Do sentences with unaccusative verbs involve syntactic movement? Evidence from neuroimaging

Abstract

This study focuses on the neural processing of English sentences containing unergative, unaccusative and transitive verbs. We demonstrate common responses in bilateral superior temporal gyri in response to listening to sentences containing unaccusative and transitive verbs compared to unergative verbs; we did not detect any activation that was specific to unaccusatives. Our findings indicate that the neural processing of unaccusative and transitive verbs is highly similar, and very different from the processing of unergative verbs. We discuss the consequences of these results for the linguistic analysis of movement phenomena.

Keywords: fMRI; language; syntax; temporal cortex.

Figure 1.. External and internal Merge.

Figure 2.. Passive perception of speech is associated with widespread activity in bilateral dorsolateral temporal lobes.

A comparison of perception of all sentences compared to a rest condition was associated with significant activity in bilateral dorsolateral temporal cortices, extending from posterior to middle/ anterior STG and inferior frontal gyrus in the left hemisphere (a). A comparison of all speech perception conditions with a high-level baseline of (unintelligible) rotated speech revealed significant activity in bilateral mid/anterior STG in both hemispheres, premotor cortex and extra-striate visual areas in the left hemisphere (b). Activity in STG in this contrast spread anteriorly compared to the speech vs. rest condition.

Figure 3.. Perception of sentences with unaccusative and transitive verbs is associated with activity in similar areas, compared to perception of sentences with unergative verbs.

For the comparison of sentences with transitive and unaccusative verbs, mean parameter estimates were extracted from spherical regions of interest of 5-mm radius, taken from peak coordinates from the null conjunction of (unaccusative vs. unergative) and (transitive vs. unergatives). The null conjunction revealed peak coordinates in superior temporal gyri in both hemispheres (p < 0.005 uncorrected) shown in (c). The lines denote the cortical activations during perception of unaccusative vs. unergative and transitive vs. unergative. Both peaks lay in superior mid-temporal gyri. Parameter estimates for three contrasts are plotted for both of these region of interest (ROI)s. In both regions there was greater activity during perception of unaccusative and transitive sentences.

Figure 4.. The syntactic structures projected from unergative, transitive and unaccusative verbs; blue arrows indicate θ-marking relations.