Fig. 3. ROI results for pair wise comparisons in the LATL, activation averaged across subjects. On thewaveform plots, the shaded regions indicate that the difference in activity between the
showed that Complex number elicited reliable increases in activity in
BA38 [434–600 ms; p = .003], BA20 [437–600 ms; p = .01] and BA21
[400–600 ms; p = .01] (Fig. 3). Lastly, we also compared Color Modification
and Complex number composition directly. These tests did not reveal
any reliable differences between these conditions in either the earlier or
the later time window.
Thus both complex number formation and color modification modulated
the LATL, while numerical quantification did not affect it. When
the center of this effectwas visualized in the full brain contrasts, it localized
ventrally in the anterior temporal cortex for color modification,
while the center of the effectwas a bit more superior for complex number
composition (Fig. 3). Crucially though, both of these localizations are
within the range of previously reported effect centers in prior LATL composition
studies (Del Prato and Pylkkänen, 2014; Westerlund and
Pylkkänen, 2014). Thus, there is not yet any detailed characterization
as regards the precise locus of combinatory effects within the LATL,
and given the somewhat fuzzy spatial resolution of MEG, it is also unclear
whether such generalizations could be obtained with MEG.
The analyses on the AG aimed to test sensitivity for plurality in this
area. Therefore, conditions containing plural NPs (Color modification
and Numerical quantification) and singular NPs (Complex number
and Number list) were contrasted. The results did not show significant
differences for any of these contrasts. The vmPFC was reliably engaged
when composing a complex number in comparison to naming a number
list [458–600 ms; p = .01] but not when performing a numerical
quantification (we did not find any cluster of activity) or performing a
colormodification [186–223 ms; p=.22]. Lastly, no comparisons elicited
reliable clusters of activation in the LIFG (Additional Fig. 2).