A quick update

I realize I have been remiss in updating this since my last post: the reason is simply that not a great deal has happened that is worthy of reporting since then, and the introductory sections on koineization and simplification are taking place behind the scenes. But I hate to think of you all hanging around at the water cooler with nothing to discuss, so here’s a quick update.

This has been a frustrating couple of weeks. I gained a number of agreements to make recordings over the Christmas/New Year period, and deliberately didn’t follow them up then, as people are busy with their holidays. Since the New Year, then, I have been trying to reel in those who agreed to make recordings, and with an annoyingly low success rate.

I have made one recording, with a guy who for the sake of anonymity we shall call Mighty Mouse.[1] In fact, even this was in one sense a failure, because during the course of the recording he gave me information that disqualified him form being an informant: he had lived in the state of Goiás for six years. The recording session was also conducted in a place of his choosing, which was too noisy for really good quality sound. However, I am taking an optimistic view of this session, because as a dry run I found it very useful, and certain problems which arose from it I have learnt from and will correct — largely around the language itself. My interview schedule was based upon that of Sali Tagliamonte, published as an online appendix to Tagliamonte 2006, available here, with additional input from  Llamas 1999. I had my print-outs of these with me, along with my notes on them, but was relying on my ability to translate on the fly. This proved to be an unnecessary distraction, and the recording is embarrassingly full of “ums” and “ahs” from me. So since then I have spent some — rather tedious — time translating the entire schedule, with the exception of Tagliamonte’s question concerning whether the interviewee has ever had “any problems with a bear.”

A second potential informant, Mightier Mouse, agreed to make a recording, set a date, and then did not turn up. He has subsequently agreed to another date and time, which is tomorrow. Hopefully he will make it this time. The 84-year-old guy in my apartment block, Jebediah, has agreed, but I have not seen him around for a while. And several other potentials are proving elusive to pin down, and a couple simply not responding.

This is still early days, of course, and I still know few people well, but it is worrying me. During the interview with Mighty Mouse I asked him to compare Taguatinga with his home town of Patos de Minas, and he observed that Taguatinga is “muito frio” and there is no denying this. I need to be more active and inventive in extending my networks, and to this end I have — reluctantly — decided to go to church.

Those readers who know me personally will know that this is somewhat contrary to my personal belief structure, let alone morality, but there is no denying that it will be an ideal way to expand my contacts, particularly in the older age group. I have to be careful in selecting which church to attend — the Catholic is definitely out, partly for class reasons (Catholics here tend to be quite middle class) and partly because … well, best not get me started on that. But of the other churches, I’m going to try and find a smallish one, but one which looks community-oriented rather than just Sunday Prayers Plus. Cycling round the local park last Saturday, I observed a guy from the local baptist church handing out free bottles of water (along with pamphlets, of course), and so if he is there next Saturday I will stop and chat with him.

Any advice on different denominations and their tendencies towards or against community-forming activities would be gratefully received. I understand they all worship an invisible sky fairy: beyond this, and the intricacies of the C14th poverty-of-Christ debate as set out in The Name of the Rose, I am ignorant of the whole shebang.

[1] Informants are guaranteed anonymity in sociolinguistic studies and so even potential informants will be pseudonymized in this blog. [Back up]

An informant, an enclave, and a travesty

Happy New Year to you all. Given the Winterval celebrations, I was uncertain as to whether I would have a great deal to report in this post, but there are enough points on my general themes to warrant a post. Following what seems to be standard practice, I cover what I’ve been up to, some observations on the social context of Taguatinga/Brasília, and then sneak in some linguistics via an apparently innocent anecdote.

The process of network development and informant recruitment has been slowed by the festivities. However, rather surprisingly, after a shamefully sober New Year’s eve itself, I went for a smug hangover-free cycle ride the following morning (there is, apparently, a morning of New Year’s day — who knew?) and when I got rained off into a little kiosk got chatting with the girl who worked there. She turned out to be the daughter of mineiros, and was willing to be an informant. She also mentioned that her mother was from a pretty rural part of the Minas interior and will ask her if she is happy to help as well. So one or two new potential informants, much to my surprise.

A word here on the “potential” of “potential informants.” So far, I have received an indication of a willingness to be recorded from six people. However, I have yet to make any recordings. This is fairly understandable: the people I am meeting I am still doing so in very public places, and usually when they are working, so this is wholly unsuitable for recording. As such, although six would appear to be a good number for just a couple of weeks here, a couple of caveats are necessary. Firstly, I have not checked whether all of them actually meet my all of my requirements: the aforementioned mother, for instance, may not fall within one of my age bands. More seriously, an indication of a willingness to be recorded, however genuine, is not equivalent to a guarantee of recording. The complications of arranging an interview, finding a suitable place, and negotiating the somewhat relaxed Brazilian attitude towards timekeeping mean that I would be foolish to take it for granted that I have any of these in the bag. I have deliberately not made any active requests for interviews over the festive season: now that it is over, I shall start to do so.

I had been concerned before arriving here that I would be hamstrung in finding a suitable place for recording by the sheer heat in Brazil, meaning that most rooms would either have a fan or air conditioning operating, causing continual low levels of background noise. A great relief I had in the first week was the discovery that, as Brasília is on a quite elevated plateau, the temperature drops fairly rapidly once the sun has gone in, and in the evening and at nights air conditioning or fans are rarely necessary. Against this, however, is the problem of rain, which has been pretty much continuous for the last week. Finding a place free of the drumming of tropical downfall has become more important than a cool-but-aircon-less room.

The young lady mentioned above was unable to recall for me the exact town from which her mother comes but, interestingly, was certain that it was not Patos de Minas. Why is this interesting? Because almost all of my other potential informants, and a number of ruled-out but still mineiro people with whom I have spoken, are from Patos. My total number of contacts is still far too low to go around making grandiose generalizations, but I am beginning to suspect that I have landed in the middle of a patoense enclave.

If this is the case, this is quite interesting. I have, in previous posts, mentioned the original study that largely motivated this project, Bortoni-Ricardo 1985. Although I am not lacking in doubts about some aspects of it, Bortoni’s work is largely interesting precisely because it focusses upon the role of networks in linguistic change, and takes metrics of network density in the migrated community as central to her analysis. Key to this is the fact that her informants all came from the same general location within Minas Gerais, and had migrated to the same area (the then quite small town of Brazlândia) and she could thus examine whether the rural-style dense network connections were retained in the urban setting; whether and how new more typically urban connections were established; and how the urbanization of their networks impacted (or not) upon changes in her informants’ speech.

That the original migrants to Brasília should tend to bunch together is intuitively unsurprising — people tend to move to recommended areas, or those in which they are likely to know people. I started last year looking to see if I could find some more detailed analyses of migration patterns in the Brazilian context — see, for example, Wilkening, Pinto & Pastore 1968 — but largely drew a blank. The apparent presence, 50 years after the founding of the city, of a large number of migrants from Patos in a neighbourhood of North Taguatinga (and I have no evidence how far this apparent enclave spreads) certainly cannot be ignored for the purposes of my research. Although I have yet to confirm to my satisfaction that there is such an enclave, some adjustments to the research design seem in order. I shall thus extend the number of questions intended to draw out strength of network ties and how they are constructed, specifically focussing on effect of region-of-origin and include some questions concerning Patos, and attitudes to it, whether or not the interviewee is themselves patoense. I shall also be reinvigorating my search for studies of patterns of migration to Brasília and also be seeing if I can find any specific literature upon patoense speech. Suggestions for either of these will be gratefully received.

Finally, the sneaky linguistics part of the entry. This started off a bit light and silly, for the New Year, and also does not even directly illustrate one of the features I shall be examining. In fact, it almost got written up in the other place, but as I started writing, I found myself referencing some more general theoretical issues which I am thinking over, and so it ended up here.

We start from the travesty of this entry’s title, as shown below.

Bob Sponja Pizzeria

Now, other than the fact that Nickelodeon’s licensing department may well be interested in this institution, and the frankly astonishing suggestion that the gentleman concerned has anything to with pizza — it’s Crabby Patties, duh! — this sign reminded me of an issue which has long bounced around inside by head, represents an issue which causes me genuine problems in articulating complex ideas in Portuguese. The general problem is nominal compounds, and the specific is the total failure of Romance languages to be able to render the subtle humour of “SpongeBob.”

NB: The use of the k-word above makes me painfully aware that I have stormed ahead with three posts looking at quite detailed aspects of my research, without setting out some of its general principles. In brief, for now, koineization is the term linguists use for the formation of new dialects (in certain social contexts). After this post, I shall make my priority creating some pages actually setting out the basic pillars of my research — koineization and simplification.

Nominal compounds are compounds resulting in a noun-type construction, and I am particularly interested here in noun-noun compounds: where one noun is used to modify another. At this point I should admit that this is far from my area of expertise, and that my knowledge and understanding of compounds derives as much Sanskrit as from formal linguistics. The type of compound I am interested in here is what the Sanskrit grammarians called karmadhārya — the descriptive compound. This is where the right-most noun in the compound supplies the hypernym, the actual category of what is being referenced, and it is modified/qualified by the preceding noun. Thus, for instance, in English we have the word catfish, the cat-like fish. Importantly, we can modify this further in a recursive fashion with more nouns, for instance baby catfish. In German this kind of compounding is even more productive than in English and in classical Sanskrit it reached levels of “artistry” that could only be considered extreme.

The Sanskrit scholars — and I assume modern semanticists, although I know not their terminology — differentiated between karmadhārya, where the antecedents quantify or modify the final term by virtue of restriction, and tatpuruṣa in which the antecedents stand in a case relationship to the final term. Roughly speaking, we can gloss all karmadhārya with “the X that is a Y,” applied recursively. Thus baby catfish is [[the fish that is a cat] that is a baby]. Tatpuruṣa, however, require a more complex, case-related relationship to gloss them: workhouse is the house WHERE there is work (locative), violin teacher is the teacher OF the violin (genitive or instrumental). What I always find extraordinary about this kind of compounding is the productivity with which we apply it, and the accuracy with which we create and parse novel compounds for the correct type of relationship.

Note we are interested in the syntactic order here, and not whether the orthographic result is spaced, hypenated, or unspaced. In English, such orthographic considerations generally reflect the extent to which the construction is novel or established as word (“lexicalized”). Generally, we only fully concatenate where the term has been lexicalized and thus may have developed its own range of reference beyond the simple sum of the parts — this means that most fully concatenated words are often karmadhārya (although not always, cf workhouse above). In German, of course, there is far greater freedom to orthographically concatenate the words. In Sanskrit, in the original Devanāgarī, words were not separated out anyway; in Romanized representations we tend to concatenate the terms of a compound, largely because in Sanskrit grammar all bar the rightmost element remain uninflected.

Now here I get very speculative, but I have a hypothesis that the reason we find SpongeBob an amusing name is that, very subtly, it breaks the rules of karmadhārya by placing a proper noun as the final term, but with a non-detachable qualifier. Proper nouns, we intuit, refer directly and largely uniquely (whether or not this is actually the case, I leave to Frege and his successors to argue over). Where there is a need to qualify because proper nouns are not actually truly unique we only need do so until the correct reference is established, and from then on omit the qualifier. For instance, if it were necessary, I could distinguish between two of the many people called Andy in my life by referring to them as “Programmer Andy” and “Engineer Andy” (actually, they’re “Big Andy” and “Little Andy,” but I wanted to illustrate that they can be differentiated by nouns). Once I had established within a discourse established that I was referring to “Programmer Andy” I could then proceed to use “Andy” without the qualifier. So, SpongeBob amuses us with the incongruity of orthographically indicating a mandatory qualifier, but one in which the final term should have a range of one — and therefore not require qualifying.

Thus at least is my pet theory. What certainly matters here is that not all languages permit this type of right-headed noun-noun compound construction and, specifically, Romance languages including Portuguese do not. Compare SpongeBob’s name in the following languages:

French Bob l’Eponge
German SpongeBob Schwammkopf
Portuguese & Spanish Bob Esponja

As you can see, German happily supports right-headed compounding. The German retains the entirety of “SpongeBob,” presumably opaquely, because the second word Schwammkopf means “spongehead.” Note that, according to my theory, even were SpongeBob to have been rendered as SchwammBob this would not necessarily have been as amusing to a German speaker as SpongeBob is in English because, as mentioned above, the rules of orthographic representation of compounds in German permit concatenation in non-mandatory contexts. The real travesty though lies in Romance languages. The grammars of these do not support this kind of right-headed compounding, and thus completely fail to be able to even approximate the subtle humour of “SpongeBob,” rendering it simply as a name, “Bob (the) Sponge.”

What on earth does any of this have to do with my research? Whether or not my analysis of the humour content of SpongeBob is true, it is clear that, on a personal level, somewhere in my grasp of Portuguese is a half-extant expectation that you ought to be able to create right-headed compounds in the language and, unsurprisingly, I often find myself half-constructing these, and then recasting them at the last moment.

This is due to how we learn second languages, via the development of an “interlanguage” — the term given in linguistics to the idiosyncratic linguistic system a learner uses before they are fully fluent in a language. An interlanguage usually has certain distinctive features — over-generalizaton (that is, using systematic rules in cases where they do not apply, such as when a learner of English says “I goed”), avoidance strategies, simplification and — as here for me — transfer of grammatical structures from the mother tongue.

Ideas are about interlanguage are important, because interlanguage-based learning and pidgin/creole formation are closely related: as a gross generalization, the process(es) known as “pidginization” essentially reflect incomplete, overly rapid or otherwise reduced interlanguage development. And ideas about pidgin/creole formation, and in particular simplification, inform my research into dialect contact.

One aspect I shall be examining is whether we see grammatical simplicity emerging in the New Town scenario. A definition of grammatical simplicity is going to have to be a topic of a subsequent post — it is far from an easy concept to define, let alone measure; but for now it is enough to observe that whereas in ordinary dialect change we might expect socially marked grammatical forms to be slowly eradicated from a dialect, in the case of migrant koines, such as are created in of New Towns (built from scratch and therefore populated entirely by immigrants or their descendants), we may see a competing pressure towards selection of the simplest grammar. Specifically, in Taguatinga, rural grammars from the interior of Minas Gerais, being simpler (for now, let us accept this on an intuitive level — they have less conjugations and less requirement for agreement), may be being preserved in the speech of subsequent generations. This is one of the major points of investigation of my studies, and so SpongeBob — and his pitiful translation — have served, at least, to help me work my way round to a discussion of it.

Mineiro identity and first informants

I intend to usually update this blog on Sundays. This posting is therefore somewhat late, caused by (a) little of consequence happening until yesterday, (b) the fact that the topic upon which I started writing the linguistics content turned out to be extremely lengthy, and will need to be split up between a number of other posts, and (c) having lost two days of my life to Brazilian bureaucracy in order to obtain a temporary ID card. Apologies to all who were expecting to see this on Sunday.

I have now been in Taguatinga for ten days, renting a small room in a small apartment in Taguatinga North, CNB 13, and have spent the time getting to know the area, and starting to set up some networks. As my research will focus on the speech of migrants from Minas Gerais and their children, I have also been watching out for signs of mineiro identity in the town.

There does seem to be a strong mineiro identity still here. A great number of the cafés and por-kilo restaurants identify themselves as selling mineiro food[1] or incorporate the words mineiro or related terms into their names, and I have also seen a number that identify as caipira. As I start to try and form some networks of contacts, I have been visiting these places a fair amount. Two of them in particular bear noting.

Patureba restaurantThe first is a buffet restaurant called “Patureba.” The word patureba is a demonym for people originating from Patos de Minas, a town in the triângulo mineiro (see map). However, the “correct” demonym is patoense; patureba is considered derogatory. Similarly, a small café-bar near my flat is called o chapeau mineiro, “The Mineiro Hat.” Both the owner of the café and its sign sport a hat of a distinctive shape, similar to that associated (in my mind, at least) with cigar-toting Texan oilmen. The active branding of such institutions with perjorative indicators of rurality suggest that the somewhat yokelish stereotype of the rural mineiro[2] has been, not without a certain humour, repossessed by the mineiro community as markers of identity.

The behatted owner of the latter institution in fact also comes from Patos de Minas and, more importantly, agreed yesterday to be recorded, getting me my first informant within ten days of arriving in Taguatinga. Shortly after this, my landlord informed me that an elderly gentleman who lives on the same floor, and who I’ve chatted to a few times, was himself mineiro and happy to be recorded. This gentleman is actually well above my top age band in my design: he is 84 and was therefore at least 34 when Brasília was founded, and corresponds to the “adult” group in Bortoni-Ricardo 1985.[3] Although I still think it unlikely that I could gather enough 75+ informants to correspond to Bortoni-Ricardo’s adult group, I shall certainly be consulting with him, and may well use him as an exemplar or reference point.

A third route to informants also opened up yesterday — although at something of a price. One of the first things I did here was join a gym, which here are quite sociable institutions. Taguatinga is hardly touristy, and one can imagine very few foreigners passing through for long enough to join a gym, and so the novelty of a foreigner in their gym (coupled, no doubt, with my innate charm) meant that I quickly got chatting with a number of the staff and regulars there, and have been invited to join the staff for their weekly football session. There are, apparently, three mineiros on the staff, so hopefully at least one of them will fall within my age bands. But football! Oh dear Lord, the humiliations I am prepared to put myself through for my work. This is Brazil — land of Pele and Ronaldo and Ronaldinho and Kaká — and I? I have not played football since I was 16, and was not spectacularly good at it then. (I wasn’t terrible: I seem to recall I made a half-decent defender largely by failing to draw fine — or indeed any — distinctions between rugby and football tackling techniques.) Next Wednesday, in the name of firming up my network and meeting new people, I shall expose my sporting failure to members of the most footie-mad nation on the planet.

There is little more to report, other than a mildly interesting shift in my own speech: I have changed my pronunciation of coda r to the “strong” r, in which—

An observant reader interrupts: Hang on a moment. “Little more to report,” indeed. You did this last time. You mentioned a small final point, which turned out to involve a linguistic discussion of many paragraphs. That, coupled with the introductory comment indicating most posts are going to have linguistic content leads me to wonder—

Me: Yup. Bang on. Buckle up, you’re in for a discussion of Brazilian Portuguese rhoticity.

Standard Brazilian Portuguese is generally described as having two rs — the “weak r” and the “strong r.” These contrast in words such as caro/carro (“expensive”/“car”). The weak r here is pronounced something like Anglo-English r (that is, English of England). The strong r has a variety of realizations, all of the type that linguists call uvular (produced at the back of the mouth). The most familiar of these is something like the r of French.

This distinction only holds inside words, between vowels. At the start of words r is always strong, and as the second consonant of a cluster, such as tr or br it is always weak. Where it gets interesting, though, is in the position linguists call “coda” — at the end of a syllable (or word).

Now, it may be necessary at this point to let other Anglo-English speakers understand that it is possible to pronounce an r here. People from England are generally what is referred to as non-rhotic here: in, for instance, the word “car”, we simply say it with a long a; we do not pronounce an actual r of any type. But if you are English, think of the Scottish or American pronunciations, and you’ll realize it is possible to articulate an r at the end of a syllable. We just don’t.

The Brazilians have a huge range of pronunciations here. They, too, can drop it entirely (particularly common at the end of words, and even more likely when that word is an infinitive). They can also produce either the weak or the strong r in any of the variants above. There are also regional pronunciations including a retroflex articulation (made with the tongue curled right back in the mouth), which is associated with rurality and highly stigmatized.

The pronunciation of coda r is going to be — did you guess? — one of the variables I examine. My own pronunciation, as I say, has shifted. Whereas previously I produced the English-style approximant I have been increasingly concerned that, unsurprisingly given I speak non-rhotic English, the r has just tended to disappear, without me really noticing I am dropping it. Whilst this is one of the pronunciations in Brazilian Portuguese, from the point of view of my own command of the language, I would prefer to avoid optional deletions such as this. This is not a judgement upon the social significance of such speech, but rather simply so that I am comfortable that I am not deleting in inappropriate contexts, or accidentally misrepresenting the word in my own lexicon.

What started through natural accommodation to my landlord — who has a strong coda r, and I speak with him more than anyone else at the moment — I have subsequently consciously taken up through deliberate monitoring of my speech, and shifted to producing the strong coda r. For me, of course, this is now marked, but that is partly the point: by changing to a strong r I am ensuring to my own satisfaction that I have the underlying structure of the word correct.

Plus, to be honest, I kinda like the way it sounds as well.

[1] Characterized by a preference for pork over beef and chicken, as well as a local delicacies such as pequi, with which I have had something of a run-in. [Back up]

[2] The stereotype of the rural mineiro is exemplified here (note the hat in the caricature). The joke translates (badly, or it is a bad joke) as:

Three paulistas wanting to impress a mineiro:

First paulista: — I have plenty of money, I’ll buy the Vale do Rio Doce [a mining company].

Second paulista: — I’m rich, I’ll buy Fiat Automobiles.

Third paulista: — I’m a millionaire, I’ll buy Usiminas [a steel producer].

And the three wait to see which the mineiro will speak to. He adjusts his hat on his head, takes a puff of his straw cigarette [tobacco hand-rolled in a corn husk], spits, swallows, pauses … and says:

— I’m not selling.

[Back up]

[3] Bortoni-Ricardo’s study was conducted in the early 1980s, and forms a major springboard for mine. She divided her informants in two groups: “adults” (aged 25 and up) and “youth.” Although she was largely concerned with the adult language change — how those in the adult group had “urbanized” their speech — I decided in my project design not to include an age band corresponding to her “adult” group simply because this would involve a minimum age of 75. [Back up]