Language unlimited: The science behind our most creative power by David Adger (review)

David Adger is the author of one of the most influential linguistics textbooks ever written. His introduction to generative linguistics, Core syntax: A minimalist approach (Adger 2003), has impacted a whole generation of young linguists, including the author of this review—who is therefore excited about having been invited to review A's most recent book, Language unlimited: The science behind our most creative power.

This book is not an updated version of his previous linguistics textbook; it is rather an introduction to the philosophy behind generative linguistics. A states this clearly somewhere in the middle of his book: 'This book isn't about how to do linguistics, it's about what the nature of language is, so I'm only going to scratch the surface of why the [linguistic] structures look as they do' (173).

Fair enough. As a reviewer, I would thus like to look in more detail at A's general points about 'the nature of language', rather than delving into the linguistic details that A uses in some of the chapters to illustrate basic characteristics of human language. Those details range from his creation of the language of the Warigs (known from the British TV series Beowulf; Ch. 5), over uncontroversial facts about binding theory (Ch. 7), to fine-grained dialectal data (Ch. 10). A refers to the technical literature in his footnotes, and as a reviewer I think that meticulously discussing and evaluating these aspects of the book would not do justice to the book's overall messages and mission, which is to convey to a broad audience—whose training is not necessarily in linguistics—what the core of our amazing capacity called human language is.

In this review, I would therefore like to focus on three fundamental points that A makes in his book: (i) human uniqueness (language qualitatively distinguishes us from other species), (ii) Merge is the source of human creativity (the syntactic operation Merge allows us to use language in a limitless fashion and is the source of building new worlds of ideas), and (iii) the limits of grammar (many social and emotional aspects of language are never co-opted into our grammar). In what follows, I spend most of the limited space that I have in this review on discussing and evaluating those three points because they are the main messages of the book. In line with this, I do not provide the reader with detailed chapter-by-chapter summaries because I often think that those consume too many paragraphs in book reviews. Accordingly, the following single paragraph should suffice to indicate to the reader how the book's ten chapters are organized. After this outline, I turn to the evaluation of A's three fundamental points about the nature of language.

In Ch. 1, A argues that language imposes structure on our experience and thus cannot be learned from mere experience and pattern learning, and that it is different from the mere use of symbols and signals (Ch. 2). Rather, language requires what A calls 'a sense of structure' (Ch. 3), which is more or less Chomsky's famous concept of universal grammar. Ch. 4 elaborates on the idea of universal grammar (read: an innate biological endowment for language) by discussing the phenomenon of homesign languages where deaf children create a language without receiving the relevant input from their hearing parents. Chs. 5, 6, and 7 sketch some universal features of language (the observation that one can come up with impossible patterns that are nowhere found in the world's languages; the fact that language structure involves only discrete and no gradual categories; and the idea that hierarchy and not order is crucial for language structure). Ch. 8 builds on those three universal features and demonstrates that AIs like Amazon's Alexa don't really learn languages based on human principles of language structure. Ch. 9 introduces the basic formal apparatus for generating human language structure by...