Methods and Procedures of Lexicological Analysis

    Methods and Procedures of Lexicological Analysis

It is commonly recognised that acquaintance with at least some of the currently used procedures of linguistic investigation is of considerable importance both for language learners and for prospective teachers as it gives them the possibility to observe how linguists obtain answers to certain questions and is of help in the preparation of teaching material. It also helps language learners to become good observers of how language works and this is the only lasting way to become better users of language.

The process of scientific investigation may be subdivided into several stages.

1 Observation is an early and basic, phase of all modern scientific investigation, including linguistic, and is the centre of what is called the inductive method of inquiry.

2 The next stage after observation is classification or orderly arrangement of the data obtained through observation. For example, it is observed that in English nouns the suffixal morpheme -er is added to verbal stems (speak + -er, writ(e) + -er, etc.), noun stem’s (village + -er, London + -er, etc.), and that -er also occurs in non-derived words such as mother, father, etc. Accordingly all the nouns in -er may be classified into two types — derived and simple words and the derived words may be subdivided into two groups according to their stems. It should be pointed out that at this stage the application of different methods of analysis is common practice.1

3 The following stage is usually that of generalisation, i.e. the collection of data and their orderly arrangement must eventually lead to the formulation of< a generalisation or hypothesis, rule, or law.In our case we can formulate a rule that derived nouns in -er may have either verbal or noun stems. The suffix -er in combination with adjectival or adverbial stems cannot form nouns (cf. (to) dig — digger but big — bigger).

4 One of the fundamental tests of the validity of a generalisation is whether or not the generalisation is useful in making reliable predictions. For example, proceeding from the observation and generalisation discussed above we may ‘predict’ with a considerable degree of certainty that if a new word with a suffix -er appears in modern English and the suffix is added to a verbal stem, the word is a noun denoting an active doer (cf., e.g., the new words of the type (moon-)crawler, (moon-)walker (lunar-)rouer which appeared when the Soviet moon car was launched.1 Moreover we may predict if we make use of statistical analysis that such words are more likely to be coined than the other types of nouns with the -er suffix.

5 Any linguistic generalisation is to be followed by the verifуing process. Stated, simply, the linguist is required, as are other scientists, to seek verification of the generalisations that are the result of his inquiries. Here too, various procedures of linguistic analysis are commonly applied.

The methods and procedures briefly discussed below are as follows: 1. Contrastive analysis, 2. Statistical methods of analysis. 3. Immediate Constituents analysis, 4 Distributional analysis and co-occurrence, 5. Transformational analysis, 6. Componential analysis, 7. Method of semantic differential.

All methods of linguistic analysis are traditionally subdivided into formalised and non-formalised procedures.

Contrastive Analysis

   Contrastive linguistics as a systematic branch of linguistic science is of fairly, recent date though it is not the idea which is new but rather the systematisation and the underlying principles. It is common knowledge that comparison is the basic principle in comparative philology. However the aims and methods of comparative philology differ considerably from those of contrastive linguistics. The comparativists compares languages in order to trace their philogenic relationships. The material he draws for comparison consists mainly of individual sounds, sound combinations and words, the aim is to establish family relationship. The term used to describe this field of investigation is historical linguistics or diachronic linguistics.

Comparison is also applied in typological classification and analysis. This comparison classifies languages by types rather than origins and relationships. One of the purposes of typological comparison is to arrive at language universals — those elements and processes despite their surface diversity that all language have in common.

Contrastive linguistics attempts to find out similarities and differences in both philogenically related and non-related languages.

Linguistic scholars working in the field of applied linguistics assume that the most effective teaching materials are those that are based upon a scientific description of the language to be learned carefully compared with a parallel description of the native language of the learner.1

Contrastive analysis can be carried out at three linguistic levels: phonology, grammar (morphology and syntax) and lexis (vocabulary). In what follows we shall try to give a brief survey of contrastive analysis mainly at the level of lexis.

Contrastive analysis is applied to reveal the features of sameness and difference in the lexical meaning and the semantic structure of correlated words in different languages.

It is commonly assumed by non-linguists that all languages have vocabulary systems in which the words themselves differ in sound-form but refer to reality in the same way. From this assumption it follows that for every word in the mother tongue there is an exact equivalent in the foreign language. It is a belief which is reinforced by the small bilingual dictionaries where single word translations are often offered. Language learning however cannot be just a matter of learning to substitute a new set of labels for the familiar ones of the mother tongue.

Firstly, it should be borne in mind that though objective reality exists outside human beings and irrespective of the language they speak every language classifies reality in its own way by means of vocabulary units. In English, e.g., the word foot is used to denote the extremity of the leg. In Russian there is no exact equivalent for foot. The word нога denotes the whole leg including the foot.

Classification of the real world around us provided by the vocabulary units of our mother tongue is learned and assimilated together with our first language. Because we are used to the way in which our own language structures experience we are often inclined to think of this as the only natural way of handling things whereas in fact it is highly arbitrary. One example is provided by the words watch and clock. It would seem natural for Russian speakers to have a single word to refer to all devices that tell us what time it is; yet in English they are divided into two semantic classes depending on whether or not they are customarily portable. We also find it natural that kinship terms should reflect the difference between male and female: brother or sister, father or mother, uncle or aunt, etc. yet in English we fail to make this distinction in the case of cousin (cf. the Russian — двоюродный брат, двоюродная сестра). Contrastive analysis also brings to light what can be labelled problem pairs, i.e. the words that denote two entities in one language and correspond to two different words in another language.

Compare, for example часы in Russian and clock, watch in English, художник in Russian and artist, painter in English.

Each language contains words which cannot be translated directly from this language into another. For example, favourite examples of untranslatable German words are gemütlich (something like ‘easygoing’, ‘humbly pleasant’, ‘informal’) and Schadenfreude (‘pleasure over the fact that someone else has suffered a misfortune’). Traditional examples of untranslatable English words are sophisticated and efficient.

This is not to say that the lack of word-for-word equivalents implies also the lack of what is denoted by these words. If this were true, we would have to conclude that speakers of English never indulge in Shadenfreude and that there are no sophisticated Germans or there is no efficient industry in any country outside England or the USA.

If we abandon the primitive notion of word-for-word equivalence, we can safely assume, firstly, that anything which can be said in one language can be translated more or less accurately into another, secondly, that correlated polysemantic words of different languages are as a rule not co-extensive. Polysemantic words in all languages may denote very different types of objects and yet all the meanings are considered by the native speakers to be obviously logical extensions of the basic meaning. For example, to an Englishman it is self-evident that one should be able to use the word head to denote the following:

whereas in Russian different words have to be used: голова, изголовье, сторона, головка, etc.

The very real danger for the Russian language learner here is that having learned first that head is the English word which denotes a part of the body he will assume that it can be used in all the cases where the Russian word голова is used in Russian, e.g. голова сахара (‘a loaf of sugar’), городской голова (‘mayor of the city’), он парень с головой (‘he is a bright lad’), в первую голову (‘in the first place’), погрузиться во что-л. с головой (‘to throw oneself into smth.’), etc., but will never think of using the word head in connection with ‘a bed’ or ‘a coin’. Thirdly, the meaning of any word depends to a great extent on the place it occupies in the set of semantically related words: its synonyms, the constituents of the lexical field the word belongs to, other members of the word-family which the word enters, etc.

Thus, e.g., in the English synonymic set brave, courageous, bold, fearless, audacious, valiant, valorous, doughty, undaunted, intrepid each word differs in certain component of meaning from the others, brave usually implies resolution and self-control in meeting without flinching a situation that inspires fear, courageous stresses stout-hearted-ness and firmness of temper, bold implies either a temperamental liking for danger or a willingness to court danger or to dare the unknown, etc. Comparing the corresponding Russian synonymic set храбрый, бесстрашный, смелый, мужественный, отважный, etc. we see that the Russian word смелый, e.g., may be considered as a correlated word to either brave, valiant or valorous and also that no member of the Russian synonymic set can be viewed as an exact equivalent of any single member of the English synonymic set in isolation, although all of them denote ‘having or showing fearlessness in meeting that which is dangerous, difficult, or unknown’. Different aspects of this quality are differently distributed among the words making up the synonymic set. This absence of one-to-one correspondence can be also observed if we compare the constituents of the same lexico-semantic group in different languages. Thus, for example, let us assume that an Englishman has in his vocabulary the following words for evaluating mental aptitude: apt, bright, brilliant, clever, cunning, intelligent, shrewd, sly, dull, stupid, slow, foolish, silly. Each of these words has a definite meaning for him. Therefore each word actually represents a value judgement. As the Englishman sees a display of mental aptitude, he attaches one of these words to the situation and in so doing, he attaches a value judgement. The corresponding Russian semantic field of mental aptitude is different (cf. способный, хитрый, умный, глупый, тупой, etc.), therefore the meaning of each word is slightly different too. What Russian speakers would describe as хитрый might be described by English speakers as either cunning or sly depending on how they evaluate the given situation.

The problem under discussion may be also illustrated by the analysis of the members of correlated word-families, e.g., cf. голова, головка, etc. head, heady, etc. which are differently connected with the main word of the family in each of the two languages and have different denotational and connotational components of meaning. This can be easily observed in words containing diminutive and endearing suffixes, e.g. the English word head, grandfather, girl and others do not possess the connotative component which is part of the meaning of the Russian words головка, головушка, головёнка, дедушка, дедуля, etc.

Thus on the lexical level or to be more exact on the level of the lexical meaning contrastive analysis reveals that correlated polysemantic words are not co-extensive and shows the teacher where to expect an unusual degree of learning difficulty. This analysis may also point out the effective ways of overcoming the anticipated difficulty as it shows which of the new items will require a more extended and careful presentation and practice.

Difference in the lexical meaning (or meanings) of correlated words accounts for the difference of their collocability in different languages. This is of particular importance in developing speech habits as the mastery of collocations is much more important than the knowledge of isolated words.

Thus, e.g., the English adjective new and the Russian adjective новый when taken in isolation are felt as correlated words as in a number of cases new stands for новый, e.g. новое платье — a new dress, Новый Год — New Year. In collocation with other nouns, however, the Russian adjective cannot be used in the same meaning in which the English word new is currently used. Compare, e.g., new potatoes — молодая картошка, new bread — свежий хлеб, etc.

The lack of co-extension may be observed in collocations made up by words belonging to different parts of speech, e.g. compare word-groups with the verb to fill:

to fill a lamp — заправлять лам- to fill a truck — загружать ма-

ny              шину

to fill a pipe — набивать трубку to fill a gap — заполнять пробел

As we see the verb to fill in different collocations corresponds to a number of different verbs in Russian. Conversely one Russian word may correspond to a number of English words.

For instance compare тонкая книга — a thin book тонкая ирония — subtle irony тонкая талия — slim waist

Perhaps the greatest difficulty for the Russian learners of English is the fact that not only notional words but also function words in different languages are polysemantic and not co-extensive. Quite a number of mistakes made by the Russian learners can be accounted for by the divergence in the semantic structure of function words. Compare, for example, the meanings of the Russian preposition до and its equivalents in the English language.

(Он работал) до 5 часов              till 5 o'clock

(Это было) до войны              before the war

(Он дошел) до угла              to the corner

Contrastive analysis on the level of the grammatical meaning reveals that correlated words in different languages may differ in the grammatical component of their meaning.

To take a simple instance Russians are liable to say the *news are good, *the money are on the table, *her hair are black, etc. as the words новости, деньги, волосы have the grammatical meaning of plurality in the Russian language.

Of particular interest in contrastive analysis are the compulsory grammatical categories which foreign language learners may find in the language they are studying and which are different from or nonexistent in their mother tongue. These are the meanings which the grammar of the language “forces” us to signal whether we want it or not.

One of the compulsory grammatical categories in English is the category of definiteness/indefiniteness. We know that English signals this category by means of the articles. Compare the meaning of the word man in the man is honest and man is honest.

As this category is non-existent in the Russian language it is obvious that Russian learners find it hard to use the articles properly.

Contrastive analysis brings to light the essence of what is usually described as idiomatic English, idiomatic Russian etc., i.e. the peculiar way in which every language combines and structures in lexical units various concepts to denote extra-linguistic reality.

The outstanding Russian linguist acad. L. V. Sčerba repeatedly stressed the fact that it is an error in principle if one supposes that the notional systems of any two languages are identical. Even in those areas where the two cultures overlap and where the material extralinguistic world is identical, the lexical units of the two languages are not different labels appended to identical concepts. In the overwhelming majority of cases the concepts denoted are differently organised by verbal means in the two languages. Different verbal organisation of concepts in different languages may be observed not only in the difference of the semantic structure of correlated words but also in the structural difference of word-groups commonly used to denote identical entities.

For example, a typical Russian word-group used to describe the way somebody performs an action, or the state in which a person finds himself, has the structure that may be represented by the formula adverb followed by a finite form of a verb (or a verb + an adverb), e.g. он крепко спит, он быстро /медленно/ усваивает, etc. In English we can also use structurally similar word-groups and say he smokes a lot, he learnsslowly (fast), etc. The structure of idiomatic English word-groups however is different. The formula of this word-group can be represented as an adjective + deverbal noun, e.g. he is a heavy smoker, a poor learner, e.g. “the Englishman is a slow starter but there is no stronger finisher” (Galsworthy). Another English word-group used in similar cases has the structure verb to be + adjective + the infinitive, e.g. (He) is quick to realise, (He) is slow to cool down, etc. which is practically non-existent in the Russian language. Commonly used English words of the type (he is) an early-riser, a music-lover, etc. have no counterparts in the Russian language and as a rule correspond to phrases of the type (Он) рано встает, (он) очень любит музыку, etc.1

Last but not least contrastive analysis deals with the meaning and use of situational verbal units, i.e. words, word-groups, sentences which are commonly used by native speakers in certain situations.

For instance when we answer a telephone call and hear somebody asking for a person whose name we have never heard the usual answer for the Russian speaker would be Вы ошиблись (номером), Вы не туда попали. The Englishman in identical situation is likely to say Wrong number. When somebody apologises for inadvertently pushing you or treading on your foot and says Простите\ (I beg your pardon. Excuse me.) the Russian speaker in reply to the apology would probably say — Ничего, пожалуйста, whereas the verbal reaction of an Englishman would be different — It’s all right. It does not matter. * Nothing or *please in this case cannot be viewed as words correlated with Ничего, Пожалуйста."

To sum up contrastive analysis cannot be overestimated as an indispensable stage in preparation of teaching material, in selecting lexical items to be extensively practiced and in predicting typical errors. It is also of great value for an efficient teacher who knows that to have a native like command of a foreign language, to be able to speak what we call idiomatic English, words, word-groups and whole sentences must be learned within the lexical, grammatical and situational restrictions of the English language. .

Statistical Analysis

An important and promising trend in modern linguistics which has been making

progress during the last few decades is the quantitative study of language phenomena and the application of statistical methods in linguistic analysis.

Statistical linguistics is nowadays generally recognised as one of the major branches of linguistics. Statistical inquiries have considerable importance not only because of their precision but also because of their relevance to certain problems of communication engineering and information theory.

Probably one of the most important things for modern linguistics was the realisation of the fact that non-formalised statements are as a matter of fact unverifiable, whereas any scientific method of cognition presupposes verification of the data obtained. The value of statistical methods as a means of verification is beyond dispute.

Though statistical linguistics has a wide field of application here we shall discuss mainly the statistical approach to vocabulary.

Statistical approach proved essential in the selection of vocabulary items of a foreign language for teaching purposes.

It is common knowledge that very few people know more than 10% of the words of their mother tongue. It follows that if we do not wish to waste time on committing to memory vocabulary items which are never likely to be useful to the learner, we have to select only lexical units that are commonly used by native speakers. Out of about 500,000 words listed in the OED the “passive” vocabulary of an educated Englishman comprises no more than 30,000 words and of these 4,000 — 5,000 are presumed to be amply sufficient for the daily needs of an average member of the English speech community. Thus it is evident that the problem of selection of teaching vocabulary is of vital importance.1 It is also evident that by far the most reliable single criterion is that of frequency as presumably the most useful items are those that occur most frequently in our language use.

As far back as 1927, recognising the need for information on word frequency for sound teaching materials, Ed. L. Thorndike brought out a list of the 10,000 words occurring most frequently in a corpus of five million running words from forty-one different sources. In 1944 the extension was brought to 30,000 words.2

Statistical techniques have been successfully applied in the analysis of various linguistic phenomena: different structural types of words, affixes, the vocabularies of great writers and poets and even in the study of some problems of historical lexicology.

Statistical regularities however can be observed only if the phenomena under analysis are sufficiently numerous and their occurrence very frequent. Thus the first requirement of any statistic investigation is the evaluation of the size of the sample necessary for the analysis.

To illustrate this statement we may consider the frequency of word occurrences.

It is common knowledge that a comparatively small group of words makes up the bulk of any text.3 It was found that approximately 1,300 — 1,500 most frequent words make up 85% of all words occurring in the text. If, however, we analyse a sample of 60 words it is hard to predict the number of occurrences of most frequent words. As the sample is so small it may contain comparatively very few or very many of such words. The size of the sample sufficient for the reliable information as to the frequency of the items under analysis is determined by mathematical statistics by means of certain formulas.