Evidentiality (review)

Since the mid-1980s there has been a steadily increasing interest within linguistics in the notional domain of evidentiality. The interest has manifested itself in a number of journal articles, a few monographs (e.g. Floyd 1999), and a handful of anthologies (Chafe & Nichols 1986, Guentchéva 1996, Johanson & Utas 2000, Dendale & Tasmowski 2001, and Aikhenvald & Dixon 2003). With a few exceptions (e.g. Anderson 1986 and Willett 1988), the studies have focused on the description of evidentials and evidentiality in one language or a small group of related languages. Aikhenvald is the first to attempt a crosslinguistic approach to evidentiality in the format of a monograph, and there should be no doubt that the result marks the culmination in the study of the domain.

A’s monograph is intended to serve ‘both as a sourcebook for further typological studies, and as a textbook’ (xi). It is based on a study of over 500 languages and presents numerous relevant examples from a great many languages, including languages on which A has done fieldwork herself. She offers a typology of evidentiality systems, meanings, and coding types, and a study of the relations, both diachronic and synchronic, with other dimensions of meaning and grammar, and even cognition and culture. The sheer size of this endeavor (and its 450 pages), but also its depth and intellectual riches, will guarantee its success as a sourcebook for further typological work on evidentiality.

The most important result of A’s crosslinguistic study has to do with the notional definition of evidentiality. A demonstrates beyond all doubt that the narrow conception of evidentiality in terms of the notion of ‘source of information’ is linguistically highly relevant. In other words, evidentiality defined narrowly as the indication of source of information is a linguistic notional domain, or ‘category’, ‘in its own right’ (7). The main argument for this consists in the many examples of narrow grammatical evidential paradigms, that is, morphosyntactically delimited systems of two or more grammatical items, the primary meanings of which can all be described as specifications of source of information. To be sure, it is evident from A’s monograph that evidentiality is related to other linguistic domains. Evidential systems, or singular evidentials, may interact with, derive from, or extend into categories such as person, tense, aspect, and mirativity. In particular, they are often closely related to ‘epistemic modal’ categories concerned with ‘degree of certainty’, and A acknowledges that ‘[e]videntials are part of the linguistic encoding of epistemology’ (186). But she argues that this does not mean that the core of evidentiality is concerned with ‘degree of certainty’. The fact demonstrated by A that evidentiality is a domain in its own right, which can be narrowly and precisely defined, should be taken into account by anyone concerned with the external relations of the domain and possible generalizations including it.

There are many other important results in A’s monograph. For instance, it provides a good overview of the semantic paths to and from evidential meaning (Ch. 9 and Ch. 5, respectively), it demonstrates that grammatical evidentials are not necessarily mutually exclusive (87–96), and it presents a strong argument against the view that evidentials cannot be negated (256–57). Moreover, it demonstrates the usefulness of a number of distinctions that are linguistically relevant in general. Among these are (i) the distinction (implied in Chs. 4 and 5) between primary or encoded meaning and strategy or extension (which are obviously diachronically linked), (ii) the distinction (drawn upon in Chs. 2 and 3) between systematic and ‘scattered’ coding, (iii) the distinction (implied on pp. 315–24) between employing the code and ‘manipulating’ it, and (iv) the distinction (made, for example, on p. 10) between lexical and grammatical expression of the source of information. These distinctions (some of which are classical structuralist heritage) also do an excellent job of structuring A’s text and enhancing the readability of it.

It should be noted, though, that some of the distinctions and notions are not made clear...