Current Approaches to Syntax is the third handbook published in De Gruyter's series Comparative Handbooks of Linguistics. Each volume in the series focuses on a designated linguistic theme, and each contains a collection of studies that are written and structured in accordance with the overall design features specified by the volume editors. The aim is to make cross-linguistic and cross-theoretical comparison possible. The handbook under review is an exceptional item in this series (so far) inasmuch as the target of comparison is not a selected set of natural language phenomena, but the syntactic theories that study such phenomena. It is thus a metatheoretical enterprise that invites the reader to study and reflect upon the variation attested across different syntactic theories, and upon the process of syntactic theorizing.
… a systematic presentation of a set of contemporary syntactic theories by defining parameters of comparison, applying these parameters consistently and conspicuously across the chapters, and probing into their metatheoretical foundations.
(Chapter 1, pp. 15-16.)
These parameters of comparison are (i) the goals of the theory, (ii) the data selected for syntactic theorizing, (iii) the conceptual tools employed in syntactic descriptions, and (iv) the (self-)evaluation of the theory in terms of empirical adequacy, consistency, and simplicity. In addition, the editors asked the linguist authors of the syntax chapters to provide an analysis of an English sentence in the theoretical model that they use. This sentence was the same for everyone. Chapter 21, written by András Kertész and Csilla Rákosi, concludes the volume with the message that metatheoretical reflection is not only an indispensable tool to improve syntactic theorizing, but it may also contribute to the enrichment of our knowledge of natural language syntax.
Part I of the book contains thirteen chapters that each describe individual syntactic theories. These chapters are written by prominent practitioners of the respective approaches. Part II includes six chapters on the metatheoretical foundations, plus Chapter 21 by Kertész & Rákosi, which concludes the volume. The metatheoretical studies focus on issues that have figured prominently in philosophical discussions with a bearing on linguistic metatheory, and that have, in part, also received attention within the linguistic community in the last decades. These issues include, among others, the data problem with all its ramifications, the presence and the proper treatment of inconsistencies at different levels of syntactic theorizing, or various aspects of the complex task of comparing and evaluating syntactic frameworks.
The editors and the authors of the individual chapters (especially of those in Part II of the handbook) invite and, indeed, inspire the linguistic reader to reflect upon both the practical and the theoretical aspects of their work. This is exactly what the author of the current review, a practicing linguist himself, sets out to do here: I make some inspired notes over the contemporary syntactic landscape, discussing and commenting on pertinent locations in the handbook.
The central objective of the handbook is the scientific investigation of the field of syntactic theorizing, and the editors themselves take care to avoid any bias in the selection and the presentation of the particular syntactic theories that figure in the book. The chapters that describe these theories in Part I follow each other in the alphabetical order of the (first) authors. The emerging list is the following: Cognitive Grammar, Construction Grammar, Simpler Syntax, Lexical Functional Grammar, the Decathlon Model, the Minimalist Program, the Parallel Architecture, Usage-based Grammar, Optimality-theoretic Syntax, the Functional Discourse Grammar approach to syntax, Head-Driven Phrase Structure Grammar, Dependency Grammar, and Combinatory Categorial Grammar. This list is not fully comprehensive, as some contemporary or near-contemporary approaches are acknowledged to have been left out (see Chapter 1, p. 2), but the selection is still representative. In his review of the handbook, Imrényi (2019) expresses his criticism concerning what he considers to be an underrepresentation of functional or cognitive approaches in the volume, and of usage-based grammars in general. It is true that generative grammars form a major part of this selection, and Chomskyan generative grammar in particular is a central and a recurrent point of reference in the metatheoretical chapters in Part II. But this attention to generative grammar(s) does reflect fairly on the current state of affairs if relative prominence of a theory or of a paradigm is measured in terms of the number of its practitioners, or of the publications or conference talks that are couched in the framework. Though the volume itself reports on little specific data in support of the previous claim, topic selection within the volume certainly shows no bias in this weighted sense of prominence.
Having said that, the thirteen syntactic theories in Part I are not equally familiar to everyone in the linguistic community. In fact, an average linguist would probably not be widely knowledgeable about more than half of these linguistic theories. To serve the needs of both the general linguistic community and of the wider potential readership, it would have been useful to include some brief encyclopaedic summary where the major tenets, the founders or the most important figures of the approach, key publications, theory specific conferences and journals, dedicated websites, or similar useful information are listed. Much of this information, including comments on the history of the respective theories, can be found in the pertinent chapters, but it is not always there or it is not presented in a transparent manner. Admittedly, the handbook is not meant to be a reference book on contemporary syntactic theory in this sense. Still, such an encyclopaedic section would have been a useful addition to the otherwise carefully constructed auxiliary compartment including the author, language and subject indices.
… To identify differences between two objects, some shared properties must be identified as points of reference. There are two sources of necessarily shared features of syntactic theories. One is their subject matter: they are all about syntax; the other is their conceptual status: they are theories.
(Chapter 1, p. 2.)
But it is not immediately obvious what this conceptual status exactly is so that we can regard individual syntactic theories as discreet entities populating the domain of syntactic theorizing. Note in this respect that it is explicitly stated in Chapter 1 (p. 3.) that the notions theory, approach, or framework are used interchangeably in the book. In other words, these three terms are taken to be equivalent. I have the feeling that approach or framework are used in linguistic practice when we are not entirely certain that what is described is a theory in the proper, purportedly well-defined sense of the term ‒ if there indeed exists such a well-defined understanding of what a theory is. The term research programme often seems to be used in the book as a similar, somewhat less official alternative to the Kuhnian notion of paradigm.
It is also interesting to observe that the naming practice of the syntactic theories under investigation varies in a related manner. Most of these theories aim at being general theories of language, even if syntax may receive core attention in certain cases. Yet seven of these approaches simply call themselves grammars. Each theory is equipped with tools to develop any arbitrary number of grammars even for the same language, as is often emphasized in more formal-minded approaches like Head-Driven Phrase Structure Grammar (HPSG) or Lexical Functional Grammar (LFG). So in that technical sense of the word, a grammar is a specific model of a language, and not a theory. The rest of the approaches discussed in the book call themselves some sort of syntax, model, program or architecture, but nobody calls themselves a theory. Perhaps all this variation in terminology could be dismissed as irrelevant, but it may as well reflect upon some underlying uncertainty about the exact ontological status of one's chosen framework, or about how linguists themselves relate to and conceive of this framework.
The central problem addressed by Combinatory Categorial Grammar (CCG) is the nature of the mapping between sound and meaning.
(Chapter 14, p. 389: Combinatory Categorial Grammar)
As its very name suggests, CG aims at offering a cognitively plausible theory of grammar.
(Chapter 2, p. 24: Cognitive Grammar)
Whereas in mainstream Generative Grammar non-productively derived constructions have tended to be regarded as exceptional phenomena outside of the core interests of the grammarian, Construction Grammar treats idiosyncratic uses as essential and integral parts of language, worthy of study on their own right. Yet other contributors of the volume make it explicit that they share many of their research goals with other approaches:
Like many other approaches to language, the goal of CxG is to arrive at an explicit, contradiction-free, and generalization-prone model of natural language which has the widest possible empirical coverage, from the highly idiomatic and rigid to the fully productive and compositional.
(Chapter 3, p. 50: Construction Grammar)
The goal of research is similar to the goals of Construction Grammar and hence also includes many of the goals of Mainstream Generative Grammar (GB, Minimalism and variants thereof): we want to understand language as a cognitive system, we want to understand which properties are common to all languages and how languages may vary, we want to understand how natural language can be acquired and processed.
(Chapter 12, p. 319: HPSG)
If we abide by this description alone, then there indeed does not appear to be much substantial difference between HPSG and some of the other frameworks discussed in the volume.
The above HPSG quote has another interesting feature: the HPSG goals of research are claimed to be similar to the goals of Construction Grammar. Müller (to appear) in fact makes an even stronger claim by stating that "HPSG is a Construction Grammar", though he adds in a footnote attached to this statement that HPSG is also a Generative Grammar in the early Chomskyan sense of being explicit and formalized. Chapter 18 of the handbook under review argues at length that Generative Grammar and Construction Grammar are two distinct, co-existing and competing research paradigms. So where does HPSG belong? The current Wikipedia entry for this framework lists HPSG as an instance of the category generative linguistics, and I think this is the kind of categorization that most linguists would assume, unless otherwise instructed. But even the above quote taken from Chapter 12 does not make it very clear where HPSG should belong. Yet we are told in Chapter 18 that Construction Grammar and Generative Grammar are two distinct paradigms. If that is the case, how come HPSG is not an easy fit in either?
Similar concerns could be raised for some other theories discussed in the book, or for specific versions of some of these theories. Certain versions of Minimalism, for example, share some constructivist traits that are not present in mainstream Generative Syntax. More recent developments in LFG urge us to enter the metaphorical bush of language data that are "spontaneously produced by speakers, and form quantitative generalizations based on concepts such as conditional probability and information content" (Bresnan 2016: 599). Bresnan contrasts this bush with the garden of the kind of orderly data that generative grammarians traditionally work with. And again, LFG is otherwise listed as a member of the generative paradigm. One could add a long list of many similar examples: my point here is simply that if one digs deeper, it becomes much less obvious how different theories of syntax relate to and differ from each other. Or, to be more precise, it is less obvious that they are so fundamentally different from each other in each compartment where a strong difference is otherwise usually assumed. I may also add to this the anecdotal observation that theoretical differences are very often backgrounded at thematic workshops and conferences these days, where many of the speakers try to use a converging language of linguistic discussion even if they come from different theoretical backgrounds.
This is not to downplay the importance of the distinctions between different theories of syntax, because these are obviously real. But on analogy with some linguists regarding linguistic data as noisy and less well-behaving than traditionally thought of, syntactic theories, as objects of metascientific inquiry, may too be considered to form a complex web of partially overlapping entities. Nevertheless, the real proof of the pudding is in the eating, and differences between the competing frameworks become perhaps the most visible if it comes to representing data. For a linguist, an attribute-value matrix of an HPSG practitioner or an f-structure of someone from the LFG community is an immediate and conclusive proof that we are confronted with the respective theories, and much the same can be said of the kind of cognitive models used in Cognitive Grammar. It is therefore an attractive feature of the handbook that the editors asked the authors of the thirteen syntactic chapters to provide an analysis of an English sentence. This sentence was the following: After Mary had introduced herself to the audience, she turned to a man she had met before. As any linguist will know, this particular sentence includes many challenges for a syntactician, but it is less interesting for us now what these challenges are. Much more interesting is the way in which the authors of the individual chapters execute this task, and I invite the reader to take a closer look at these analyses to get a real feel of the theories they are not familiar with. In fact, while some of the authors develop a full-fledged analysis for the sample sentence, others only do a partial description or perhaps make no real attempt at providing even a fragmentary analysis. It may sound surprising to an outsider that a syntactic theory is not well-equipped (or rather: not yet sophisticated enough) to provide for a comprehensive analysis of any arbitrary piece of language, but this task is rarely considered to be one that has much practical importance ‒ even if it is an explicitly stated goal of a theory to predict for any possible string of words whether it is part of the language studied or not. The following bon mot comes to my mind at this point, which I heard from a leading computational linguist (whose identity shall not be revealed here): Linguists do not write grammars, they write articles. There is something profoundly true about this, linguists tend to find more exciting challenges than to write comprehensive, but predictably shallow descriptions of languages.
MP, in other words, sharpens divisions that have been latent in the discipline for 60 years and prizes the kind of work that is less knee deep in descriptive detail than most work in GG. And (surprise, surprise) some don't like this. And this malignity has prevented many from evaluating MP on its own terms. That is too bad, for when so evaluated, the achievements have been reasonably impressive. [...] On its own terms, MP has been very productive. Unfortunately, many have missed this. The fault does not lie with MP.
(Chapter 7, pp. 212)
But to mention an example from the other side, Chapter 16 includes a lengthy list of quotes mostly from Chomsky which are meant to show the inconsistent nature of Chomskyan theorizing. Together with the respective commentary, this passage comes close to a sort of a learned stand-up performance for the cognoscenti. But this what we linguists often are, and I only wonder what it is like in other disciplines.
This review has necessarily been only a very fragmented reflection on a volume rich in linguistic and metalinguistic coverage. The handbook provides a very comprehensive outlook on what syntactic theory is now, and especially, on what it has been in recent decades. I recommend it especially for readers who have some hands-on experience in linguistic research and theorizing. They will surely find much food for thought therein, and they will definitely find themselves more equipped with the knowledge that is required to perform "metatheoretically reflected object theoretical research" (Chapter 21, p. 578.). The main message of the handbook is that such reflection is to the benefit of all.
Imrényi, András. 2019. Kertész András – Moravcsik Edit – Rákosi Csilla szerk., Current Approaches to Syntax. A Comparative Handbook. Magyar Nyelv 115(4). 479–485.
Müller, Stefan. to appear. HPSG and Construction Grammar. In Stefan Müller, Anne Abeillé, Robert D. Borsley, Jean-Pierre Koenig (eds). Head-Driven Phrase Structure Grammar: The Handbook. Language Science Press.