Howard Jackson Leamers' Dictionaries in Contrast: Langenscheidt and Longman Abstract. Monolingual dictionaries ahned at advanced learners of English as a foreign language have been available since the Oxford Advanced Learner's Dictionary was first published in 1948; now there are at least three major contenders .in the market. The first such dictionary for learners of German has only recently (1993) appeared: the Langenscheidts Großwörterbuch Deutsch als Fremdsprache. The Langenscheidt is brought, in this article, in contrast with the 1987 edition of the Longman Dictionary of Contemporary English, which it most closely resembles. The contrastive study takes the user's perspective and examines the dictionaries from the point of view of the assumptions that are made in respect of users' prior knowledge and skills needed to extract all the information from the dictionaries. 1. lntroduction When the Longman Dictionary of Contemporary English was published in its first edition in 1978, it entered as a competitor in the English learners' dictionary market to the long-established Oxford Advanced Learner's Dictionary, then in its third edition (1974). A. S. Homby, with his colleagues E. V. Gatenby and H. Wakefield, had prepared the Advanced Learner's Dictionary, or the Idiomatic and Syntactic English Dictionary as it was originally called, while working as teachers of English as a Foreign Language in Japan before the Second World War (McArthur 1989). Interestingly, this dictionary was brought out during the Second World War by the Japanese publisher Kaitakusha, and then taken up by Oxford University Pressafter the war, appearing as A Learner's Dictionary of Current English in 1948. For the next thirty years it was the only dictionary of its kind and it achieved a world-wide reputation. lt would be a hard act to follow, and any competitor would have to make significant improvements. 1.1 LDOCE (1978) The Longman Dictionary of Contemporary English (henceforth LDOCE) claimed general inspiration from modern descriptive linguistics and emphasised two particular features as 'striking' (Randolph Quirk's word in the Preface ). The first of these was the use of a controlled defining vocabulary. All the definitions and examples in the dictionary were composed using approximately 2 000 words (and affixes), which are listed in an appendix to the dictionary. The definitions and examples were then computer-checked to ensure that they contained only words from the controlled defining vocabulary. Any words not from this source were given in FLuL 24 (1995) Learners' Dictionaries in Contrast: Langenscheidt and Longman 59 small capital letters, as a form of cross-reference. The controlled list was put together after scrutiny of frequency lists and of lists of words used in English teaching. The aim of so controlling the defining vocabulary was, in the words of the Editor-in-Chief, Paul Proctor, "the fulfilment of one of the most basic lexicographical principles that is that the definitions are always written using simpler terms than the words they describe" (LDOCE 1978: ix). To understand the explanations and examples in LDOCE, a user would need to be familiar with a restricted set of words, which they would, in any case, be likely to know from their learning of English. Tue second, 'striking' feature was the information given about the grammatical behaviour of words, and the means by which this was expressed in a series of codes. This was not so much an innovation as an improvement. Homby's Advanced Learner's Dictionary bad been rieb and innovative in its grammatical information: distinguishing countable and uncountable nouns; indicating the appropriate prepositions to be used after verbs, adjectives and nouns; and above all specifying the syntactic structures into which verbs might enter, by means of Homby's 'verb pattems' (see Homby 1954). All this information was intended to enable the user to exploit the dictionary for encoding purposes, to form 'correct' and appropriate sentences. Homby's verb pattems were indicated in the dictionary by means of a series of numbers prefaced by 'VP' (e.g. VP21, VP6c). Suchcodes were explained in a list of the verb pattems (of which there were more than 50, including subpattems) in the front matter of the dictionary. lt was difficult to leam which pattem was associated with which code, except by long and constant use. LDOCE presented a coding system, which was intended to be both more memorable and more systematic. Moreover, it reflected an up-to-date grammatical analysis of modern English, found in its similarly named Longman volume, A Grammar of Contemporary English (Quirk et al. 1972). Tue grammatical codes in LDOCE (1978) were composed mostly either of a single letter (e.g. 'P' for 'plural noun') or of a letter plus a number (e.g. 'U3' for 'uncountable noun followed by the infinitive with to'). Tue codes described the morphological and syntactic operation of nouns, adjectives and adverbs, as well as of verbs. Where possible, the codes were made mnemonic: 'C' stood for 'count noun', 'I' for 'intransitive verb', 'L' for 'linking verb', 'T' for 'transitive verb', 'D' for 'ditransitive verb'; 'O' meant 'not followed by anything', '1' meant 'followed by one or more nouns or pronouns'. But there were also many codes that bore no mnemonic relation to what they signified: 'B' stood for 'an ordinary (i.e. absolute form) adjective', 'F' for 'an adjective or adverb used after a verb', 'X' for 'a verb with one object + something else'; '2' meant 'followed by the infinitive without to', '4' meant 'followed by the -ing form', and so on. Clearly, the numbers bad less potential for being mnemonic than the letters. This coding scheme was a distinct improvement over that of the Oxford Advanced Learner's Dictionary (1974) in two ways: it brought the grammatical information into a unified coding system for all word classes that need morphological and syntactic operations described; FLuL 24 (1995) 60 Howard Jackson and it had the potential for being more easily leamed, with a large number of the codes relating in some obvious way to the syntactic operation they signified. Grammatical coding is not the only feature that makes a leamer's dictionary usable. Equally important are: the range of words included, the treatment of multiward lexemes (phrasal verbs and idioms); the quality of the definitions; information about style, register and collocations; the use of examples and pictorial illustrations; and so on. But the grammatical coding in LDOCE (1978) received considerable comment in reviews and discussions of the dictionary (e.g. Ellegärd 1978, Herbst 1984), probably because this is the most significant area in which leamers' dictionaries provide information either not found or not found so extensively or systematically in other dictionaries, with the purpose of offering help to leamers in producing and checking their own written language output. 1.2 LDOCE (1987) The Dictionary Department at Longman took due note of the comments by reviewers and of the experience reported by users of the first edition of LDOCE, many of which had been solicited by their own research projects. In 1987 a new edition was published: it retained the restricted defining vocabulary, but it presented a thoroughly revised scheme of indicating syntactic operation, with a minimum of coding. In addition, more attention was paid to the examples and to collocations, and a series of 'Language Notes' of a pragmatic nature were included on topics such as 'Apologies', 'Invitations and Offers', 'Politeness', as well as on some grammatical and lexical topics such as 'Articles', 'Collocations', 'Idioms', 'Prepositions'. Tue scheme for indicating syntactic operation in the new edition of LDOCE retained some of the more obvious and widely accepted letter codes from the first edition, such as 'U' for 'uncountable', 'I' for 'intransitive', 'T' for 'transitive'. The one less obvious retention here was 'F' for predicative adjectives, and presumably intended tobe mnemonic of 'following'. The other 'codes' (if they can be called such) in the scheme are more explicitly suggestive of the syntactic operation to which they refer: '[+ to-v]' means that a ward can be followed by an infinitive verb with to; '[+ (that)]' means that a ward can be followed by a clause beginning with that, but the that may be omitted; '[not in progressive forms]' means precisely what it says! For someone with the requisite grammatical knowledge, the scheme is now almost fully transparent and represents only a small leaming effort in order to extract the information on syntactic operation. In the same year, 1987, the Collins COBUILD English Language Dictionary appeared, to rival the Oxford and Longman products, with a number of innovative features. lt was based on the systematic computer investigation of a corpus of texts; its definitions were füll sentences, designed to read like teacher explanations in the classroom; grammatical and other technical data were separated from the main entries into an adjacent colurnn; the examples were all 'real English', derived from the corpus. By now the Oxford Advanced Learner's Dictionary was looking de- FLuL 24 (1995) Leamers' Dictionaries in Contrast: Langenscheidt and Longman 61 cidedly dated. Hornby having died, the editorship passed to Tony Cowie, who instituted a thorough revision, taking account of all the debate that bad been engendered on the nature of learners' dictionaries. A fourth edition of the Oxford Advanced Leamer's Dictionary appeared in 1989, with a new layout and the abandonment of the 'verb patterns' that bad been the means of giving syntactic information in the previous editions. 1.3 LGDAF (1993) The Langenscheidts Großwörterbuch Deutsch als Fremdsprache (henceforth LGDAF) bad its genesis in the rnid-1980s at the University of Augsburg, where an Anglist (Dieter Götz), a Romanist (Günther Haensch) and a Germanist (Hans Wellmann) cooperated in producing the guidelines for a learners' dictionary of German, with due consideration of the experience of making such dictionaries for French and English. The project was taken up by the Langenscheidt publishers in 1989 and the first edition reached the bookshops in 1993: the first German learners' dictionary. While account was no doubt taken of the whole learners' dictionary experience in French and English, LGDAF appears to have been modelled most closely on LDOCE, which is published in Germany by Langenscheidt. The general layout of the page matches that of LDOCE; a number of the illustrations have been taken from LDOCE; and a number of the principles underlying LDOCE are evident here, such as: attention to straightforward definitions, though not (at least not stated) with a restricted defining vocabulary; attention to collocations and examples, to illustrate meaning and use; systematic and comprehensive information on morphological and syntactic operation, the latter by means of "Strukturformeln" though these are more akin to the scheme used in the fourth edition of the Oxford Advanced Leamer's Dictionary (1989) than that in LDOCE (1987). LGDAF lacks the pragmatic 'Language Notes' of LDOCE; the equivalent tables in LGDAF are devoted largely to grammatical matters (adjective inflections, demonstrative pronouns, norninalised adjectives and participles). German is different from English in morphology and syntax, and it is appropriate that attention should be directed at the linguistic features that learners particularly need to know about. 2. Reference Skills There are any of a number of approaches that could be taken in attempting to bring two dictionaries into contrast, from a general survey of their layout and content to a detailed analysis of some particularly relevant or striking feature (the nature and use of examples, perhaps). The approach to be taken in this article takes the perspective of the dictionary user's prior knowledge and skills that appear to be assumed by the compilers of LDOCE and LGDAF. One way of formulating this FLuL 24 (1995) 62 Howard Jackson approach is by means of the following questions: What does a learner need to know, in terms of understanding linguistic terminology, in order to be able to use successfully, or to extract the maximum information from, the dictionaries under consideration? What reference, or more specifically dictionary, skills does a learner need in order to access the information in these dictionaries? The assumption made by both dictionaries is that users will come with experience of using a bilingual dictionary. LDOCE envisages that a learner will progress from using a bilingual dictionary to using a monolingual learners' dictionary (Quirk in the Preface to LDOCE 1987). LGDAF sees the monolingual learners' dictionary as "die ideale Ergänzung zum zweisprachigen Wörterbuch" (p. vii). Since such dictionaries are aimed at 'advanced' learners, certain assumptions can be made about linguistic knowledge and previous experience of using dictionaries. However, as Henri Bejoint found when he surveyed bis students at Lyon (Bejoint 1981), most of them did not make use of the grammatical information available in their monolingual dictionaries. General dictionary skills relate primarily to the practice of decoding, especially reading. There are particular, encoding, skills that the user of a learner's dictionary needs to develop. 2.1 General dictionary skills As part of our general education, we are taught, or at least acquire, skills in looking up information in reference books: the use of alphabetised lists, contents, indexes; finding a place by paragraph, section, chapter and page numbers or headings; understanding the use of abbreviations, such as 'p' for 'page', 'ff for 'following'; and so on. Same of these skills are relevant to using dictionaries, but a dictionary demands more of a user. The list of headwords in a dictionary is in alphabetical order, but the ward that is being searched for may not have headword status. lt may be contained within the entry for another ward: this is the case very often for words derived by suffixation though not, of course, for those derived by prefixation as well as for compounds, idioms and other multi-ward lexemes. A dictionary user must develop strategies for finding items, based on experience and, perhaps, lexical knowledge. For the user of a foreign language dictionary, whether bilingual or monolingual, where lexical knowledge and experience is more lirnited, there may be further strategies to develop. For example, the first acquaintance with a verb may be in its past tense or past participle form. If it is irregular in its inflections, it may be listed as a headword and cross-referenced to the füll entry at the infinitive headword (e.g. gefunden, gegangen, gegeben in LGDAF). But if it is inflected regularly, the ward will need to be converted to its appropriate base or infinitive form, which may not always be obvious: gefastet is the past participle inflection of fasten, but gefährdet of gefährden, andfasziniert offaszinieren. There is, then, a skill in using a dictionary associated with finding the entry for the ward that you wish to look up. FLuL 24 (1995) Learners' Dictionaries in Contrast: Langenscheidt and Longman 63 There is also a skill associated with reading dictionary definitions. One of the innovations of the Collins COBUIW (1987) dictionary was to use füll sentences for definitions, because traditional definitions based on a phrase were thought to pose some problems of comprehension for language learners. Both LDOCE and LGDAF use a traditional style of defining, which is more economical on space, and for which users have, arguably, already developed some skill in making sense of. Economy of space is also responsible for a further skill that dictionary users must develop: the interpretation of the abbreviations used in dictionaries. H. G. and F. W. Fowler, the first editors of the Concise Oxford Dictionary (1911), claimed to have abbreviated as much as they thought could be tolerated by the potential users of their dictionary, as a way of saving space. Abbreviation occurs mainly in the grammatical and stylistic labels used in dictionary entries, e.g. word class labels (n, v, adj, prep, etc.), but it also inc1udes devices like the swung dash, used to stand for a headword within its entry. 2.2 Particular learner skills Tue general dictionary skills are largely to do with locating the item that you wish to look up and with interpreting the entry in order to apply its information to understanding a word that has been encountered, notably in the process of reading. They are concerned with decoding tasks. Learners' dictionaries also, perhaps even primarily, have encoding tasks in view. So, there is a further set of skills that a learner needs to develop in order to access and apply dictionary information in the encoding process, principally writing. First of all, a learner needs to be able to make sense of the grammatical terminology and to translate the description that it embodies into appropriately structured sentences. A grammatical terminology will reflect a particular model of grammatical description, as LDOCE (1978) reflects Quirk et al. (1972). If the terminology is not entirely farniliar to a user, or if a coding system is used, there may be some learning and skill development to be undertaken, by consulting the guide to using the dictionary, perhaps something that Bejoint (1981) found few of his students had done. Tue developments reflected in LDOCE and LGDAF attempt to make the grammatical terminology transparent and learnable, but a level of knowledge and understanding is required for the information to be accessed and applied successfully. Tue same point essentially applies to register and style labels. A learner needs some skill in interpreting labels such as 'derogatory' and 'pompous' (LDOCE) or 'pejorativ' and 'vulgär' (LGDAF), not to mention the distinction between 'oldfashioned' and 'old use' (LDOCE) or between 'veraltend' and 'veraltet' (LGDAF). While grammatical terminology may be related to an established grammatical model, style and register labels do not obviously derive from a well-rehearsed pragmatic or sociolinguistic theory. Tue guide to the dictionary is the only recourse. FLuL 24 (1995) 64 Howard Jackson lt must be stressed that such information should be of considerable benefit to a learner, who, having acquired the necessary knowledge and skill, can be more confident in using words so marked more appropriately. Tue same is true of the more extensive structural semantic information to be found increasingly in both native-speaker and learners' dictionaries. Both LDOCE and LGDAF are concemed to provide information, though in different ways, on collocations and on synonyms and antonyms. Same understanding of structural semantics and skill in being able to make use of this information is required on the part of a dictionary user. 3. Using LDOCE (1987) lt should be emphasised that the edition of LDOCE under consideration is the revised version of 1987, with the modified scheme of grammatical coding. We will consider it from three perspectives concemed with the assumptions made about a user's skill and knowledge: finding words, figuring out the abbreviations, understanding the terminology. 3.1 Finding words LDOCE devotes the first four pages of its 'Quick Guide to Using the Dictionary' to this topic: it recognises this as an important issue for dictionary users, and especially for learners, who cannot be assumed to have such extensive experience of and intuitive knowledge about the language, even though their conscious knowledge about the language may well be much greater than that of most native speakers. Nowhere is this perhaps more relevant than in identifying a ward that has only been heard, and not been seen in its orthographic form. This can be a problem for native speakers: few would guess that in phthiriasis the initial ph is silent and that the second i is pronounced / ai/ . Tue vagaries of English spelling are wellknown, with the one-to-many correspondence between sound and letter presenting difficulties for dictionary users. LDOCE has gone some way towards recognising this difficulty by producing a 'Pronunciation and Spelling Table' on one side of a separate laminated card (the other side contains the grammar coding): the table gives a key ward to illustrate the spelling of each phoneme (of Received Pronunciation) and a number of other words (up to six in some cases) illustrating further common spellings. This is a recognition that there may be a lack of experience or skill on the part of a user that needs to be aided, and it is probably as far as such aids can go without producing a dictionary where headwords are based on pronunciation. Such a table may not help much with working out that / sju: doua: keiik/ translates orthographically into pseudo-archaic, but it will be a substantial aid in finding a ward that has only been heard and not seen. In common with all dictionaries, LDOCE assumes that the user is familiar with the regular inflections of English verbs, nouns and adjectives. This is a reasonable FLuL 24 (1995) Leamers' Dictionaries in Contrast: Langenscheidt and Longman 65 assumption, since to assume otherwise, and have a separate cross-referenced entry for every inflectional variant, would produce an unwieldy dictionary. In any case, this is likely to be conscious and familiar knowledge for an advanced learner of the language. lrregular inflectional variants, e.g. mice, taught, do have separate entries, with c,ross-reference to their base form. These are the items that are likely to cause difficulty, so that the assumptions about users' pri? I" knowledge of inflections is appropriate. We turn now to derived and compound words. All compounds, whether written open (post office), hyphenated (post~free) or solid (postman), are entered as separate headwords, so that they are located by following the alphabetical arrangement of headwords. The same is true of words derived by prefixation (befriend, disagree, reopen): they occur in the appropriate place in the alphabetical sequence. Words derived by suffixation are treated in two ways: if they are considered to warrant separate definition, they have their own entry; but if they are undefined, because th~ir meaning can be deduced from the root, then they are entered as run-ons under the root, with a word-class label and if considered appropriate pronunciation and/ or examples. So, removal is a headword, but removable is a run-on under remove (without pronunciation or examples); transformer is a headword, but transformable and transformation are run-ons under transform, the latter with pronunciation and examples. This reflects common practice in modern native-speaker dictionaries; LDOCE perhaps tends to define more derived words, and so give them headword status, than a native-speaker dictionary. Similarly, LDOCE, in keeping with general Longman practice, gives a separate entry for each ward class represented by a particular spelling, and not just for homonyms and homographs. So, commission has two entries, one as noun and the other as verb; silver has three, as noun, adjective and verb. The effect of these measures is to maximise the number of headwords in the dictionary and so ease the task of locating a desired ward. A further set of items that are sometimes difficult to locate in a dictionary are multi-ward lexemes. In English, this means especially phrasal verbs and idioms, two areas where a learners' dictionary would do well to make few assumptions about prior knowledge or skill on the part of users. In respect of phrasal verbs, LDOCE lists them clearly, each with its own subentry, at the end of the entry for the single verb lexeme. So, turn has subentries for eleven phrasal verbs, from turn against to turn upon. In respect of idioms, they are listed under the entry for the first 'main' ward in the phrase, as a numbered sense or meaning of the ward, after all the other senses. So, under nose are found five idioms, as senses 5 to 9, including get up someone's nose, put someone's nose out of joint and under someone's (very) nose. At the very end of the entry, there are cross-references to other idioms that contain nose, e.g. pay through the nose cross-referenced to pay. The LDOCE compilers seem to have taken into consideration the difficulties that a learner might have in locating words and phrases in the dictionary, and they have provided appropriate help, beyond what might be expected in a native-speaker dictionary. FLuL 24 (1995) 66 Howard Jackson 3.2 Figuring out abbreviations Apart from the grammatical codes, which will be considered in 3.3, LDOCE lists some 65 abbreviations under the heading 'Short Forms and Labels' on the inside back cover. Tue list includes: word-class labels (adj, n, prep) and other grammatical terms (fern, pass, pl); labels for regional varieties (AmE, dial, NZE, ScotE), languages (Fr, Ger, Lat), style and status (derog, euph, infml, sl), and register (bibl, med, tech); and general dictionary terms (abbrev, esp, sbdy, usu). When combined, these can produce label series such as: "moving picture n fml, esp. AmE for MOVIE"; "mucky adj infml or dial", "nosy parker n BrE derog infml". This is a modest number of abbreviations in a dictionary. Most of them will be familiar to advanced learners of the language. Tue few that are not familiar do not present a very great learning task. There may be a greater problem in understanding a few of the terms that the abbreviations stand for, such as 'appreciative', 'oldfashioned', 'pompous', though these are given some explanation in the 'Full Guide to Using the Dictionary' (p. F 46). Tue compilers of LDOCE seem to have judged the skill of users in figuring out abbreviations at about the right level. 3.3 Understanding the terminology A dictionary represents a lexical description of a sample of the vocabulary of a language, possibly written, as in the case of learners' dictionaries, with a particular group of users in mind. A lexical description uses terms from general descriptive linguistics, which will reflect a formal or informal model of language. Lexical description does not create its own terms: word class categories and terms such as 'transitive' and 'passive' come from grarnmar, 'collocation' and 'synonym' come from semantics, 'dialect' and 'informal' come from sociolinguistics or pragmatics, and so on. What can a dictionary compiler assume about the extent and sophistication of a user's linguistic knowledge? A learner's dictionary sets itself the task of providing information that will enable the user to encode grarnmatically correct and contextually appropriate sentences. lt must therefore provide an adequate description of the relevant grammatical, semantic and pragmatic features of each lexical item. This description must be presented in appropriate terminology. The best that a dictionary can do is to use terminology that is widely known and accessible. LDOCE uses the grarnmatical terminology of Quirk et al. (1972, 1985), which could be characterised as a development, under the influence of modern descriptive linguistics, of traditional grammar. Consequently, the general approach and much of the terminology are likely to be familiar to language learners. This is the case, for example, with the word-class labels, though the terms 'determiner' and 'predeterminer' may not be familiar to some users, since these are relatively recent terms. Items in these word classes were previously considered members either of the adjective class (articles, demonstratives, numerals, quanti- FLuL 24 (1995) Leamers' Dictionaries in Contrast: Langenscheidt and Longman 67 fiers) or of the pronoun dass (possessives). Tue terms are not used by current British native-speaker dictionaries, except for Collins English Dictionary (1979), which uses the term 'determiner'. However, LDOCE is not consistent in its designation of the possessives 'my', 'your', etc.: in the entries for these items, they are labelled 'determiners', but in the usage note under the entry for 'father' they are referred to as 'possessive pronoun'. Tue terms used to refer to inflections (e.g. plural, 3 rd person singular present, past participle, superlative) are standard and are legitimately assumed to be known by users. Tue syntactic terms also derive, for the most part, from this pool of standard terminology. For nouns, the terms 'countable' and 'uncountable', coded 'C' and 'U', are now well-established, even though not yet taken up by native-speaker dictionaries. Adjectives with restricted occurrence are coded 'A' for 'attributive' and 'F', presumably for 'following'. Tue 'F' code is, rather confusingly, explained in two ways: in the 'Full Guide to the Dictionary' (p. F 40) it is glossed as 'used only after a verb'; but in the summary table on the inside front cover and the laminated insert it is said to mean 'used only following the noun that it describes'. This could equally well apply to an adjective like galore, which, however, is marked in the dictionary simply as 'after n'. Clearly, F refers to adjectives like afraid, which are restricted to what is called 'predicative' position (Quirk et al. 1972: 247), after a copular verb. lf the distinction between 'attributive' and 'predicative' (both terms are entered in LDOCE) is familiar to users, it is likely tobe in those terms: using 'F' and glossing it in an unclear way is not helpful. Verbs are coded 'I' (instransitive), 'T' (transitive) or 'L' (linking): 'transitive' and 'intransitive' are familiar terms, 'linking' may not be. Most native-speaker dictionaries make a distinction only between transitive and intransitive uses of verbs; the exception is the more linguistically sophisticated Collins English Dictionary (1979), which marks verbs like be, seem and sound as 'copula'. 'Copula' is the term that is used in Quirk et al. (1972: 820, 1985: 1171), though the term 'linking verb' is also mentioned. However, if the concept is known to a user of LDOCE, it is more likely that the term 'copula' is familiar rather than the term 'linking verb'. lt seems as if the coding system is driving the terminology, both in respect of 'L' and of 'F': the more familiar terms, 'copula' and 'predicative' could not be coded by 'C' and 'P', because 'C' is already needed for 'countable' and 'P' for 'plural'. Tue general account of the syntax of verbs uses terminology that is either standard (e.g. direct/ indirect object, clause, complement) or descriptive of surface syntax (e.g. to-infinitve, that-clause, clause with wh-). Tue problematical term here is wh-, which could be an unfamiliar term but is not properly explained in the 'Full Guide' (p. F 42/ 43), and even less so in the summary table: "[+wh-] shows that a word can be followed by a word beginning with wh- (such as where, why or when) or by how." There is no indication that the wh-word will introduce a clause, nor that this clause could be infinitive. One of the examples in the entry for advise is marked [+obj+wh-] and it contains an infinitive clause after the wh-word: "She advised us where to eat." No distinction is drawn either between indirect questions FLuL 24 (1995) 68 Howard Jackson (which could be introduced by if/ whether) and nominal relative clauses. That may be justified: it is often a difficult distinction to draw in practice and it is fair to assume that many users will be unfamiliar with it; these are in any case functional terms and LDOCE generally limits itself to the more familiar categorial terms. However, even as a categorial term, whis not clearly explained, although it is likely to be the least familiar of the syntactic elements indicated. Finally, we should note that LDOCE uses the term 'progressive aspect' for the 'be + -ing' verb forms, rather than 'continuous' or the more categorial 'expanded form'. 'Progressive' has, perhaps, the widest currency, although it may not be the most familiar term to students coming from some English teaching traditions. In general, we can say that LDOCE's assumptions about students' familiarity with grammatical terminology are reasonable, with the exceptions that we have noted in respect of 'F', 'L' and wh-. lt does not demand a great amount of new learning or revision of terms already leamt. While there is a pool of well-established terminology for grammar, the same is not the case for semantics and pragmatics. The labels used by LDOCE to mark particular style and usage restrictions are explained in the 'Full Guide' (p. F 45/ 46); 'Language Notes' and 'Usage Notes' are couched in non-technical language. Perhaps less thought has been given to some of the collocational information. For example, the verb benefit is said to be "esp. of an action or event", while laudable is said to be "esp. of behaviour, actions, etc". lt is not clear whether terms like 'action' and 'event' are being used herein a technical sense (cf. Jackson 1990: 8) or informally. lf the latter, what does a user understand by 'action or event', and if the former, an unwarranted assumption is perhaps being made about prior semantic knowledge. 4. Using LGDAF (1993) One of the features of recent learners' dictionaries is the attention paid to the 'Guide to the User'. LDOCE (1987) has two such guides: a 'Quick Guide' of some thirteen pages, based on examples and framed in terms of dictionary operations ('Choosing the right form of a word', etc); and a twenty-four-page 'Full Guide', more tightly packed and explaining all the features, abbreviations and labels used in the dictionary. LGDAF has an eighteen-page 'Hinweise für den Benutzer', which gives an extensive description of the features of the dictionary, though with a minimum ~of exemplification and written in an impersonal style: the user is not addressed, as they are in both guides in LDOCE (1987). 4.1 Finding words Like LDOCE, the first section of 'Hinweise für den Benutzer' in LGDAF, entitled ~Wo findet man was? ', is devoted to instruction on how to find items in the dictionary. Generally, learners of German do not have the difficulty in deducing the FLuL 24 (1995) Leamers' Dictionaries in Contrast: Langenscheidt and Langman 69 written form of a word from its spoken form that learners of English experience. So there is no 'Pronunciation and Spelling Table' in LGDAF. The only uncertainty rnight be the way in which umlauted vowels are treated in the alphabetical ordering: LGDAF follows the user-friendly tradition of listing words containing them as if the vowel was not umlauted; so, fällen comes between fallen and fallenlassen. To be a successful dictionary user, ,a learner of German needs to know more about inflections than a learner of English does. There is, for example, no simple regular/ irregular distinction for noun plurals, as there is in English; there is a greater variety of verb inflections; and adjectives are inflected. No noun plurals are given separate entry in the dictionary: this is reasonable, since, on the one band learners should have some experience of plural inflections, and on the other the relation of plural to singular form is usually clear and would in any case appear in close proximity to the singular form in an alphabetical listing (cf. Ball/ Bälle, Maus/ Mäuse, Turm/ J'ürme). The same is true of adjective inflections. For verbs, the distinction between regular and irregular is relevant. LGDAF assumes that users can recognise regular present and past tense and past participle forms of verbs, including those ending in -ieren or formed with inseparable prefixes (e.g. be-, er-, ver-) that do not take a geprefix in the past participle. However, in the case of irregular (strong) verbs, irregular 3 rd singular present, past and past participle forms are listed as separate entries and cross-referenced to their full entry at the infinitive. lrregular verbs are also listed in Anhang 5 (pp. 1180-1182), as are the English ones in LDOCE Appendix 6 (1987: B23-B26). Adictionary, even a learners' dictionary, has to make assumptions about users' ability to relate inflectional variants to the citation form in the dictionary. Some outline of the processes involved could, nevertheless, have been usefully included in the 'Wie findet man was? ' section of the 'Hinweise für den Benutzer'. What this section does inform the user about is the treatment of compound and derived words, as well as of idioms. An advanced learner will have some familiarity with the compounding possibilities of German, and LGDAF seeks to build on that familiarity. After the symbol 'K-' for 'Komposita', within an entry, it lists, without definition, the common compounds having the headword as first element, and after '-K' those that have the headword as second (last) element. Unless these latter are also listed under their first element, the user will have difficulty finding them. Not all of them are: Boulevardpresse is under Presse but not Boulevard; Flugsand is under Sand but not Flug, Rudersklave is under Sklave but not Ruder. Compounds with a meaning that can no longer be deduced from that of its parts are separate headwords. Words derived by suffixation are treated similarly: if without definition, then listed at the end of an entry after 'II hierzu', otherwise as separate headwords. Prefixes, and indeed derivational suffixes, have separate entries; the ones for prefixes are more extensive and more likely to be found, by a user searching for a prefixed derivative. Not only are the various meanings and functions of prefixes listed, but they are also copiously exemplified, with the e: X: amples then treated as headwords at their appropriate place in the alphabetical FLuL 24 (1995) 70 Howard Jackson sequence. LGDAF seems to have made reasonable assumptions about users' possible prior knowledge and experience of word formation, and has provided significant help to extend it. The treatment of idioms is not so helpful. An idiom is listed only once, and the rules for which word it will occur under are not straightforward: if there is one in the idiom, it is under the first noun, otherwise under the first adjective, or alternatively the first verb. A user could be engaged on a long search. LDOCE has a more user-friendly treatment of idioms. 4.2 Figuring out abbreviations On the inside back cover of LGDAF is a list entitled 'Verwendete Abkürzungen und Bezeichnungen'. lt contains some 100 abbreviations, and in addition some twenty unabbreviated labels. The unabbreviated labels are the following: Artikel, Demonstrativpronomen, Film, Geometrie, Indefinitpronomen, indeklinabel, Infinitiv, Interjektion, Interrogativpronomen, Konjunktion, Meteorologie, ohne Steigerung, Partikel, Partizip Perfekt, Partizip Präsens, Possessivpronomen, Relativpronomen, Seefahrt, veraltend, veraltet. lt is not clear in many of these cases why they have been left unabbreviated when other similar terms have been abbreviated. Why 'Geometrie', but not 'Gastronomie' (Gastr), 'Geographie' (Geogr) and 'Geologie' (Geol)? Why 'Artikel' and 'Konjunktion', but not 'Adjektiv' (Adj) and 'Präposition' (Präp)? We might assume that terms would be left unabbreviated if they were thought to be unfamiliar to users or where possible abbreviation could not be readily associated with the füll form. If this is the principle that LGDAF has followed, it seems to have been applied rather arbitrarily. Some abbreviations, rather confusingly, do not relate directly to their füll form: 'Agr' stands for 'Landwirtschaft', 'Jur' for 'Recht', 'Ökon' for 'Wirtschaft' or 'Volkswirtschaftslehre', 'Transp' for 'Verkehrswesen'. Rather subtly, 'lit' represents 'literarischer Sprachgebrauch', while. 'Lit' stands for 'Literaturwissenschaft'; and 'gespr' is glossed as 'gesprochene Sprache, umgangssprachlich', while 'gespr! ' means 'sehr salopp, aber noch nicht vulgär'. The abbreviations include word-class labels (Adj, Präp, Subst, Zahladj), other grammatical terms (Akk, Pl, präd), labels for regional varieties (CH, nordd, südd), style and status (euph, gespr, pej, vulg), a more extensive (nearly 50) set of register labels (Aero (for Luftfahrt), EDV, Mil, Rel, Soz); and general dictionary abbreviations (Abb, bes, mst, o.ä., od). There is also a series of abbreviations for the indefinite article (e-e for eine, e-m for einem, and so on) and for jemand G-d, j-n for jemanden, j-s for jemandes, and so on). LGDAF makes a greater demand than LDOCE on the user in respect of abbreviations. To be sure, the abbreviation list on the inside back cover is easily referred to, but having to look up too many abbrevations can be dispiriting and detract from the more essential information that a user needs to access in the dictionary. The FLuL 24 (1995) Learners' Dictionaries in Contrast: Langenscheidt and Longman 71 compilers of LGDAF have perhaps here assumed too great a knowledge and skill on the part of users. 4.3 Understanding the tenninology Unlike LDOCE, LGDAF does not specifically derive its grammatical terminology from a published grammatical description, nor is it well described in the 'Hinweise für den Benutzer'. For example, a number of word dass (Wortarten) labels - 'Substantiv', 'Verb', 'Adjektiv' and 'Adverb' have an extensive description because users need to be informed about the treatment of their inflections. Tue remaining word dasses are simply listed (in 5.4, p. XVII): Artikel, Demonstrativpronomen, Fragewort, Indefinitpronomen, Interjektion, Konjunktion, Partikel, Personalpronomen, Possessivpronomen, Relativpronomen, Zahladjektiv. Tue term 'Präposition' has been omitted from this list, presumably an oversight. Tue term 'Fragewort' is not actually used in the dictionary: items like Was? and Wer? are marked 'Interrogativpronomen'; and items like Warum? and Wie? are marked 'Adv' for 'Adverb'. Besides these inconsistencies, it is a considerable assumption that users will know what is meant by each of the word dass labels listed in paragraph 5.4. lt may well not be immediately dear what is included in the dass of 'Partikel' or what is meant by the term: items like doch and wohl are induded in the adverb dass by Collins German Dictionary (1981). Moreover, a distinction is made in the entries for these items between · 'betont' and 'unbetont', which is not explained in the 'Hinweise'. There is also some uncertainty about how items are treated, which would attract the label 'determiner' in LDOCE. LGDAF treats the possessives (mein, etc.) in both their pronoun and determiner roles as 'Possessivpronomen'. Helbig/ Buscha (1980: 230, 316) treat them differently, assigning determiner possessives to the dass of 'Artikelwörter'. Similarly, Helbig/ Buscha distinguish between demonstrative pronouns and demonstrative 'artides' (1980: 223, 316), and treat items like jeder, mancher as artides: LGDAF treats these as 'Demonstrativpronomen' and 'Indefinitpronomen' respectively. Some of these shortcomings may be compensated for by the darity of the entries and the rich exemplification. However, if terminology is not understood, the user does not have access to the füll range of information that the dictionary offers. Some terms are explained in the 'Hinweise', such as 'produktiv' in respect of affixes (p. XI) and 'attributiv' and 'prädikativ' in respect of the use of adjectives (p. XVII). But there are some instances where assumptions are made about users' knowledge of terms and no explanation is given. Can it be assumed, for example, that a user knows the term 'Konjunktiv' and the distinction between 'Konkunktiv I' and 'Konjunktiv II' (e.g. the entries for brächte and gäbe)? What is the standard term to refer to the 'past tense' form of the verb: 'Imperfekt' is used in the 'Hinweise' (p. XVI) andin the dictionary entries (e.g. hielt), but 'Präteritum' is used in the table of irregular verbs (Anhang 5, p. 1180-2). A number of nouns are marked FLuL 24 (1995) 72 Howard Jackson 'Kollekt(ivbegriff)', but there is no explanation of the significance of this designation, either morphologically or syntactically. One of the features emphasised by LGDAF is its description of the syntax of verbs by means of 'Strukturformeln'. Verbs are designated as: 'Vi' (intransitive), 'Vimp' (impersonal verb), 'Vr' (reflexive verb), 'Vt' (transitive verb), 'Vt/ i' (transitive verb that can omit its object). This classification is explained in the 'Hinweise' (p. XXI f), as is the scheme of 'Strukturformeln' (pp. XXIII-XXVI), with copious exemplification. The formulae use 'etw.' to stand for a non-human noun, various forms of jema,nd G-d, j-n, etc) to indicate a human noun, items such as irgendwohin, irgendwann to indicate adverbial elements. Tue forms of jema,nd indicate the case required; if there is possible confusion with etwas, the case is given (e.g. 'etw. (Akk)'), Brackets indicate optional (facultativ) elements. For example, "konspirieren .. j-d konspiriert mit j-m (gegen j-n/ etw.)", "senden 2 .. G-m) etw. s•... 3 .. j-n/ etw. irgendwohin s." Once the general principles behind the 'Strukturformeln' have been understood and the common forms of abbreviation remembered, it is a transparent scheme and one that does not place a great burden on the user's memory. What the scheme does not give information about, however, is the range of syntactic elements. that can fill the various slots in the formulae for any particular verb; in particular, there is little information, except in the examples; about the nominal clauses that are possible after verbs. The entry for berichten, with the formula 'G-m) etw. b.' contains the 'NB': 'Das Akkusativobjekt ist nie ein Subst.' lt is not clear what a user is meant to make of this statement. Occasionally such information is given; under 'fragen Vr 5' the formula 'sich f., ob/ warum/ wie .. .' is given. In the discussion of 'Strukturformeln' in the 'Hinweise', nominal clauses are not mentioned, though a couple of 'zu + Infinitiv' appear in the examples. Are users expected to already know this kind of information, which is comprehensively and explicitly provided in LDOCE, or are they expected to look elsewhere for it? lt would seem appropriate to fmd it in a leamers' dictionary that aims to enable its users to encode grammatical sentences. As in LDOCE, terminology relating to semantic and pragmatic matters is explained in the 'Hinweise' of LGDAF. The information presented here is rather more extensive than in LDOCE. Synonyms (marked by a double swung dash : : : : ), antonyms (marked by a doubleheaded arrow H), and collocations (enclosed between angle brackets <>) are routinely given in entries. Stylistic and usage labels are given in respect of: 'Stilebene' (e.g. gespr, geschr, vulg), 'Einstellung des Sprechers' (e.g. euph, pej, iron), 'Sprache und Zeitbezug' (e.g. veraltet, hist), 'Sprache und Region' (e.g. CH, nordd, ostd), and 'Fachwortschatz' (nearly 50 labels). FLuL 24 (1995) Leamers' Dictionaries in Contrast: Langenscheidt and Longman .73 5. Conclusion Learners' dictionaries are aimed at a wide constituency of users, who may speak any of a wide variety of first languages and who may have beeil taught by any of a variety of methods. Users may also possess varying levels of reference skills and have received more or less instruction in using either a bilingual or a monolingual dictionary. Users may possess varying amounts of linguistic knowledge and be familiar to varying degrees with linguistic terminology. A dictionary compiler must pitch the demand made on the user at a level that will not alienate the less compe-' tent nor patronise the more competent of the intended user group. From our comparison, it would seem that LDOCE and LGDAF have pitched their demand on users at slightly different levels: LDOCE at a lower level than LGDAF. Tue English dictionary . .is more generous with its explanations in its guides to the use of the dictionary, lises fewer abbreviations and potentially unfamiliar linguistic terms, though one or two inconsistences were noticed; it makes it easier to find idioms, provides a pronunciation to spelling guid~, and gives _füll information on the syntactic operation of verbs, including the use of nominal clauses. Tue German dictionary gives a füll selection of compound and derived words and is more generous with its register labelling; but it makes a greater demand in its use of abbreviations, uses terminology that lacks explanation, has some inconsistencies in use of terminology, and, despite its transparent system of structural formulae, does· not provide a comprehensive account of the syntactic operation of verbs; the 'Hinweise für den Benutzer' could be usefully expanded, and expressed in a more user-friendly tone. Dictionarie_s are subject to a process pf continual revision and improvement. ·Tue existence of direct competitors is a spur to that process. Dictionary publishers are more aware, especially in the case of specialised dictionaries, of the nature and needs of their intended users. They should be encouraged to monitor who uses their dictionary and what use they make of it, and adjust their presentation accordingly. Longman have followed up LDOCE (1987) with the Longman Dictionary of English Language and Culture (1992) and the Longman .Language Activator Dictionary (1993) and indeed a third edition of LDOCE (1995). We look forward to the next edition of the Langenscheidt Großwörterbuch Deutsch als Fremdsprache and further offshoots. FLuL 24 (1995) 74 Howard Jackson References BE.JOINT, Henri (1981): "The Foreign Student's Use of Monolingual English Dictionaries". In: Applied Linguistics 2.3, 207-222. Collins English Dictionary (1979, 1986, 1991). Glasgow: Harper Collins. Collins German Dictionary (1981). 'fERRELL, P. [et al.] (eds.). Stuttgart: Collins & Klett. Concise Oxford Dictionary (1911). F0WLER, F. G./ F0WLER, H. W. (eds.). Oxford: Oxford University Press. ELLEGÄRD, A. (1978): "On Dictionaries for Language Learners". In: Moderna Sprak 72.3, 225-242. HELBIG, Gerhard/ BUSCHA, Joachim (1980): Deutsche Grammatik, Ein Handbuch für den Ausländerunterricht. Sixth Edition. Leipzig: VEB Verlag Enzyklopädie. HERBST, Thomas (1984): "Bemerkungen zu den Pattemsystemen des Advanced Learner's Dictionary und des Dictionary ofContemporary English". In: GöTZ, Dieter/ HERBST, Thomas (Hrsg.): Theoretische und Praktische Probleme der Lexikographie. München: Max Hueber Verlag, 139-165. H0RNBY, A. S. (1954): A Guide to Patterns and Usage in English. Oxford: Oxford University Press. JACKS0N, Howard (1990): Grammar and Meaning. London: Longman. Langenscheidts Großwörterbuch Deutsch als Fremdsprache. Das neue einsprachige Wörterbuch für Deutschlernende. Dieter GöTZ, Günther HAENSCH, Hans WELLMANN. In Zusammenarbeit mit der Langenscheidt-Redaktion. Berlin: Langenscheidt 1993. Longman Dictionary of Contemporary English (1978). PROCTOR, Paul (ed.). London: Longman. Longman Dictionary of Contemporary English (1987). SUMMERS, Della (ed.). London: Longman. Longman Dictionary of English Language & Culture (1992). London: Longman. Longman Language Activator Dictionary (1993). SUMMERS, Della (ed.). London: Longman. MCARTHUR, Tom (1989): "The Background and Nature of ELT Learners' Dictionaries". In: TICK00, M. L. (ed.): Learners' Dictionaries: State of the Art, SEAMEO Regional Language Centre, 52-64. Oxford Advanced Learner's Dictionary (1948, 1963, 1974, 1989). H0RNBY, A. S. [et al.] (eds.). Oxford: Oxford University Press. QUIRK, Randolph [et al.] (1972): A Grammar of Contemporary English. London: Longman. QUIRK, Randolph [et al.] (1985): A Comprehensive Grammar of the English Language. London: Longman. FLuL 24 (1995)