András Kertész. 2017. The Historiography of Generative Linguistics. Tübingen: Narr. 210 S.

There is a difference between history and historiography. Whereas the former describes what happened and aims to explain aspects of these events, the latter develops a framework for writing history. Such a framework serves as the basis for selecting relevant events and explaining the links between them. In this brief monograph, Kértesz aims to give a critical overview of the historiographic literature about Chomsky’s generative linguistics and to show that his own framework is better (p. 11). As such, the book operates at three levels: historical, historiographic and meta-historiographic.

The book consists of seven chapters and a preface. The preface, which strangely is missing from the table of contents, introduces the topic and aims of the book. Chapter 1, “Introduction”, presents Kuhn’s (1970) theory of science, because it has played such an important role in the historiography of generative linguistics. Chapter 2, which takes up almost half of the book, presents 22 analyses of the history of generative linguistics, each corresponding to a framework for writing history. Chapter 3 evaluates these frameworks. They are classified on the basis of their perceived deficiencies. The conclusion is that none of them is good enough. In chapter 4, Kertész sets out his own model, the p-model. Chapter 5 applies the p-model to one of the questions of the historiography of generative linguistics, how Chomsky’s (1957) theory should be interpreted in relation to the Post-Bloomfieldian theories of Zellig Harris and Charles Hockett. Chapter 6 outlines some of the questions that still have to be addressed from within the p-model. Chapter 7 concludes that the p-model is an important contribution both to historiography and to linguistics.

Chapters 2 and 3 are grouped together as “Part I: Meta-historio­graphical overview”. Although chapter 2 is by far the longest chapter, the discussion of the individual analyses is rather superficial. The order of presentation is determined by the work of Chomsky for which they claim that it is revolutionary or not. This approach to classification leads to some decisions that are somewhat arbitrary, because especially the more thorough analyses take into account several of Chomsky’s works. An additional constraint on the scope of the discussion is that it is narrowly focused on finding the answers to three questions:

1. What are the basic terms of the discussion?

2. What is the central hypothesis?

3. What is the framework?

The answer to (1) is a list of central concepts introduced in the framework under discussion. The answer to (2) is whether there is a revolution or not. The answer to (3) is, for instance, the framework of Kuhn (1970) or a competing one. By focusing only on a brief answer to these questions, many of the subtleties that qualify a good analysis are not noticed, because they cannot be represented as answers to the questions (1–3). A further problem is that in the selection and presentation of theories, Kertész seems to favour quantity over quality. He lumps together careful historical accounts of the whole of Post-Bloomfieldian and generative linguistics, such as Matthews (1993), with personal accounts of a short historical period, such as Lakoff (1989). After the initial presentation of the approaches and their answers to the questions (1–3), further generali­zations are made. Various tables and tree representations serve to abstract away from even more details. In chapter 3, Kertész uses just 14 pages to demonstrate that none of the 22 analyses is adequate, so that a new framework is necessary. Obviously, this does not do justice to the more careful analyses, but the author seems more interested in creating space for his own framework than in doing justice to previous ones.

Part II, “Towards a new historiography”, starts with chapter 4 and covers the remaining chapters. Together, these chapters are 65 pages, i. e. considerably less than the 81 pages of chapter 2. The scope of part II is to present, exemplify and evaluate Kertész’s own framework. There are a number of striking features in this part of the book. One is that Kertész does not present a solution to the problem for which he takes the other analyses to account. Instead, he applies his p-model only to a single issue in the historiography of generative linguistics. He attributes this to “the genre of this short monograph” (p. 138), but as a reviewer I cannot help but wonder why he chose this genre if it is not very well suited to his purpose.

Another striking point is that in the whole of Part II, very few references to the analyses in Part I are made. The reason is that the p-model approaches the history of generative linguistics from a rather different angle to any of the 22 analyses discussed in chapter 2. Whereas the theories in chapter 2 take social interaction, rhetoric or the overarching view of the nature of language as their starting point, often bringing in the other factors as well, Kertész’s p-model is based on a narrow interpretation of argumentation. The p in p-model stands for plausibility. The focus is on a scalar variant of truth value assigned to individual statements. As opposed to classical logic, which has a binary concept of truth with only two values, 0 and 1, Kertész adopts a version of logic in which 0 and 1 are the extremes of a continuum.

What we have in this short monograph is therefore two rather independent parts. Part I claims to analyse and evaluate previous approaches, but in fact is laid out so as to create the space for the author’s new model. Part II only gives a general taste of the nature and application of this model, suggesting rather than demonstrating its power. Taken together, this results in a rather strange hybrid.

Arguably, Part I is at the same time too long and too short. The presentations of the analyses are too short to be a basis for a fair comparison. The exclusive focus on questions (1–3) listed above does not do justice to the fact that many of the approaches have a different aim. Some of the works discussed are historical in orientation, but not all with the same focus. Matthews (1993) is a history of ideas whereas Newmeyer (1986) relates ideas to people operating in a particular social environment. Other works are not historical at all. Lappin et al. (2000) make a brief point about the intellectual climate in which the Minimalist Program emerged. Ten Hacken (2007) aims to explain the differences between Chomskyan linguistics and some competing research programmes. By neglecting these aims, Kertész cannot do justice to any of them. This conflicts with Kertész’s repeated claims that he aims for his presentation to be “unbiased, balanced and impartial” (e. g. p. 11).

At the same time, the discussion in Part I is too long to be just an introduction to the presentation of a new framework. If it were shorter, such an introduction could be acceptable as a way of indicating what all of these models get wrong. In such a presentation, it would not be necessary to take too many details into account, as long as the subsequent presentation of the new framework is convincing. However, the sheer number and the systematic presentation of the many approaches suggest an exhaustive treatment. By imposing the criteria he used for the development of his own theory as evaluation criteria for the analyses in chapter 2, Kertész’s evaluation is inherently biased towards his own theory. If Part I were shorter and Kertész’s stated aim had been to present his own theory, this would not be a major problem. The brevity of Part II, however, especially in relation to the length of Part I, makes an inter­pretation of Part I as an introduction to the presentation of the new theory hard to sustain. The large number of approaches in chapter 2 and the harsh and summary criticism in chapter 3 raise expectations that are not fulfilled in Part II, where Kertész presents his own theory with a remarkable degree of self-indulgence.

The least one can say is, therefore, that the book under review is unbalanced. The discussion of approaches in Part I is unsatisfactory as it stands. It could have been a book of its own if expanded to take into account the aims of individual approaches. Such a book would require a rather different attitude than Kertész shows here. Rather than trying to reduce the analysis to answers of pre-set questions, an effort would have to be undertaken to actually understand what each of the approaches was aiming for. Alternatively, Part I could have been shortened to become a brief introduction to a new theory.

The presentation in Part II is also unsatisfactory as it stands. It might qualify for a rather long article suggesting work in progress, but then without the preceding chapters 1–3. Alternatively, it could have been expanded into a book of its own, with a much-reduced version of Part I as an introductory chapter. The single case study in chapter 5 would then have to be supplemented with other case studies. Some of the open questions in chapter 6 should have been treated in more detail. In particular, there should be an argument why it is not the history of ideas or the social environment or the model of language study, but the details of argumentation that determine a proper analysis of the history of generative linguistics.

In this monograph, Kertész proposes a new model as a framework for the historiography of generative linguistics. Its emphasis on argumentation implies that he takes the validity of individual arguments as the central problem of historiography. This assumption is different from other historiographic approaches, which rather focus on identifying the background in social interaction and the underlying model of language as a basis for explaining the nature and outcomes of linguistic discussions. A problem with this monograph is that Kertész does not address this opposition sufficiently. What he shows is that if one assumes that assessing the plausibility of individual statements is the central issue of the historiography of a scientific domain, the p-model is an attractive option. What he fails to demonstrate is why this view of historiography should be preferable over the ones based on social interaction or the underlying model of language.