Language Choice

J. K. Buda

Although most of the world's population can speak only one language, a sizeable minority is able to communicate in two or more. Of the world's 750 million speakers of English, for example, only 300 million use it as their first or native tongue. The remaining 450 million speak it as a second or third language. Whenever speakers of two or more languages come together, a decision has to be made about which of these languages is to be used. It may be thought that the factors affecting choice of language are few and simple, but such is not the case. Often no satisfactory explanation can be given as to why speakers make the choices they do.

Sociolinguists have long been fascinated by the phenomenon of bilingualism and the complex language switching patterns that often accompany it. Many bilingual speakers are able to switch from language to language with ease, sometimes in mid-sentence. Attempts to define such patterns have not, however, met with much success. Research reports on the subject are cluttered with such obscure terms as 'diglossia', 'domain', 'code-switching' and 'ethnolinguistic vitality', but reduced to the level of a layman's understanding, the less than original conclusion would seem to be that choice of language is dictated primarily by the milieu in which the speaker finds himself.

It is only fair to say that research into this subject has been hampered by the inherent unreliability of the information-gathering methods employed, the two most common being analysis of tape-recorded speech events, and questionnaires in which subjects are asked about their linguistic behaviour patterns.

The problem with the first of these methods is that the presence of a researcher, or even the suspicion that the conversation is being recorded, is usually sufficient to affect spontaneity. Covert recording of speech events is of course possible, but severely limits the range of available speech event environments.

Questionnaires, likewise, are inherently unreliable. Subjects may not be fully conscious of their own language usage patterns, or may wish to portray them in a socially or culturally favourable light. In every language survey of India, for example, two or three thousand people (0.0004% of the population) invariably claim that Sanskrit is their first language.

Most research has focused on bilingual subjects, but true bilingualism — native-speaker fluency in both languages — is not a prerequisite for the exercise of language choice. Anyone who can speak two or more languages well enough to communicate his or her thoughts and emotions is free (if circumstances allow) to exercise choice.

Indeed it may be possible to gain a few more insights into the phenomenon of language choice by deliberately not focusing on bilingual subjects, and by widening the scope of investigation to include all polyglots — the term 'polyglot' here being used in its widest sense, meaning someone who can speak several languages, though not necessarily with the same degree of fluency.

It is with this in mind that the following comments are offered, not as definitive conclusions based on exhaustive and systematic research, but as a few tentative conjectures drawn from personal experiences and observations.

It may help to clarify the problem if we begin by suggesting that language choice is subject to two categories of factors: preferences and constraints.

A simple model of language choice might recognize the presence of only one factor in each category. It might be thought, for example, that someone with reasonable fluency in several languages would nevertheless choose to speak his mother tongue wherever possible, that being the language with which he is most familiar and comfortable. This natural tendency would be constrained by only one factor: linguistic congruity. A native speaker of Hungarian, for instance, might prefer to use that language wherever possible, but if he should himself in an environment where there are no other Hungarian speakers — in a foreign country, for example — he would be obliged to use his second or third language.

A simple model such as the above does, in fact, serve to explain a large number of language-choice events, but it is woefully inadequate in defining many others. Polyglots do not always prefer to use their native language, and the constraint of congruity does not apply to situations where the participants in a conversation share a knowledge of several languages.

A number of factual anecdotes may help to illustrate the complexity of the problem.

  1. A group of people attending a party are holding a conversation in Polish. The Scottish hostess approaches the group to ask how everyone is enjoying the party, and the conversation switches to English. Presently the hostess leaves to talk with other guests, and the group continues to talk in English for several minutes. After a short pause in the conversation one of the members of the group reverts to Polish, and the rest of the group follows suit.
  2. A foreigner visiting a Japanese department store approaches one of the shop assistants and asks for something in Japanese. The customer's Japanese is grammatically correct and well pronounced, and the assistant has no difficulty in understanding what the person has said. Nevertheless she chooses to reply in English. For several minutes the conversation continues in two languages, the customer speaking in Japanese, and the assistant persisting in a use of English.
  3. On a flight from Honolulu to Tokyo, two young Oriental ladies are discussing, in elementary accented Japanese, the friends they will meet and the shopping they will do when they get to Japan. Their conversation is interrupted when a cabin attendant comes along and asks them, in English, whether they would care for some light refreshment. Both of them answer in fluent unaccented American English, and after the cabin attendant has gone to attend to their order, they switch back to Japanese.

A simple single-element preference/constraint model cannot fully explain the language choices illustrated in these three examples.

It does not always follow that a bilingual or polyglot will prefer to speak his native language, or will feel most comfortable using it. Some polyglots prefer, for a number of reasons, to use their second or third language. The children of immigrants, for example, often feel antipathy towards the language of their parents' homeland, and prefer to use that of their host country, even though they may not yet be completely fluent in that language. As they grow older, the language of the host country usually becomes dominant, leading in some instances to a partial or even complete loss of the heritage tongue. Small children are particularly sensitive to peer rejection, and those of ethnic minorities are prone to over-emphasize their conformity with group standards of dress, behaviour, and language. In many cases, however, the period of late adolescence brings a new awareness of cultural and ethnic identity, and the mother tongue may acquire a strong symbolic value.

The sons of, for example, Sikh immigrants to the United Kingdom or Canada may, during their years in elementary or middle school, strive to suppress their ethnic and religious identity in the interests of group solidarity, but a growing awareness of discrimination may eventually lead them to turn away from a society which they feel is not prepared to accept them, and seek to redefine their identity by a return to the dress, religion, and language of their parents or grandparents. In the case of Sikh males this would entail adopting the distinctive Sikh unshorn hair style, the wearing of, among other things, a turban and steel bracelet, and the preferential use of the Punjabi language.

The children of Caribbean immigrants may exhibit the same pattern of rejection and subsequent espousal of a specific variety of English, as opposed to a discrete language. They may consciously suppress the British, Canadian, or American speech patterns they have acquired over the years — from birth in the case of those born in the host country — and attempt instead to mimic those of the Caribbean island from which their family came.

The same reaction can also be seen among the children of groups which do not experience the same degree of discrimination as Asian, African, or Caribbean minorities living in Western countries. The children of East European immigrants — usually indistinguishable from the community at large — may still feel the attraction of their heritage culture as they grow older, not because they feel rejected by the society in which they have been raised, but because of the narcissistic appeal of identifying with a disparate group.

Assuming that the two ladies in the third of our examples were second- or third-generation Japanese living in Hawaii (a not unreasonable assumption in the circumstances), it may have been this kind of cultural and ethnic awareness that led them to conduct their conversation in a language which was clearly not their strongest.

Some polyglots may feel that one or other of the languages they speak (not necessarily the dominant one) is better suited to the expression of certain ideas or emotions, and they may prefer to choose that language whenever the situation warrants. Certain languages may be perceived as 'soft', better able to express emotions and feelings, whilst other 'hard' languages may be thought more capable of expressing concepts and ideas.

Language choice may also be affected by utilitarian considerations. A speaker may feel that use of a particular language will place him in an advantageous position either within a group or within a wider social context. If his antagonists in a discussion or argument are less fluent than he is, this will clearly serve to give him a valuable edge. If the use of a specific language is perceived as socially advantageous — the use of English in India is an obvious example — this would be another factor contributing to its preferential use.

In some cases a speaker might perceive an advantage in the use of a language with which he is less than familiar — to feign ignorance, for instance, or to cut short an unwelcome dialogue.

The perceived advantage does not have to be vis--vis other individuals. It may be for purely personal considerations that a person chooses to speak a particular language. A student of a foreign language may prefer to use that language whenever possible, with the sole intention of improving his ability. This may have been the reason behind the persistent unilateral use of English by the shop assistant in the second example.

We have already mentioned the most common constraint upon the use of a specific language, that of congruity. Communication is only possible if both speakers share the same language, and there is little to be gained from addressing someone in a language which they do not understand. The only possible exception might be the early stages of direct-method language teaching.

A related constraint is that of linguistic etiquette. There is an almost universal taboo upon the use of a language which might exclude one or more members of a group from a discussion, even if the subject of that discussion has no direct relevance to that person or persons. A group of several French speakers may, for example, be discussing plans for a farewell party for one of their workmates who is about to retire. Another person, one who does not work at the same company, who does not know the gentleman in question, and who will not be invited to the party, joins the group. This newcomer, moreover, does not speak French. It is now incumbent upon the group to continue their discussion in a language which that person can understand — English or German, perhaps.

Having to change the language of the discussion to one which may be a second or third language for a majority of the members can, of course, have a stultifying effect upon the course of the discussion, making it more difficult to express thoughts and ideas. In this case, however, the exclusion constraint takes precedence over the language preference of the group majority. In extreme cases the requirement for a common language might force all of the speakers to adopt second or third languages — if the newcomer in the example given above were a native speaker of Arabic, for example.

The search for a common language may sometimes prove unsuccessful, and a group will have to choose the language which allows the participation of the greatest number of people. Before switching to that language, however, it is considered polite to apologize to the person or persons who will be excluded, and to obtain their consent.

The first of the three examples given earlier would appear to be a good illustration of this particular constraint. The members of the group at the party were all of Polish descent, and Polish was the language they chose to speak among themselves. When they became aware of the presence of their hostess, they felt obliged to continue their conversation in English, opening the way for her possible participation. The interesting variation shown in this example is the way the conversation continued in English for some time after the hostess had moved on to chat with other guests. Most of the members of the group had lived in the United Kingdom for many years, and were fluent in English. There was no linguistic pressure to switch back to Polish at the earliest possible opportunity, and perhaps some of the members may have felt that such a hasty return would have been rather indecorous.

One other constraint deserves to be mentioned. Various psychological factors sometimes preclude an easy and spontaneous switch to another language. It may be that two or more speakers have become accustomed to conversing among themselves in a particular language, and feel extremely uncomfortable having to change to another. Yet other speakers may find the act of language-switching difficult in itself. Whereas some bilinguals or polyglots can shift with consummate ease between languages, even in mid-sentence, there are those whose language choice is reactionary, a reflexive response to the language in which they are addressed, or the language environment in which they find themselves.

One interesting example of the latter constraint was furnished by an Italian consular official who once told the author that he could speak twelve languages, and a thirteenth which he refused to use. That thirteenth was Serbo-Croatian, a language he had been forced to use on summer visits to elderly maiden aunts in Yugoslavia, and which for him was forever associated with some of the most dreary and unpleasant memories of his childhood.

At the end of the Second World War, the diplomat was employed by the Italian national railway, and given the task of coordinating the repatriation of thousands of Allied soldiers. One day a Polish soldier visited his office, and the conversation was conducted in German, the language they had used on many previous occasions. After several minutes of discussion, the Polish soldier remarked that he had heard that the Italian could speak Polish, and asked him if this were true. The official confirmed that he could indeed speak Polish, and after a moment's hesitation, the conversation continued in German. Several minutes later the Polish soldier once again asked if the Italian could really speak Polish, and the answer was the same. Towards the end of the conversation the soldier repeated his question for a third time, clearly dissatisfied with he answers he was getting. When the Italian answered 'yes' for the third time, the Polish soldier asked him rather bluntly, "Well then, why don't you?" The Italian replied, "Because you're speaking German, that's why."

From what we have said so far, it should be clear that a large number of factors influence language choice, and many factors may work either with or against each other, producing a complex web of interaction which makes the task of describing any one language-choice event extremely difficult. The suggested division of factors into two categories may be of some help, but it must be recognized that some factors may exhibit the characteristics of both preferences and constraints, making it impossible to locate them satisfactorily in either of the two categories. It is difficult to determine at which point a negative preference — an intense dislike of a particular language, for example — becomes a constraint. A positive constraint, likewise, may exhibit some of the characteristics of a preference. A more useful approach might be to try and plot these factors somewhere along an axis which represents a continuous gradation between positive forces directing choice towards a specific language, and negative forces directing choice away from that language.

Finally, it needs to be recognized that language choice events do not exist in a vacuum. Language is, after all, a medium for interaction and communication between people, and the use of language will reflect the infinite complexity of human relationships.

Reprinted from Otsuma Review, No. 24, 1991