The left inferior frontal gyrus and the resolution of unimodal vs. cross-modal interference in speech production: A transcranial direct current stimulation study

Abstract

Most neurobiological models of spoken word production propose that multiple lexical candidates are activated in left posterior temporal cortex during word retrieval. Some accounts also propose a role for the left inferior frontal gyrus (LIFG) in selecting the correct word from among these candidates. Evidence for both proposals has come from the picture-word interference (PWI) paradigm, in which participants name pictures (e.g., RABBIT) while ignoring a distractor word. Categorically related distractors (e.g., horse-RABBIT) slow naming latencies compared to unrelated words (e.g., violin-RABBIT), an effect known as semantic interference, whereas congruent distractors (e.g., rabbit-RABBIT) facilitate naming, but the precise conditions in which these effects occur remain a matter of debate. Although the neuroimaging evidence for left temporal cortex engagement in this paradigm is robust, the evidence for LIFG involvement is more equivocal, particularly for semantic interference. A key factor distinguishing LIFG involvement in neuroimaging studies is distractor modality, i.e., activity is more consistently reported for auditory distractors. We therefore applied online anodal transcranial Direct Current Stimulation (atDCS) to LIFG and left posterior temporal cortex in a three-way, cross-over, sham-controlled PWI paradigm involving either written (Experiment 1) or auditory (Experiment 2) distractors. Surprisingly, compared to sham, atDCS over posterior temporal cortex significantly slowed the congruent facilitation effect with written distractors, but did not modulate the semantic interference effect, while atDCS to LIFG did not significantly influence either effect. We also failed to observe any significant modulatory effects of atDCS with auditory distractors. The present results only partially support left temporal lobe engagement during PWI and provide no evidence for LIFG involvement. We recommend future PWI studies systematically investigate different electrode montages in tDCS protocols.