Translation, Brains and the Computer by Bernard Scott

Author:Bernard Scott
Language: eng
Format: epub, pdf
Publisher: Springer International Publishing, Cham

5.2 Recent Views of the Cerebral Process

Recent neuroimaging-based studies in the past decade have begun to reveal a considerably more nuanced picture of the cortical language process. For one, brain areas hitherto not particularly known to be recruited for language are increasingly being recognized, areas such as the subcortical basal ganglia, the premotor cortex, and the middle temporal gyrus (see Postscript 4-C). And the sharp division between syntax in the Broca area and semantics in the hippocampus is giving way to a better understanding of their interaction, a recognition being arrived at even by the authors of the 2003 project described above.15

For example, in a subsequent study by Optiz and Friederici (2007), while confirming that complex violations involving long-distance dependencies and embeddings (called “hierarchical dependencies”) are handled by the Broca area, these researchers now propose that the hippocampus responds to “violations of local dependencies” (Optiz and Friederci 2007). In effect, the semantically-oriented hippocampus is now also seen to be making simple grammaticality judgments, i.e., judgments about syntax. This is clearly a new development.

A study by Hauser et al. (2012), and that includes Optiz as a co-author, again takes up the question of the two types of knowledge—rule-based versus association-based—and which of them best accounts for human ability to make grammaticality judgments.16 The object, again, is to investigate the interplay of these two knowledge types and the neuronal correlates that they recruit when engaged in grammaticality judgments. The study employs the BROCANTO artificial language, but with the difference that participants now are instructed in the underlying rules beforehand.

This 2012 study is still in the generativist spirit of the original 2003 BROCANTO study, but nevertheless offers some telling new insights. For one, they found that the two forms of knowledge—rule-based and similarity-based—operate in parallel and compete in processing efficiency. For another, they found that participants differ individually in their use of the two types of knowledge, a finding that documents a subjective factor in cerebral language processes.17

Not unexpectedly, individuals who rely on similarity are seen to recruit the hippocampus, whereas participants who rely on rules are seen to activate the left prefrontal areas (Broca) and the premotor cortex. In the case of similarity-prone participants, neuroimaging data show that after initial recruitment of the hippocampus, an activation shift occurs to the frontal area and premotor cortex. These theory-guided researchers still interpret this shift to mean that the word-specific combinations formed in the hippocampus are subsequently fed to areas believed to generate a syntactic grasp of these combinations. Syntactic knowledge, they still propose, is what enables correct grammaticality judgments.18 The assumption is still that the hippocampus is not capable of generating the kind of abstract knowledge requisite for grammaticality judgments because it is “item-specific,” too “superficial” therefore.19

A 2013 natural language study at the University of Iowa offers a considerably more robust view of the hippocampus (Kurczek et al. 2013). The investigation shows that impairments to the hippocampus disrupt an individual’s ability to relate pronouns to their antecedents, thus suggesting a distinctly grammatical, non-trivial syntactic role for this semantics-oriented reticulum.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.