This article investigates the influence of contextual pressures on the evolution of overspecification, i.e. the degree to which communicatively irrelevant meaning dimensions are specified, in an iterated learning setup. To this end, we combine two lines of research: In artificial language learning studies, it has been shown that (miniature) languages adapt to their contexts of use. In experimental pragmatics, it has been shown that referential overspecification in natural language is more likely to occur in contexts in which the communicatively relevant feature dimensions are harder to discern. We test whether similar functional pressures can promote the cumulative growth of referential overspecification in iterated artificial language learning. Participants were trained on an artificial language which they then used to refer to objects. The output of each participant was used as input for the next participant. The initial language was designed such that it did not show any overspecification, but it allowed for overspecification to emerge in 16 out of 32 usage contexts. Between conditions, we manipulated the referential context in which the target items appear, so that the relative visuospatial complexity of the scene would make the communicatively relevant feature dimensions more difficult to discern in one of them. The artificial languages became overspecified more quickly and to a significantly higher degree in this condition, indicating that the trend toward overspecification was stronger in these contexts, as suggested by experimental pragmatics research. These results add further support to the hypothesis that linguistic conventions can be partly determined by usage context and shows that experimental pragmatics can be fruitfully combined with artificial language learning to offer valuable insights into the mechanisms involved in the evolution of linguistic phenomena.

Sound systems vary dramatically in their lower-level details as a result of cultural evolution, but the presence of systematic organisation is universal. Why does variation pattern differently at these two levels of abstraction, and what can this tell us about the cognitive mechanisms that underpin human acquisition of speech? We explore an evolutionary rationale for the proposal that human learning extends to, and is perhaps even specialised for, making inferences at the higher-order level of abstraction. The ability to infer systematicity from distributional cues, by identifying signatures of structural homogeneity and anticipating subtle exceptions, can bootstrap lower-level learning, and is not subject to the moving target problem, a major evolutionary objection to specialisation in speech cognition. We examine this idea from a statistical perspective, by studying the representational assumptions that underpin generalisation among concurrent phonetic category induction problems. We present a probabilistic model for jointly inferring individual sound classes and a system-wide blueprint for the balance of shared and idiosyncratic structure among these classes. These models lead us to an evolutionary conjecture: culture pushes cognitive adaptation up the hierarchy of abstraction in learning

This article investigates the terminology and the processes underlying the fundamental historical relations between words in linguistics ( cognacy ) and genes in biology ( homology ). The comparison between linguistics and biology shows that there are major inconsistencies in the analogies drawn between the two research fields and the models applied in phylogenetic reconstruction in linguistics. Cognacy between words is treated as a binary relation which is either present or not. Words, however, can exhibit different degrees of cognacy which go beyond the distinction between orthologous and paralogous genes in biology. The complex nature of cognacy has strong implications for the models used for phylogenetic reconstruction. Instead of modeling lexical evolution as a process of cognate gain and cognate loss, we need to go beyond the cognate relation and develop models which take the degrees of cognacy into account. This opts for the use of evolutionary models which handle multistate characters and allow to define potentially asymmetrical transition tendencies among the character states instead of time-reversible binary state models in phylogenetic approaches. The benefit of multistate models with asymmetric transition tendencies is demonstrated by testing how well different models of lexical change perform in semantic reconstruction on a lexicostatistical dataset of 23 Chinese dialects in a parsimony framework. The results show that the improved models largely outperform the popular gain–loss models. This suggests that improved models of lexical change may have strong consequences for phylogenetic approaches in linguistics.

Numerals have fascinated and mystified linguists, mathematicians and lay persons alike for centuries. The productive use of numerals (in languages where this happens) exploits recursivity to give rise to what we call the ‘the number line’. While the smaller numerals 1–10 have enjoyed intense scrutiny, the typological study of the formation of the higher numerals has received comparatively less attention. This article contains a comprehensive typological account of how languages in the Indo-European language family code numerals beyond 10 (10–99, 100s, 1,000s), the morphemes involved, and how these are ordered. We use this dataset from eighty-one Indo-European languages with phylogenetic comparative methods to propose diachronic reconstructions of these patterns in the Proto-Indo-European language. Our findings indicate that small numerals (11–19) show the widest cross-linguistic variation, and that higher numerals exhibit more consistency in both component parts and their ordering. Additionally, we show statistical evidence of correlations between the ordering of base and atom morphemes and other word order patterns (noun-postposition, noun-genitive, and verb-object order).

If languages do indeed evolve then they must show the three crucial aspects of an evolving system: variation of traits, inheritance of those traits, and the differential survival—that is selection—of those traits (Lewontin 1970). We know that languages vary (otherwise the fields of linguistic typology and sociolinguistics would be boring). We know that variation is passed through speech communities and inherited from parent language to its daughters (otherwise historical linguistics would likewise be boring). However, whether linguistic traits are selected for is much less clear (Ramsey and De Block 2016). Nor is it clear just where selection might operate (Dediu et al. 2013). Despite these issues, many possible candidates for selective effects have been proposed in the linguistic literature. Selection might operate at the level of the speech act . For example, the more lexical items are used, the more stable they are over time, which may be consistent with frequency-dependent selection (Pagel et al. 2007; Calude and Pagel 2011), while the length of these more frequent words appears to be shorter to increase communication efficiency (Piantadosi et al. 2011). Alternatively, perhaps rate of speech is constrained as there …

Does (the presence or complexity of) tone inversely correlate with dryness of climate? The authors (Everett et al.) suggest that the absence of ambient humidity in the air negatively correlates with the presence of (complex?) lexical tone, partly because of the effect that dry air has to increase the difficulty in achieving precise articulatory targets. There are two main problems with the argumentation used. 1. Conflating ‘tone’ with ‘pitch’ or ‘fundamental frequency’, and mistaking ‘complexity’ with a syllable domain for tone assignment; 2. conflating ‘dry climate’ with the absence of humidity. The authors are not guilty in an absolute sense of these problems, acknowledging that there are complications. Their reliance on pitch contrasts as a proxy for tonal category contrasts, and the use of air humidity rather than (easily available) climate information for the ranges of different languages means that the authors are dealing with ephemeral correlations between proxy features. In the next two sections, I will critique the use of tone primarily to refer to distinctions realised by pitch, and the use of humidity as a powerful explanatory for the existence of tone categories. The authors acknowledge that ‘many non-pitch phenomena are associated with the production of tone, including ancillary laryngealization and duration influences’, but go on to claim that ‘the heightened role of F0 (and therefore pitch) in languages with complex tone is evident in the fact that its fine-grained modulation is required on every or almost every syllable, in contrast to pitch accent languages’. First, this …

Speaking Our Minds: Why human communication is different, and how language evolved to make it special , by Thom Scott-Phillips, Palgrave MacMillan (2014), 212 pp, £24.99. Language evolution is a notoriously complex problem. As such, the central topic of how and why language evolved tends to split into a number of questions that are often tackled separately. Why are humans the only species to have such a complex communication system? How were the first communicative conventions established, and how did they develop into languages that take on the forms we observe? What explains our motivation to communicate, and what pressures allow language to remain evolutionarily stable? In Speaking Our Minds , Thom Scott-Phillips offers an account that aims to address all of these questions. The central argument is that language is made possible by an evolutionarily novel form of communication that is uniquely human: ostensive-inferential communication. In this form of communication, signallers provide evidence for their intentions, and receivers interpret this evidence to infer the meaning the signaller intends to convey. This ability relies on sophisticated forms of social cognition which, in turn, arose in our species as a consequence of selection pressure for the ability to better track others’ intentions, as the size and complexity of social groups increased. Scott-Phillips marshals a broad range of evidence in support of this view, from non-human primate communication to the creation, learning and use of artificial languages by human participants in the lab. An unstated assumption in both mainstream and evolutionary linguistics is that language is fundamentally a code, enriched and made flexible by our inferential abilities. The key insight underpinning the book is that this mainstream view is wrong. Specifically, it is backwards. Ostensive-inferential communication makes language possible, and the conventional codes we are able to create as a result make it flexible and productive. This is not just a terminological quibble. The implication is that the traditional view rests on a mechanistically wrong account … casilvey{at}uchicago.edu