Researchers believe that the prefrontal cortex acts as a dynamic filter for the brain. Dynamic filtering is selecting needed information for a current task from all the information streaming through the frontal cortex. This is why the prefrontal cortex acts as a dynamic filter, it must sort through the information and pick only that which is currently relevant.
Thompson-Schill et al. (1997) wanted to study the dynamic filtering hypothesis so they had subjects generate verbs associated with presented nouns. In other words, if the person saw a “cat” they might say “meow” or “nap” or something else. Thompson-Schill et al. assigned subjects to a high or low noun-verb selection condition. In the high condition, subjects were shown nouns with many associated verbs (e.g., a ball is shown and subject could produce “bounce,” “hit,” “kick,” or “throw”) whereas in the low condition subjects are shown a noun with only one (in most cases) related verb (e.g., a chair is shown and subject says “sit”).
They conducted this experiment to see if the inferior frontal cortex is associated with just semantic memory (basically long-term-memory-type information) or if it an area that supports working memory processes (retrieving information from semantic memory and working with it—in this case filtering through it for relevant associations). The experimenters found that the inferior frontal cortex (IFC) was more activated in the high-selection conditions than in the low-selection condition. If it had not been more activated then it would merely have been a semantic, long term, memory-related area. Because of the higher activation it was concluded that the IFC was associated with working memory, specifically pulling relevant information from semantic memory. It acts as a filtering mechanism, a dynamic filter.
To confirm this finding Thompson-Schill et al. (1998) conducted another study with brain-damaged patients. They selected subjects with lesions of the IFC. They found that in the high-selection condition these patients failed to produce any verbs 15% of the time. But in the low-selection condition these lesioned subjects performed the same as control subjects. The researchers concluded that because those with IFC lesions could not generate verbs to go along with displayed nouns when there were possibly many to choose between, the lesion caused a deficit in selection. It was not a semantic deficit, but a working memory one. They could not decide what verb to use and so they said nothing. This provided neuropsychological evidence for the IFC acting as a dynamic filter for the frontal cortex, at least as far as semantic information is concerned.
Thompson-Schill et al. (1999) also conducted another study where they looked at the temporal lobe in addition to the IFC. They replicated their previous experiment with one key difference. They had the subjects complete two generative trials (one of naming an action verb like in previous experiments and the other was naming an associated color) with the same list of nouns shown the second time for the subjects. Some repeated the first association task and others did the color one the second time. They found that IFC activation increased when the association task changed but temporal lobe activation decreased the second trial for both association conditions. Gazzaniga, Ivry, and Mangun (2002) sum it up best, “The fact that the decrease was observed is consistent with the idea that semantic attributes, be they relevant or irrelevant to the task at hand, were automatically activated upon presentation of the nouns” (p. 522). This is further evidence that the IFC acts as a dynamic filter for the frontal lobes.
[Note: Contact me if you would like the references cited in my post].