Abstract

The cognitive revolution in the middle of the last century has transformed the ways in which we study the human mind. Curiously, when it comes to language there is a growing behaviourist trend, where it is regarded as an acquired skill, not unlike the way in which Large Language Models (LLMs) work. In contrast, linguists in the Generative Grammar tradition consider the faculty of language to be a computational system within the mind, part of the human biological endowment. This means that biological aspects of language, in particular evolution, development, and (neural) mechanisms, are open to investigation. I will discuss recent work on ‘comparative linguistics’, particularly the behavioural, neural and cognitive parallels between human language and birdsong, and what we can and cannot conclude from it. The current behaviourist view of language has led to the rapid rise of LLMs, although these AI models are actually not about language at all. Natural language appears to be unique to the human mind, and has no parallels either in animal or artificial intelligence.