Fala rural

Here’s a link to the first in a series of pretty amusing cartoons concerning Tonin, a young rural mineiro who gets taken by ninjas to follow his destiny.


It is full of stereotypes, and young Tonin and his parents speak with strong mineiro accents. The subtitles are nice, reflecting the pronunciation, for example:

Noís tem doze fi! Se é pra iscoiê um, iscói o menorzim.

“We have twelve children. If you’re going to choose one, choose the youngest.”

This is chock full of ruralisms! Should you care I spotted (and I am sure a native speaker would pick more):

  • Noís for nós. This addition of an i prior to a syllable-final s is not just limited to mineiro speech; it is fairly common throughout Brazil. It is interesting that, in the audio track, the pronunciation of this is actually closer to “noysh”. This pronunciation of syllable-final s as sh is typical not just of the mineiro accent but also that of Rio de Janeiro. It can be represented orthographically — as noíx — but they have chosen not to. This tells us a little about which features of the speech are considered emphatically mineiro, and thus worth highlighting for humour, and those which are seen as just “ordinary” variations.
  • Tem here is the third person singular form of ter, “to have,” yet it is being applied to “we” (noís). This lack of agreement is one of the very distinctive features of vernacular speech, and one which looks very creolized (to reference the previous post).
  • Fi is impressive: it is a substantial reduction of filhos, “sons, children” (remember my gender/markedness post?). How one gets from filhos to fi is a combination of the mineiro tendency to drop unstressed vowels at the end of the word, the rural pronunciation of lh (usually pronounced as a palatal lateral, which doesn’t exist in English, but sounds to our ears more-or-less like ly) as the semi-vowel y and probably the fact that the singular filho would have been used anyway (lack of agreement again). That is:
    • for “standard” filhos (“feelyus”) the farmer would have said singular filho;
    • the lh is pronounced y, which merges with the preceding ee (=“fiyu”);
    • the word-teminal u is dropped (=“feey”)
  • Se é pra iscoiê um means, more-or-less, “If you are to choose one.” However the exact meaning is actually quite difficult for me to parse due to the non-standard grammar, compounded by the (perfectly standard) omission of a pronoun. The é (“is”) is, in standard terms, “wrong.” It could be that it applies to the ninjas, (that is, we would gloss the sentence as “If [you] are”), in which case it fails to agree, as per the tem above. However, it could also be just an existential “it is” — the sentence starts “If it is that”. Either way, after se (“if”) the standard grammar would expect the subjunctive (seja or sejam in the plural) — which you may recall in a previous post I noted was generally absent in the rural vernaculars, and so it is here.
  • Iscoiê and iscói are both renderings of the rural pronunciation of forms of the verb escolher, “to choose.” The initial e is raised to i (sounds which in English we would generally represent with eh and ee respectively) and, as for filhos above, the lh has become a semi-vowel i.
  • Finally, in menorzim, we see the characteristic addition of the diminutive zinho, with (as before) the equally distinctive deletion of the final vowel.

There’s plenty more, both in this episode and others. Worth a chuckle, possibly mainly for Portuguese-speakers though!

The perils of fieldwork, part three and a quick history lesson

Having kept you all waiting for an unseemly long time, here is the third part in my tales of peril in the field, and this one — depending on your proclivity towards counterfactual speculation — is truly a little scary.

It’s taken me so long to write up that the event is now a good few weeks back, as it happened whilst I was visiting Picinguaba. I went with a couple of English friends (one resident in Picinguaba, the other visiting her) and a Brazilian chap to a waterfall, near Ubatumirim. Visiting waterfalls is a popular alternative to a beach visit: there is considerably more shade, and usually a pleasant stroll through the forest to get to the falls. A pleasant stroll indeed was had, and then a few hours lolling around and sunbathing and reading and generally being indolent.

We started back, and Gian (the Brazilian guy) was in the front, which was fortunate as but a couple of meters down the path he stopped, held us back, and warned, “Be careful, there is a snake there.” Coming up next, I looked at where he was pointing a saw, well, nothing. It took me a goodly period before I could make it out, and there is no way I would have spotted it just walking past, but there was, indeed, a smallish brown/yellow snake curled up on the path. Camouflage works really rather well actually out there in the wild, it would appear.

Anyway, Gian told us to be very careful. This was a jararaca or a jararacuçu (c is a “hard” c/k sound, whereas ç is pronounced as s; the stress is on the penultimate and the last syllable respectively, so these are said ja-ra-RA-ka and ja-ra-ra-ku-SU). Whichever it was, he warned us, the snake was very venemous. Can I admit to you that a little voice in my head pshawed him? I mean, it was clearly a biter rather than a squeezer, but it wasn’t that big, and more importantly we know — don’t we, fellow city-dwellers? — from nature documentaries that Nature She Warns of Venom. Poisonous creatures are brightly coloured or otherwise flouncy and exhibitionist: I cite you the coral snake, the lion fish and the magnificently-named Pfeffer’s Flamboyant Cuttlefish, an animal which clearly spends far too much time in the dressing room. So it was clear to me that this well-camouflaged brownish-yellowish thing couldn’t really be properly poisonous. Nevertheless, it could at the least leave two unwelcome punctures in my shin, and might sting a bit, so I gave it, as did everyone else, a wide berth.

When we arrived back at Picinguaba I was chatting with my friend Peter, and I mentioned this snake to him. He seemed impressed, and reiterated the line that they were very dangerous. Pshaw, the doubting voice in my head went, but slightly less confidently so than it had previously. On the pretext of establishing whether it was a jararaca or a jararacuçu I got out my laptop and started to look it up online …

… and found that it’s a pit viper. That’s a genuinely rather venemous snake, and aggressive to boot. (Can snakes be anything to boot?) Indeed, of the jararacuçu — which, from the colouring and geographical distribution, I now think it was; however both are of the Bothrops genus — the first scholarly article I found said that it is “one of the most dreaded snakes of Brazil.” I appealed to reason. According to most online sources they grow over 2m in length, and this one couldn’t have been more than 50cm. So it couldn’t be that bad, could it? Peter demurred. The small ones, he claimed, are more aggressive and more likely to inject their entire venom reservoir. The pshaw-voice in my head was oddly muted by now. Perhaps, after all, this local knowledge should be respected.

So there you are, an exciting brush with death. Now how, I guess you are wondering, will Stuart manage to inject a little linguistics into this post?

Other than appreciating your little “inject” pun there — nice one — I have to answer: actually, in two ways. First a romp through the history of Brazilian Portuguese and the língua geral and then, time permitting, a chat about syllable templates and simplification.

We’ll be talking, as so often in my posts, about language contact and about dialect contact. Of course, languages and dialects (together referred to as “lects”) are — except in the most isolated communities (did you watch that Trudgill lecture in the previous post?) — always in contact with each other. But we’re interested in the exceptional circumstances of rapidly changing societies, or those which have undergone a sudden shift, where contact means more than just gentle osmosis between existing lects but the actual emergence of new forms. Where languages contact we generally call this creolization and the result a creole, where it is dialects koineization, and the result a koine. Although in general much ink is split in discussing the difference between a dialect and a language, for our purposes there is no need as the key difference is provided for us in the very concepts of creolization and koineization: the former are the processes of contact between two or more non-mutually understandable systems, the latter the processes of contact between mutually understandable systems.

Now those of you who have any exposure to Romance languages — and I’m guessing that’s most of you — may have noticed that jararacuçu and jararaca don’t seem very Latiny, and you would be right. They are both borrowings into Portuguese from the língua geral, and so have their origins, ultimately, in the indigenous Tupi languages. A full history of the emergence of this language would take up far too much of this post, and anyway the Wikipedia articles here and here are actually pretty good. But here is a precis of what is important to me:

Prior to the European invasions, a group of closely-related languages from a family known as Tupi-Guarani were spoken along the Brazilian coast. Amongst these one, known in English as Old Tupi, served as a lingua franca between the various tribes. As the member languages of Tupi-Guarani are reportedly fairly close to each other, it is likely that this language emerged through a koineization process — although that of an regional koine as opposed to a migrant koine (see Siegel 1985 for a discussion of the differences). The early Portuguese colonists learnt this language — giving rise to the língua geral (common tongue). It is a little unclear whether or not Old Tupi underwent extensive change in the creation of the língua geral. (In fact, it should be noted that there were two línguas gerais, that of the northern, Amazonian region, a descendent of which is still used as an indigenous língua franca today and that of the southern area centred (colonially) on São Paulo, which is now extinct and about which considerably less is known.)

At this time — the late 16th and the 17th century — Jesuit missionaries were largely in control of colonial education in Brazil. The Jesuits were an enlightened bunch — as far as medieval Catholics go — and learnt língua geral, translated works (especially, of course, prayer books), and encouraged it in general, with the result that, given the early colonists were overwhelmingly male and therefore mainly married indigenous women, it became effectively the primary language of the colonial settlements within a generation.

During this period some terms from língua geral were borrowed into Portuguese and, unsurprisingly, these included names for flora and fauna which were not native to Portugal. Jararaca and jararacuçu were two of these. We should note that simple borrowing of terms does not creolization make, and the situation described above is not one that would be considered typical of creolization; indeed if any creolization will have taken place it will have been within língua geral and not Portuguese.

Why is this? Well, the social aspect most strongly identified with creolization is imperfect language-learning: that is, the sub-optimal acquisition of languages either due to substantial adult language learning or due to insufficient exposure, or both. It is the structural effects of such which is generally presumed to give rise to the typological features common throughout creoles. A prototypical creolization scenario might be (African) slave-laboured plantations where newly-arrived slaves (who did not even necessarily have their native languages in common) needed to learn the language of the plantation owners. Portuguese, in this early colonial Brazil, does not appear to have been imposed upon the indigenous folk; rather their koine became the primary language of intercultural exchange and even of the colonists themselves: the bandeirantes — explorer–slavers who penetrated inland from the coastal colonies — were língua geral-speakers. As the colonists largely married indigenous women, and the transmission of língua geral to the next generation will therefore have been in the hands of existing speakers of Old Tupi, we cannot even presuppose extensive creolization within língua geral itself — although our ignorance of this language and the Old Tupi progenitor prohibit us from making too decisive declarations either way.

So the early site of language contact was not — as many may assume — an Afro-Portuguese creole. Until 1645, there appears to have been a relatively stable diglossic situation in Brazil. This does not prohibit the transfer of some Tupi forms into Portuguese, but this would have been through “admixture” rather than creolization.

So what of African influence? Initially there were relatively few African slaves traded into Brazil — the colonists having preferred to enslave the native population. This situation was later inverted as over the course of the next two centuries until final prohibition as late as 1888, Brazil traded a shocking 35% of all slaves ever traded. In 1600 there were just 14,000 Africans in Brazil, largely in plantations and mills. What is unclear is what language these slaves will have had to learn — whether Portuguese or língua geral. There is evidence, according to Holm, that Portuguese or a Portuguese creole was spoken in these plantations.

All had changed by the middle of the 18th century, due to the discovery of mineral wealth — and gold! — by the bandeirantes, largely located in the region west of São Paulo which became known as Minas Gerais — “general mines.” The gold wealth brought a new influx of colonists from Portugal — speaking European Portuguese (modified somewhat in the intervening centuries by some French influence) and unwilling to learn the língua geral. In 1759 the Jesuits were expelled from Brazil and, shortly afterwards, official tolerance of língua geral was retracted and the language soon banned. The enslavement of indigenous folk was on the wane — largely because many tribes had got wise to the Portuguese and retreated deeper into the forests — and so the new gold-hunting European Portuguese brought with them another surge in African slavery (by 1770 half the population were African slaves). All of this — a newly-imposed colonial language which itself was receiving a new European influence, the mix of African slave languages, the banned língua geral — creates a far more conducive situation to creolization than that which we have previously seen. Thus, if, as many scholars claim, vernacular Brazilian Portuguese had its origin in a creolized form (now decreolizing, that is, converging back with the higher-status norms), it is here, in Minas Gerais in the 18th century where the creolization will have taken place.

Now those of you still with me (hello? is this thing on?) and who keep alert may have pricked up your ears a bit at that last sentence — because Minas Gerais is the state from which my migrants to Brasília come, and this is not a coincidence. One of the overall theoretical thrusts of my thesis is simplification, which is postulated to occur in both language and dialect contact. The vernaculars from Minas, therefore, should show strongly creole-like simplifications — and they do. These vernaculars are now — not just in Brasília, but in general through media and migration — in contact with other, more “standard” Portuguese forms.

The more wide-spread contact that is going on throughout Brazil we might expect to follow a broadly-described process known as dialect levelling in which regional or stigmatized forms disappear from the speech community, in favour of less marked, so-called “standard” forms, and this indeed is documented. But in Brasília (or, more accurately, the surrounding satellite cities) we have a classic migrant koineization scenario: a mix of dialects from throughout Brazil, with no local forms, as the cities are newly built. In this situation, as in language contact, theorists predict simplification. The problem, however with this view is that so far it has always been conducted post hoc: an existing form already considered a koine is examined and those features that appear simplified highlighted.

This is, of course, partly because saying that simplification will occur does not tell you exactly where it will. It is impossible to predict that any particular form will simplify in a novel manner, simply that in comparing a koine either diachronically with its progenitors or synchronically with other similar-status dialects of the same language one would expect to see overall a higher level of simplicity in the koine.

However for my research I am following the idea that although we cannot predict that any given form will spontaneously simplify it seems a reasonable supposition that in a simplifying environment, where there are multiple available forms for a dialect to settle on, the emergent dialect will settle on the simplest. That is, the same forces which provide a pressure to what I shall call novel simplification will also provide a pressure to selective simplification.

And now, I hope, any of you left at the end of this post will understand why I am examining the dialects of Minas Gerais in particular, and not, say, those of the northeast. Brazilian vernaculars are perhaps too often amalgamated by academics into one single “popular Brazilian Portuguese” and then words are thrown around as the extent of creole-style restructuring that has occurred. Different levels may have occurred in different parts of the country: what is certain (to me, at least) is that both the history of the language and the typological features of mineiro speech provide strong support for it being creolic in origin. Correspondingly, in the koineization processes taking place over the last fifty years in Brasília, my hypothesis is that — regardless of whether we see any novel simplification (I shall be very pleased if I do — a possible candidate has already arisen, although I need far more data before I am confident enough to do anything more than dangle hints here), we certainly should see the widespread use of mineiro forms which one would otherwise (in dialect levelling) expect to be declining in influence.

To be continued: I meant to talk also in this post about the actual nature of simplification, particularly as exhibited by my snaky terms. However I have already probably gone on too long, so will save that for a future post.


Brazilians, it would appear, are more than passingly fond of football. Who knew?

I am now co-opted as a lifetime and devoted supporter of Botafogo, and last week watched them play, and lose on penalties to, Flamengo (boo, hiss). Football already features in my research to a certain extent, because one of the questions I ask informants is which team they support and then, if they respond positively, questions about how they are performing, who are the best players, etc. This is partly because it is a good topic to get people talking on, but also because it is useful for getting some third person plural forms (that is, sentences using nós, “we” and its conjugations) and, finally, the answer goes into the mix of variables which will be used to calculate indexes of regional identity (as discussed here).

But the games themselves are wonderful for linguistic observation — sadly, of course, totally unrecordable. Other than slews of such profanities as would make your little ears burn, what one observes a huge amount of is imperatives, and these are really interesting to me.

Imperatives are instructions or orders, such as Go!, Get in there! and, of course, Shoot! In English we just use the bare infinitive, but many languages have their own set of conjugations for the imperative. In Portuguese, these are derived from the subjunctive conjugations. What the subjunctive is, and when it is used is not too important — and as English barely has it, there is little to compare it with. All you need to know is that the subjunctive forms, and therefore their imperative counterparts, are considered “difficult.” People from the lower end of the socio-economic scale often do not have good command of these forms, and avoid using them, or simply use the present indicative — that is, the “normal” present tense.

Imperatives themselves I shall not be examining from the analytical viewpoint, due to the problems in eliciting sufficient data. In the relatively controlled and one-to-one environment of a sociolinguistic interview it is highly unlikely that many spontaneous imperatives will occur, and drawing them out by other means (such as for instance, asking “What would you say to X to get them to do Y?”) is, to my mind, too artificial. See my comments on a relevant paper here for more on this.

To return to the match: throughout the game there were passionate calls to the players to shoot, expressed (fortissimo) as chuta!. This recurred again and again, and caught my attention. It is clearly derived from a borrowing of “shoot” and descriptive grammars tend to attest that all new verbs in Brazilian Portuguese take the -ar conjugation. I later checked with a number of people, all of whom independently confirmed that there is, indeed, a verb chutar, “to shoot.” Now, in the -ar conjugation, the imperative form would be expected to be (orthographically) chute! I note “orthographically” because for the vast majority of Brazilians, word-final written e is pronounced like British “ee”, and this has a corresponding effect that a preceding t becomes “tch.” Thus chute is pronounced “shootchee” whereas chuta, the simple present tense form, is pronounced “shoota.” The difference in the articulation of the t is very useful because, in the fraught and noisy surroundings vowels can easily be missed. However, I was confident that the watchers were, indeed, largely using the present tense rather than the imperative form produce their directives.

I cannot draw on this observation with anything more than anecdotal force and, as any proper scientist will tell you, the plural of anecdote is not data. I am probably not even going to look at the grammatically-related subjunctive mood because similar problems of elicitation arise. However, overall, this shows that non-standard verb forms are still alive and well in Taguatinga, a fact that cheers me greatly.