Language contact, whether between similar dialects or typologically different languages, almost always influences change in one or more of the languages in question. Linguistic borrowing of any sort is closely associated with certain socio-historical and socio-cultural factors, namely those of conquest and immigration. In such situations, there is usually a dominant language which has a higher status in the speech community. However, language contact does not always come about only due to such sociological factors. Monolingual communities, while comparatively rare, can and do borrow linguistic elements from languages which have a high status globally. Take, for instance, the relatively recent influx of anglicisms into many languages (e.g. German).
Here I’m going to consider the nature of language contact mainly from a diachronic perspective (i.e. how contact induces language change). I won’t deal with topics like (synchronic) bilingialism, second language acquisition or pidgins and creoles. For a good introduction to all the issues surrounding language contact, you are advised to consult Yaron Matras’ 2009 book (see references). Instead, I’ll focus on how language contact influences change in various components of the grammar, and I’ll examine the claim that it is lexical items that are borrowed first and most readily, while morpho-syntactic elements are most resistant to borrowing. Most of the examples will come from the history of English, for a couple of reasons. First, most people reading this blog post will be native speakers of English, and it might interest them to learn about the history of their language. In addition, English has undergone language contact with both typologically similar and less similar languages during its history – in the first instance, Norse, in the second, Norman French.
It’s certainly true that lexical items are very readily borrowed. Over the last century, for example, speakers of German have borrowed a good deal of lexis from English, particularly in certain semantic fields such as business and trade. These include lexical items for which German words already exist, such as die Show (alongside die Vorstellung), as well as what are known as pseudo-loans, that is, loans that are indeed borrowed but refer to something different from in the source language, such as der Smoking (a dinner jacket) and das Handy (a mobile phone). In trying to account for the reasons why such lexical items are borrowed, researchers often appeal to the notion of prestige, which is actually remarkably difficult to define, but refers to something like high status within a particular speech community. We know that prestige is a factor here because there is no need to borrow a good deal of the lexis, given the existence of native words for the same items. It’s worth noting, too, that notions of prestige are not static, and indeed German has borrowed extensively from other high-status languages during its history, especially Latin and French. This indicates that social factors play a significant role in linguistic borrowing, but also that these factors are subject to change.
Are there restrictions on the kind of lexis that usually gets borrowed? The short answer is yes, it tends not to be basic vocabulary that gets borrowed. This is probably resistant to change because it gets used a lot more frequently than non-core vocabulary. In the time of extensive borrowing from French into English following the Norman conquest, a vast number of French lexemes were borrowed into English, such as visible, table, and arrive, to name but a few. However, the basic vocabulary, such as kinship terms, numbers, colours, and so on, remained Germanic (English, if you recall, shares a common ancestor with other Germanic languages such as German, Dutch, Frisian, and the Scandinavian languages). Some items of core vocabulary were borrowed, but these items were (and still are) associated with a higher register or status: think of items of food that we get from French such as pork and veal.
Despite the fact that the lexicon is most readily borrowed across linguistic systems, other components of the grammar can also be affected. For example, the French loans we have just discussed also influenced the phonological component of English speakers. McMahon (1994) shows that, in Old English, [v] and [f] were allophones of the same phoneme /v/. That means that [f] and [v] didn’t distinguish meaning in any way, each one only appeared in a specific phonological environment, and they were in what’s called complementary distribution (more on that here). For example, word initially, you never got [v] in Old English, in that instance the phoneme would always be articulated as [f]. But many French loans were pronounced with initial [v], such as veal, virtue and visible. This contributed to what’s known as a phonemic split, where the speech sounds [v] and [f] ceased to be allophones of the same phoneme, no longer in complementary distribution. They split into two phonemes, /v/ and /f/, which is what we have in Modern English.
Similarly, the morphological component can also be affected, and there are also examples of this from French borrowing into English during the Medieval period. Speakers reanalyzed -ible/-able in words such as terrible and visible, as a productive derivational suffix (see this post for more info on what this means), which was then extended freely to non-French lexemes such as enjoyable. By ‘reanalysis’, we mean that the hearer understands the meaning or form of a linguistic element to be different from that of the speaker, in this case that -ible/-able is a suffix that can be used freely also with native adjectives. And ‘extension’ refers to the fact that the new structure or meaning is extended to new contexts, like here with enjoyable.
Likewise, syntactic borrowing is also possible, and Harris & Campbell (1995) list it as one of the three mechanisms of syntactic change. For example, the comparative Spanish construction mas … que was borrowed into Pipil, a language spoken in El Salvador, as mas … ke. Unlike the effects of language contact we witnessed in the phonological and morphological components above, syntactic borrowing requires no reanalysis, at least in this example, according to Harris & Campbell (1995). In the [v]/[f] and -ible/-able examples above, both reanalysis and extension were the mechanisms of the change: the contact was just the facilitator. Harris & Campbell (1995) think syntactic borrowing is different, in that it is its own independent mechanism of change, not requiring reanalysis or extension. This is not an uncontroversial position, however!
While borrowing of lexis is usually widespread in situations of contact, the other components of the grammar are often much less affected. However, when the contact languages are typologically similar, that is, if they stem relatively recently from a common ancestor and are (almost?) mutually intelligible, then this observation does not hold. For example, the Viking invasions influenced English in manifold ways, which is clear when one compares Old English with Middle English. While it is a matter of some debate as to whether Norse and Old English were mutually intelligible dialects during at least some of the period that we associate with the Viking settlement, it is certainly the case that the two varieties were much more similar than, say, Old English and Norman French. So we find, for example, that some items of the core vocabulary of Norse make their way into English, pertaining e.g. to the weather (such as sky) and body parts (e.g. skin). (As an aside, did you ever notice that shirt and skirt descend from the same Germanic word? Skirt was borrowed later from Norse.) Similarly, grammatical function words like the pronouns their and them were borrowed into English, likewise the demonstrative the. Consequently, the phonemes /ð/ and /θ/ became part of English’s phonological inventory. It is suspected that the languages’ typological similarity facilitated these changes. Why and how this happened, however, is a matter of debate (and a topic for a future blog post, perhaps).
Thomason & Kaufman (1988) account for this observation by suggesting that languages can only borrow structural elements if they correspond to ‘tendencies of development’ of the language into which the borrowing is happening. To my mind, this abstracts languages away from its speakers somewhat and suggests (even if only by the back door) that there are determined ‘pathways’ of change for specific languages in their development, which I’m not sure I totally sign up for in any teleological sense. Otherwise, it seems a bit like their view is just a restatement of the fact that typologically similar languages borrow readily across all components of the grammar.
An alternative explanation is based on the generative principle that the lexicon is less highly organised than other components of the grammar. That is, phonological and morpho-syntactic phenomena are less likely to be borrowed because they involve multiple lexical items and multiple conceptual levels. For example, the borrowing and reanalysis of /v/ as a distinct phoneme from /f/ in English affects the underlying level as well as the surface structure. However, where linguistic components are already very similar, such as is the case with Old Norse and Old English, structural borrowing can be facilitated more easily, as the rules and representations as well as any underlying forms are likely to be much more similar.
To conclude, then, we’ve seen that lexis is most commonly borrowed, whereas other components of the grammar are much less affected when the languages in question are typologically quite distinct. This is, however, not usually the case when languages are typologically similar. Some ways of accounting for this were also offered, in particular that the lexicon is less organised than the other components. But what does ‘less organised’ really mean? And how ‘similar’ do languages need to be to facilitate grammatical borrowing? Is this always the case? These are some of the questions linguists are still trying to answer.
Harris, A. C., & L. Campbell. 1995. Historical syntax in cross-linguistic perspective. Cambridge: Cambridge University Press.
Matras, Y. 2009. Language Contact. Cambridge: CUP.
McMahon, A. 1994. Understanding Language Change, Cambridge.
Thomason, S. & T. Kaufman. 1988. Language contact, creolization, and genetic linguistics (first ed.), Berkeley: University of California Press