30 July 2015

Language following

Last week, I was at a party in Edinburgh to mark Peruvian independence day. As I was leaving, I heard someone refusing a drink because "tengo que manejar" -- "I have to drive".

Funnily enough, I've had a couple of discussions recently about that very word "drive". It all started with a discussion on a Welsh-language Facebook group. The traditional word presented there was gyrru, whereas people often tend to use the term dreifio, presented there as an Anglicism. Strangely enough, the very next day, I ran into an old university classmate of mine, Carwyn, who was up from Wales to visit a conference in Edinburgh. When I asked him which word he would use to say "drive", his answer was "probably the wrong one", which I immediately took to mean dreifio.

I explained to him why I felt that dreifio was less of an Anglicism than gyrru.

How so?

This is a phenomenon that I call "dictionary following", for wont of a better term. (If there's a widely-accepted alternative name, please do let me know in the comments.) It's a peculiar form of language change that minority languages seem particularly prone to undergoing, where a word-form in one language gets locked to the changing meaning of a single equivalent in another language.
Edit: An Cionnfhaolach over at IrishLanguageForum.com tells me that this transferrence of all meanings for a word in one language to a similar word in another is called a "semantic loan".

In this case, the dictionary word gyrru is a word that means to spur animals onwards -- it's "drive" as in "driving cattle": what drovers do. The modern sense of "drive" comes via the idea of forcing carthorses forward, and thus the English word has broadened.

Across Europe, the equivalent word often evolved analogously. The French and Italian equivalent term is actually to "conduct" a car, and in Spanish, you either "conduct" or "handle" your car -- which is where manejar comes into the equation (manejar = manage = handle; mano = hand).

It's too easy to focus on the grammatical and lexical items as being the characteristics of a language, but if that is not underpinned by idiomatic usage and unique linguistic metaphors, then it doesn't feel like a complete language; and for me at least, much of the joy of learning and speaking that language is lost.

So for me, I'm happier to adopt the English "drive" morpheme into languages like Gaelic and Welsh than to adopt the English metaphor with a native room and claim that this is somehow "purer".

20 July 2015

Undefined article error

No, Blogger isn't on the blink, that's the intended title of the article.

The error in question is the continued use of classical terminology for grammatical articles: specifically the terms definite article and indefinite article. For over a decade, I tried to reconcile the grammatical feature with the common sense of the words "definite" and "indefinite" -- i.e. certain and uncertain -- but it made no sense at all.

It wasn't until I started discussing grammar in foreign languages that I clicked what I'd been missing all along -- the terms we use are basically a mistranslation of classical terminology.

The English word definite has diverged drastically from its etymological roots, but this is not true in the Romance languages on mainland Europe. When the French say défini or the Spanish say definido, what they are actually saying is defined.

That's right, the definite article is really the defined article, which means the indefinite article must be the undefined article. From that perspective, everything seems to make much more sense.

Plenty of languages survive quite well without any articles -- they are essentially redundant as even in English, in a lot of circumstances you can drop them without losing any information in the sentence.

What I'd never got my head round was that the articles don't add any information to the sentence -- they simply act as a sort of "signpost" to information that already exists elsewhere. But most importantly, it refers to the listener's frame of reference and not the speakers.

What the definite article flags up is essentially "you know which one I mean", and the indefinite article says "you don't know which one I mean". If I say "You should go home -- the wife'll be waiting," context says I'm talking about your wife, but if I say "I should go home -- the wife'll be waiting," then you know that I'm talking about my wife. And if I say "a friend of mine is coming to visit," I'm telling you that I don't expect you to know which one I'm talking about. But in both cases, if you delete the articles, I would still make the assumption of yours/mine or that I'm not sure in the second.

Now I know that isn't very clear, but to be honest, I still haven't got this that clear in my own head.

This "signposting" idea is pretty abstract, so describing it is pretty difficult. But to be fair, it's no more abstract than the phenomenon it's describing, and the more I think about articles, the more weird and abstract they look to me. For something at first class so basic, they are incredibly complex.

I suppose I'll be working for years trying to work out the best way to teach, discuss and describe them, but for now I'll satisfy myself with using the terms defined and undefined in place of definite and indefinite, because at the very least we'll be one step closer to a meaningful definition.