Showing posts with label morphology. Show all posts
Showing posts with label morphology. Show all posts

Tuesday, August 20, 2013

Talking funny: Fula

As I was writing my conference proceedings paper last week, I had the idea to write a bit about some of the unusual (to English speakers) features of one of the languages I worked on. I'm thinking of turning this into a series, if I can come up with enough material for it--it's cool to see how diverse languages really are!

Fula (aka Fulani, Fulfulde, Pulaar) is a language of sub-Saharan Africa, spoken by somewhere between 10 and 25 million people (estimates vary); that makes it one of the top 100 most widespread languages in the world. The homeland of Fula is somewhere around Senegal in West Africa, but since the speakers of Fula were originally nomadic herders--and sometime Islamic imperialists--they spread as far east as Sudan. So dialects of Fula are now spoken in a band across the continent.

Men of the Fula-speaking Wodaabe people in Niger (photo by Dan Lundberg)
Some early European scholars studying Fula found its grammar to be so complex that some wondered (half-seriously) how the language could ever have developed naturally. What could have gotten them so flummoxed? I'm going to tell you about two linguistic features of Fula that seem pretty bizarre, from a European point of view (neither of which are actually unique to Fula).

First off, like virtually all its relatives (including Swahili, Zulu and many many others), Fula has things called noun classes. If you've studied a language like Spanish, German, or French, you're already familiar with these--they're called genders in European languages. All nouns in the language are divided into groups that each get different markings. So in Spanish, you say:
  • el pulp-o 'the octopus', a masculine noun
  • la langost-a 'the lobster', a feminine noun
The two genders get different endings (-o for masculine, -a for feminine) and different articles (the two words meaning 'the').

OK, so Fula has genders too, except instead of two or three, it has twenty-something (the exact number depends on the dialect). And the classes have nothing to do with biological sex. Here are some examples:
  • ɗem-ngal 'tongue' (class 16)*
  • nagg-e 'cow' (class 13)
  • njaayrii-ji 'open spaces' (class 25)
  • loo-ngel 'little pot' (class 3)
One cool thing about the noun classes is that you can take a single noun and put it in any one of several classes to get different nuances of meaning. Let's take the root laam 'chief':
  • laam-ɗo 'chief' (class 1)
  • laam-ɓe 'chiefs' (class 2)
  • laam-ngel 'petty chief' (class 3)
  • laam-ngum 'worthless little chief' (class 5)
  • laam-kon 'petty chiefs' (class 6)
  • laam-nga 'mighty chief' (class 7)
  • laam-ko 'mighty chiefs' (class 8)

Pretty fun, right?

OK, so if you want to learn Fula, you have to learn what noun goes in what class, and also memorize the suffixes for each class. But that's not all! Look what happens to the root meaning 'compound' in different classes:
  • wur-o 'compound'
  • gur-e 'compounds'
  • ngur-on 'small compounds'

WHOA!! Not only did the suffix change, but the first consonant of the root changed too! This is known as initial consonant mutation, and it's a feature of Atlantic, the small subfamily that Fula belongs to (it's also found in the Celtic languages). The consonants of the language are arranged in sets of up to three, called grades (for example, w, g, ng), and the first consonant of any noun or verb root switches between the consonants in a set. Each noun class requires a noun to use a certain grade of consonant. Another example is the word for 'a Fula person', which is Pull-o in the singular but Ful-ɓe in the plural.

So when you learn Fula, not only do you have to learn which noun is in which noun class and the suffix for each noun class, you also have to know which consonant grade each noun class likes! Oh, and did I mention that the noun class suffixes also undergo consonant mutation?...

If you want to get a feel for how Fula sounds, here's a recording which I guess is of a Bible story.



*The symbols ɗ and ɓ stand for a special type of d and b that are made by drawing air into the mouth instead of expelling it.

Thursday, May 23, 2013

Talking about talking: regular mice

Last month I posted about the linguistic process of analogy--where speakers change one part of their language to be more like another part. In that post, I mentioned that in general, a given irregular pattern in a language will tend to become regular over time. This raises an obvious question: if speakers tend to eliminate irregularity, then why aren't all languages totally regular by now? Where do new irregular patterns come from? The answer is actually quite simple: they're accidental.


Monday, April 8, 2013

Talking about talking: wing, wang, wung?

I'm a grad student in linguistics, so I can't help noticing cool little quirks in the different ways people talk. I've been catching up on old episodes of The Tolkien Professor podcast (HIGHLY recommended if you're at all interested in Tolkien), and I heard a seminar participant say, "I wung it" (meaning "I improvised it").

Why did he (or whoever he learned the word from) use wung over the standard English winged? If you think about it a second, you might be able to guess why. Historical linguists (people who study how language changes) call this kind of development analogy, just like those annoying SAT problems I had to do a million of: A is to B as X is to __ . Analogy is a way of making one part of your language more similar to another part.

What might be the basis of the analogy this speaker used to create wung? I think there are a few possibilities:
  1. fling - flung, sting - stung, and also similarly hang - hung. So the analogy might be "fling is to flung as wing is to wung".
  2. Verbs with three different forms, like sing - sang - sung, ring - rang - rung, also drink - drank - drunk, stink - stank - stunk.
The problem with option (2), at first glance, is that the analogy doesn't line up: since the past tense of sing is sang, and so forth, the past tense of wing should be...well, wang. I can think of a few good reasons to avoid using this form. But there might be an actual linguistic reason too.

I've noticed other people (and myself too) using the u-forms (the past participles) of words like sing in the past tense instead of the a-forms. How do these sentences sound to you?
  • We all stood up and sung the national anthem.
  • That stupid dog rolled in poo and stunk up my whole house!
To me, they're not all that bad (although I recognize them as being non-standard). I think these three-form verbs like sing might be slowly simplifying to two-form verbs, like fling. After all, it's one less form to worry about (most English verbs only have two forms anyway) and there are very similar verbs like fling to base the analogy on.

Another possible example: you're probably familiar with thunk, as in "Who'd a' thunk it?" This verb seems to have been formed by analogy to drink - drank - drunk. But again, by this analogy the past tense of think should be *thank. Have you ever heard someone say:
  • I thank about him all day long.
So, going back to wung, I think we might have two things going on: formation of a new past tense by analogy, and the beginnings of simplification of three-form verbs to two-form.

Oh, one more reason why wung is interesting: in general, historical linguists have noticed that a given irregular pattern in a language will tend to fade away over time. Based on this principle, we would expect to hear people saying things like *singed, *drinked and so forth. But in this case, we can see that if an irregular form is used frequently enough, it can actually overpower regular forms (like winged) and also different irregular forms (like thought > thunk). And it doesn't just happen in non-standard dialects--did you know the past tense of wear used to be weared?