Paired t -tests (with Bonferroni correction adjusting p to <.0125) confirmed that neural activity in pMTG was higher for high control compared to low control judgements following IFG stimulation (t =3.905, p =.001) with a near-significant effect of control demands in the scan following OP stimulation (t =1.897, p =.077).
Limitations We acknowledge that the behavioural results did not reproduce the previously reported pattern of selective disruption of high control semantic judgments ( Whitney et al., 2011 ); rather we found a main effect of TMS disruption for both types of semantic judgement that approached significance.
Moreover, neural recruitment of LIFG showed a trend-level interaction between control demands and stimulation site: the response in this region following stimulation to LIFG might have been better maintained for weaker associations that required greater control.