Language contact studies
The comparative school of linguistics, which developed after 1786 in the wake of the discovery of Indo-European, views the historical development of language as a gradual process of change, of language divergence over time. The speakers of a language spread out and become isolated from one another; dialects develop, then gradually evolve into separate languages. The metaphor of linguistic families consisting of mother and daughter languages works well for describing this type of evolutionary process.
But language history cannot be completely understood without also studying the phenomenon of language contact and mixing. This second aspect of the history of language has traditionally been studied as a sort of sociolinguistic footnote in historical linguistics (for instance, the need to exclude borrowed words from historical-comparative studies). However, as we shall see, language contact often plays a very important role in the history of languages. The history of a language involves gradual evolution as well as sudden changes brought about by various types of language contact. The metaphor of mother and daughter languages is inadequate for describing the effects of languages interacting and mixing. Many problems in historical linguistics have not been solved because the effects of language mixing have not been given proper attention. Let's look at some of these possible effects.
The languages of peoples isolated from contact with other languages--such as an island community speaking a single language (Hawaii before Captain Cook, for instance, or Iceland) tend to change very slowly. The primary types of language changes in such a situation are indeed internal spontaneous changes and gradual invention of vocabulary. It would be possible to study the historical development of Icelandic or Polynesian languages using the traditional model of comparative linguists with little attention to language contact or mixing.
Such instances of language isolation, however, are fairly rare. Even in aboriginal Australia and the Americas--areas long isolated from the main centers of human contact and mixing, tribes living in close proximity often interacted in ways that produced rapid language change.
How can significant language contact between two or more adjacent groups affect the languages involved? There are several possible outcomes.
1) Occasionally, language contact between different speech communities leads to linguistic divergence. Language contact may precipitate a deliberate attempt speakers to maintain or even increase linguistic differences. In Australia and New Guinea cases have been found where neighboring tribes have deliberately attempted to alter their languages to make them even more different from the speech of neighboring tribes. This phenomenon seems especially prevalent in societies where the very form of language is thought to have magical properties. In parts of Australia and New Guinea, sounds seem to have been deliberately dropped only because they happened to be shared by a neighboring tribe. (This is what would happen, for example, if the French were to drop their front rounded vowels just because the Germans also had these sounds.) Also, many tribes have devised special ritualistic forms of languages that use rare sounds such as ingressives or clicks; the purpose is to make the ritual language as different as possible from other forms of speech. On a smaller scale, the same desire to be different causes every generation of teenagers and other groups of English speakers to adopt special slang words and ways of speaking as a badge of their independence, of their in-groupness.
The same kind of impetus toward deliberate divergence can also be seen in belated attempts by the speakers of one language to "purify" their language by purging the effects of past borrowing. A mild example of intentional language divergence can be seen in the efforts by the French Academ of Language to expel English words from their language to preserve the "purity," of French, or recent in Kazakh efforts to replace Russianisms.
Thus, language contact may trigger a desire to be different and may motivate deliberate changes in native speech patterns (as well as belated attempts to purge the effects of past borrowing). It is unknown how great a role such deliberate linguistic tinkering has played in the creation of the world's present linguistic diversity. Perhaps intentional divergence played a significant role in the dim past of our hunter-gatherer ancestors, when groups were small and cohesive enough for this type of linguistic engineering to be successful.
2) A far more commonly encountered effect of language contact, at least in the period of recorded history, is one of language convergence--an increase in shared features between neighboring languages due to various processes of language mixing. The social contact required to produce language mixing exists in various forms, each of which has its own particular effects. Below I'll discuss five types of language convergence.
a) The first situation involves contact between two language communities who maintain a high degree social separation. Both language communities retain their native language and only borrow select items. A large number of words may be borrowed, particularly if one group acts as the source of new items of culture or environment for the other. Precisely this type of language contact accounts for the large number of Arabic loan words borrowed together with the new religion of Islam into such unrelated languages as Turkish, Farsi, and Malay.
In instances of selective borrowing, new words tend to be phonetically reshaped to fit the phonological pattern of the language recipient (for instance, when importing Islamic religious terms, such languages as Malay and Turkish dropped the Arabic pharyngeals). Occasionally, however, when large numbers of new vocabulary items are borrowed, new sounds or new sound combinations may arise in the recipient language. (Cf. [f] borrowed from Greek by the Russians, or, more recently the combinations [shm], [shl] borrowed into English from Yiddish). New word building techniques may even develop (such as the technique of compounding semantically opposing nouns such as firewater borrowed by English on the basis of Native American polysynthesis.) Usually, however, the impact of borrowing between two independent language communities is limited to new vocabulary and rarely has much influence on the grammatical structure of a language.
b) A high degree of bilingualism between speakers of adjacent languages over a prolonged period of time seems to facilitate the gradual process of language convergence that affects all levels of language, especially the phonology. This is the case in the Balkans with various distantly related Indo-European languages (Slavic, Albanian and Greek), which all share certain features of grammar, such as the evidential verb forms--acquired through centuries of contact with the Turks. Such an area of gradual linguistic convergence between non-related or perhaps distantly related languages can produce what is know as a Sprachbund (a group of unrelated languages with shared structural features). There are Sprachbunds in the Caucasus mountains (great numbers of consonant clusters; preponderance of glottalized sounds) along the northern Pacific coast (many labialized and glottalized sounds, lack of nasals).
c) A third contact-based situation that causes language convergence requires significant intermarriage or other types of family-level intermixing between speech communities. A large number of the members of one linguistic group may be incorporated into the social structure of another group in various ways: either as spouses, as servants and slaves as a result of war, or as refugees adopted as equal members of the tribe. Intermarriage frequently occurs between different speech communities, sometimes voluntarily, sometimes involuntarily on the part of the incoming spouse. When such mixing occurs on a large or prolonged scale, the effect on the language of the incorporating community often goes deeper than the mere borrowing of words. This is especially the case when large groups of women are incorporated into the community from outside--be it as wives, or as servants or slaves charged with helping raise children. If the newcomers are beyond the age of adolescence, they will necessarily learn the new language with imperfections caused by interference of their native tongue. If there are a large enough number of these non-native- speaking women, they will affect their children's speech--bringing about significant changes in a language. Children can be strongly affected when raised primarily by non-native speakers or by speakers with a particular dialectal peculiarity. This occurred on a modest scale in the American South during the pre-Civil War era, when Black English affected the pronunciation of southern upper class white children. It also occurred in the formation of the backwoods southern American dialects, which were undoubtedly influenced by the speech patterns of Cherokee women who married settlers. And it occurred to an even greater extent in the development of Dutch into Africans in the Boer Republics of present-day South Africa, when the speech of the local Bantu tribes as well as imported East Indian laborers led to radical changes in the way children spoke the Dutch of the original immigrants. Finally, it occurred to an even greater extent during the past two thousand years as the Bantu peoples displaced the native Khoisan speakers throughout southern Africa. The Bantu languages of southern Africa today contain clicks and other features of pronunciation not found in the Bantu languages of Central and north-eastern Africa. These sounds were definitely brought in by the conquered aborigines.
d) Besides causing language convergence of the types just enumerated, language contact can also result in an entire group abruptly abandoning its native speech and adopting a new language. This phenomenon is known as language shift. Language shift happens every time an immigrant learns the native language of the new country and passes it down to children in place of the old country language. If people undergo language shift on the level of individuals or isolated small groups, then the effects of language shift on the speech community as a whole will be negligible. Generally, the linguistic peculiarities of a single person or a small group of single persons will not be passed down to future generations. If, however, a whole group of people undergo language transfer en masse, perhaps because the language they are adopting is considered prestigious, then the features caused by the interference of their original, native language may be passed down to the next generation, creating a permanent ethnic dialect or even a new language.
There are two main reasons why the children of groups that undergo abrupt language shift might not dissolve homogeneously into the surrounding linguistic landscape:
1) They may be segregated--either voluntarily, as in the case of the Amish of Pennsylvania, or involuntarily as in the case of African Americans.
2) They may be more numerous than the original native speakers, and thus have the determining effect on the language of the next generation. A sudden language shift in an entire population--with the accidental language mixing that always occurs in such a situation-- can create a new, mixed language--a creole-- within the space of a single generation. This phenomenon is called abrupt creolization. Such is the case in the development of many European based creoles in tropical areas where the European population was constantly decimated by local diseases against which the natives had more resistance. Abrupt creolization is also thought to have taken place in India and southern China, where northern peoples imposed their language on peoples living in more tropical areas. The locals learned the conqueror's language, disease then left these new speakers in the majority, and their creolized version of the language became the language of the next generation.
e) The last and most radical form of language change due to language mixing involves a situation when several linguistically diverse groups are confronted with the need to communicate.
Today, for political reasons, every country has a national, or official language. In countries where many languages are spoken, one often is chosen as a common language. A language used for communication between peoples who speak different languages is called a lingua franca. English is a lingua franca in India and parts of Africa; Russian is a lingua franca in most of the former USSR; Mandarin Chinese, or Putonghua, serves the same purpose in the People's Republic of China; as does French in many parts of West Africa. In Somalia, 97% of pop. speak various dialects of Somali, but the official language is English. Swahili is a lingua franca spoken by 40 million people throughout east Africa, mostly as a second language. Sometimes the lingua franca undergoes simplification or modification among the many groups who speak it. A slightly simplified version of a lingua franca is known as a koine, from the Greek word for common. The original Koine was a modified Greek dialect spoken throughout the Mediterranean from around 4th cent BC till the 1400's. Swahili is a Bantu-based koine which developed so that Africans could communicate with Arab traders.
Sometimes, however, there is no time or opportunity for diverse groups to learn a second language--even in a modified form. For reasons of communicative expediency, therefore, people invent what is called a pidgin. A pidgin is a makeshift mixture of several different languages invented by and used by adults as a second language. Pidgin languages develop when several different linguistic groups are suddenly confronted with the need to communicate with one another, usually for economic purposes of trade (the word pidgin is supposedly a Southeast Asian corruption of the English word business). If there ever was a tower of Babel, the polylingual workers would have had to come up with a pidgin.
What kind of structure might such a pidgin have? The pidgin may be a random combination of the native languages of various speakers. Or one of the contributing languages may act as the base to which parts of the other languages are grafted; thus, there are English-based pidgins in parts of Africa and Asia, French-based pidgins in the Caribbean and West Africa, etc. Usually, the culturally dominant language of the region will supply the bulk of the vocabulary; the grammar is more of a mixture of all the contributing languages. The original lingua franca, for instance, a term which literally means Frankish tongue, was the pidgin language Sabir, a Latin based pidgin used by Europeans in the middle ages to trade with the Arab world. Two pidgin languages are used today in New Guinea as lingua francas to aid in communications between the islands several hundred different ethnic groups: these are Hiri Motu, Tok Pisin, the latter of which literally means talk pidgin. Notable pidgins in American history are Chinook jargon, Mobilian in the US southeast, and the Delaware trading language of the Colonial era.
In the creation of a pidgin, linguists speak of the process of pidginization, which means a radical structural simplification of language form. Structurally, a pidgin is not a complete language: it is language stripped of everything except the bare essentials, with vocabulary usually limited to practical or economic terms. Pidgins also show great simplification of the syntactic and phonological patterns found in the contributing languages.
Also, a true pidgin is not the native language of anybody and in this sense it is the closest thing we have to a primitive language.
Because of historical circumstances, a pidgin can come to be learned as the native language of the next generation. In such cases the pidgin quickly expands its vocabulary and becomes a full-fledged language, called a creolized pidgin, or simply creole. (The word creole derives from Portuguese 'crioulo', which meant a European born in a colonial territory. Creole later came to mean a person of racial mixture and later still came to mean a mixed language. The process whereby children acquire a pidgin as a native tongue, then expand it into a full-fledged language is known as creolization.
Thus there are two types of creoles: one occurs when children acquire a pidgin as their native language and expand it into a full-fledged language (pidgin-based creole); the other when a group abruptly adopts the language of another group and passes it on to their children, flaws and all, to form a new dialect of that language (language shift, or abrupt, creole). Language shift creoles are mixtures of two languages, whereas pidgin-based creoles are mixtures of several languages, one of which often acts as the basic source. Examples: English for Hawaiian pidgin, French in Haiti.
What happens to a creole language? It may develop into a separate language, as happened in India with the descendants of Sanskrit. Or it may converge with one of the parent languages to and become a minor dialect of that language. The latter process is known as decreolization. Colonial Black English was originally a pidgin based on southern English and several West African languages. It became a creole, the native language of the next generation, and gradually converged with southern English. One black dialect that has resisted decreolization more than general Black English is Gullah (Florida to S. Carolina coast, from West African Ngola tribe). Today there are over 200 creoles and pidgins (6 million people speak Creoles and a few million use pidgins on a daily basis).
As a result of various forms of language mixing--be it word borrowing between separate groups, the gradual influence of long-term bilingualism, the induction of foreign speakers into the group, abrupt language shift by an entire group, or a pidgin language being adopted as a native language--adjacent languages that are historically unrelated can come to resemble one another in striking ways. Historical linguistics cannot ignore language mixing in setting up the family trees of languages. The study of creoles--whether made from pidgins or developed out of language shift--could contribute greatly to comparative linguistics. The proto-language could itself be a mixture of other languages (an example might be ancient Japanese.) Without recognizing language mixing, it is impossible to trace the ancestry of such languages.
So far we have been talking about the creation of new languages and dialects. The opposite process--the extinction of a language--which is known as language death, has been far more common in the past few hundred years. Since 1400 the linguistic map of the world has changed radically due to massive language death. European languages have nearly completely replaced the aboriginal tongues of the Americas, Oceana and Australia, as well as on isolated islands like Tasmania and the Canary islands. Thus the effects of language death have outstripped the creation of new dialects and languages in the recent past. And the process of extinction of aboriginal languages continues and remnant groups lose their language.
But even dead languages have a way of living on. A language may mix with other languages before dying out. The features left behind in the surviving language represent what is known as substrate influence. The Russian language shows a substrate influence from the languages of the now extinct Finnic tribes that once inhabited north eastern Europe. The Germanic languages show some kind of aboriginal Baltic substrate influence. Celtic shows all sorts of interesting substrate features, including a VSO pattern which seem to be borrowed from languages related to Basque. Substrate influences such as these have been studied inadequately by historical linguists. No one has ever tried to use substrate features to reconstruct a dead language that has no living descendants. Yet it might be possible to do this to some degree and thus extend our knowledge of language distribution farther back in time.