Riassunto esame di Linguistica Inglese, prof. Maggioni, testo consigliato: Readings in English Lexicogrammar

Readings in English Lexicogrammar

A cura di Margherita Ulrych

Lexis and lexicography

1.0 Introduction

On EFL (English as a Foreign Language) and ESL (English as a Second Language) lexicography is a good example of a domain in which linguistic insights can be directly applied and practical advantages are readily recognized. But we should not forget that lexicographic practice can also be of service to refinements in linguistic description.

1.1 The image of the dictionary: User and use

Bilingual dictionaries are generally employed in the initial stages of learning a language. As proficiency develops, greater use is made of monolingual dictionaries; prolonged dependency on bilingual dictionaries probably tends to retard the development of second language proficiency.

The dictionary is a trusted and respected repository of facts about a language. A dictionary has institutional authority and is also a tool of learning.

Two dictionaries in particular, the Oxford Advanced Learner's Dictionary (OALD) and the Longman Dictionary of Contemporary English (LDOCE), have contributed to the development and design of dictionaries for non-native learners of English, highlighting the differences which exist between bilingual and monolingual dictionaries. Bilingual dictionaries are more generally employed in the initial stages of learning a language; as proficiency develops, greater use is made of a monolingual dictionary. Prolonged dependency on bilingual dictionaries probably tends to retard the development of second-language proficiency. Monolingual dictionaries come in many more varied shapes and sizes ('pocket dictionaries', dictionaries for children, specific purpose/technical dictionaries, etc.)

We will here examine the specific properties of those monolingual dictionaries which are designed for the EFL/ESL user.

1.2 Dictionary definitions

Major distinctions between monolingual dictionaries for native and for non-native speakers lie in the kinds of information supplied. For the non-native speaker a main aim is to supply encoding information which will allow for productive use of the language.

Ex: The Concise Oxford Dictionary (COD) gives one piece of grammatical information about the noun work (i.e. 'n' for noun), whereas the OALD marks it as an uncountable noun (i.e. 'u' for uncountable) thus preventing the production by learners of mistakes. Learner's monolingual dictionaries also often provide more detailed guidance both on matters of syntax (like the coding of verb patterns), on pronunciation, and on cultural and stylistic restrictions.

Another marked distinction is in the different definitions of words, using restricted defining vocabulary to try to write clear and unambiguous definitions. In this case the definitions are always written using more simple terms than the words they describe. Defining on restricted vocabularies are not without problems, anyway:

• There is no guarantee that they will be known by the learner (since not all the words are particularly common).
• Defining vocabularies can achieve simplicity at the expense of accuracy. On the other hand, when more words are needed, the result can sometimes be unnaturally circumlocutory.
• Defining vocabularies work rather better in the explanation of concrete rather than abstract terms.

Searching examinations have, however, revealed that advantages clearly outweigh disadvantages. It is important not to lose sight in what learners use dictionaries for: students tend to use EFL dictionaries rather as they would general monolingual dictionaries, for looking for meaning and synonyms, for checking spellings, and for decoding activities in the written medium such as translation and reading.

1.3 Grammar and the dictionary

In the design of a monolingual EFL dictionary, the nature of the syntactic information supplied by the lexicographer will be crucial. In the case of quite a common verb such as enter, learners should require detailed information about the syntactic patterns formed by the verb and its different but inseparable prepositions:

• Enter sb. as a member
• Enter sb. at public school
• Enter sth. in a diary

Additionally, learners would need to know that the passive form cannot be used with enter [room] but is permissible with other collocations (ex: the details were entered in the book); also that enter [society, room] can take both indefinite and definite articles before the noun but that enter [army, church] is more usual with the definite article.

In the face of such facts, it is clearly inadequate for a single example to be supplied without any indication of the morphological, syntactic, and stylistic patterns the phrase can generate.

For syntactic patterns, learner dictionaries such as the LDCE and the OALD have adopted coding systems. These enable learners to check, for example, object-complement relations, transitive/intransitive patterns, whether a noun is countable or uncountable, etc.

Ex: the verb determine in the OALD explicitly tells the learner how the word in each of its senses takes a particular type or types of clause pattern.

With the third definition (determine = to decide firmly, to make up one's mind) are listed, among others, grammatical codes 7A and 10, which supported by illustrative sentences, stand for the following construction types:

• VP7A Subject + vt (not) + infinitive, etc. He determined to learn Greek.
• VP10 Subject + vt dependent clause/question Have they determined where the new school will be built?

But with the fifth definition (determine = to be the fact that __s) the sentence pattern is VP6A, which means:

• VP6A Subject + vt noun/pronoun Do heredity and environment determine a man’s character?

The coding system here derives from research by Hornby in the 1940s. One of the main problems with such codings is that learners have to invest a considerable amount of time and effort in mastering a system which is in itself very complex and requires constant reference to another part of the dictionary before any clear return is shown.

More acute problems arise in the appropriate provision of entries for phrasal verbs and related idiomatic expressions. For example, the need to differentiate semantically between get sth. over, get over sth./sb., get smb. over sth. Examples need to be juxtaposed for the different meanings to be derived and for positional variation and restriction in the use of the preposition to be recorded.

She can't get her ideas over to her students. She can get over the loss of her mother. She tried to get her sister over the disappointment.

Capturing the syntactic behaviour of items involves more than positional variations and patterns. It also involves complex issues of compounds and derivatives. For example, should a composition like time-wasting be listed under the entry for time or for waste? It is clear that entering the item under the verb form emphasizes grammatical relationships, but the inexperienced dictionary user, who will probably rely on alphabetical ordering, may be easily deterred by necessary inconsistencies. Hyphenization, stress, and spelling can also be determining factors.

In the case of derivatives, decisions over points of entry can involve sharper semantic differentiation as well as issues of spelling. It is reasonable that encouragement should be listed under the entry for its root encourage rather than as a separate entry. But such is not the case with an item such as high from which the following items may be said to derive: high class; highly strung; height; etc. Should there be one entry or four separate ones? Should differences in spelling in a derived form merit separate treatment or might this too readily assume that the learner knows the different spellings involved (ex satisfy/satisfaction)?

There are also problems where the derivations reflect closely related but distinct differences of meaning.

Ex: The verb adhere has derivatives such as adherence, and adherent, but also adhesion and adhesive, which relate to different senses of the verb.

He adheres to another political party. He is an adherent of another political party. His adherence is to another political party.

These are all transformationally related (adhere + political + party) and the derivatives they contain should all appear in the same entry. Adhere in the sense of 'stick', adhesion and adhesive represent a quite distinct line of derivation they should be treated in a separate entry.

The problem with such a decision is that the retrieval of individual items may become too taxing for the learner. But the alternative strategy of grouping complex forms together because of spelling similarities can be counter-productive: it is often the simplest and most common words which contract the most complex syntactic and collocational partnership.

1.4 Fixed expressions and the dictionary

Syntactic patterns do generally exhibit stability and have been codified extensively. More complex is the question about lexical acceptability. Clines exist between fixed and stable patterns and patterns which are more indeterminate and negotiable. An appropriate sequence from fixed to less fixed was suggested with reference to clines of collocational restriction, syntactic structure, and semantic opacity.

• Collocational restriction: from unrestricted collocation (ex keep: 'keep a house, a diary, a pet, a boat, a job', etc.) to relatively restricted collocation (ex stark naked [nudo come un verme], gin and tonic, etc.)
• Syntactic structure: from flexible (ex break somebody's heart, heart-breaking, heart-breaker) to irregular (ex the more, the merrier; to hold true [restare valido])
• Semantic opacity: from transparent (ex long time, no see; honesty is the best policy) to opaque (ex to be over the moon, to smell a rat).

Such clines illustrate the importance of viewing the lexicon of a language as a repository of potential for open and creative exploitation but also as a source of stereotyped, non-transacted communication.

The Oxford Dictionary of Current Idiomatic English has the main aim to enable the learner to understand usage and put this understanding to productive use. In ODCIE the main strategy is to present information which is selected as representative of available choices.

• “Open” collocation: The entry for break down, which in one of its non-metaphorical senses collocates quite 'openly' with a range of objects grouped semantically in a category of material barrier or obstacle: Break down . . . O: wall, door, fence. From this the learner might quite suitably derive gate, barrier as appropriate collocates in the position of grammatical object (O).
• “Semi-restricted” collocation: Formed from different semantic sub-classes, the convention is to use semi-colons to separate off the different senses. Shoot up . . . rise, increase sharply . . . S: price, cost, rent; temperature, pressure. Here (S) stands for grammatical subject.
• Restricted collocation: Learners have to be warned about appropriate collocation boundaries. ODCIE uses a warning sig '!' Blow up . . . make bigger, enlarge. O: ! negative, photography, picture.
• Idioms (which allow some slight modification): An oblique stroke can be employed to reinforce the nature of the restrictions: to not have the slightest/least/faintest/foggiest idea to be in someone's good graces/books

But considerable problems remain.

• The students require information concerning the relative frequencies and currency of particular patterns;
• Style levels are notoriously variable in the area of conventionalized language. The differences in formality levels, for example, between hit the road, hit the target or hit the bottle need to be specified;
• It will be clear that the greatest problems arise at the points in the cline where patterns and collocation are not fixed, but rather 'familiar' or 'semi-restricted'.

Like the complex collocability relations contracted by ostensibly synonymic items such as small and little; and the freer the possible collocations of such synonyms, the more complex the learning becomes.

There are groups of idiom-prone items such as go, give, break, hit, take, come which have very extensive but not completely open collocation.

• Break the ice
• Break a rule
• Break a leg
• Break a cup

The question is whether numerous separate sub-entries are required or whether the basic sense of break should be explained with some clear indication that users can generate a wide range of possible meanings according to context.

Lexicographers need to resolve how far the complete range of fixed expressions is to be represented. The growth of interest in the implications to language teaching conventionalized language has led to include more fixed expressions in language course books and related materials. Among the problems here might be the ephemeral nature of allusions and catch-phrases; the domain restrictedness of certain stylistic formulae; and the fact that explanations of 'stereotypes' and 'social formulae' would need to be sufficiently detailed to allow appropriate use in the right context, but not so detailed that they become descriptions of the contexts themselves.

The emphasis on problems may in itself be dangerous since it concedes to idiomaticity and fixed expressions a problematic status and thus ignores arguments concerning the naturalness of such 'universal' relations in language.

1.5 Monosemy, polysemy and dictionary entries

The fundamental questions raised, in the problem of monosemous and polysemous words, are whether all the words treated as polysemous in dictionaries are actually polysemous and whether in the case of genuine blurring between senses lexicographers have sufficiently developed techniques for the representation of such meanings. The existence of clear-cut instances alongside indeterminate cases necessitates the construction of a cline of relatedness of meaning from monosemous to polysemous words.

In the paper 'Monosemous words and the dictionary' Moon argues that the problematic instances are words which are quasi-monosemous. Taking as an example the adjective light, Moon cites then possible contexts or collocate groups all of which arguably generate different senses of the word (1. not very great in amount, degree or intensity – a light rain was falling; 5. not strong or deep in sound – her light voice; 9. easy and not onerous – light housework; etc.). The larger the dictionary the more likely it is that such a word may be allowed to take as many meanings as the imagination of the lexicographer can produce.

Light is clearly not monosemous but it could be said to have only two main strands of meaning: (1) not great in intensity, (2) to only small degree.

Common verbs such as take, give, come, go, break and hit are monsemic but are judged as polysemic by dictionaries and linguists because their essential general meanings are confused with contextual, inferential meanings. The dilemma of the lexicographer in such cases is that words do not exist in isolation, but in variable contexts and the dictionary user (especially the foreign learner) needs to have appropriate citation. The nature of this dilemma is explored further in a paper by Stock. Stock's aim is to explore whether existing lexicographic conventions adequately represent polysemous words. Different collocational and colligational patterns entered by the node word (aka the word under analysis) can aid differentiation.

Ex: The different senses of post can be distinguished according to whether it is count/non-count (There's some post for the headmaster today = letters; There are ten posts a day in central London = collection or delivery of post). Similarly it can be differentiates from post ('wooden stake') or ('station, outpost') as a result of its collocational environments: for example, He is soon to take up an exciting new post in Malaysia.

Stock points out, however, that not all word senses can be divided up in this way. For example, the word culture, whose senses are not so divisible because the word is frequently used in a vague way with a sometimes convenient slippage for the user between its various senses (ex It's a case of cultural shock; She is a woman of great culture; Different cultures have to learn to co-exist, etc.).

The problem for the lexicographer is that the senses are sufficiently related to discount separate numbered entries in a dictionary, but sufficiently diverse to demand some kind of categorization. Stock proposes that a subdivided numbering system would enable the word to be split into its main senses (1, 2, 3, etc.) with subsenses entered as 1a, 1b, 1c, etc.

The analysis of structural relations between these differentiable senses of a lexical item can be of value for describing words in dictionaries and especially in lexicons where semantic groupings form the organizational basis. For example, the separate senses of the verb rise can be differentiated simply by cross-reference to relations of antinomy:

• Rise-set (sun/moon)
• Rise-fall (temperature)
• Rise-sink (cakes, etc.)

(However, the semantic organization of words in lexicons do not always separate out senses in this way).

Other widely discussed issues in lexicography include the use of examples in the definition of lexical terms. The lexicographer needs to decide how far to define a word's meaning and thus not compromise on accuracy or sacrifice such precision in the service of explaining clearly to the non-specialist user. In the case of presentational ordering the difficult decision for the lexicographer is to resolve – especially with polysemous words – which sense of the words to introduce first so that it serves to clarify the definition of subsequent items. With a word such as literature, for example, should the first reference be to it in its most frequent sense of creative writing, or should it be the less frequent sense of the word as written material (ex Have you got any literature on holidays in Spain?) to appear as the leading category?

1.6 Corpora, computers and lexicography

The most significant developments in lexicography in the past two decades have involved more extensive corpora of spoken and written language.