A possible anatomical substrate is direct connections between hemispheres via the corpus callosum (Barbas and Pandya, 1984). Cortical feedforward processing has been shown to be mediated by gamma and theta oscillations while feedback processing is mediated by alpha/beta oscillations (Bastos et al., 2015; Buschman and Miller, 2007; van Kerkoerle et al., 2014). Our results showed increases in interhemispheric theta and high-beta synchrony, and decreases in alpha/low-beta, during the presumed time of memory trace transfer. One possibility is that interhemispheric communication occurs both in frequency bands used by feedforward (theta) and feedback (beta) processing. Perhaps this allows flexible interaction of interhemispheric signals with either feedforward or feedback signals. Another possibility is that higher frequency cortical communication—typically in the traditional gamma band—is shifted slightly downward in frequency (to 40 Hz and lower) to maintain reliable transmission despite the time delays imposed by long callosal axons. Of course, there are other possible routes for interhemispheric transfer other than direct prefrontal callosal connections. Signals might be transferred at some higher level region and then conveyed via feedback to PFC, or conceivably transfer could happen at lower levels and feed forward to PFC. Ultimately, resolving this issue will likely require causal manipulation of connectivity.
