the same phonological representations that are used in speech recognition, that these networks are in the STS, and that the dorsal sensory-motor circuit is the pathway that allows for frontal articulatory mechanisms to actively maintain these representations over a delay. This is essentially Fuster’s concept of working memory, i.e., an active state of “long term memory.”
This question of the relation between the proposed dorsal auditory stream and working memory was taken up by TB West alum Brad Buchsbaum. Brad and I did a handful of experiments which first identified area Spt and the circuit we now believe supports sensory-motor integration for speech and some aspects of music, and which supports verbal working memory. See Buchsbaum et al. 2001 (Cognitive Science, 25:663-78) and Hickok et al. 2003 (JoCN, 15: 673-82) for the first empirical studies on this topic. Brad has continued to push this issue with several more empirical papers. Check ‘em out, they’re all good.
But, here’s the real reason for this post: Brad has a new review/theory paper out in JoCN with Mark D’Esposito that does a really nice job of summarizing the history of the neural search for the phonological store, and developing the theory and empirical argument behind the link between phonological working memory and auditory-motor integration. The main claim is that the phonological store does not correspond to a single brain region, but emerges from the interaction of systems involved in speech perception and production. I like the way he thinks! Definitely a must-read, so check it out! Brad's an entertaining writer as well, so it's a fun read.
Buchsbaum & D’Esposito (2008). The search for the phonological store: From loop to convolution. J of Cog. Neurosci., 20:762-78.