Language Evolution and Computation Bibliography

Fluid Construction Grammar (FCG) is an open-source computational grammar formalism that is becoming increasingly popular for studying the history and evolution of language. This demonstration shows how FCG can be used to operationalise the cultural processes and cognitive mechanisms that underly language evolution and change.

Human language, as well as birdsong, relies on the ability to arrange vocal elements in new sequences. However, little is known about the ontogenetic origin of this capacity. Here we track the development of vocal combinatorial capacity in three species of vocal learners, combining an experimental approach in zebra finches (Taeniopygia guttata) with an analysis of natural development of vocal transitions in Bengalese finches (Lonchura striata domestica) and pre-lingual human infants. We find a common, stepwise pattern of acquiring vocal transitions across species. In our first study, juvenile zebra finches were trained to perform one song and then the training target was altered, prompting the birds to swap syllable order, or insert a new syllable into a string. All birds solved these permutation tasks in a series of steps, gradually approximating the target sequence by acquiring new pairwise syllable transitions, sometimes too slowly to accomplish the task fully. Similarly, in the more complex songs of Bengalese finches, branching points and bidirectional transitions in song syntax were acquired in a stepwise fashion, starting from a more restrictive set of vocal transitions. The babbling of pre-lingual human infants showed a similar pattern: instead of a single developmental shift from reduplicated to variegated babbling (that is, from repetitive to diverse sequences), we observed multiple shifts, where each new syllable type slowly acquired a diversity of pairwise transitions, asynchronously over development. Collectively, these results point to a common generative process that is conserved across species, suggesting that the long-noted gap between perceptual versus motor combinatorial capabilities in human infants1 may arise partly from the challenges in constructing new pairwise vocal transitions.

The word for “sky” in the indigenous Saaroa language of Taiwan is laŋica. Across the South China Sea in the Philippines, the speakers of Ilonggo use laŋit, whereas, on the far-flung islands of the Pacific, Hawaiians say lani and Rarotongans and New Zealand Maori raŋi (1). Systematic sound correspondences between many such words tell us that these languages have evolved from a common ancestor to form part of the Austronesian language family. By meticulously comparing the sounds of words across many languages, linguists can learn about the genealogical relationships between languages and the people who speak them, how sounds change through time and even how long-extinct ancestral languages would have sounded. In PNAS, Bouchard-Côté et al. (2) automate this process by using probabilistic models of sound change to trace the evolution of thousands of words across more than 600 Austronesian languages.

We report on the emergence of functional flexibility in vocalizations of human infants. This vastly underappreciated capability becomes apparent when prelinguistic vocalizations express a full range of emotional content--positive, neutral, and negative. The data show that at least three types of infant vocalizations (squeals, vowel-like sounds, and growls) occur with this full range of expression by 3-4 mo of age. In contrast, infant cry and laughter, which are species-specific signals apparently homologous to vocal calls in other primates, show functional stability, with cry overwhelmingly expressing negative and laughter positive emotional states. Functional flexibility is a sine qua non in spoken language, because all words or sentences can be produced as expressions of varying emotional states and because learning conventional "meanings" requires the ability to produce sounds that are free of any predetermined function. Functional flexibility is a defining characteristic of language, and empirically it appears before syntax, word learning, and even earlier-developing features presumed to be critical to language (e.g., joint attention, syllable imitation, and canonical babbling). The appearance of functional flexibility early in the first year of human life is a critical step in the development of vocal language and may have been a critical step in the evolution of human language, preceding protosyntax and even primitive single words. Such flexible affect expression of vocalizations has not yet been reported for any nonhuman primate but if found to occur would suggest deep roots for functional flexibility of vocalization in our primate heritage.

One of the oldest problems in linguistics is reconstructing the words that appeared in the protolanguages from which modern languages evolved. Identifying the forms of these ancient languages makes it possible to evaluate proposals about the nature of language change and to draw inferences about human history. Protolanguages are typically reconstructed using a painstaking manual process known as the comparative method. We present a family of probabilistic models of sound change as well as algorithms for performing inference in these models. The resulting system automatically and accurately reconstructs protolanguages from modern languages. We apply this system to 637 Austronesian languages, providing an accurate, large-scale automatic reconstruction of a set of protolanguages. Over 85% of the system’s reconstructions are within one character of the manual reconstruction provided by a linguist specializing in Austronesian languages. Being able to automatically reconstruct large numbers of languages provides a useful way to quantitatively explore hypotheses about the factors determining which sounds in a language are likely to change over time. We demonstrate this by showing that the reconstructed Austronesian protolanguages provide compelling support for a hypothesis about the relationship between the function of a sound and its probability of changing that was first proposed in 1955.

It is generally difficult to define reasonable parameters and interpret their values in mathematical models of social phenomena. Rather than directly fitting abstract parameters against empirical data, we should define some concrete parameters to denote the sociocultural factors relevant for particular phenomena, and compute the values of these parameters based upon the corresponding empirical data. Taking the example of modeling studies of language competition, we propose a language diffusion principle and two language inheritance principles to compute two critical parameters, namely the impacts and inheritance rates of competing languages, in our language competition model derived from the Lotka–Volterra competition model in evolutionary biology. These principles assign explicit sociolinguistic meanings to those parameters and calculate their values from the relevant data of population censuses and language surveys. Using four examples of language competition, we illustrate that our language competition model with thus-estimated parameter values can reliably replicate and predict the dynamics of language competition, and it is especially useful in cases lacking direct competition data.

The search for ever deeper relationships among the World’s languages is bedeviled by the fact that most words evolve too rapidly to preserve evidence of their ancestry beyond 5,000 to 9,000 y. On the other hand, quantitative modeling indicates that some “ultraconserved” words exist that might be used to find evidence for deep linguistic relationships beyond that time barrier. Here we use a statistical model, which takes into account the frequency with which words are used in common everyday speech, to predict the existence of a set of such highly conserved words among seven language families of Eurasia postulated to form a linguistic superfamily that evolved from a common ancestor around 15,000 y ago. We derive a dated phylogenetic tree of this proposed superfamily with a time-depth of ∼14,450 y, implying that some frequently used words have been retained in related forms since the end of the last ice age. Words used more than once per 1,000 in everyday speech were 7- to 10-times more likely to show deep ancestry on this tree. Our results suggest a remarkable fidelity in the transmission of some words and give theoretical justification to the search for features of language that might be preserved across wide spans of time and geography.

Language serves as a cornerstone for human cognition, yet much about its evolution remains puzzling. Recent research on this question parallels Darwin's attempt to explain both the unity of all species and their diversity. What has emerged from this research is that the unified nature of human language arises from a shared, species-specific computational ability. This ability has identifiable correlates in the brain and has remained fixed since the origin of language approximately 100 thousand years ago. Although songbirds share with humans a vocal imitation learning ability, with a similar underlying neural organization, language is uniquely human.

Networks of interconnected nodes have long played a key role in Cognitive Science, from artificial neural networks to spreading activation models of semantic memory. Recently, however, a new Network Science has been developed, providing insights into the emergence of global, system-scale properties in contexts as diverse as the Internet, metabolic reactions, and collaborations among scientists. Today, the inclusion of network theory into Cognitive Sciences, and the expansion of complex-systems science, promises to significantly change the way in which the organization and dynamics of cognitive and behavioral processes are understood. In this paper, we review recent contributions of network theory at different levels and domains within the Cognitive Sciences.

Language serves as a cornerstone for human cognition, yet much about its evolution remains puzzling. Recent research on this question parallels Darwin's attempt to explain both the unity of all species and their diversity. What has emerged from this research is that the unified nature of human language arises from a shared, species-specific computational ability. This ability has identifiable correlates in the brain and has remained fixed since the origin of language approximately 100 thousand years ago. Although songbirds share with humans a vocal imitation learning ability, with a similar underlying neural organization, language is uniquely human.

What evolutionary events led to the emergence of human cognition? Although the genetic differences separating modern humans from both non-human primates (for example, chimpanzees) and archaic hominins (Neanderthals and Denisovans) are known, linking human-specific mutations to the cognitive phenotype remains a challenge. One strategy is to focus on human-specific changes at the level of intermediate phenotypes, such as gene expression and metabolism, in conjunction with evolutionary changes in gene regulation involving transcription factors, microRNA and proximal regulatory elements. In this Review we show how this strategy has yielded some of the first hints about the mechanisms of human cognition.

We consider a simple computational model of the evolution of a quantitative trait and its phenotypic plasticity based on directional and positive frequency-dependent selection in order to explore whether and how leaning might facilitate evolution under the dynamics that arise from communicative interactions among individuals. In the model, each individual expresses, at many different times in its lifetime, its real-valued trait depending on the probability distribution determined by its own genotypes. In communicative interactions between two individuals, the contribution of an interaction to the fitness is high when their trait values are close to each other as well as large, which represents the positive frequency-dependent and directional components of selection, respectively. The iterative interactions allow individuals to acquire a more adaptive trait pair through trial and error. Under the stochastic evolution process with the limited number of individuals, we show that learning allows the population to avoid getting stuck in the global but low optimum of the innate and individual-level fitness landscape via both aspects of the components of selection, and brings about the successful evolution by increasing the genetic variation of the population. We also analyze how such an effect of learning can be realized by measuring the degree of the two different contributions for increasing the adaptivity and similarity of communicative traits, respectively. We show that this effect of learning arises from these different types of contributions depending on the biological and environmental conditions such as the mutation rate and the duration of communicative interactions. We further show the condition for the complete genetic assimilation to occur.

Did Neanderthals have language? This issue has been debated back and forth for decades, without resolution. But in recent years new evidence has become available. New fossils and archeological finds cast light on relevant Neanderthal anatomy and behavior. New DNA evidence, both fossil and modern, provides clues both to the relationship between Neanderthals and Homo sapiens, and to the genetics of language. In this paper, I review and evaluate the available evidence. My conclusion is that the preponderance of the evidence supports the presence of some form of language in Neanderthals.

The Homeric epics are among the greatest masterpieces of literature, but when they were produced is not known with certainty. Here we apply evolutionary-linguistic phylogenetic statistical methods to differences in Homeric, Modern Greek and ancient Hittite vocabulary items to estimate a date of approximately 710–760 BCE for these great works. Our analysis compared a common set of vocabulary items among the three pairs of languages, recording for each item whether the words in the two languages were cognate – derived from a shared ancestral word – or not. We then used a likelihood-based Markov chain Monte Carlo procedure to estimate the most probable times in years separating these languages given the percentage of words they shared, combined with knowledge of the rates at which different words change. Our date for the epics is in close agreement with historians' and classicists' beliefs derived from historical and archaeological sources. The Homeric epics are among the greatest masterpieces of literature. The Iliad's story of the Trojan Wars tells us that the epics were almost certainly produced sometime after the 12th century BCE – if indeed the wars were ever fought – but the question is how much later? Herodotus thought considerably later: Writing in the Histories Book II.53 around 450 BCE, he stated that Homer ‘lived, as I believe, not more than 400 years ago’. The most commonly accepted date among modern classicists, drawing on historical, literary and archaeological analyses, is around the mid-8th century BCE 1, 2, although some authors propose a more recent 7th century BCE date 3. Here, we investigate whether formal statistical modelling of languages can help to inform this historical question. In particular, we investigate whether evolutionary-linguistic statistical methods can be usefully applied to differences in Homeric, Modern Greek and ancient Hittite vocabulary items to provide a date for these great works.

Zipf's law on word frequency and Heaps' law on the growth of distinct words are observed in Indo-European language family, but it does not hold for languages like Chinese, Japanese and Korean. These languages consist of characters, and are of very limited dictionary sizes. Extensive experiments show that: (i) The character frequency distribution follows a power law with exponent close to one, at which the corresponding Zipf's exponent diverges. Indeed, the character frequency decays exponentially in the Zipf's plot. (ii) The number of distinct characters grows with the text length in three stages: It grows linearly in the beginning, then turns to a logarithmical form, and eventually saturates. A theoretical model for writing process is proposed, which embodies the rich-get-richer mechanism and the effects of limited dictionary size. Experiments, simulations and analytical solutions agree well with each other. This work refines the understanding about Zipf's and Heaps' laws in human language systems.

It is well-known that word frequencies arrange themselves according to Zipf's law. However, little is known about the dependency of the parameters of the law and the complexity of a communication system. Many models of the evolution of language assume that the exponent of the law remains constant as the complexity of a communication systems increases. Using longitudinal studies of child language, we analysed the word rank distribution for the speech of children and adults participating in conversations. The adults typically included family members (e.g., parents) or the investigators conducting the research. Our analysis of the evolution of Zipf's law yields two main unexpected results. First, in children the exponent of the law tends to decrease over time while this tendency is weaker in adults, thus suggesting this is not a mere mirror effect of adult speech. Second, although the exponent of the law is more stable in adults, their exponents fall below 1 which is the typical value of the exponent assumed in both children and adults. Our analysis also shows a tendency of the mean length of utterances (MLU), a simple estimate of syntactic complexity, to increase as the exponent decreases. The parallel evolution of the exponent and a simple indicator of syntactic complexity (MLU) supports the hypothesis that the exponent of Zipf's law and linguistic complexity are inter-related. The assumption that Zipf's law for word ranks is a power-law with a constant exponent of one in both adults and children needs to be revised.

This study examines the intergenerational transfer of human communication systems. It tests if human communication systems evolve to be easy to learn or easy to use (or both), and how population size affects learnability and usability. Using an experimental-semiotic task, we find that human communication systems evolve to be easier to use (production efficiency and reproduction fidelity), but harder to learn (identification accuracy) for a second generation of naïve participants. Thus, usability trumps learnability. In addition, the communication systems that evolve in larger populations exhibit distinct advantages over those that evolve in smaller populations: the learnability loss (from the Initial signs) is more muted and the usability benefits are more pronounced. The usability benefits for human communication systems that evolve in a small and large population is explained through guided variation reducing sign complexity. The enhanced performance of the communication systems that evolve in larger populations is explained by the operation of a content bias acting on the larger pool of competing signs. The content bias selects for information-efficient iconic signs that aid learnability and enhance usability.

Grammatical agreement means that features associated with one linguistic unit (for example number or gender) become associated with another unit and then possibly overtly expressed, typically with morphological markers. It is one of the key mechanisms used in many languages to show that certain linguistic units within an utterance grammatically depend on each other. Agreement systems are puzzling because they can be highly complex in terms of what features they use and how they are expressed. Moreover, agreement systems have undergone considerable change in the historical evolution of languages. This article presents language game models with populations of agents in order to find out for what reasons and by what cultural processes and cognitive strategies agreement systems arise. It demonstrates that agreement systems are motivated by the need to minimize combinatorial search and semantic ambiguity, and it shows, for the first time, that once a population of agents adopts a strategy to invent, acquire and coordinate meaningful markers through social learning, linguistic self-organization leads to the spontaneous emergence and cultural transmission of an agreement system. The article also demonstrates how attested grammaticalization phenomena, such as phonetic reduction and conventionalized use of agreement markers, happens as a side effect of additional economizing principles, in particular minimization of articulatory effort and reduction of the marker inventory. More generally, the article illustrates a novel approach for studying how key features of human languages might emerge.

Understanding the patterns and causes of differential structural stability is an area of major interest for the study of language change and evolution. It is still debated whether structural features have intrinsic stabilities across language families and geographic areas, or if the processes governing their rate of change are completely dependent upon the specific context of a given language or language family. We conducted an extensive literature review and selected seven different approaches to conceptualising and estimating the stability of structural linguistic features, aiming at comparing them using the same dataset, the World Atlas of Language Structures. We found that, despite profound conceptual and empirical differences between these methods, they tend to agree in classifying some structural linguistic features as being more stable than others. This suggests that there are intrinsic properties of such structural features influencing their stability across methods, language families and geographic areas. This finding is a major step towards understanding the nature of structural linguistic features and their interaction with idiosyncratic, lineage- and area-specific factors during language change and evolution.

The ASJP (Automated Similarity Judgment Program) described an automated, lexical similarity-based method for dating the world’s language groups using 52 archaeological, epigraphic and historical calibration date points. The present paper describes a new automated dating method, based on phonotactic diversity. Unlike ASJP, our method does not require any information on the internal classification of a language group. Also, the method can use all the available word lists for a language and its dialects eschewing the debate on ‘language’ vs. ‘dialect’. We further combine these dates and provide a new baseline which, to our knowledge, is the best one. We make a systematic comparison of our method, ASJP’s dating procedure, and combined dates. We predict time depths for world’s language families and sub-families using this new baseline. Finally, we explain our results in the model of language change given by Nettle.

Our goal of this study is to characterize the functions of language areas in most precise terms. Previous neuroimaging studies have reported that more complex sentences elicit larger activations in the left inferior frontal gyrus (L. F3op/F3t), although the most critical factor still remains to be identified. We hypothesize that pseudowords with grammatical particles and morphosyntactic information alone impose a construction of syntactic structures, just like normal sentences, and that “the Degree of Merger” (DoM) in recursively merged sentences parametrically modulates neural activations. Using jabberwocky sentences with distinct constructions, we fitted various parametric models of syntactic, other linguistic, and nonlinguistic factors to activations measured with functional magnetic resonance imaging. We demonstrated that the models of DoM and “DoM+number of Search (searching syntactic features)” were the best to explain activations in the L. F3op/F3t and supramarginal gyrus (L. SMG), respectively. We further introduced letter strings, which had neither lexical associations nor grammatical particles, but retained both matching orders and symbol orders of sentences. By directly contrasting jabberwocky sentences with letter strings, localized activations in L. F3op/F3t and L. SMG were indeed independent of matching orders and symbol orders. Moreover, by using dynamic causal modeling, we found that the model with a inhibitory modulatory effect for the bottom-up connectivity from L. SMG to L. F3op/F3t was the best one. For this best model, the top-down connection from L. F3op/F3t to L. SMG was significantly positive. By using diffusion-tensor imaging, we confirmed that the left dorsal pathway of the superior longitudinal and arcuate fasciculi consistently connected these regions. Lastly, we established that nonlinguistic order-related and error-related factors significantly activated the right (R.) lateral premotor cortex and R. F3op/F3t, respectively. These results indicate that the identified network of L. F3op/F3t and L. SMG subserves the calculation of DoM in recursively merged sentences.

We propose a simple model for genetic adaptation to a changing environment, describing a fitness landscape characterized by two maxima. One is associated with “specialist” individuals that are adapted to the environment; this maximum moves over time as the environment changes. The other maximum is static, and represents “generalist” individuals not affected by environmental changes. The rest of the landscape is occupied by “maladapted” individuals. Our analysis considers the evolution of these three subpopulations. Our main result is that, in presence of a sufficiently stable environmental feature, as in the case of an unchanging aspect of a physical habitat, specialists can dominate the population. By contrast, rapidly changing environmental features, such as language or cultural habits, are a moving target for the genes; here, generalists dominate, because the best evolutionary strategy is to adopt neutral alleles not specialized for any specific environment. The model we propose is based on simple assumptions about evolutionary dynamics and describes all possible scenarios in a non-trivial phase diagram. The approach provides a general framework to address such fundamental issues as the Baldwin effect, the biological basis for language, or the ecological consequences of a rapid climate change.

Languages evolve over space and time. Illuminating the evolutionary history of language is important because it provides a unique opportunity to shed light on the population history of the speakers. Spatial and temporal aspects of language evolution are particularly crucial for understanding demographic history, as they allow us to identify when and where the languages originated, as well as how they spread across the globe. Here we apply Bayesian phylogeographic methods to reconstruct spatiotemporal evolution of the Ainu language: an endangered language spoken by an indigenous group that once thrived in northern Japan. The conventional dual-structure model has long argued that modern Ainu are direct descendants of a single, Pleistocene human lineage from Southeast Asia, namely the Jomon people. In contrast, recent evidence from archaeological, anthropological and genetic evidence suggest that the Ainu are an outcome of significant genetic and cultural contributions from Siberian hunter-gatherers, the Okhotsk, who migrated into northern Hokkaido around 900–1600 years ago. Estimating from 19 Ainu language varieties preserved five decades ago, our analysis shows that they are descendants of a common ancestor who spread from northern Hokkaido around 1300 years ago. In addition to several lines of emerging evidence, our phylogeographic analysis strongly supports the hypothesis that recent expansion of the Okhotsk to northern Hokkaido had a profound impact on the origins of the Ainu people and their culture, and hence calls for a refinement to the dual-structure model.

Recently there has been a movement in the social sciences to apply theories and methods developed originally in evolutionary biology to understand human behavior and cultural change. In this article, I highlight an ever growing branch of this research that analyzes cultural data using a suite of techniques based on evolutionary trees. These phylogenetic methods fall into two categories: (a) “tree-building” that is, inferring the historical relationships between units, and (b) comparative analyses that is, mapping other traits onto the tips of these trees to make inferences about trait evolution. Using lexical data I infer the phylogenetic relationships between 82 Indo-European languages. The estimated divergence time of these languages agrees with earlier studies in supporting a hypothesis that places the homeland of the Indo-Europeans somewhere in Anatolia 8000-9500 BP. Using data from Austronesian-speaking societies I give examples of the use of comparative methods to (i) infer the system of land tenure in ancestral societies, (ii) examine the ecological correlates of violence in the Polynesian and Micronesian societies, and (iii) test hypotheses about the mode and tempo of sociopolitical evolution. I examine future areas of cultural phylogenetic research in the evolutionary social sciences and argue that the kinds of studies introduced in this article hold the potential for fruitful collaborations between cultural phylogeneticists and cross-cultural researchers.

The naming game (NG) describes the agreement dynamics of a population of N agents interacting locally in pairs leading to the emergence of a shared vocabulary. This model has its relevance in the novel fields of semiotic dynamics and specifically to opinion formation and language evolution. The application of this model ranges from wireless sensor networks as spreading algorithms, leader election algorithms to user-based social tagging systems. In this paper, we introduce the concept of overhearing (i.e., at every time step of the game, a random set of Nδ individuals are chosen from the population who overhear the transmitted word from the speaker and accordingly reshape their inventories). When δ = 0 one recovers the behavior of the original NG. As one increases δ, the population of agents reaches a faster agreement with a significantly low-memory requirement. The convergence time to reach global consensus scales as logN as δ approaches 1.

In this paper, we present an experiment that integrates a semiotic investigation with a dynamical perspective on embodied social interactions. The primary objective is to study the emergence of a communication system between two interacting individuals, where no dedicated communication modalities are predefined and the only possible interaction is very simple, non-directional, and embodied. Throughout the experiment, we observe the following three phenomena: (1) the spontaneous emergence of turn-taking behaviour that allows communication in non-directional environments; (2) the development of an association between behaviours and perceptive categories; (3) the acquisition of novel meaning by exploiting the notion of complementary set theory.

Language transfers information on at least three levels; (1) what is said, (2) how it is said (what language is used), and, (3) that it is said (that speaker and listener both possess the ability to use language). The use of language is a form of honest cooperation on two of these levels; not necessarily on what is said, which can be deceitful, but always on how it is said and that it is said. This means that the language encoding and decoding systems had to evolve simultaneously, through mutual fitness benefits. Theoretical problems surrounding the evolution of cooperation disappear if a recognition system is present enabling cooperating individuals to identify each other – if they are equipped with “green beards”. Here, I outline how both the biological and cultural aspects of language are bestowed with such recognition systems. The biological capacities required for language signal their presence through speech and understanding. This signaling cannot be invaded by “false green beards” because the traits and the signal of their presence are one and the same. However, the real usefulness of language comes from its potential to convey an infinite number of meanings through the dynamic handling of symbols – through language itself. But any specific language also signals its presence to others through usage and understanding. Thus, languages themselves cannot be invaded by “false green beards” because, again, the trait and the signal of its presence are one and the same. These twin green beards, in both the biological and cultural realms, are unique to language.

It is generally accepted that second language (L2) acquisition becomes more difficult as one grows older and that success in adult L2 acquisition is highly variable. Nevertheless, humans in language contact situations have to cope with intergroup communication. This article examines the ways society has responded to this challenge. It describes eleven strategies or workarounds that humans use to deal with this problem: developing lingua francas, forming Sprachbunds, practicing exogenous marriage, undergoing language shift, learning the target language imperfectly, developing a pidgin or creole, shifting the learning burden to children, relying on individual differences in aptitude among learners, developing the art and practice of translation, developing machine translation technology, and providing instruction. The eleven strategies can be understood as involving four broad societal responses to the burden of SLA: to simplify the learning task, to create a simpler language, to leave it to a special set of learners, and to provide special training. Finally, the article offers some speculations about the evolutionary reasons for the difficulties experienced in adult L2 acquisition.

The word-stock of a language is a complex dynamical system in which words can be created, evolve, and become extinct. Even more dynamic are the short-term fluctuations in word usage by individuals in a population. Building on the recent demonstration that word niche is a strong determinant of future rise or fall in word frequency, here we introduce a model that allows us to distinguish persistent from temporary increases in frequency. Our model is illustrated using a 10^8-word database from an online discussion group and a 10^11-word collection of digitized books. The model reveals a strong relation between changes in word dissemination and changes in frequency. Aside from their implications for short-term word frequency dynamics, these observations are potentially important for language evolution as new words must survive in the short term in order to survive in the long term.

This article studies a biased version of the naming game in which players located on a connected graph interact through successive conversations to bootstrap a common name for a given object. Initially, all the players use the same word B except for one bilingual individual who also uses word A. Both words are attributed a fitness, which measures how often players speak depending on the words they use and how often each word is pronounced by bilingual individuals. The limiting behavior depends on a single parameter: phi = the ratio of the fitness of word A to the fitness of word B. The main objective is to determine whether word A can invade the system and become the new linguistic convention. In the mean-field approximation, invasion of word A is successful if and only if phi > 3, a result that we also prove for the process on complete graphs relying on the optimal stopping theorem for supermartingales and random walk estimates. In contrast, for the process on the one-dimensional lattice, word A can invade the system whenever phi > 1.053 indicating that the probability of invasion and the critical value for phi strongly depend on the degree of the graph. The system on regular lattices in higher dimensions is also studied by comparing the process with percolation models.

Despite a burgeoning science of cultural evolution, relatively little work has focused on the population structure of human cultural variation. By contrast, studies in human population genetics use a suite of tools to quantify and analyse spatial and temporal patterns of genetic variation within and between populations. Human genetic diversity can be explained largely as a result of migration and drift giving rise to gradual genetic clines, together with some discontinuities arising from geographical and cultural barriers to gene flow. Here, we adapt theory and methods from population genetics to quantify the influence of geography and ethnolinguistic boundaries on the distribution of 700 variants of a folktale in 31 European ethnolinguis- tic populations. We find that geographical distance and ethnolinguistic affiliation exert significant independent effects on folktale diversity and that variation between populations supports a clustering concordant with European geography. This pattern of geographical clines and clusters paral- lels the pattern of human genetic diversity in Europe, although the effects of geographical distance and ethnolinguistic boundaries are stronger for folk- tales than genes. Our findings highlight the importance of geography and population boundaries in models of human cultural variation and point to key similarities and differences between evolutionary processes operating on human genes and culture.

There is disagreement about the routes taken by populations speaking Bantu languages as they expanded to cover much of sub-Saharan Africa. Here, we build phylogenetic trees of Bantu languages and map them onto geographical space in order to assess the likely pathway of expansion and test between dispersal scenarios. The results clearly support a scenario in which groups first moved south through the rainforest from a homeland somewhere near the Nigeria–Cameroon border. Emerging on the south side of the rainforest, one branch moved south and west. Another branch moved towards the Great Lakes, eventually giving rise to the monophyletic clade of East Bantu languages that inhabit East and Southeastern Africa. These phylogenies also reveal information about more general processes involved in the diversification of human populations into distinct ethnolinguistic groups. Our study reveals that Bantu languages show a latitudinal gradient in covering greater areas with increasing distance from the equator. Analyses suggest that this pattern reflects a true ecological relationship rather than merely being an artefact of shared history. The study shows how a phylogeographic approach can address questions relating to the specific histories of certain groups, as well as general cultural evolutionary processes.

As in biological evolution, multiple forces are involved in cultural evolution. One force is analogous to selection, and acts on differences in the fitness of aspects of culture by influencing who people choose to learn from. Another force is analogous to mutation, and influences how culture changes over time owing to errors in learning and the effects of cognitive biases. Which of these forces need to be appealed to in explaining any particular aspect of human cultures is an open question. We present a study that explores this question empirically, examining the role that the cognitive biases that influence cultural transmission might play in universals of colour naming. In a large-scale laboratory experiment, participants were shown labelled examples from novel artificial systems of colour terms and were asked to classify other colours on the basis of those examples. The responses of each participant were used to generate the examples seen by subsequent participants. By simulating cultural transmission in the laboratory, we were able to isolate a single evolutionary force—the effects of cognitive biases, analogous to mutation—and examine its consequences. Our results show that this process produces convergence towards systems of colour terms similar to those seen across human languages, providing support for the conclusion that the effects of cognitive biases, brought out through cultural transmission, can account for universals in colour naming.

A key aim in biology and psychology is to identify fundamental principles underpinning the behavior of animals, including humans. Analyses of human language and the behavior of a range of non-human animal species have provided evidence for a common pattern underlying diverse behavioral phenomena: words follow Zipf's law of brevity (the tendency of more frequently used words to be shorter), and conformity to this general pattern has been seen in the behavior of a number of other animals. It has been argued that the presence of this law is a sign of efficient coding in the information theoretic sense. However, no strong direct connection has been demonstrated between the law and compression, the information theoretic principle of minimizing the expected length of a code. Here we show that minimizing the expected code length implies that the length of a word cannot increase as its frequency increases. Furthermore, we show that the mean code length or duration is significantly small in human language, and also in the behavior of other species in all cases where agreement with the law of brevity has been found. We argue that compression is a general principle of animal behavior, that reflects selection for efficiency of coding.

An in-depth analytic study of a model of language dynamics is presented: a model which tackles the problem of the coexistence of two languages within a closed community of speakers taking into account bilingualism and incorporating a parameter to measure the distance between languages. After previous numerical simulations, the model yielded that coexistence might lead to survival of both languages within monolingual speakers along with a bilingual community or to extinction of the weakest tongue depending on different parameters. In this paper, such study is closed with thorough analytical calculations to settle the results in a robust way and previous results are refined with some modifications. From the present analysis it is possible to almost completely assay the number and nature of the equilibrium points of the model, which depend on its parameters, as well as to build a phase space based on them. Also, we obtain conclusions on the way the languages evolve with time. Our rigorous considerations also suggest ways to further improve the model and facilitate the comparison of its consequences with those from other approaches or with real data.

The study of the transmission of sign languages can give novel insights into the transmission of spoken languages1 and, more generally, into gene–culture coevolution. Over the years, several papers related to the persistence of sign language have been reported. All of these studies have emphasized the role of assortative (non-random) mating by deafness state (ie, a tendency for deaf individuals to partner together) for increasing the frequency of recessive deafness, and hence for the persistence of sign language in a population.

Studying language evolution has become resurgent in modern scientific research. In this revival field, approaches from a number of disciplines other than linguistics, including (paleo)anthropology and archaeology, animal behaviors, genetics, neuroscience, computer simulation, and psychological experimentation, have been adopted, and a wide scope of topics have been examined in one way or another, covering not only world languages, but also human behaviors, brains and cultural products, as well as nonhuman primates and other species remote to humans. In this paper, together with a survey of recent findings based on these many approaches, we evaluate how this multidisciplinary perspective yields important insights into a comprehensive understanding of language, its evolution, and human cognition.

The paper argues that the structure, derivation and evolution of syntax is given by the sequence (elements, concatenation, embedding). We discuss the implications of this sequence for language and the numeral system in general and on the evolution of language in particular. A four-stage model of the evolution of syntax, broadly compatible with several earlier scenarios, is proposed. The four stages are (1) signs, (2) increased number of signs, (3) commutative concatenation, and (4) noncommutative concatenation. We support the model by showing that its stages can be adaptive per se, which could explain why they evolved. We also identify two preconditions for maintaining the stages: stage (2) depends on the ability to conceptualize asymmetric relations between concepts and the adaptiveness of stage (3) depends on cultural constraints on linguistic interpretation.

We propose a novel account for the emergence of human language syntax. Like many evolutionary innovations, language arose from the adventitious combination of two pre-existing, simpler systems that had been evolved for other functional tasks. The first system, Type E(xpression), is found in birdsong, where the same song marks territory, mating availability, and similar “expressive” functions. The second system, Type L(exical), has been suggestively found in non-human primate calls and in honeybee waggle dances, where it demarcates predicates with one or more “arguments,” such as combinations of calls in monkeys or compass headings set to sun position in honeybees. We show that human language syntax is composed of two layers that parallel these two independently evolved systems: an “E” layer resembling the Type E system of birdsong and an “L” layer providing words. The existence of the “E” and “L” layers can be confirmed using standard linguistic methodology. Each layer, E and L, when considered separately, is characterizable as a finite state system, as observed in several non-human species. When the two systems are put together they interact, yielding the unbounded, non-finite state, hierarchical structure that serves as the hallmark of full-fledged human language syntax. In this way, we account for the appearance of a novel function, language, within a conventional Darwinian framework, along with its apparently unique emergence in a single species.

Children learn their native language by exposure to their linguistic and communicative environment, but apparently without requiring that their mistakes be corrected. Such learning from “positive evidence” has been viewed as raising “logical” problems for language acquisition. In particular, without correction, how is the child to recover from conjecturing an over-general grammar, which will be consistent with any sentence that the child hears? There have been many proposals concerning how this “logical problem” can be dissolved. In this study, we review recent formal results showing that the learner has sufficient data to learn successfully from positive evidence, if it favors the simplest encoding of the linguistic input. Results include the learnability of linguistic prediction, grammaticality judgments, language production, and form-meaning mappings. The simplicity approach can also be “scaled down” to analyze the learnability of specific linguistic constructions, and it is amenable to empirical testing as a framework for describing human language acquisition.

Learning theory has frequently been applied to language acquisition, but discussion has largely focused on information theoretic problems—in particular on the absence of direct negative evidence. Such arguments typically neglect the probabilistic nature of cognition and learning in general. We argue first that these arguments, and analyses based on them, suffer from a major flaw: they systematically conflate the hypothesis class and the learnable concept class. As a result, they do not allow one to draw significant conclusions about the learner. Second, we claim that the real problem for language learning is the computational complexity of constructing a hypothesis from input data. Studying this problem allows for a more direct approach to the object of study—the language acquisition device—rather than the learnable class of languages, which is epiphenomenal and possibly hard to characterize. The learnability results informed by complexity studies are much more insightful. They strongly suggest that target grammars need to be objective, in the sense that the primitive elements of these grammars are based on objectively definable properties of the language itself. These considerations support the view that language acquisition proceeds primarily through data-driven learning of some form.

Quantitative phylogenetic methods have been used to study the evolutionary relationships and divergence times of biological species, and recently, these have also been applied to linguistic data to elucidate the evolutionary history of language families. In biology, the factors driving macroevolutionary processes are assumed to be either mainly biotic (the Red Queen model) or mainly abiotic (the Court Jester model) or a combination of both. The applicability of these models is assumed to depend on the temporal and spatial scale observed as biotic factors act on species divergence faster and in smaller spatial scale than the abiotic factors. Here, we used the Uralic language family to investigate whether both ‘biotic’ interactions (i.e. cultural interactions) and abiotic changes (i.e. climatic fluctuations) are also connected to language diversification. We estimated the times of divergence using Bayesian phylogenetics with a relaxed-clock method and related our results to climatic, historical and archaeological information. Our timing results paralleled the previous linguistic studies but suggested a later divergence of Finno-Ugric, Finnic and Saami languages. Some of the divergences co-occurred with climatic fluctuation and some with cultural interaction and migrations of populations. Thus, we suggest that both ‘biotic’ and abiotic factors contribute either directly or indirectly to the diversification of languages and that both models can be applied when studying language evolution.

A new model called Naming Game with Multiple Hearers (NGMH) is proposed in this paper. A naming game over a population of individuals aims to reach consensus on the name of an object through pair-wise local interactions among all the individuals. The proposed NGMH model describes the learning process of a new word, in a population with one speaker and multiple hearers, at each interaction towards convergence. The characteristics of NGMH are examined on three types of network topologies, namely ER random-graph network, WS small-world network, and BA scale-free network. Comparative analysis on the convergence time is performed, revealing that the topology with a larger average (node) degree can reach consensus faster than the others over the same population. It is found that, for a homogeneous network, the average degree is the limiting value of the number of hearers, which reduces the individual ability of learning new words, consequently decreasing the convergence time; for a scale-free network, this limiting value is the deviation of the average degree. It is also found that a network with a larger clustering coefficient takes longer time to converge; especially a small-word network with smallest rewiring possibility takes longest time to reach convergence. As more new nodes are being added to scale-free networks with different degree distributions, their convergence time appears to be robust against the network-size variation. Most new findings reported in this paper are different from that of the single-speaker/single-hearer naming games documented in the literature.

A hallmark of the human language faculty is the use of syntactic rules. The natural vocalizations of animals are syntactically simple, but several studies indicate that animals can detect and discriminate more complex structures in acoustic stimuli. However, how they discriminate such structures is often not clear. Using an artificial grammar learning paradigm, zebra finches were tested in a Go/No-go experiment for their ability to distinguish structurally different three-element sound sequences. In Experiment 1, zebra finches learned to discriminate ABA and BAB from ABB, AAB, BBA, and ABB sequences. Tests with probe sounds consisting of four elements suggested that the discrimination was based on attending to the presence or absence of repeated A- and B-elements. One bird generalized the discrimination to a new element type. In Experiment 2, we continued the training by adding four-element songs following a ‘first and last identical versus different’ rule that could not be solved by attending to repetitions. Only two out of five birds learned the overall discrimination. Testing with novel probes demonstrated that discrimination was not based on using the ‘first and last identical’ rule, but on attending to the presence or absence of the individual training stimuli. The two birds differed in the strategies used. Our results thus demonstrate only a limited degree of abstract rule learning but highlight the need for extensive and critical probe testing to examine the rules that animals (and humans) use to solve artificial grammar learning tasks. They also underline that rule learning strategies may differ between individuals.

Global linguistic diversity is rapidly declining. As the world becomes less linguistically diverse, it is becoming culturally less diverse as well as the world’s tribes and languages are dying out or being assimilated into modern civilization because their habitats are being destroyed. At the same time the world is experiencing a substantial decline in biodiversity. The extinction of languages is part of the larger picture of near total collapse of the worldwide ecosystem, and languages are vital parts of complex local ecologies that must be supported if global biodiversity is to be maintained.

Maynard Smith and Szathmáry (1995) propose a series of major transitions in the evolutionary history of life. Their work provides a rich framework for thinking about replication. They identified the importance of language in this light, but language is a new system of replication in more than one sense: it is both an enabler of cultural replicators with unlimited heredity, and also a new kind of evolutionary system itself. Iterated learning is the process of linguistic transmission, and it drives both language change and the transitions to qualitatively new kinds of linguistic system. By seeing language as an evolutionary system, the biggest payoff we get may be the ability to take biologists’ insights into the evolution of life and apply them to the evolution of language.

Human vocalization systems are characterized by complex structural properties. They are combinatorial, based on the systematic reuse of phonemes, and the set of repertoires in human languages is characterized by both strong statistical regularities—universals—and a great diversity. Besides, they are conventional codes culturally shared in each community of speakers. What are the origins of the forms of speech? What are the mechanisms that permitted their evolution in the course of phylogenesis and cultural evolution? How can a shared speech code be formed in a community of individuals? This chapter focuses on the way the concept of self-organization, and its interaction with natural selection, can throw light on these three questions. In particular, a computational model is presented which shows that a basic neural equipment for adaptive holistic vocal imitation, coupling directly motor and perceptual representations in the brain, can generate spontaneously shared combinatorial systems of vocalizations in a society of babbling individuals. Furthermore, we show how morphological and physiological innate constraints can interact with these self-organized mechanisms to account for both the formation of statistical regularities and diversity in vocalization systems.

The complex inter-relationships between genetics and linguistics encompass all four scales highlighted by the contributions to this book and, together with cultural transmission, the genetics of language holds the promise to offer a unitary understanding of this fascinating phenomenon. There are inter-individual differences in genetic makeup which contribute to the obvious fact that we are not identical in the way we understand and use language and, by studying them, we will be able to both better treat and enhance ourselves. There are correlations between the genetic configuration of human groups and their languages, reflecting the historical processes shaping them, and there also seem to exist genes which can influence some characteristics of language, biasing it towards or against certain states by altering the way language is transmitted across generations. Besides the joys of pure knowledge, the understanding of these three aspects of genetics relevant to language will potentially trigger advances in medicine, linguistics, psychology or the understanding of our own past and, last but not least, a profound change in the way we regard one of the emblems of being human: our capacity for language.

This chapter discusses language processing during conversation. In particular, it considers why taking part in a conversation is more straightforward than speaking or listening in isolation. We argue that conversation is easy because speakers and listeners automatically align with each other at different linguistic levels (e.g., sound, grammar, meaning) which leads to alignment at the level of interpretation. This alignment process is reflected in the repetitiveness of dialogue at different levels and occurs both on the basis of local mechanisms of priming and more global mechanisms of routinization. We argue that the latter process may tell us something about both acquisition of language and historical processes of language change.

This paper proposes a new approach to simulating language evolution; it expands on the original work done by Lee and Zadeh on Fuzzy Grammars and introduces a Type-2 Fuzzy Grammar. Ants in an Ant Colony Optimization algorithm are given the ability of embedding a message on the pheromone using a Type-2 Fuzzy Grammar. These ants are able to gradually adopt a foreign language by adjusting the grades of membership of their grammar. Results that show the effect of uncertainty in a language are given.

Complex adaptive systems consist of a large number of interacting agents. Agents are goal-directed, cognitive individuals capable of perception, information processing and action. However, agents are intrinsically “bounded” in their rational understanding of the system they belong to, and its global organization tends to emerge from local interactions, resulting in a coordination of the agents and their actions. This coordination minimizes conflict or friction, while facilitating cooperation or synergy. The basic mechanism is the reinforcement of synergetic interactions and the suppression of conflictual ones. As a result, the system as a whole starts to behave like an integrated cognitive “superagent”. The author presents several examples of this process of spontaneous coordination that leads to distributed cognition, including the emergence of a shared vocabulary, the development of standard referential expressions, the evolution of transmitted ideas (memes) towards more stereotypical forms, and the aggregation of diverse experiences into collective decisions, in which the system as a whole is more intelligent than its individual components. These phenomena have been investigated by means of multi-agent computer simulations and social psychological experiments.

Like an increasing number of linguists and other scholars especially interested in the evolution and/or the ontogenetic development of language, the author claims that languages are complex adaptive systems (CAS). These have been characterized as reflecting complex dynamics of interactive agents, experiencing constant instability, and in search for equilibrium in response to changes in the ecologies of their usage. Putatively, thanks to self-organization, transitional moments of apparent stability obtain during which patterns and systems emerge, and evolutions obtain from the alternations of periods of instability and stability in seemingly unpredictable ways. The author addresses the issues of the many interpretations of ‘complexity’ applying to language(s), of the description of the interactive agents that produce the above characteristics, of the emergence of complexity in language(s) from the point of view of language evolution, of the kind(s) of evidence that support(s) the various interpretations of ‘complexity’ that are conceivable, of the way in which complexity in language compares with complexity in other non-linguistic phenomena, and of the causes of the “chaos” which prompts languages to reorganize themselves into new systems.

In the present thesis, first, the principle of center-embedded (CE) recursion is explained. Next, we briefly discuss animal studies on recursion learning, followed by a section with theories and experimental evidence about human learning. Here, the complexity of the principle is contrasted with pragmatic learning strategies. Then, we discuss the features of the input that might help recursion learning. Finally, we discuss the methodological issues regarding the use of artificial language to study aspects of natural language learning.