Register Specification in the Learner's Dictionary

Yoshihiko Ikegami

One of the important features of the new edition of the Longman Dictionary of Contemporary English which will be welcomed by Japanese teachers (as well as students) of English is its detailed specification of register. English teaching in Japan has traditionally been sadly deficient in this respect, and Japanese students often have problems with formality in particular they often use words that are much too formal for the situation that they are in. For example, an advanced student might use very formal discourse markers such as moreover or furthermore in a simple written description of their job or interests, without realizing that these words are more typically used in formal presentations of arguments or more formal descriptions.

This rather sad state of affairs derives largely from the characteristically decontextualized way in which English is taught and learned in Japan. For most students, the classroom with a non native teacher and a more or less standardized textbook authorized by the Ministry of Education is the only situation in which they encounter English at all. They may come to learn quite a large number of English words and phrases (typically by associating them with the apparently corresponding word or phrase in Japanese), but in the absence of concrete situations in which they are exposed to a variety of registers manipulated by native speakers of English, their interest in matters related to register is necessarily very low. This general neglect of register in foreign language teaching is somewhat paradoxical in view of the fact that the Japanese language is very sensitive to different situational types and participant relationships, and speakers of Japanese are constantly involved in choosing appropriate registers in their daily communicative interaction through language (a well known case of which is the variety of different ways of speaking 'politely'). One may even suspect that Japanese speakers feel 'liberated' when they come to a Western language like English and are easily misled into thinking that in English all words and phrases are equal in terms of formality.

Register labels are clearly indispensable to the maximal usefulness of learner's dictionaries. There are, however, a number of questions about register specification confronting contemporary English lexicographers. I would like to discuss these questions with reference to the recently published new editions of three dictionaries designed for use by students of English as a foreign language, namely the Longman Dictionary of Contemporary English (LDOCE), the Cambridge International Dictionary of English (CIDE), and the Oxford Advanced Learner's Dictionary (OALD).'

Firstly, there is the question of what labels are used for marking register. There are several labels which the three dictionaries we are considering provide in common: 'formal', 'informal', 'slang', 'taboo', 'British', 'American' (or 'US'), 'Scottish' and 'Australian'. There is plenty of agreement in the way in which the two core terms 'formal' and 'informal' are defined, although minor differences may be noted, for example how far the correlation between 'formal'/'informal' and 'written'/'spoken' is emphasized, or whether 'politeness' is referred to in relation to the 'formal' register. There seems to be a little more uncertainty about the label 'slang', depending on whether it is conceived of as primarily located very low down the scale of informality, or as being used by a restricted group of speakers. There is still more uncertainty about the label 'taboo'. The three dictionaries agree in so far as they assign this label to words and phrases with obscene connotations. They do not always agree, however, on whether a word merits the ‘taboo' label. For example, CIDE marks the word honk as 'taboo slang', while OALD marks it as 'slang jocular', and LDOCE marks it as 'BrE slang humorous'. OALD marks piss as 'taboo slang', while CIDE marks it as 'slang', and LDOCE marks it as ‘informal' end begins the definition with the phrase "an impolite word meaning..." OALD also has a separate label 'offensive'. CIDE and LDOCE prefer to build the word 'offensive' into their definition of the individual word meaning.

Secondly, there are quite a few cases in which the labels used by the three dictionaries are not the same, but the intended notions are apparently the same. This seems to apply to 'jocular' (OALD 'intended to be funny') and 'humorous' (CIDE 'intended to make people laugh or smile', LDOCE 'used in a joking way'); 'archaic' (OALD 'no longer in current use') and 'old use' (CIDE 'used before the 20th century but not now', LDOCE 'used in earlier centuries'); 'dated' (OALD 'passing out of usage, old fashioned', CIDE 'sound(s) old-fashioned, not modern') and 'old fashioned' (LDOCE 'used earlier in this century, but would sound old fashioned today'); 'dialect' (OALD 'used in particular regions', LDOCE 'used in a particular part of Britain or the US') and 'regional' (CIDE 'used mostly by people from particular areas'); and 'technical' (OALD 'used by people who specialize in a particular field or fields', LDOCE 'used by doctors, scientists, or other specialists') and 'specialized' (CIDE 'used especially by people with special knowledge about a particular area of work or subject of study'). One point to be noted about these equivalent terms is that the three dictionaries agree in distinguishing between two stages along the chronological axis: 'having passed out of use' and 'passing out of use'. The labels for the latter are often used to mark those words and phrases which were at one time commonly used as slang but are now out of date. Another point worth mentioning is that 'derogatory' (OALD 'show(ing) that the user feels disapproval or scorn') and 'disapproving' (CIDE 'showing that the speaker has a bad opinion or feelings') can also be considered equivalent, while LDOCE opts instead to build the derogatory connotation into the definition, so as to avoid using a label which the student may not understand: e.g. acquisitive '... too much desire...', pedantic'...too much attention...' (this point will be discussed further).

Thirdly, there are labels which are used in one or two of the three dictionaries, but not in all three. The label 'spoken' is used uniquely by LDOCE. Words and phrases are labelled 'spoken' in LDOCE because they are much more frequent in the spoken corpus than in the written corpus. The question arises as to what the distinction is between 'spoken' and 'informal'. Apart from the possibility that a 'spoken' piece of language can be 'formal' as well as ‘informal', or that 'informal' speech may be presented in 'written' as well as 'spoken' form, it is interesting to note that the label 'spoken' in LDOCE is conveniently applied to mark those more or less fixed phrases frequently used in spoken language whose constituent words need not individually be marked as 'spoken'. The significance of this use of the label 'spoken' in LDOCE is twofold: firstly, it takes note of the important fact that a phrase as well as a word may need register specification (i.e. the register for a whole phrase may not be the same as the sum of the registers of its constituent words) and secondly, it demonstrates the usefulness of recognizing an intermediate stage between 'free expressions' and 'idioms'. In LDOCE, 'spoken phrases' are neither entirely 'free expressions' (because they are used in more or less fixed forms), nor entirely 'idioms' (because the meaning of the whole phrase is still more or less deducible from the meanings of the constituent words). Such 'spoken phrases' have failed to receive special lexicographical treatment for the reason that they are not exactly 'idioms'. If, on the other hand, they are simply presented as examples in the form of illustrative sentences, their special status may not be noticed. Compare the treatment of the phrase be an angel in the three dictionaries:

Worthy of note among the other labels uniquely employed by one, but not by all three of the dictionaries, are 'not technical' in LDOCE and 'not standard' in CIDE. These labels serve to remind the students that there are alternative expressions which may be used, depending on the context.

We now turn to the question of how individual words are characterized in terms of register labels and how much agreement in this respect can be found among the three dictionaries. Actually, we find quite a few cases of disagreement, even taking into account the fact that the labels provided by the three dictionaries are by no means the same. On the basis of checking the entries under A, the following general observations can be offered.

The three dictionaries agree most often in the assignment of the label 'formal'. Such cases include words like abjure, ablution, abnegation, abode, abominate, above mentioned, abrogate, accede, accoutrement, acquaint (with sth), acrimony, actuate, acuity, antipathetic, apparel, apprehend ('arrest'), ascend, assent, atone, attain, and attire. There are also cases where the label 'formal' is assigned by only one or two of the three dictionaries. For example, words like abdicate ('refuse to accept responsibility'), aberrant, abroad (in the 'rumour' sense), abstruse, accommodate ('get used to a new situation'), affirm etc are characterized as 'formal' by two of the three, and words like abstract (verb), absent (verb), abstemious, accommodate ('give someone a place to stay'), afflict, and antipathy are characterized as 'formal' by only one of the three dictionaries, i.e. LDOCE.

On the other hand, there are cases at the other extreme (although by no means many) in which three different labels are given by the three dictionaries. Thus abroad ('outdoors') is marked as 'archaic, rhetorical' (OALD), 'literary, old use' (CIDE) and 'old use' (LDOCE); acme has no label (OALD), is 'literary' (ClDE, and is formal' (LDOCE); alas (adverb) is marked as 'dated, rhetorical' (OALD), 'old use, dated' (CIDE), and 'formal, old use' (LDOCE); anew is marked as ‘rhetorical' (OALD), 'formal' (CIDE), and 'literary' (LDOCE); apprehend ('understand') is marked as 'dated, rhetorical' (OALD), 'formal' (CIDE), and fold use' (LDOCE); arcane has no label (OALD), is 'formal' (CIDE), and is 'literary' (LDOCE); and asunder is marked as 'dated, formal' (OALD), 'formal, literary' (CIDE), and 'literary' (LDOCE). Two points at least emerge from the picture reviewed above: firstly, the judgement tends to differ for words of minor rather than major importance, and secondly, 'literary' tends to be a highly uncertain label throughout.

The last point that I would like to discuss is the density of register label assignment. As was mentioned in a note, register marking seems rather sparse in COBUILD. Among the three dictionaries we are comparing, there is also a difference in density. LDOCE seems to mark register most densely. There are quite a number of cases in which a register label is given in LDOCE but not in OALD or CIDE. These include aboriginal, abstract (verb), abyss, accompaniment (noun 3), accustomed (adj 2), acquiescent, afflict, amplify (verb 2, 3) anchorite, antipathy, ardent, astray (2a), avenge, and awhile. Examples of the reverse case happening (i.e. marked both in OALD and CIDE but not in LDOCE) are very few e.g. abominable and acerbic. If we take into account the usefulness of information on register, the detailed specification found in LDOCE is certainly a feature to be welcomed.

By way of conclusion, let me add a few general comments on register specification. Firstly, there is the question of how far we may extend the notion of register. It is easy to see that there exists a certain ambiguity at its periphery. Consider again the label 'derogatory/ disapproving' already discussed. As we have seen, our three dictionaries are divided as to whether the notion associated with this label is to be presented as register or as part of the meaning defined for the word. There is nothing particularly strange about this, since both the register and the word meaning serve the user of the language as a constraint on the proper use of the word. The constraint on use imposed by the register 'derogatory/ disapproving' may in effect be nearly the same as the constraint on use imposed by the semantic feature 'too (much)' incorporated in the definition of the meaning of the word.

The question still remains as to what the essential difference is between register and word meaning. As a first approximation, I submit that the two notions concern 'how to say' and 'what to say' respectively. Since an option on 'how to say' necessarily entails an option on 'what to say' (in language, 'different forms, different meanings', as Bolinger said and as any literary critic would agree), it is natural that there is always a certain ambiguity between the two notions. Granting that the boundary between the two can be fuzzy, we can as a next step concentrate on the domain of 'how to say', and ask what are the factors that condition the speaker's choice of a particular way of saying something. Again a certain amount of ambiguity will be unavoidable, but the basic idea here will be the speaker's attitude (on which any decision on 'how to say' is ultimately dependent). A convenient subdivision can then be made between the speaker's attitude (I) towards what s/he is going to say, (ii) towards his or her hearer(s), and (iii) towards the circumstance in which s/he speaks. (Recalling Jakobson's well known scheme of classification of the functions of language, we may also want to add the speaker's attitude (iv) towards the (linguistic) code s/he is going to use.) Most of the register labels will still have to be characterized as involving more than one of these factors, but certain labels are easily seen as involving mainly one or two of them. 'Approving' and 'disapproving/derogatory', for example, refer primarily to (i). Surviving traditional terms like 'poetic' and ‘literary' will have to be reinterpreted in terms of the speaker's attitude (i.e. when the speaker opts to express himself 'poetically' or in terms of a 'literary work'). The reinterpretation presupposes a prior definition of 'poetry' or 'literature' as a text genre, but it serves to bring all the register labels together under the unifying classificatory principle of 'the speaker's attitude'.

One final point to be noted about register marking is that register concerns not only single words, but also phrases and sentences. In other words, it is not necessarily the case that once a word is marked with a particular register, it so functions in all of its uses. In this respect, the traditional lexicographical practice of marking register only for individual words, or at best for particular meanings of a word, is clearly inadequate. The new practice, introduced in LDOCE 3, of citing a phrase as a separate sense when it contains the entry word and has its own distinct meaning, as well as of citing 'spoken phrases' as prominent entries, is certainly commendable.

[1] The new edition of Cobuild Dictionary of the English Language is not taken into consideration because it does not go as far as the other three in its specification of register.

Back to top