A History of English Reflexive Pronouns: Person, Self, and Interpretability (review)

This extremely ambitious contribution both to the study of reflexivity and to the theory of diachronic syntax is characterized by three main attributes. First, van Gelderen pursues a distinctly holistic approach to diachronic syntax. Within a minimalist framework, the grammar of reflexivity throughout the different stages of English is related to other subsystems of grammar such as word order, verbal agreement, pro-drop, and the case system. In short, vG ‘links the changes in reflexives to the transformation of English from a synthetic to an analytic language’ (1). In minimalist terms, this development translates as the introduction or activation of functional categories that check specific (uninterpretable) person and number features. As a consequence of this holism, large portions of the book are dedicated to the discussion of issues that are, on the face of it, not genuinely germane to reflexivity. Second, vG’s approach is reductionist insofar as a wide range of facts is explained in terms of very few principles of grammar. This reductionism entails a relatively high amount of strong hypotheses, both theory-internal and in general. The third central characteristic of this book is the excellent empirical basis on which it is built. With great accuracy, vG provides plenty of examples from several authors, taking into account dialectal differences and idiosyncrasies of specific texts. The rich empirical material renders the book highly valuable also for those readers who are not primarily interested in the theoretical argument.

The central theoretical concern of this book is the attempt to explain the historical development of reflexives from Old English to Modern English in terms of pronominal feature composition. vG takes it that the referential status of a pronoun is directly and exclusively dependent on its feature specification. This amounts to abolishing specific rules for reflexivity such as the binding conditions and to reducing anaphoricity phenomena in English to conditions on chain formation. vG sets out from Reinhart & Reuland’s (1993) chain condition which requires that every chain have exactly one link that is fully specified for φ-features and structural case. Accordingly, pronominals can form a chain with a given antecedent if and only if they are underspecified in terms of φ-features or lack structural case. However, vG considerably modifies Reinhart & Reuland’s account insofar as she assumes that the referentiality of a pronoun is sensitive to the (un)interpretability of its features: ‘only by being checked, i.e., as I argue, by having Uninterpretable features, can an element function referentially’ (15). This assumption requires the parameterization of the interpretability of φ-features: For example, it is claimed that person features are interpretable in some languages (French) while they are uninterpretable in others (Modern English).

In the somewhat heterogeneous introductory chapter, vG prepares the theoretical ground for the discussion to follow (e.g. feature checking in recent generative grammar, interpretability, binding, structure of pronouns). Furthermore, she provides some basic information about Old English morphology and justifies the composition of her corpus. The main discussion starts in Ch. 1 with a presentation of the reflexivity facts of Old English. vG makes a number of descriptive generalizations across various dialects and authors and underpins them with quantitative analyses. As is well-known, OE simple personal pronouns can be interpreted reflexively. Self (/seolf/sylf) is optionally used to specify a pronoun as reflexive. In Late OE, self occurs somewhat more frequently in the third person than with speech act participants. vG explains the ability of OE pronouns to be bound in a local domain by their exhibiting inherent case. Since inherent case is interpretable, OE pronouns can form a chain and thus co-refer with a local antecedent.

Ch. 2 constitutes the theoretical core of this book and contains the most central arguments. It starts with an account of the expression of reflexivity in Early Middle English. The above-mentioned tendency of Late OE to mark third person reflexives with self persists and increases. Nevertheless, simple pronouns...