Two of the main theoretical approaches to the evolution of language are biolinguistics and usage-based approaches. Both are often conceptualized as belonging to seemingly irreconcilable ‘camps.’ Biolinguistic approaches assume that the ability to acquire language is based on a language-specific genetic foundation. Usage-based approaches, on the other hand, stress the importance of domain-general cognitive capacities, social cognition, and interaction. However, there have been a number of recent developments in both paradigms which suggest that biolinguistic and usage-based approaches are actually moving (...) closing together. For example, theoretical advancements such as evo-devo and complex adaptive systems theory have gained traction in the language sciences, leading to changed conceptions of issues like the relative influence of “nature” and “nurture.” In this paper, we outline points of convergence between current minimalist biolinguistic and usage-based approaches regarding four contentious issues: a) modularity and domain-specificity, b) innateness and development, c) cultural and biological evolution, d) knowledge of language and its description. We show that across both paradigms, researchers have come to increasingly embrace more complex views of these issues. They also have come to appreciate the view that biological and cultural evolution are closely intertwined, which leads to an increased amount of common ground between minimalist biolinguistics and usage-based approaches. (shrink)

The article reconstructs the main lines of three hypotheses in the current literature concerning the evolutionary pace which characterized the natural history of human language: the “continuist” and gradualist perspective, the “discontinuist” and evolution-free perspective, and the “punctuationist” view. This current debate appears to have a long history, which starts at least from Darwin’s time. The article highlights the similarities between the old and the modern debates in terms of history of ideas, and it shows the current limits of each (...) of these perspectives. In the final part, we present an alternative approach which considers human language not as a single trait, but as the result of a mosaic of different elements, some of which recently evolved, and some others gradually evolved and phylogenetically ancient. When matched with a tree thinking and comparative perspective, this view suggests that expressions of symbolic and linguistic behaviors in other human species could have preceded the more complex and systematic behaviors showed by Homo sapiens. (shrink)

In their recent book, Ladyman and Wiesner (What is a complex system?, Yale University Press, 2020) delineate the bounds of the exciting interdisciplinary field of complexity science. In this work, they provide examples of generally accepted complex systems and common features which these possess to varying degrees. In this paper, I plan to extend their list to include the formal study of natural language, i.e. linguistics. In fact, I will argue that language exhibits many of the hallmarks of a complex (...) system, specifically a complex biological system. Thus, my aim is to advocate contra the the ‘Minimalist Program’ (Chomsky, The minimalist program, MIT Press, Cambridge, 1995), which motivates simple underlying mechanisms (i.e. Merge) in their idealisations, that biolinguistics should embrace a ‘Maximalist Program’ in which multiple subfields contribute component explanations to an emerging whole. (shrink)

To what degree can language be acquired from linguistic input alone? This question has vexed scholars for millennia and is still a major focus of debate in the cognitive science of language. The complexity of human language has hampered progress because studies of language–especially those involving computational modeling–have only been able to deal with small fragments of our linguistic skills. We suggest that the most recent generation of Large Language Models (LLMs) might finally provide the computational tools to determine empirically (...) how much of the human language ability can be acquired from linguistic experience. LLMs are sophisticated deep learning architectures trained on vast amounts of natural language data, enabling them to perform an impressive range of linguistic tasks. We argue that, despite their clear semantic and pragmatic limitations, LLMs have already demonstrated that human-like grammatical language can be acquired without the need for a built-in grammar. Thus, while there is still much to learn about how humans acquire and use language, LLMs provide full-fledged computational models for cognitive scientists to empirically evaluate just how far statistical learning might take us in explaining the full complexity of human language. (shrink)

Howson famously argues that the no-miracles argument, stating that the success of science indicates the approximate truth of scientific theories, is a base rate fallacy: it neglects the possibility of an overall low rate of true scientific theories. Recently a number of authors has suggested that the corresponding probabilistic reconstruction is unjust, as it concerns only the success of one isolated theory. Dawid and Hartmann, in particular, suggest to use the frequency of success in some field of research \ to (...) infer a probability of truth for a new theory from \. I here shed doubts on the justification of this and similar moves and suggest a way to directly bound the probability of truth. As I will demonstrate, my bound can become incompatible with the assumption specific testing and Dawid and Hartmann’s estimate for success. (shrink)

The topic of this dissertation is the nature of linguistic competence, the capacity to understand and produce sentences of natural language. I defend the empiricist account of linguistic competence embedded in the connectionist cognitive science. This strand of cognitive science has been opposed to the traditional symbolic cognitive science, coupled with transformational-generative grammar, which was committed to nativism due to the view that human cognition, including language capacity, should be construed in terms of symbolic representations and hardwired rules. Similarly, linguistic (...) competence in this framework was regarded as being innate, rule-governed, domain-specific, and fundamentally different from performance, i.e., idiosyncrasies and factors governing linguistic behavior. I analyze state-of-the-art connectionist, deep learning models of natural language processing, most notably large language models, to see what they can tell us about linguistic competence. Deep learning is a statistical technique for the classification of patterns through which artificial intelligence researchers train artificial neural networks containing multiple layers that crunch a gargantuan amount of textual and/or visual data. I argue that these models suggest that linguistic competence should be construed as stochastic, pattern-based, and stemming from domain-general mechanisms. Moreover, I distinguish syntactic from semantic competence, and I show for each the ramifications of the endorsement of a connectionist research program as opposed to the traditional symbolic cognitive science and transformational-generative grammar. I provide a unifying front, consisting of usage-based theories, a construction grammar approach, and an embodied approach to cognition to show that the more multimodal and diverse models are in terms of architectural features and training data, the stronger the case is for the connectionist linguistic competence. I also propose to discard the competence vs. performance distinction as theoretically inferior so that a novel and integrative account of linguistic competence originating in connectionism and empiricism that I propose and defend in the dissertation could be put forward in scientific and philosophical literature. (shrink)