Editor's Note: The following reply addresses content in the review for The Philosophy of Theoretical Linguistics: A Contemporary Outlook (Nefdt, 2024; Cambridge University Press) which can be read in its entirety here: https://linguistlist.org/issues/36/243/.

As a writer of a monograph and a scholar in general, it is always pleasing to see someone engage thoroughly with your work. This feeling holds whether or not that person was particularly laudatory or critical. In fact, critical reviews have immense merit. However, when someone chooses to produce an unbalanced picture of your work, emphasising its faults without so much as a glance at its potential merits, then some nuance and balance needs to be restored. Thus, I’m writing this review of the review, hoping to not only defend my book but to also evince the balance and nuance which its target neglected.

In a recent review, Begley lists a number of grievances with my new book The Philosophy of Theoretical Linguistics: a contemporary outlook (Cambridge University Press 2024). In fact, that’s really all he does. Specifically, he lists grievances with some of the framing of the book and its philosophy. He is rather silent on the main aims of the book, such as providing a philosophy of linguistics, in an eclectic manner hitherto unattempted. He very briefly and purely descriptively mentions the pivotal chapters on syntax, pragmatics, and phonology. He claims the chapter on syntax is “largely well executed” then immediately gripes that a particular definition was used before it was provided a few lines later.

But this chapter did so much more! It described and analysed not only the core scientific claims of Minimalism in generative syntax but found commonalities with dependency grammar, construction grammar, the parallel architecture, head-driven phrase structure grammar and model-theoretic approaches as well as lexical functional grammar. It compared these frameworks in terms of questions of possible languages without syntax and, more importantly, it asked what is the central scientific project in syntactic theory. Many people might not agree with my analysis or conclusions but surely a review of a book which spends over a tenth of its words on a topic should receive more than a nod and quibble?

Earlier on in the review he discusses the subsection of the introduction I devote to formal approaches and finite state automata (FSA). He states that “the question regarding what these “machines” are goes overlooked”. Again, this is a short subsection of a chapter that discussions formal approaches, social scientific approaches, describes each chapter of the book and introduces my general view of theoretical linguistics. And yet, I do discuss what are:

Every automaton is an accepting machine. It can be paired with a particular formal grammar and language.
Furthermore, what it accepts are sequences of strings of a certain complexity. (pg 11)

I compare them to phrase structure grammars, include a footnote on the automata associated with those, and have a figure devoted to FSA! Nevertheless, Begley is correct that my critique of why more complex grammars (like unrestricted Turing machines and the like) are inadequate as grammars of nature language (and its cognition) does not mention issues of computational tractability. I’ve written many times on these issues and my worry was that including such material in such a short subsection would lead to a perfunctory level description (although I do mention computational intractability and parsing in FLT on pg 30). Still, I concede that this is an omission and I should have had the skill to do better there.

Then Begley takes me to task for a throwaway comment about platonism (which I gather is an important topic for him). He questions who the “hard-line platonists” are that I claim would be satisfied with linguistics if all the facts about formal languages were settled. He then defends Katz and recommends his own work as a tonic. But perhaps I didn’t mention the referents of my term because I was being intentionally vague so as not to pick any individuals out? I wanted to gesture at a possible position, i.e. someone who believes that languages are abstract objects and linguistics is mathematics simpliciter. Maybe no such individual exists. Again, this is Begley’s fourth paragraph on the introduction.

His comments on my second chapter are not entirely fair, in my view. He describes me as saying that ontological debates are absent from science. I do no such thing nor do I endorse the irrelevance of ontology (in fact, that would denigrate much of my own work including my first book, Nefdt (2023)). What I do say is the following:

It’s interesting to note that these ontological debates are largely absent in other scientific domains. No one in the cognitive neuroscience of memory asks how individual tokens or memories are related to some platonistic concept of Memory, divorced from its instantiations. I thank Wolfram Hinzen for this particular observation. (pg 30)

If it wasn’t clear, this is a footnote. The important words are “these ontological debates” referring to debates about the abstracta which individual languages are meant to instantiate. I stand by the actual meaning of the claim being made here. Again a few lines later, he states that I hold that Michael Devitt’s position “lacks overt ontological commitment” but Begley’s fetish for footnotes disappears here given that my claim is immediately linked to a numeral pointing to a note about Devitt’s ontological baggage.

Begley then goes on to bash my modal metaphysical scaffolding for the question of what a possible human language is. He doesn’t discuss my axioms or their plausibility but takes issue with my using S5 and my “loose” characterisations. While I agree that I could have been more careful in places (even though I describe my account as “dipping my feet into modal logic”, I do not agree that I needed to do a deep dive into modal metaphysics. This is the case simply because that is not the point of the chapter. So to say “[t]his is unusual for a purported work of philosophical logic” is interesting since the work involves no such purport. I actually state exactly this before supplementing my use of the S5 system in accordance with my view.

We’ll skip over the details of my chapter on semantics, since Begley does, to focus on his issue with my naturalistic stance. In the chapter, I suggest that there are two ways to do metasemantics, one is via traditional metaphysics, and the other is via metascientific reflection. Begley considers the former to be “compulsory” and the naturalism I endorse to be “anti-philosophical”. I can’t enter into a debate about very different approaches to philosophy (and science) here. I’ve advocated for scientific naturalism in other places (as has Quine, Maddy, Ladyman, Ross, Machery, and many others) and indeed I assumed it here. Perhaps my dismissal of abstract metaphysics was too harsh or unfair for Begley’s liking. I’ll own that. I’ll also own the shorthand which Begley considers conflation in my discussion of meaning representations. But as the reader who made it this far might have anticipated, his characterisation is far from charitable. In fact, I clearly state that “[t]here are methodological issues such as whether semantic value is best modelled as truth conditions, dynamic context-change potentials, or numerical vectors representing collocational data” (pg 84) prior to slipping into shorthand. So attributing me with a view that confuses vectors (which are data structures, as Begley helpfully points out) with actual meanings is not only unfair but misleading. Again, this is an area in which I have notable publications.

He states that I make subword level semantics out to be a new trend. Clearly my references from works published over thirty years ago are not enough to dispel this impression. He then states “I would have expected some of that history to be at least mentioned here” after adding some references to earlier work. But again, in my Preface I mentioned that I would not be directly engaging in historical reflection except in specific cases. Begley might think this case called for it but since I never indicated that these issues were new as he claims, I disagree.

This metareview is getting lengthy as Begley’s mischaracterisations and uncharitable readings seem too many to address. So I’ll mention one more clear case and then conclude with my entreaty to readers of his review and potential readers of my book.

In the computational linguistics chapter, again there is an argument not discussed in the review, Begley makes many accusations. He accuses me of ‘commonly’ misunderstanding the black-box problem in NLP (despite the fact that I have actually contributed to the literature), infelicities, a typo, and who knows what else. But there’s a point at which he claims that I use quantum mechanics as a foil to the explanation-first approach of the philosophy of science and generative linguistics. He then adds that my example is “somewhat self-defeating” as one “might nonetheless aim for the ideal of explanation and understanding, although one temporarily settles for shutting up and calculating if that is all one can do.” Fair enough. But there is a big controversy as to whether this was indeed the position or how engineering successes were implicated in this history, Einstein versus Bohr, Kuhn’s involvement etc. If only I had referenced some text that discusses this or flagged these issues. Oh wait, I did both in a footnote linked to the paragraph.

Begley’s final evaluation of my book is unsurprisingly uncharitable and yet he suggests that “[a]nother one hundred pages might have been needed to rectify” many of the issues with pacing and pitch. This reminds me of the joke describing disgruntled patrons of a restaurant in the Catskill Mountains who complain that the food is horrible…and in such small portions. It seems clear to me that Begley did not like the book and that, of course, is his right. However, using mischaracterisaton and what seemed to be angry, misleading arguments to push his personal agenda seems less appropriate (especially in print). There were many things I could have done better than I did. I agree that there were pacing issues and some topics might have needed more attention (and others less). But I stand by my work and the merits of offering a “song-bird’s eye view” of theoretical linguistics with a particular philosophical lens. I hope that other readers can find those merits should they choose to read the book.

Ryan M. Nefdt, PhD
University of Cape Town and University of Bristol

Linguistic Field(s): General Linguistics
Linguistic Theories