Affix hopping: a new Olympic discipline or innovative brewing technique? Anglo Frisian brightening: a meteorological phenomenon occurring over the North Sea? Cranberry morph: a result of genetically-modified gardening? Well, no, as it turns out. This Dictionary ‘provides concise and clear definitions of all the terms any undergraduate or graduate student is likely to encounter in the study of linguistics and English language or in other degrees involving linguistics, such as modern languages, media studies and translation.’ It has approximately 3000 entries. In some cases, there is more than one definition of a term, e.g. three for ‘declarative’, five for ‘domain’ and three for ‘ergative’ (one of these ‘condemned by some linguists’). Many entries include examples from English and other languages; among the other languages, Russian (transliterated) and Turkish are particularly well represented - some might say over-represented. As well as terminology entries, there are entries for 246 languages, and for key figures in the history of linguistics such as Jespersen, Labov, Sapir etc.

There are ‘extended entries on some terms’. ‘Antipassive’, ‘case’, ‘consonant’ and ‘modality’, for example, get more than a page each, and the entry for ‘cardinal vowels’ includes vowel tables and a vowel quadrilateral diagram. Other entries, however, are unhelpfully minimal. ‘Checked’, for example, is ‘[a] distinctive feature associated with ejective and implosive consonants’, but what exactly is this feature? ‘Critical applied linguistics’ is ‘[t]he use of applied linguistics in projects for the transformation of society’; this would be clearer if at least one example of such a project was given. ‘Idiolect’ is defined as ‘[a] distinctive way of speaking associated with a single individual’; this is insufficient, since ‘way of speaking’ could be interpreted to mean voice quality, for example. Entries such as these are particularly unfortunate since, generally, the Dictionary does not give references to other works for further reading. Occasional other points require clarification; for example, interpretation of ‘Smaller forms [than words] can be mentioned but not used’ assumes understanding of technical senses of ‘mention’ and ‘use’ which are not included in the Dictionary.

Although the information given, especially in the longer entries, is generally reliable, there are exceptions; here are a few examples: In the entry for ‘dialect’, RP is adduced as an example of a dialect associated with a particular social background, but RP is an accent, not a dialect. The description of ‘indirect speech’ is oversimplified to the point of being inaccurate. Infinitives in English are not always ‘signalled by to preceding a verb.’ The definition of ‘nuclear tone’ is inaccurate. In the entry for ‘imperfect’, one of the Russian sentences is wrongly translated, and does not illustrate the point intended. The illative form of Finnish talo is not talon, but taloon.

In some entries, one sense of a term is included, but another is missing. For example, ‘apostrophe’ is defined as ‘[a]n interruption in a speech or written text to address some person, usually not present, or some inanimate object’, but its use as the name of a punctuation mark is not included. ‘Rhotic’ is defined as ‘[t]he class of ‘r-sounds’, but its use in the term ‘rhotic accent’ is omitted. ‘Diction’ is defined as ‘[t]he choice of vocabulary to create different types of literary text,’ but its more familiar meaning, relating to clarity of speech, is not mentioned. ‘Fossilization’ is included as a diachronic process in language, but not as a process in language acquisition.

The entries for languages are for the most part limited to the language family which the language belongs to, the number of speakers it has and the country or countries where it is spoken. Sometimes other information is included, often for no apparent reason, e.g. for ‘Aymara’, ‘Aymara-speaking people have been in close contact with their Quechua-speaking neighbours for many centuries.’ The entry for ‘Japanese’ describes the writing systems used by this language. ‘Finnish’ is ‘[v]ery closely related to Estonian,’ but no reciprocal information is given under ‘Estonian’.

One unexpected but welcome feature is the inclusion of etymologies for some terms. This is particularly useful for Latin-derived terms such as ‘adnominal’, since the elements ‘ad’ and ‘nominal’ can easily be related to other English vocabulary. Coverage is patchy, however; there is no etymology for ‘denominal’, for example. The etymologies are perhaps less helpful in the case of Greek-derived terms such as ‘aposiopesis’, ‘hypocoristic’, ‘oxymoron’ and ‘paronomasia’. It would also have been useful to include pronunciation transcriptions of terms such as ‘deictic’, ‘diaeresis’ and the names of less-familiar languages.

‘Blindness’ and ‘deafness’ are obviously not specifically linguistic terms, but the entries for these terms briefly explain their relevance to language acquisition. Similarly, the entries for ‘default’, ‘ethnomethodology’, ‘hierarchy’, ‘prestige’, ‘rate’ and ‘sampling’ exemplify how these concepts apply to linguistics. However, some entries include no indication of any particular application in linguistics - e.g. ‘adventitious’, ‘autobiography’, ‘calculus’, ‘cochlea’, ‘essentialism’, ‘generalization’, ‘hearsay’, ‘induction’, ‘necessary condition’, ‘null set’, ‘peer group’, ‘standard deviation’, ‘T-test’ and ‘white noise’. The same applies to quite a number of terms related to computing, such as ‘byte’, ‘database’, ‘domain’ (sense 3), ‘domain name’, ‘icon’, ‘information extraction’, ‘information retrieval’ and ‘network’. ‘Epidemiology’ - ‘The evidence-based study of factors that affect the health or illness of a population. It forms the basis of interventions made for the public health’ - must surely be an intruder from a different dictionary.

There is other evidence that the text was not prepared for publication with sufficient care. The table of examples printed under the entry for ‘consonance’, for example, actually belongs under ‘consonant mutation’. The entry for ‘front’ repeats, in reverse order, the tables of primary and secondary vowels given under ‘cardinal vowels’. The entries for ‘cluster’ and ‘consonant cluster’ more or less duplicate each other; one of these entries should simply have been replaced by a cross-reference to the other. The second sense of ‘expletive it’ has nothing to do with the word ‘it’, and should appear as a separate entry ‘expletive’. The entry for ‘level’ is poorly laid out, with the second of the two senses beginning mid-way through the second of the two paragraphs. Similarly, ‘voice’ has three senses and six paragraphs; senses 2 and 3 begin in the middle of paragraphs, making them hard to locate at a glance.

In the table of ‘phonetic symbols for English vowels’ (they are actually phonemic, not phonetic) the symbol /ɪ/ as in ‘pit’ is used in diphthongs except the one in ‘high’, which is given as /ai/, i.e. including the symbol /i/ as in ‘see’. Even so, in the entry for ‘diphthong’, ‘day’ is transcribed as /dei/; here, too, ‘cow’ is transcribed with the Greek alpha: /kαʊ/. No length marks are used in the long vowel symbols listed, and yet the transcription of the example word ‘paw’ includes a length mark: /pɔ:/. There are several incorrect transcriptions in ‘morph’: ‘/s/, /z/ and /z/’ as allomorphs of [Plural], cups /cʌps/, logs /lɔgz/, pillows /piləʊz/. Under ‘gradual opposition’ and ‘homographs’, /ε/ is used in place of /e/. The entry for ‘assimilation’ has incorrect symbols in place of /ə/, /ʌ/ and /əʊ/. There are other instances of incorrect symbols, as well as wrongly-sized symbols.

There are many examples of missing italicisation, bolding and punctuation. Under ‘dialectology’, there are two different senses; the first is introduced by the figure ‘1’, but there is no corresponding ‘2’. The definition of ‘orality’ is prefaced by ‘1’ although there is only one sense. Alphabetical ordering is disrupted in the sequence of entries ‘hierarchical structure’ - ‘HFC’ - ‘hiatus’ - ‘hierarchy’. In ‘morphophonemic alternation’ and elsewhere, reks is given as if it is an orthographical form, rather than the pronunciation of Latin rex. Diacritics are missing from ‘Bokmål’ (under ‘Norwegian’) and ‘zählen’/'zählte’ (under ‘strong verb’).

Overall, this Dictionary can be recommended to its intended readership with reservations. The definitions are mostly clear and accurate, and are most useful when exemplified. The substantial cross-referencing between entries is very helpful. On the other hand, some entries are too brief, or otherwise misleading. Inconsistencies in the use of pronunciation symbols, in particular, are likely to confuse novices.