The term etymology, in the volume under review as in most modern discussions, is defined as the investigation of word histories generally, and also as an account of the origin and historical development of individual words. As a scholarly practice, etymology has traditionally been concerned most especially (but not exclusively) with ‘those areas where there is a doubt about a stage in a word's history, or where the documentary record fails us’ (p. 284; similarly, pp. 1, 291). Words and lexicalized phrases and their origins and shifting meanings are, obviously enough, of central concern to professional lexicographers, historical linguists and philologists. But they have intrigued non-specialist scholars and lay persons with a very wide variety of interests and intellectual backgrounds.

The author of this work is principal etymologist for the Oxford English Dictionary and has lectured, broadcast and published regularly on English etymology. He addresses this practical introduction ‘to anyone who has taken the important first step of realizing that etymologies are the result of research, and would like to discover something about the nature of that research, and the principles and methodologies which underlie it’ (p. ix). The take-away point for readers entirely new to the field is that the compressed etymologies that one finds in many standard dictionaries are the product of extensive research, which can be difficult and fraught with uncertainties. Durkin has further attempted ‘to frame this book so that it is addressed most centrally to someone who has an interest in historical linguistics . . . Etymology is a part of this wider field, and anyone's understanding of etymology will be greatly enriched by at least some acquaintance with the broader concerns of the discipline as a whole’ (p. ix). Most of the data are drawn from English. The narrow focus is offset by the opportunity for readers to avail themselves of current etymological research for the third edition of the OED, with which the author is personally involved. Though not unprecedented internationally (cf. Seebold Reference Seebold1981), a textbook dedicated specifically to etymology is very much to be welcomed.

The book consists of nine substantive chapters and a brief conclusion, which are followed by a glossary, suggestions for further reading, references, a general index and an index of word forms. The first chapter lays out some basic concepts (such as cognacy, language families, comparative and internal reconstruction), provides the rationale for the study of etymology and describes what an etymologist does. Oddly, perhaps, the term etymon, which would seem to fall into the category of ‘basic concepts’, is defined quite precisely in the glossary but not upon its first (p. 18) and subsequent occurrences; nor is it entered in the index. The histories of English friar and sad work well as Paradebeispiele. The presentation of Grimm's and Verner's Laws, as an illustration of the relationship between sound change and etymology and a demonstration of the comparative method, is clear and should be comprehensible to a general readership, though a chart or list of phonetic symbols used in the book might have been helpful. The establishment of manual ‘involving the hand, operated by the hand’ in English (from Latin manuālis ← manus ‘hand’, via French) at the expense of Middle English handly and semantically narrowed handy ‘skilled with the hands, convenient to handle or use’ is a nice example of how heavy borrowing can alter diagrammatic patterns in word formation. Durkin takes pains to alert the neophyte to the existence of false cognates, such as English care beside Latin cūra ‘care, concern, guardianship, administration’, Latin deus ‘god’ beside Greek θos ‘god’, which show similarities in form and meaning but obtain from different roots, and the ‘etymological fallacy’, that is, the belief that knowledge of a word's origin and particularly its original meaning should inform present-day usage.

Chapter 2, entitled ‘What is a word? Which words need etymologies?’, deals with the linguistic unit that is central to lexicography and etymology. Durkin discusses the analysis of linguistic forms into words, showing how none of the conventional criteria (meaning, prosody, orthography, morphotactics) can be strictly applied. I was surprised to find no mention of a long-established working definition of word as ‘a minimum free form’ (Bloomfield Reference Bloomfield1933: 178). For present purposes, Durkin uses word rather loosely in the sense of lexeme (pp. 41, 296) and refers to words by their citation forms (lemmata, though Durkin eschews this term), that is, the word forms that conventionally represent the canonical forms of lexemes in dictionaries.

Lexical innovation and integration may (but need not) progress along a developmental continuum: nonce formation (ad hoc coinage by individuals) > institutionalization (maintenance of semantic transparency, conventional use within a speech community ‘in a given context or with a fairly specific meaning’) > lexicalization (‘opaque – in meaning, or composition, or both’) (p. 49). Durkin is, of course, free to define his terms as he deems appropriate, but one would appreciate coherence and consistency in return. A statement that ‘institutionalized words . . . remain (at least relatively) transparent’ (p. 49) does not square with an allusion to ‘words with a non-predictable, institutionalized meaning’ a few pages later (p. 59). Nor does it help that the definition of institutionalized in the glossary (p. 292) refers to social norms with no mention of semantic transparency. Lexicalization (at least as I understand the term) refers to the expression of a concept by means of a word or the creation of a lexical unit out of a group of words, the meaning of which is beyond the sum of its parts. The concept ‘act of untrammeled power, enforcement of power’ is lexicalized in Afrikaans in a relatively transparent complex word (kragdadigheid) but not in English. (South African English has borrowed the Afrikaans word.) Strictly speaking, conventionalization and compositionality are conceptually independent phenomena. The personal indiscretion of a politician has given American English hiking the Appalachian Trail, a new, non-compositional, abruptly lexicalized euphemism for ‘sneaking off with one's lover’.Footnote ⁴

While it may be feasible to compile an etymological dictionary that includes all of the extant words of a philological language (e.g. Lehmann Reference Lehmann1986 for Gothic), this is obviously not practical for living languages, the lexicons of which are infinitely extendible. Moreover, productive word-forming patterns, i.e. those capable of forming new words (e.g. affixation in un-, pre-, re-, -ism, -ly), can generate an enormous number of one-off forms that are deployed within the context of particular utterances and are not stored in the mental dictionaries of the interlocutors. With these considerations in mind, Durkin sets the following priorities for etymological coverage: (i) monomorphemic words (acorn, lord, lady are informative examples discussed in this chapter); (ii) complex words containing semantically opaque bound elements (so-called ‘cranberry morphs’) that are obscure or foreign in origin (e.g. lukewarm, cranberry; cf. Low German kranebere, Standard German Kranbeere) or native fossils (e.g. winsome, raspberry); (iii) words that are formally analyzable but semantically non-compositional (e.g. penknife, handsome); (iv) idioms (e.g. to cut a caper ‘to make a playful, skipping movement, to act ridiculously’); and (v) ‘any word which has a form which is not explicable by the productive word-formation processes of the language’ (p. 59).

Chapter 3, ‘Are words coherent entities?’, examines homonymy, polysemy, formal and semantic merger (or near-merger), and split in word form. Although one might cavil at the quirky title and the choice of merger and split over convergence and divergence (given the close association of the former with historical phonology), this chapter is quite solid. Durkin cautions that relations of polysemy and homonymy fluctuate over time and can work in both directions. The English word crane in the meaning ‘machine for the hoisting and lowering of heavy objects’ is a metaphorical extension of crane ‘a type of long-legged, long-necked bird’. However, it is debatable whether contemporary speakers feel a connection between these senses, and the OED and other dictionaries treat them under separate lemmata. With regard to semantic convergence, Durkin recalls Bloomfield's (Reference Bloomfield1933: 436) suggestion that since the loss of medial *-h- and of post-tonic vowels in English has made ear ‘spike or head of corn’ (< OE ēar < *ahur < PGmc. *ahuz) homophonous with ear ‘organ for hearing’ (< OE ēare < PGmc. *auzō- or perhaps *auzān), and since ‘the meanings have some resemblance, ear of grain has become a marginal (transferred) meaning of ear of an animal’. As Durkin points out, Bloomfield recognized that semantic proximity is not subject to precise measurement; linguists and lexicographers will insist on describing such forms as homophonous pairs. Bloomfield regarded meaning as very difficult to investigate. Yet, he opined that ‘the only type of semantic expansion that is relatively well understood, is what we may call the accidental type: some formal change . . . results in a locution which coincides with some old form of not too remote meaning’ (Reference Bloomfield1933: 436). Historiography aside, Durkin does not find Bloomfield's example entirely convincing, but the larger point is that ‘the actual history of words may be totally obscure to speakers of a language, and certainly it is very feasible that new, historically unjustified links in meaning may become established between etymologically unrelated homonyms’ (p. 79) or for that matter near-homonyms. An instructive example would be the controversies that have surrounded the etymologically innocuous adjective niggardly ‘stingy, miserly’ (< ON hnøggr, cf. OE hnēaw ‘mean, stingy’, Dutch nauw ‘strict, narrow, close’), which is probably irredeemably tainted by its formal likeness to an unrelated, deeply offensive epithet.

The next five chapters survey processes of word formation (chapter 4), lexical borrowing and imposition, the latter as a consequence of language shift and code switching (chapters 5 and 6), the effects of sound change on lexis (chapter 7) and semantic change (chapter 8). The approach is preponderantly traditional, though Durkin does take pains to show how insights from grammaticalization theory, variationist accounts of change in progress, cognitive linguistics, pragmatics, and contemporary research on transferability in language contact (viz. the borrowing scale of Thomason & Kaufman Reference Thomason and Kaufman1988: 74–6; Thomason Reference Thomason2001: 70–1) are significant for etymological research. In his preface Durkin writes that he has ‘tried to pay particular attention to those areas which are important for etymology but which receive relatively little attention in most introductory books on historical linguistics’ (p. ix). With the exception of an approach that stresses connections between word histories and the history of material culture – metonymically deriving its name from the journal Wörter und Sachen (1909–42) – virtually all of the fundamental concepts of language change surveyed herein find ample treatment in the textbooks that I have used in courses for beginning and advanced students (respectively Trask & Millar Reference Trask and Millar2007; Hock Reference Hock1991). Lacking is a discussion of so-called Wanderwörter (a living example of which is vuvuzela ‘stadium horn’, courtesy of the 2010 FIFA World Cup tournament in South Africa), which can pose special problems for etymology, and macro-comparative reconstruction of global etyma (e.g. Ruhlen Reference Ruhlen1994), a subject of fierce controversy on which the reviewer takes no stand. On the positive side, it is commendable that chapter 9 on onomastics and etymology is there at all.

The value of these chapters – indeed, of the work as a whole – lies in the illustrative etymologies themselves, which overall are well chosen and often interesting in their own right. Inevitably, specialists will take issue with matters of presentation, fact, or detail, particularly in regard to the enduring cruxes. To cite but one instance: the date and location of the first appearance of English big – c. 1300 in northern texts – and its probable earliest meaning of ‘sturdy, strong, mighty’ (> ‘large’) are suggestive of Old Norse as a source. Durkin notes the regional use of bugge ‘mighty man’ and bugga ‘rich, wealthy, powerful’ in modern Norwegian, but asserts that there is no obvious etymon recorded in the early stages of any of the Scandinavian languages. What he fails to mention is that there is some evidence from Old Danish and Old Swedish for an ON by-name Buggi ‘fat man’ (Fellows Jensen Reference Fellows Jensen1968: 68) or ‘powerful man’ (Peterson Reference Peterson2002: 40), which lends slight support to a Norse connection, if one stipulates that Middle English big(ge) continues a variant showing umlaut of the tonic vowel and subsequent unrounding. The picture clouds over again if big(ge) has a connection to by-names or surnames of the form Bigge, Bigga (which is by no means certain), the earliest attestations of which occur in southern counties, especially Kent, from the first half of the eleventh century (pp. 278–9). I agree that these data are inconclusive, though in etymologizing this word, one must determine whether congeners in other Germanic dialects are extant but masked by accretions of phonological, morphological and/or semantic change. Davis (Reference Davis2000) proposes an etymology that traces English big back to Proto-Germanic times. This hypothesis, too, may be wrong, but it does take the requisite heuristic step.

Etymology concentrates on the history of single words (and morphemes). ‘Chaque mot a son histoire’ – the anti-neogrammarian aphorism variously attributed to Jules Gilliéron or Hugo Schuchardt – conveys something fundamental about the enterprise, mutatis mutandis, insofar as ‘the etymologies in any dictionary are a set of unique word histories, resulting from the interplay of manifold factors’ (p. 209). Etymology is thus a part of lexicography as well as historical linguistics. As an independent field, lexicography takes what it needs from linguistics and can ignore the rest. The Oxford guide to etymology is a worthy introductory text for students and general readers whose needs and interests fit this profile.

The book does not have the theoretical grounding and precision that would make it suitable for use in a linguistics curriculum. Definitions of basic terms are all too often soft and oversimplified. Syntax, for example, is casually described as ‘the study of the meaning relations between words within a sentence’ (p. 2); the statement ‘monomorphemic words are composed of only a single morpheme or meaningful unit’ (p. 43) is in conflict with the strange definition of morpheme in the glossary: ‘a minimal meaningful unit within a complex word’ (p. 293); infixes, we learn, ‘interrupt a morphological base’ (pp. 47, 288); disyllabic words containing an initial open syllable are characterized as words that have ‘only one medial consonant’ (pp. 63, 184); defining grammaticalization as ‘the process by which words develop increasingly grammatical meanings and functions over time’ (p. 224) isn't very helpful, though one does encounter this tautology in the specialized literature, too (e.g. Heine & Kuteva Reference Heine, Kuteva and Wray2002: 377). The unwary reader might infer from the extract on p. 247 that dynamic modality (ability, physical possibility) is a synonym for, rather than a value of, root (non-epistemic) modality – the other value of which, deontic modality, is not so identified three lines later.

This volume is handsomely designed and produced, with very few technical glitches beyond an ugly misprint on p. 182 (‘Gimm's’ for ‘Grimm's’). My wish for its readers, if I may borrow American poet John Ciardi's signature sign-off from his weekly etymological rambles on radio during the 1980s: Good words to you.