Did the Celts "bastardize" English?
Linguist and author John McWhorter thinks so. In his book Our Magnificent Bastard Tongue, he presents an “untold story of English”, the story of its being shaped by people who spoke a Celtic language. This is a new story, one which is just beginning to emerge from recent scholarship into the history of the English language. Until recently, the consensus has been that English has had little (if any) influence from Celtic languages. And if one looks at the English vocabulary, this is indeed so. There are, of course, various recent borrowings from Celtic languages, such as whisky (from the Scottish Gaelic uisge-beatha ‘water of life’) and strontium (from the Scottish Gaelic Sròn an t-Sìthein, name of a mountain, near which the element was discovered), but the “old” Celtic borrowings, with the exception of place names, are rare. This led many scholars to follow Baugh & Cable’s (1993: 85) conclusion that “outside of place-names the influence of Celtic upon the English language is almost negligible”. But it turns out that all these years we’ve been looking in the wrong place.
Instead of the vocabulary, we should be looking at the sound system and especially the grammar of English. According to the Wikipedia article on Old English, “why these [distinctive Celtic] traits appear to be restricted to syntax and do not include vocabulary is not clear”. However, an explanation for this is fairly obvious. The conquerors (in this case, Anglo-Saxons) have never had any practical need to borrow words from the language of the conquered (in this case, from the Celtic languages that were spoken in the British Isles before the Anglo-Saxon conquest). Words are almost always borrowed consciously: one looks for a good, expressive word in one’s own language and if one is not found, one turns to another language for a loan. But there is another type of borrowing, a more subconcious borrowing. That is what happens when grammatical properties of one language diffuse into another.
In the case of the Celtic people under the Anglo-Saxon rule, many native Celtic speakers learned English in informal situations (there were no ESL classes back then!). This learning was rather incomplete: they learned just enough to get by. Moreover, most of this learning was done by adults who were not as good at picking up grammatical traits of English as babies acquiring English natively, from birth would be. As a result of this massive, imperfect second language learning, Celtic speakers ended up speaking English “with an accent”. Children, including those who were not exposed to any Celtic language, were often exposed to this “broken” English and this is the language they will end up speaking too. But this “brokeness” was mostly not in the area of vocabulary — after all one can always improve one’s vocabulary later in life — but in the area of grammar.
What specific elements of the English grammar can be said to be of Celtic origin? One well-studied example of a trait now thought to derive from Celtic is the so-called do-support. In English, a “dummy” auxiliary do must be used to form negative and interrogative sentences; hence, we say John does not swim (instead of *John swims not or *John not swims) and Does John swim? (instead of *Swims John?). Other Germanic languages (German, Dutch, Norwegian, Afrikaans, etc.) do not have this construction (note that the English use of do-support is different from the so-called emphatic do, as in John does swim!, which is found in other Germanic languages). Nor do Romance or Slavic languages have anything resembling do-support. In fact, it’s a rather rare quirk, cross-linguistically.
We now think that English picked it up from its Celtic neighbors. Unlike Germanic languages, Celtic languages like Cornish have had do-support since before English started using it. It is used in Celtic exactly as in English: to express tense and agreement in negative and interrogative sentences.
Several objections have been raised for the Celtic hypothesis on the origin of the English do-support. First, it has been suggested that the direction of borrowing could have been the opposite, that the Celtic languages borrowed this phenomenon from English rather than the other way around. The problem with that objection is that Breton has do-support too, even though Breton speakers moved across the Channel into northwestern France in 400s and 500s, too early to borrow do-support from English. Another objection concerns timing too: do-support appears in English too late (around 1300s) to be modeled on Celtic. In other words, if the influence of Celtic on English happened through children exposed to “broken” English of Celtic speakers who learned English imperfectly, we expect the do-support to have diffused into English in a matter of one or two generations, not nearly a millenium. The objection to this objection goes as follows: do-support, like other properties of the “bastardized” English might have remained a colloquial, low-register property for a long time, being finally absorbed into the English language when the Celtic speakers themselves had been absorbed into the Anglo-Saxon linguistic community.
Whether this is the correct analysis for the do-support phenomenon remains to be seen. However, many other linguistic peculiarities of English have been recently analyzed as having Celtic origin. One such peculiarity is expressing possession internally to the noun phrase, for instance The queen cut off [the king’s head] rather than something like The queen cut off [to the king] [the head], which sounds rather odd in English but is perfectly good in German. Another potentially Celtic-based peculiarity of English is the use of the same form for both a reflexive (e.g., John keeps talking to himself) and an intensifier (e.g., The president himself came yesterday). Also possibly of Celtic origin are the so-called it-clefts, as in It was John who had the best Halloween costume.
And it has also been proposed that certain properties of the English sound system are Celtic in origin, including its diphthongs and the use of voiced fricatives like /v/ to differentiate meaning. As with the syntactic properties discussed above, those potententially Celtic-based properties involving sounds did not appear in English immediately upon the Anglo-Saxons’ contact with Celtic speakers. In fact, it was not until several centuries later that /v/ emerges as a full-fledged phoneme of English. So in the next posting, we will consider an alternative theory for the appearance of /v/ as a phoneme in English.