Response to Skeide and Friederici: The myth of the uniquely human “direct” dorsal pathway

In their comment on our recent article [1], Skeide and Friederici [2] claim “that there are important data undiscussed by Bornkessel-Schlesewsky et al., strongly supporting the view that there are clear qualitative, not mere quantitative differences between [human and nonhuman primate] species both with respect to the intrinsic functional connectivity of frontal and temporal cortices, and their direct structural connection via a dorsal white matter fiber tract.” This obviously refers to work by Friederici and colleagues [3] emphasizing the functional importance of a direct connection between posterior superior temporal cortex (pSTC) and Brodmann area (BA) 44 in humans and its absence in monkeys. This assumption is largely based on DTI data from three macaque brains (two of them formalin-fixed postmortem) [4], according to which fibers tracked between BA 44 and a parieto-temporal region-of-interest terminate in inferior parietal cortex. Skeide and Friederici summarize: “while BA 44 is intrinsically connected to the pSTC in the human brain, this direct functional connection is absent in the macaque brain.” This claim has been proven wrong by recent data [5], substantiating earlier results [6]. The recent data, to which we referred in our article, unequivocally show a direct connection between BA 44 and pSTC in the macaque brain (cf. Figures 7 and 18 in [5]) using fluorescent retrograde anatomical tracing, which remains the gold standard for demonstrating structural brain connectivity. Absence of evidence from DTI cannot overrule positive evidence from direct tracing.

Thus, one of the main pillars of the human exceptionalist view of a “qualitative, not mere quantitative difference” between humans and other species is crumbling, nor can we see any convincing evidence for qualitative intrinsic functional connectivity differences of temporal cortices, as claimed by [2]. On the contrary, we strongly believe that the organization of temporal cortex –– in vision and audition –– is exemplary for the fact that the monkey brain is an excellent model for the human brain.

Despite unequivocal anatomical evidence, Skeide and Friederici defend the assumption of an anatomically distinct dorsal subpathway in humans to support their claim of a combinatorial mechanism (“complex syntax”) that is unique to humans [7]. However, as we have argued previously, the notion of complex syntax is both theoretically empty (e.g., because there is no operationalized cut-off between a simple and a complex sentence) [8] and empirically problematic [1,9].

This leaves the argument that the difference between humans and nonhuman primates lies in the intrinsic organization of frontal cortices, which is exactly the hypothesis put forward in our article. Whether the difference is “qualitative, not mere[ly] quantitative” remains to be discussed and depends both on the interpretation of “qualitative” and on whether qualitative differences can emerge from quantitative differences in neural computation, as we have argued. Although it may seem daring to claim that a complex function like language could have emerged from simpler mechanisms without divine intervention, there are many such examples of extreme non-linearity in physics and biology: e.g., the avalanche effect, which has enabled modern semiconductor electronics. Nowhere in our article did we claim “that cross-stream interaction is the only and necessary condition for the emergence of language”. Rather, we firmly believe that multiple developments during evolution, all following the “avalanche principle”, must have contributed to the emergence of a complex trait like language [10].