Sunday, May 15, 2016

When I say "I", I mean "you": public service hortatives in French

A lot of languages - Indonesian, for instance - make a rather useful distinction between two 1st person plural pronouns: "we (including you)" and "we (excluding you)". A few languages, such as Nivkh, extend this distinction to the singular, sort of, having a dual 1st person inclusive pronoun "I and you" alongside a singular 1st person exclusive pronoun "I" (and no other duals). But a 1st person singular inclusive pronoun, strictly speaking, is a contradiction in terms: it would have to be a pronoun referring to only one person which included both the speaker and the addressee.

Or is it? There are a couple of ways in which this apparent contradiction could be resolved. The most obvious would be if you had a special pronoun used only when the speaker was also the addressee; but, as such a form would be used only in talking to oneself, it would be unlikely to catch on enough to become part of the language. Less obviously, however, you could have a singular pronoun being used in a sufficiently vague way to refer to both the speaker and the addressee (but not to an uninvolved third person.)

Soon after moving to France, I realised that, in public announcements, this is in fact what French does with its 1st person singular pronoun je. The realisation was prompted by a poster in a medical insurance office saying, in big letters, something like:

Je choisis le générique, je ne fais pas d'avance de frais.
(I choose generic drugs, I pay no advance.)

This was clearly not a piece of self-observation someone had put up; rather, it was intended to tell us "Choose generic drugs, and pay no advance". Over the following days, I noticed that concealed exhortations of this form were everywhere: Oui je vote (Yes I vote), En car comme en voiture, je boucle ma ceinture (In a coach as in a car, I buckle my seatbelt), ... All easily understandable as conveying the message is "I do this, and so should you". But in English, you consistently cast such messages in the imperative, with no "I" at all: "Please take a moment to cast your vote in this important election" or "Buckle up, it's the law", and so on. One obvious side effect is that the slogan "Je suis Charlie" has at least one reading directly accessible to French speakers but not to English speakers who understand it word for word: namely, "I am Charlie, and you'd better be Charlie too".

The difference between the two languages in this respect is at the level of pragmatics, for now. But if such hortatives become sufficiently common in French, one could well imagine the construction grammaticalising further and even eventually becoming distinct from ordinary 1st person marking. In that case, we might end up with a true 1st person singular inclusive pronoun: a pronoun that simultaneously means "I" and "you", while taking strict singular agreement. Give it another 500 years...

Are you familiar with another language that does this?

Wednesday, May 04, 2016

Some thoughts on racism in Algeria

(Regular readers be warned: this post has nothing to do with linguistics; it's justified only by a tenuous link to my fieldwork.)

New York Times readers today had the dubious privilege of an editorial by Kamel Daoud on racism in Algeria. The topic certainly needs attention, even if the New York Times is hardly the most effective place to address it in. Unfortunately, he addresses it with the same broad-brush, narrative-forcing, emotional vagueness that usually characterises his editorials (with bits of outright distortion: Echourouk "Islamist"? Algerians who won't "shake hands with blacks"?). He claims that Algerians are racist on the basis of religion rather than colour, then belatedly notices that there have been conflicts with Muslim black migrants too, and "explains" this by suggesting that they are seen as insufficiently Muslim. We get quotes from a few Algerian racists, but no migrants' voices, and no sign at all of the group most obviously relevant to a framing in racial terms: black Algerians.

In many Saharan oases - including Tabelbala, where I did most of my PhD fieldwork - black people are in the majority. Even in the north, you find small villages of black people, and of course larger communities in the big cities. Kamel Daoud mentions anti-migrant riots in Ouargla and Bechar: both those Saharan towns have massive Algerian black communities. Contrary to Kamel Daoud's analysis, such groups certainly do experience racism, though in a much milder form. In the south, people assume their ancestors were slaves, in a region where people routinely claim status and allies based on genealogy. In the north, their colour makes them visible outsiders, in a context where people regularly blame social decay on "outsiders" immigrating from ten or twenty kilometres away. Unlike in America, however, they are not particularly stereotyped as criminal (though black immigrants sometimes are). In the north they tend to be stereotyped as stupid, but in the south their conspicuous relative educational success makes that image hard to maintain. Socialism and Islam, however, are equally vehement in their condemnation of such racism, and after independence the Algerian state took this issue seriously, stamping out the remnants of slavery and emphasising universal equality; everyone today at least knows they're not supposed to be racist, though that doesn't necessarily stop them.

Of course, race is in the eye of the beholder. In Tabelbala, almost everyone is black by the standards of other parts of Algeria. By their own standards, however, the situation is a bit different: anyone with the slightest tinge of known Arab or Berber ancestry counts as white, leaving only a few families to be considered as black. Until the 20th century, the former were landowners, while the latter were sharecroppers or slaves. The indistinguishability of their skin colours does not stop the former from being viciously racist about the latter when annoyed with them.

I don't claim to understand the riots in Ouargla and Bechar in any detail, but two points are noteworthy. The first is that they did not attack Algerian black people: they attacked black immigrants. To an Algerian, that may seem almost too obvious to mention - but the NYT's audience is not particularly Algerian, and has rather different baseline assumptions. The second is that they happened in a wider context of rising tensions in the Sahara over the past five years or so, including especially the ever-worsening cycle of sectarian riots in Ghardaia. It would be very useful to have a serious analysis of what's driving this rising intolerance, in the one part of Algeria that largely escaped violence throughout the 1990s. But for that, the NYT would have to call in a real journalist.

Sunday, May 01, 2016

Yuck: a borrowing from Arabic into Berber?

One of my son's first words is [x:::], "yuck!" - his attempt to pronounce the Algerian Arabic baby-talk item kəxx(i) كخّ "yuck". I was surprised to learn recently that this word goes back well over a millennium: a hadith in Sahih Muslim records its use in addressing Ali's son Hasan, then a child:
أخذ الحسن بن علي تمرة من تمر الصدقة فجعلها في فيه فقال رسول الله صلى الله عليه وسلم : كخ كخ ارم بها أما علمت أنا لا نأكل الصدقة (link)
Al-Hasan son of Ali took a charity date and put it in his mouth. So the Messenger of God, peace be upon him, said: "Kax, kax, throw it away; don't you know that we do not eat alms?"

Variants of this word (kxx, kexx, kexxa, kəxx) are very widespread in North Africa, not just in Arabic but in Berber too, as you can see from the Barefoot Linguist's Baby Talk database: it's used in Siwi, in Kabyle, in Tarifiyt, and in Senhaja. In Europe, on the other hand, it's far from universal; in fact, I don't know that it's even attested. That suggests that independent parallel innovation is unlikely. /x/ is a perfectly normal phoneme within Arabic, but in Berber it's rare in inherited roots and unlikely to be reconstructible for proto-Berber; all of the Berber languages listed there as having this word are intensely influenced by Arabic. That makes it unlikely that it's a common retention from proto-Afro-Asiatic. The most obvious conclusion is that kəxx has been borrowed from Arabic into Berber. Other cases of the borrowing of baby-talk is certainly attested, but this example seems particularly striking for the word's sheer frequency.