A Cross-linguistic Perspective, A. Y. Aikhenvald

Grammars in Contact A Cross-Linguistic Perspective

Alexandra Y. Aikhenvald

1. Why can languages be similar?

Languages can be similar in many ways; even if not counting such universal facts like in every language there are similar purposes for saying things (commands, questions etc), languages may be similar either due to genetic inheritance (historical linguistic makes sure if two languages had been ‘relatives’) or due to areal contact (comparative linguistics). Alternatively changes may just occur due to geographic proximity. Every language in a history has undergone a certain amount of influence from their neighbours. It is just necessary to remember that ‘borrowing’ as such is not one-sided, it’s a very broad process and so the term ‘borrowing’ should be used in a broad sense, firstly, the borrowing itself can be unilateral (from one source) and multilateral (from many sources) and the result may be very different—many things can be borrowed but not in the same way.

2. How languages affect each other: the effects of language contacts

If one language is significantly different from its proven genetic relatives, language contact is the ‘usual suspect’. Cantonese has features not found in most Sinitic languages—these can be explained only by nongeneric influence. Yet examples aside, language contact results may be very different—phonetic, phonological, morphological, syntactic, and especially pragmatic. Plus as any language needs to assimilate the ‘borrowings’, it’s hard to understand later when something was borrowed, to what extend, and within which frames. The result is layered languages: the inherited ‘core’ is underneath ‘layers’ of innovative influence from outside. Following scheme shows on which level it is possible to recognise borrowing and which levels reveal genetic structures:

more similar to genetic relatives

Inflectional (or core) morphology (form/function) Core lexicon Syntactic construction types Discourse structure Structure of idioms more similar to neighbouring languages

Yet extreme ‘layering’ of languages, with features and forms diffusing back and forth over thousands of years, results in obscuring the erstwhile genetic relationships and making it impossible to ‘peel off’ the actual layers. In this case linguists should not be afraid of saying that ‘they don’t know and are never likely to know’ whether certain similarities are due to genetic origin or due to geographical diffusion. Only some classification is possible: 



Intertwined languages—a result of a combination of special sociolinguistic circumstances with semiconscious efforts to ‘create a language’, in which different parts of grammar and lexicon come from different languages. Media Lengua has a Spanish vocabulary and a Quechua grammatical system; Gypsies in Britain speak a language with Romani lexicon and English grammar ( such languages are also known as ‘mixed’ ‘layered’ are languages in which there are just borrowings present just as discussed above, most of the languages are layered

*plus there may be ‘conscious engineering’—changes are brought about by language reforms within a language—Modern Hebrew, Estonian Geographic distance plays role—a language may be influenced partly by the other language—that is there is an influence in one region, there’s no such influence in another.

3. Contact-induced change and its mechanisms

This section could be started with the question ‘which grammatical features can be borrowed?’ Languages borrow forms and patterns. Borrowed forms can include a lexeme, a pronoun, an affix, a phoneme, or intonation pattern. Phonological features for example include nasalization. Borrowings in semantic patterns mean that something is changed regarding lexis—For example, in Nigerian Arabic ‘head of house’ means ‘roof’—just as in the other non-Arabic languages spoken in the area—while speaker of any other Arabic variety would understand the expression ‘head of the house’ as ‘head of the household’. Another good example regarding lexical borrowing would such a modern word as a ‘skyscraper’—it’s just the same in English, French, Portuguese, and Russian. Another question: ‘What changes in language contact?’ Well, diffusion may involve contact-induced gain, or loss, of a form, or of a pattern. The original and the diffused form, or pattern, can coexist in the language, with— or without—some functional differentiation. Or a hybrid form may be created.  

Borrowing of a grammatical system—Ayacucho, Quechua, and Tagalog have adopted the subsystem of Spanish gender marking and agreement. Adding a term to an existing system—some forms already existing in the language may become stronger—Basque and Israeli Hebrew are becoming more analytic under Indo-European influence (analytic tendency had been there before). Further system-altering changes involve case-marking patterns—Estonian is thought to have developed prepositions (which now coexist with postpositions) under the influence of Indo-European languages. Contact-induced changes may result in creating a new somewhat marginal subsystem within a language; unassimilated loans are likely to produce‘loan phonology’—it goes according to the rules of the model language and we understand it: Russian джунгли from the English ‘jungle’ [dž]-sound and Latin endings in English cactus→cacti. When the words are totally assimilated though speakers don’t feel the sounds that once had been alien as ‘strange’; Russian ‘ф’-sound and English word ‘very’ which had once been French.

From here on it could be discussed ‘How foreign forms and patterns make their way into a language?’ Once borrowed, a form or a pattern is likely to diverge from what it was in the source language, in terms of its formal adaption, and also its semantics and function. 

Enhancement of an already existing feature. If languages in contact share a category or a construction, language contact may increase its frequency or its productivity. Pre-existing analytic tendencies in Basque and Israeli became more pronounced under the influence of Indo-European languages.  Extension by Analogy. An existing structure can develop additional meanings, matching the ones in a contact language: this is why pronominal plural in Basque was extended to all pronouns, to match the Spanish pattern.  Reinterpretation and reanalysis. In Acadian French the English loan ‘back’ takes on the role of the French ‘re’ with such verbs as revenir ‘to come back’  Areally included grammaticalisation. This is a process whereby a lexical item is grammaticalised to express a category or a meaning in a target language. The basic paths are: a. The target language follows the same grammaticalisation path as the influencing language. For example the verbs ‘go’ and ‘carry’ in Basque developed the same aspectual meanings as corresponding verbs in Spanish (voy a hacer) b. Alternatively lexical item can be grammaticalised to create an ‘equivalent’ o Grammatical accommodation. This is, a native morpheme can be reinterpreted on the model of the syntactic function of a phonetically similar morpheme in the source language. The marker of possession –pal in Pipil, a Uto-Aztecan language was originally a noun, as in nu-pal ‘mine’, mupal ‘yours’, and so on. Now to achieve the Spanish ‘para’ o Loan Translations. These involve mostly adhoc word-for-word or morpheme-per-morpheme translations from one language into another—Estonian läbi hammustama from German durch beißen; välja kannatma from aushalten 4. Making diffusion possible

We cannot predict with full assurance which way a language will change. Nor can we postulate universal ‘constraints’ on language change. We can only think of possible scenarios:

o Tendency to achieve word-for-word and morpheme-per-morpheme intertranslatability. For example the use of Italian derivations in Maltese. French derivations in Middle English. o The existence of a perceivable ‘gap’ facilitates diffusion. For instance, Australian languages had no ‘conventionalised counting systems’, no numbers for counting, and the Aborigines started using either the counting practices of European invaders or made up ‘numbers’ using their own words. In the same way a language called Likpe lacked plural forms and developed them following a Ewe mould model. o Typology changes. Pipil developed impersonal meaning of a 3rd person under the influence of Spanish. Se dice que…. There are of course much more subtypes of the changes as to the contact-induced linguistic changed, but further, what does it all depend on: Sociolinguistic parameters in language contact. The impact of a prestigious 2nd language in a predominantly monolingual community typically results in an abundance of loanwords, but hardly any structural influence—English loans in Japanese as prime example; *Estonian word-for-word ‘translations’ from German ‘aushalten’ and *the way English was ‘formed’ under the influence of the French. Another sociolinguistic factor is the speakers’ attitude. When speakers of some language are conscious of the influence, they can decide whether they want it or not—in some cases, yes, as they think it’s prestigious, like discussed above, in some cases they can decide they don’t want such an influence and try to ‘ban’ it suggesting or even ‘forcing upon’ different language use. Kemal Atatürk

decided that Turkish should get rid of the Arabic loans—some of these being too antique—replacing them with native words, at the same time he had nothing against the Indo-European loans. Johannes Aavik designed Estonian that way saying that the language should be less ‘German-like’ and more ‘elegant’, he actually made language more analytic stressing it analytic tendencies. (*In his newer research about the language Mati Hint stresses that the major danger is not the occasional loan words but foreign forms in grammar.) Balanced and displacive language contact. The language contact itself may be very different of course; it was briefly touched upon above, more details in the flowing table:

Parameters Relationships between languages Linguistic effects Results

Balanced contact Displacive contact Roughly equal, or involving Dominance, unstable a traditional hierarchy; stable Rise in complexity; gain of Loss of patterns; potential patterns simplification Language maintenance Potential replacement of one language with another

The net result of language contact. How long does it take to acquire the necessary layers in the language contact, we don’t know; for the contact between Ewe and Likpe is only about 300-400 years and the language contact between the Roman languages and Basque is about two millennia. The result is that languages become convergent—more similar even if they don’t share many forms in common. For instance there’s Portuguese spoken by Amazonian Indians and Portuguese spoken by Sri Lanka inhabitants—the two variations of Portuguese are different; they have some distinct forms. Languages in contact acquire new common grammar.

5. What can we conclude?

Languages reflect sociolinguistic history of their speakers; and language attitudes influence the outcome of language contact, as do relationships between languages within a contact situation. But as there are multiple scenarios, it’s important to see this diffusion from multiple perspectives, looking at various scenarios systematically we come to understand how language come to share aspects of their grammars. 6. How is this volume organised?

Firstly, we look at possible different ‘layers’. Then comes looking for reasons—whether the influence was balanced or not. Then the role of lingua franca (most optimal and most probable variants that will dominate one way or another) Then, the language speakers’ attitude: the Pennsylvania German speakers don’t mind speaking a humble ‘hybrid’ variety of their language; but a fairly prescriptive as far as their own language. In the end there are various regional, historical, and cultural aspects that need

to be taken into account. In Ameka’s words ‘a holistic understanding of language change requires multiple perspectives’.

**** Biography: (Russian) Алекса́ндра Ю́рьевна Айхенва́льд (англ. Alexandra Aikhenvald, род. 1 сентября 1957, Москва) — российский и австралийский лингвист, доктор наук (2005), специалист по типологии и полевой лингвистике. Труды по грамматической типологии, ареальной лингвистике, семитским, берберским, аравакским, папуасским и др. языкам; активная организаторская и издательская деятельность.

Дочь известного диссидента, поэта, переводчика и критика Ю. А. Айхенвальда, правнучка известного критика Ю. И. Айхенвальда. Окончила Отделение структурной и прикладной лингвистики МГУ (1979), работала в Отделе языков Института востоковедения РАН. Кандидатская диссертация (1984) на тему «Структурная и типологическая классификация берберских языков». С 1989 г. в Бразилии, занималась полевыми исследованиями языков аравакской группы и ряда других малоизученных и исчезающих языков бассейна Амазонки. С 1994 г. в Австралии, в университетах Канберры и Мельбурна. В настоящее время — заместитель директора Научно-исследовательского центра по лингвистической типологии университета им. Ла Троба (Мельбурн); занимается полевыми исследованиями папуасских языков Новой Гвинеи.

Автор 10 монографий (в том числе грамматических описаний современного иврита, южноамериканских исчезающих языков баре и тариана), организатор большого числа конференций; под ее редакцией (также совместно с Р. Диксоном) выпущено более 10 сборников статей по типологиии и ареальной лингвистике. Основные работы: Современный иврит. М., 1990. Bare. München, 1995. Classifiers. Oxford, 1993. Language contact in Amazonia. Oxford, 2002. A grammar of Tariana. Cambridge, 2003. Evidentiality. Oxford, 2004.