Non-Rigid Syntax in Homo Sapiens and Elsewhere

Non-Rigid Syntax in homo sapiens and Elsewhere


A uniquely human ability concerns the understanding and producing of embedded sequences. This ability characterizes Language, but also our ability to create symphonies and do abstract algebra, among other things. The underpinning of such ability is thought to be the presence in the sapien mind the ability for supraregular grammars, combinatorial systems that are beyond simple FSAs. However, not all conceivable grammars are attested within human languages. Rules based on rigidly fixed distances, that is a grammar in which a certain word must occur after a fixed number of intervals before another word, is not part of the human linguistic computation system. Distances between words are specified in terms of dependency nodes in a hierarchical tree, that is they are relative, non-rigid positions – i.e. non-rigid syntactic dependency (NRSD). The left inferior frontal gyrus (IFG) (Broca's area) has been found to be involved in the computation of non-rigid but not of rigid syntax in the language domain. Here, two important questions can be asked. First, is such a capacity completely unique to humans, particularly in light of descent with modification? And second, is the neural activity underlying this non-rigid architecture language-specific, since analogues can be found in other cognitive domains?

 Jiang et al. [1]  show that macaque monkeys can be trained to produce time-symmetrical embedded spatial sequences whose formal description requires supra-regular grammars or a push-down stack automaton. In their study, the monkeys are reported to have spontaneously generalized the learned grammar to novel sequences of similar, and even longer length. Further, they could generate hierarchical sequences formed by an embedding of two levels of abstract rules. This, however, requires extensive training. The researchers further report that compared to monkeys even preschool children learned the grammars much faster . They conclude that while supra-regular grammars are accessible to nonhuman primates through extensive training, human uniqueness may lie in the speed and learning strategy with which they are acquired.

This conveniently allows us to move on to our second question regarding the domain-specificity of the neural activity supporting the computing of NRSD. A recent study by Tettamanti and colleagues [2] used functional magnetic resonance imaging (fMRI) to investigated NRSD and RSD related neural activity in sixteen healthy native speakers of Italian in the visuo-spatial domain, and compared their results with previous studies where NRSD and RSD were categorically contrasted. Their data reflected that both in the visuo-spatial and in the language domain the activation of Brodmann Area 44 of the left IFG was associated with NRSD but not RSD. They conclude that in the human brain one single "grammar without words" serves different higher cognitive functions.

So, where does that leave generative linguistics given that uniqueness of human language syntax has always been its cornerstone? While it is tempting to conclude that the whole of modern linguistics has been falsified, and such claims have been made [3], such conclusions are hastily reached. First, as Jiang and colleagues aptly point out the monkeys requires significant effort to produce a fraction of what children do spontaneously. Thus, as Chomsky has often noted [4]–[6], the spontaneous creativity of children in the face of impoverished stimuli remains to be explained. Second, for a theory of such computational ability to meet the demands of explanatory adequacy [7], [8] it must be able to go beyond merely describing the nature of said system, and causally account for its architecture. For instance, one must be able to account for the fact that there are three binding principles that make specific statements and not other conceivable ones, and explain what factors shape such principles. Whatever they turn out to be, afterall, they do underlie the fact that any children can trigger any language on mere exposure. Since we use our brains for all cognitive functions, then, the neural explanation of such convergence of grammatical structures must follow. Finally, it must be noted that Jiang and colleagues used a production/behavioral task, while the Tettamanti study used neural tools. Similar experimentation needs to be undertaken to uncover the neural mechanisms underlying complex supraregular computations in other species. Such research is likely to illuminate both human linguistics and animal cognition studies by highlighting exactly why and how our species surpasses others in such abilities. After all, human language in its fullness is constituted of a number of sub-parts fitting together, including semantic systems and neural control required for articulation of speech forms. It is always possible that brain structures, such as prefrontal cortex regions, that are present in other species but are greatly enhanced in humans support the full complexity of human languages. Thus, in conclusion, it can be readily claimed that such developments as discussed in the two currently cited studies [1,2] do nothing to dampen generative inquiry, but rather provide exciting oppurtunities to conduct converging research within and across species to better understand both human and animal cognition, unique and different in their own rights.

[1] X. Jiang, T. Long, W. Cao, J. Li, S. Dehaene, and L. Wang, “Production of Supra-regular Spatial Sequences by Macaque Monkeys,” Curr. Biol., vol. 28, no. 12, pp. 1851-1859.e4, 2018.

[2] M. Tettamanti et al., “Syntax without language: neurobiological evidence for cross-domain syntactic computations,” Cortex J. Devoted Study Nerv. Syst. Behav., vol. 45, no. 7, pp. 825–838, Aug. 2009.

[3] P. Ibbotson, “Evidence Rebuts Chomsky’s Theory of Language Learning,” p. 16.

[4] N. Chomsky and R. DiNozzi, Language and mind. Harcourt Brace Jovanovich New York, 1972.

[5] N. Chomsky, “On certain formal properties of grammars,” Inf. Control, vol. 2, no. 2, pp. 137–167, 1959.

[6] N. Chomsky, Cartesian Linguistics, 3 edition. Cambridge University Press, 2012.

[7] L. Rizzi, “The concept of explanatory adequacy.,” p. 16.

[8] N. Chomsky, “A review of BF Skinner’s Verbal Behavior,” Language, vol. 35, no. 1, pp. 26–58, 1959.