
Full text loading...
A hallmark of human language is that we combine lexical building blocks retrieved from memory in endless new ways. This combinatorial aspect of language is referred to as unification. Here we focus on the neurobiological infrastructure for syntactic and semantic unification. Unification is characterized by a high-speed temporal profile including both prediction and integration of retrieved lexical elements. A meta-analysis of numerous neuroimaging studies reveals a clear dorsal/ventral gradient in both left inferior frontal cortex and left posterior temporal cortex, with dorsal foci for syntactic processing and ventral foci for semantic processing. In addition to core areas for unification, further networks need to be recruited to realize language-driven communication to its full extent. One example is the theory of mind network, which allows listeners and readers to infer the intended message (speaker meaning) from the coded meaning of the linguistic utterance. This indicates that sensorimotor simulation cannot handle all of language processing.
Article metrics loading...
Full text loading...
Data & Media loading...
Download Supplemental Figures 1-13 as a PDF (also reproduced below).
Download Captions for Supplemental Tables 1-4 (PDF).
Download Supplemental Table 1 (XLSX).
Download Supplemental Table 2 (XLSX).
Download Supplemental Table 3 (XLSX).
Download Supplemental Table 4 (XLSX).
Supplemental Figure 1. The adaptation effect for syntax repetition: in left MTG, LIFG, and supplementary motor area, there was a repetition suppression effect for repeated compared with novel syntactic structures, not only for repetitions within modality (production-production; comprehension-comprehension) but also between modalities (production-comprehension; comprehension-production). For details, see Segaert et al., 2012.
Supplemental Figure 2. Schematic representation of the brain showing regions with reliably reported activations for sentences compared to non-sentential stimuli (top) and sentences compared to words (bottom). The left posterior inferior frontal gyrus is further subdivided into Brodmann areas 44 (above black line), BA45 (below black line, above AC-PC line) and BA47 (below AC-PC line). Green regions indicate a reliable number of reports. Numbers in dots refer to the number of studies reporting the corresponding region as activated. Pink regions indicate no reports in 53 studies. See contrasts 1 and 2 in Supplemental Table 2 for further details. See columns 1 and 2 in Supplemental Table 1 for the studies entered in the two contrasts. (AC = anterior commissure, PC = posterior commissure)
Supplemental Figure 3. Reliably reported activations for sentence reading (top) and sentence listening (bottom) compared to non-sentential stimuli. See legend to Figure 2 and contrasts 3 and 4 in Supplemental Table 2 for further details. See columns 3 and 4 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 4. Reliably reported activations for passive sentence reading (top) and sentence listening (bottom) compared to non-sentential stimuli. See legend to Supplemental Figure 2 and contrasts 5 and 6 in Supplemental Table 2 for further details. See columns 5 and 6 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 5. Reliably reported activations for sentences with higher syntactic (top) or semantic (bottom) processing demands (violations, ambiguities, or complexity) compared to less demanding sentences. See legend to Supplemental Figure 2 and contrasts 7 (Supplemental Table 3) and 15 (Supplemental Table 4) for further details. See columns 7 and 15 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 6. Reliably reported activations for sentences with higher syntactic demands compared to sentences with higher semantic processing demands (top) and for sentences with higher semantic processing demands compared to sentences with higher syntactic processing demands (bottom). See legend to Supplemental Figure 2 and contrasts 8 (Supplemental Table 3) and 16 (Supplemental Table 4) for further details. See columns 8 and 16 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 7. Reliably reported activations for sentences with syntactic (top) or semantic (bottom) violations compared to correct sentences. See legend to Supplemental Figure 2 and contrasts 9 (Supplemental Table 3) and 17 (Supplemental Table 4) for further details. See columns 9 and 17 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 8. Reliably reported activations for sentences with different kinds of semantic violation. See legend to Supplemental Figure 2 and contrasts 18 and 19 in Supplemental Table 4 for further details. See columns 18 and 19 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 9. Reliably reported activations for sentences with syntactic (top) or semantic (bottom) ambiguities compared to unambiguous sentences. See legend to Supplemental Figure 2 and contrasts 10 (Supplemental Table 3) and 20 (Supplemental Table 4) for further details. See columns 10 and 20 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 10. Reliably reported activations for syntactically (top) or semantically (bottom) complex sentences compared to simpler sentences. See legend to Supplemental Figure 2 and contrasts 11 (Supplemental Table 3) and 21 (Supplemental Table 4) for further details. See columns 11 and 21 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 11. Reliably reported activations for sentences with different kinds of syntactic complexity manipulation. See legend to Supplemental Figure 2 and contrasts 12 and 13 in Supplemental Table 3 for further details. See columns 12 and 13 in Supplemental Table 1 for the studies entered in the two contrasts.
Supplemental Figure 12. Reliably reported activations for sentences with different kinds of semantic complexity manipulation. See legend to Supplemental Figure 2 and contrasts 22, 25, and 26 in Supplemental Table 4 for further details. See columns 22, 25, and 26 in Supplemental Table 1 for the studies entered in the three contrasts.
Supplemental Figure 13. Voxel-based analysis of 31 contrasts from 28 studies reporting posterior temporal cortex activation (y-coordinate ≤ -36) for syntactically or semantically demanding sentences compared to less demanding sentences. The left posterior temporal and frontal activation foci (y and z-coordinates of the single most significant activation maximum of each study) of this set of contrasts are projected onto the 51 mm sagittal plane of Talairach and Tournoux (1988). The centers of the ellipses represent the mean coordinates of the local maxima, the radii represent the standard deviations of the distance between the local maxima and their means. The mean z-coordinates of posterior temporal activations of semantically demanding sentences (orange ellipse, mean z = 0.9, SD = 8.6) were significantly more ventral than were those of syntactically demanding sentences (blue ellipse, mean z = 10.2, SD = 5.7; independent t-test, df = 29, t = 3.47 , p = 0.002, two-sided). There was no significant difference in the y-dimension (t < 1). The mean posterior IFG activations of semantically demanding sentences were significantly more ventral and rostral than were those of syntactically demanding sentences (semantically demanding sentences: mean z = 4.2, SD = 10.8, mean y = 22.2, SD = 6.3; syntactically demanding sentences: mean z = 15.0, SD = 5.6, mean y = 14.8, SD = 8.0; independent t-test on z-coordinates, df = 26, t = 2.72 , p = 0.011, two-sided; independent t-test on y-coordinates, df = 26, t = 3.396 , p = 0.002, two-sided).
(GFm, GFi = middle and inferior frontal gyri; BA = Brodmann area; GTs, GTm, GTi = superior, middle and inferior temporal gyri; STS, ITS = superior and inferior temporal sulci; Gsm = supramarginal gyrus)