Recursive syntactic pattern learning by songbirds

Abstract

Humans regularly produce new utterances that are understood by other members of the same language community 1 . Linguistic theories account for this ability through the use of syntactic rules (or generative grammars) that describe the acceptable structure of utterances 2 . The recursive, hierarchical embedding of language units (for example, words or phrases within shorter sentences) that is part of the ability to construct new utterances minimally requires a 'context-free' grammar 2, 3 that is more complex than the 'finite-state' grammars thought sufficient to specify the structure of all non-human communication signals. Recent hypotheses make the central claim that the capacity for syntactic recursion forms the computational core of a uniquely human language faculty 4,5 . Here we show that European starlings (Sturnus vulgaris) accurately recognize acoustic patterns defined by a recursive, self-embedding, context-free grammar. They are also able to classify new patterns defined by the grammar and reliably exclude agrammatical patterns. Thus, the capacity to classify sequences from recursive, centre-embedded grammars is not uniquely human. This finding opens a new range of complex syntactic processing mechanisms to physiological investigation.The computational complexity of generative grammars is formally defined 3 such that certain classes of temporally patterned strings can only be produced (or recognized) by specific classes of grammars (Fig. 1). Starlings sing long songs composed of iterated motifs (smaller acoustic units) 6 that form the basic perceptual units of individual song recognition 7-9 . Here we used eight 'rattle' and eight 'warble' motifs (see Methods) to create complete 'languages' (4,096 sequences) for two distinct grammars: a context-free grammar (CFG) of the form A 2 B 2 that entails recursive centre-embedding, and a finite-state grammar (FSG) of the form (AB) 2 that does not ( Fig. 2a, b; 'A' refers to rattles and 'B' to warbles).We trained 11 European starlings, using a go/nogo operant conditioning procedure, to classify subsets of sequences from these languages (see Methods and Supplementary Information). Nine out of eleven starlings learned to classify the FSG and CFG sequences accurately (as assessed by d', which provides an unbiased measure of sensitivity to differentiating between two classes of patterns), but this task was difficult (Fig. 2c). The rate of acquisition varied widely among the starlings that learned the task (303.44 ± 57.11 blocks to reach criterion (mean ± s.e.m.), range 94-562 blocks with 100 trials per block), and was slow by comparison to other operant song-recognition tasks 7 .To assess the possibility that starlings learned to classify correctly the motif patterns described by the CFG and FSG grammars through rote memorization of the training exemplars, we further (Fig. 3a). The mean d' over the first 100 trials with new stimuli (roughly six responses to each exemplar) was 1.08 ± 0.50, which is significantly better than chance performance (d' = 0). Over th...

Empirical approaches to the study of language evolution

“…The most important distinction at present appears to be between the levels of regular or Bfinite state^grammars, some types of which are accessible to multiple animal species, and supra-regular grammars that go beyond this (including both context-sensitive and context-free grammars, sometimes termed Bphrase structure grammars^). At present, there are no convincing studies showing the successful acquisition of a supra-regular grammar in a nonhuman animal: for every claimed success (Abe & Watanabe, 2011;Gentner et al, 2006;Rey, Perruchet, & Fagot, 2012), there is a convincing critique (van Heijningen et al, 2009;Beckers et al, 2012;Poletiek et al, 2016). The results of such research have been summarized by Fitch and Friederici (2012), and further explored in important recent work by and Wang, Uhrig, Jarraya, and .…”

Mirror neurons and the social nature of language: The neural exploitation hypothesis

“…They are also able to classify new patterns defined by the grammar while excluding agrammatical patterns. The authors of this study (Gentner, Fenn, Margoliash, & Nusbaum, 2006) concluded that the capacity to classify sequences from recursive, centre-embedded grammars is not uniquely human.…”

Parsing recursive sentences with a connectionist model including a neural stack and synaptic gating

“…Since they did not establish dependencies between words their CFG sentences could be parsed by counting recursion. Likewise, Gentner, Fenn, Margoliash & Nusbaum (Gentner et al, 2006) claimed that their subjects (starlings) learnt CER using the same kind of structures as Fitch and Hauser (Fitch and Hauser, 2004). This in turn elicited some critique: Corballis (Corballis, 2007a;Corballis, 2007b) called attention to the fact that the sentences used could be parsed by simple counting, while others (De Vries et al, 2008;Perruchet and Rey, 2005) showed that in experimental situation similar to that of Fitch & Hauser (Fitch and Hauser, 2004) even human subjects used alternative strategies to solve the tasks.…”