ROIs were therefore taken from the literature and effects of lexical category or semantics were investigated by two-way ANOVAs. Previous work targeting both lexical category differences (noun–verb) and semantic dissociations (living–nonliving, animals–tools, etc.) was exploited in defining ROIs (Bedny et al., 2008, Chao et al., 1999, Martin and Chao, 2001 and Martin and Weisberg, 2003; see Vigliocco et al. 2011). Bedny et al. (2008) reported a significant effect of lexical category but NOT of the only semantic variable they focused on, motion—related semantic word properties. This lexical category effect was seen in three ROIs, click here where verbs evoked greater activity than nouns: left temperoparietal junction (TPJ: coordinates −58,
−48, 22), superior temporal sulcus (STS: −57, −55, 12) and anterior superior temporal sulcus (aSTS: −57, −41, −1). However, using the same ROIs to scrutinise the present data set, we could not observe any concordant significant effect, and, more generally, not any main effect or interaction of either Lexical category or Semantics (all F-values <0.5). Their left STS ROI revealed a trend towards a lexical category difference which, though weak and far below significance (F(1, 17) = .422, p > .525), somewhat resembled that reported by Bedny, with numerically greater activity for verbs
(see Fig. 2, Part A). No significant effect of lexical category appeared in either the left temperoparietal junction ROI (F(1, 17) = .400,
p > .536) or the left SCH 900776 solubility dmso anterior superior temporal sulcus ROI (F(1, 17) = .105, p > .750); in these cases, any numerical differences in favour of verbs were also absent in our data, in favour of a numerical contrast in the opposite direction. The combination of all three Bedny et al. regions (TPJ, STS and aSTS) also failed to reveal a significant effect of lexical category or semantics. Although, PD184352 (CI-1040) in our present analysis, activation maxima did not arise in the left STS in the contrast of all experimental words against baseline, we chose two coordinates located in the cluster of STS activation which were closest to Bedny et al.’s original anterior and posterior STS regions (see Fig. 2B). These coordinates, too, failed to replicate the verb advantage reported by Bedny and colleagues in left STS and showed no effect of lexical category. The present study was therefore unable to replicate the noun/verb difference in haemodynamic responses previously reported in left middle-temporal cortex. So far, analysis of all ROIs from the previous literature failed to reveal effects of lexical category. We did, however, observe a main effect of lexical category in analysis of two left frontal-insula regions (one more anterior at MNI coordinates −27, 33, 11, the other more posterior at −27, −3, 23) suggested by Martin et al.’s (1996) results of a positron emission tomography (PET) experiment investigating the naming of visually depicted animals and tools (F(1, 17) = 6.