Unconnected speech?

Unconnected, yes. I wonder: what do we mean when we talk about *connected* speech? We must mean that there is at least one other kind of speech, which is not connected, so that it makes good sense to talk about its connectedness at all. But I would very much like to know who uses it, unless we’re perhaps talking about the so-called one-word stage in child language development, when spoken utterances appear to consist of single words, or expecting speech to and from interlocutors who look like this:

“Klaatu... barada... nikto...”

Image © The Day the Earth Stood Still, 1951, via Wikipedia

Qualifying speech with the modifier connected also means that we somehow take “connected speech” as a special case of speech – or it wouldn’t need qualification by means of a dedicated adjective. This is the same kind of reasoning which identifies some people through the qualifier multilingual, thereby leaving it understood that there’s no need to identify in any special way whoever is not multilingual, because there’s nothing special about their lingualism. In the same way that monolingualism came to represent default lingualism, unconnected speech represents default speechiness. One language at a time is desirable linguistic behaviour, and so is one word at a time (whatever the word word might mean, incidentally, since nobody has ever come up with a satisfactory definition of what a “word” might be).

I wonder why. It could be that the only way we might hope to identify the words of a language is by looking at them (assuming, in turn, that we do know what “a language” might be, which is another big linguistic mystery). If you listen to a language you never heard before, chances are you’ll have serious trouble attempting to single out its words (assuming, in turn, that all spoken languages have words, which is yet another moot question). If you see a spoken language, you may have better luck. Printed representations of speech, for those languages which have them, may show spaces separating what in some of them we’ve come to call words. Others won’t, because speech and whatever we choose to call its components cannot be adequately represented in print. It’s like attempting to represent a landscape in speech. It’s like putting a girdle on things. A picture may be worth a thousand words, but pictures of words tell you very little about the thousand different ways they are pronounced, even for those languages which may share printed representations that you recognise.

Take my language students, who mostly come to me after years of traditional vocabulary + grammar language learning, where “vocabulary” means lists of words (for what “grammar” means here, see my next post). They keep insisting that speech forms like wanna and doesn’t, or j’sais pas and t’as vu, or fàchavor and tá bem, are “bad” language. They keep reminding me that even native speakers of their new languages tell them that they use their language “better”, because they learned it the “proper” all-words-in way, whereas natives tend to become “lazy” when speaking – more on which in a future post, too. And I might as well confess that some students thought better of having me as a teacher, given my tendency to attempt to wrestle pens and paper and books off their hands and concentrate on training speaking and listening. This for students who come to me because they, or their own language students, are unhappy about matters of intelligibility from and to users of their new languages.

I don’t blame them. In the textbooks that they were taught by, and taught to abide by, wannas and tá bens are either glossed over or treated in special chapters, whose titles include the phrase “connected speech” and which come after all the chapters dealing with speech forms which apparently need no special treatment and so must be the “real” speech forms. But how do you learn to understand and use a language by first spending chapters and years memorising and spelling out citation forms of visually unconnected words? To me, the disconnect between language teaching and language use is the problem: not that you say things like gonna and you’d and perhaps write them too, but that so many learners are not told that people say and write these things because this is how people speak their languages.

Next time, as promised, I’ll deal (sorry, I mean I will deal) with the “grammar” part of the traditional vocabulary + grammar language teaching methods.

© MCF 2012

Next post: Teaching *about* languages. Wednesday 10^th October 2012.

Vocal versatility and vocal fossilisation

Vocal versatility, described as the ability to make your vocal tract do whatever you want it to do, is usually discussed in connection with professional voice users. In contrast, vocal fossilisation, described as the inability to make your vocal tract move beyond what you’ve grown used to move it for, is usually discussed in connection with language learners. This makes it sound like vocal tract users neatly divide into distinct subspecies, skilled and unskilled, respectively. The point I wish to make here is that vocal versatility and vocal fossilisation are related, because as far as vocal uses are concerned, we’re all pros.

The first observation is that we all come equipped with the same vocal tract model. Since all languages are equally difficult to pronounce – or equally easy, if your outlook on life tends towards optimism – because each language has a signature sound to it, the second observation is that the way we sound relates to the uses to which we put our vocal equipment, rather than to the equipment itself. In the literature on language learning, the (mortifying) label fossilisation stands for ‘routine vocal behaviours’. From learners, as said. For some reason, the word doesn’t apply to petrified accent models that the corporate textbook industry continues churning out, as I’ve argued before.

Routine behaviours are automated and taken for granted to such a degree that you come to believe that they cannot be characterised as specific behaviours at all, and so that there is nothing that can be changed about them because they’ve never changed, as far as you can remember. But fossilised behaviours, vocal or otherwise, are in fact acquired behaviours. For language learning, the issue is then to identify the steps through which we all learned to condition our natural vocal versatility in order to sound proficiently fossilised in at least one language. We could also call this the ways in which we learned to speak with intelligible accents.

Let me try to explain what I mean with an analogy: dancing.

Image © Tannon Weber (Wikimedia Commons)

Getting our steps right involves training muscles and coordinating their movements to match specific rhythms. There is a very similar choreography going on in our vocal tract whenever we speak and, like actors and opera singers, we learners need a choreographer, whom we could also call our language teacher, to help us get our vocal movements right. Teaching you how to get things right doesn’t mean teaching you the technical jargon used to describe vocal tract actions, which is familiar to language teachers. You don’t need to know a third conditional by name either, in order to use it appropriately – an issue that I’ll address some other time. Teaching you means making you aware of what you do and what you can do, when you speak those languages you’re comfortable speaking, so that you become aware of what you need to do, in order to sound the way you want to sound in your new languages: you’ll need “a guided tour of your vocal tract”, and you can treat yourself to a preview of what this feels like in Chapter 5 of my book The Language of Language.

As with dancing lessons, the age at which you start your vocal training programme is irrelevant, and so is the alleged brain shutdown which is allegedly restricted to language learning. Learning means instructing your brain to work in ways that it hasn’t worked before. With competent guidance, and lots, and lots, and lots of practice, your brain will follow suit because that’s what brains do. One day you’ll wake up in the morning to find out that your vocal tract remembers things that you don’t remember teaching it to do, and that you had no idea it could do. But it could do them.

The training of your vocal skills through awareness of your vocal skills is routinely available to professional voice users. But unlike these professionals, who only need to give the impression that they can speak the languages that they’re speaking onstage, we amateurs learn languages because we need to speak them in real life. This is also why we need real-life guides to assist us in our learning: if we learn to samba and to speak from printed images, we’ll samba and speak like printed images.

The next post deals with a strange conception which, to my mind, could only have become a standard conception in matters of language teaching and learning if one assumes that language teaching and language learning proceed, by default, through printed materials.

© MCF 2012

Next post: Unconnected speech? Saturday 29^th September 2012.

Vocal intelligibility

Making yourself intelligible involves awareness that you may not be intelligible, just like making yourself presentable involves awareness that you may not be presentable. This kind of awareness arises from exposure to different people and different situations, that is, from exposure to different intelligibilities.

As children, we develop our linguistic skills largely unaware that we are making ourselves intelligible, in the sense that we would not be able to explain what we are doing in so many words (there is a very significant difference between what you do and awareness of what you do, whether you’re using a mobile phone or speaking a language, more on which in a coming post). Nevertheless, monitoring and constructing intelligibility is exactly what typical language acquisition involves: we progressively learn to attune our inbuilt speech production and speech reception equipment (our vocal tracts and our ears) to uses which satisfy the speech reception and speech production counterparts, respectively, of those around us.

The key factor here, to me, is “those around us”. When children eventually end up sounding like those around them, that is, when they end up making everyday linguistic sense to and from those around them, their acquisition process is deemed complete (or “perfect”, as some analysts might prefer). It seems to me that the same applies to language learners across the board, because you learn a language in order to use it, and using a language means making it work for and with those around you. Barring disorder, we are all intelligible to someone and someone is intelligible to us, which means that intelligibility is not a feature of the speaker, or of the listener, but of what both end up negotiating in order to make sense. Just like there are no “ideal” speakers, there are no “ideal” listeners either – something to which I’ll come back soon too.

Intelligibility is also a feature of the here and now, because speaking and listening are bound by real-life settings, in place and time. One of my multilingual friends, who uses English for work-related purposes, has developed fluent understanding of Texan English from his Texan business partners. But only in one-to-one situations. When two (or more) Texans meet in his presence, all hell breaks loose, as he describes it – and not just because they eventually start talking about football (N.B.: not “soccer”) teams and other Texan entities unknown to him. Besides their vocabulary, they also change their accent and their overall ways of expressing themselves in English. They do this not because they want to exclude my friend (though some of us may sometimes deliberately want to adopt similar strategies for this purpose), but because it’s only natural to switch among the different ways of making ourselves intelligible that we’ve learnt to navigate along our lives. We all do this, we all can do this – if we so wish. Perhaps monolingual speakers, of English and other languages, will have similar stories to share?

My friend could also learn to understand and produce Texan in-house vocal ways – if he so wished. Users’ wishes are the reason why I believe that sticking to the one-standard-fits-all policies which go on guiding production of traditional language teaching materials makes little sense. I’m not saying that we should strive to prepare as many teaching materials as there are varieties of languages: this is as unrealistic a goal as attempting to make sense of multilingualism through cumulative descriptions of the number and the combinations of particular languages involved in each multilingual setting, as I noted before. I’m saying that textbook standards are best used as guidelines for what learners actually need. In some cases, the book-prescribed accent may match the learners’ needs. In other cases, learners may end up becoming unintelligible, for their purposes, precisely because they were trained to reproduce intelligibility in varieties of their new languages which fail to serve the reasons why they decided to learn a new language in the first place. And maybe this is one of the reasons why so many of us routinely get bad press about our “non-native” uses: maybe we’re just being differently native?

Since there are no ideal language users, there can be no ideal language uses either, unless by “ideal” we mean ‘flexible’: accommodation to accent variability is the key to intelligible speech production and perception, as I’ve argued in a paper titled ‘Multilingual accents’. So why not start in the classroom, because we have to start somewhere and because classrooms are where this whole business of language learning starts for so many of us? The next post has some more to say about this.

© MCF 2012

Next post: Vocal versatility and vocal fossilisation. Wednesday 19^th September 2012.

“Good”, “standard”, and other intriguing language qualifiers

We can start discussion of this topic with a little survey: what does it mean, to you, to speak “good X”, or “standard X”, where X stands for the name of a language that you speak? And what does it mean, to you, to be a competent, or proficient, or good user of a language? You can also ask your friends to answer these questions, and have some fun collating the results.

You will have fun, I promise you. Attempting to describe qualifiers like “good” and “competent” in connection with uses and users of language is extremely entertaining, in that you can spend your whole life trying to find “the” answer to these questions. It’s not just that these labels have all come to mean the same: I can safely guess that your survey will show, for example, that good X means standard X and that both mean correct X, or that competent users of X are proper and/or native-like or even accentless users of it, or vice versa. It’s mostly that these labels are judgemental – just think of what their opposites mean, on which you can also conduct a revealing survey. To a linguist like me, judgement values about language are interesting as expressions of personal opinions, not as expressions of linguistic facts, which is what linguists busy themselves with.

In this spirit, I once suggested a project topic to my class of beginner linguistics students in Singapore, where they were to survey what Singaporeans understood by labels like good English and good Singlish. The former label was readily accepted as a viable survey question, but the latter drew baffled silence. Singlish is a native Singaporean language which, according to official Singaporean takes on the matter, is neither native nor a language: it’s just ‘bad English’, a statement which is about as accurate as stating that Principense, say, is ‘bad Portuguese’. The students were reacting to my apparent ignorance in attempting to collocate an adjective like “good” with something that is as inherently “bad” as Singlish. So I decided to speak some Singlish, and the students again stared blankly at me – those who did not burst out laughing, that is. “That is not Singlish!”, some of them finally giggled. “It is”, I insisted, “it’s bad Singlish.” I think I was able to drive my point home, because the discussion of their survey results on both questions turned out to be extremely interesting.

The thing is that some uses of language have become associated with prestige, another judgemental label which has nothing to do with linguistic facts, and thereby assumed as the only “proper” uses of language. This is why standardised varieties of different languages also became synonymous with the labels identifying those languages by name, sometimes in ways that users of those languages find it hard to recognise, let alone implement in their everyday life.

What users of X do use, that fails to meet “the” standard X, is thus dubbed bad X, or improper X, or accented X. Multilingual mixes, that I’ve addressed several times before, are a favoured target of language guardians. But monolingual uses are fair game too, whether in grammar, prosody or vocabulary. So-called “contracted” forms (another intriguing label to which I’ll come back soon), for example, like aren’t and they’re, are also bad language, and so is what many of us call “slang”, a word which we often use even without knowing exactly what it means (yet again), but to which we nevertheless attribute overall negative connotations. You can do another survey, to check out what it means to say “That’s slang”. But if you do, don’t tell your informants about this newly published book, titled precisely Slang. I haven’t had a chance to read it yet, but its subtitle, The people’s poetry, and a look inside seemed to me to show that Michael Adams agrees with my definition of what lingualism is all about: it’s about what people do with their languages.

Persuasions and practices based on ill-defined judgemental labels don’t help us understand what’s going on and what’s required in language learning, for children and adults alike. They merely create the illusion that the labellers know what they’re talking about, which is probably the reason why they go on impacting language education policies. The articles collected in Multilingual, Globalizing Asia. Implications for Policy and Education give an appreciation of current language policies, in multilingual Asia. And Rosina Lippi-Green’s book, English with an Accent. Language, Ideology, and Discrimination in the United States, explains the role played by policy makers, schooling and even Disney cartoons in perpetuating myths about language uses as tenets of what she calls “standard language ideology”.

In particular, such persuasions and practices have little to do with fostering linguistic intelligibility which, to me, is the end purpose of learning to socialise through learning languages. I’ll come back to this matter next time.

© MCF 2012

Next post: Vocal intelligibility. Saturday 8^th September 2012.

Teaching to the standard

Language courses are routinely identified by the name of the languages that they claim to teach you. Things like Advanced Course in Turkish or Learn Swahili in Three Weeks are common sights on textbook covers and internet sites.

We learners might then be excused for thinking that we are learning a language, when what we are in fact presented with is one particular variety of that language. What we call “languages” are about as invariant as what we call “human beings”, so course creators and sellers might in turn be excused for omitting mention of which variety their creations and products reflect. But we language learners might think it courteous, to say the least, to be informed about what exactly is contained in what we’re consuming, just like we also prefer to know what exactly is contained in that appealing chocolate bar at the convenience shop.

Chocolate bar standards, however, arise from bottom-up preferences dictated by consumer markets, whereas language standards differ in two respects: they are dictated top-down and they serve offer rather than demand. Language course contents do not stem from a belief that everyone will be equally well-served by the standard language varieties on standard offer either: rather, it simply is unprofitable, for writers and publishers, to provide language materials which are tailor-made to learners’ needs or, for schools, to change textbook adoption policies, which routinely involve the use of (read ‘being stuck with’) the same materials for several years, precisely because shorter-term adoption contracts are too expensive.

The limited offer geared to user needs in language teaching matches the limited offer in speech-language diagnostic and assessment tools, for monolinguals and multilinguals alike. Like speech-language therapists, language teachers may find themselves required (read ‘forced’) to work with language varieties which they themselves do not use outside of professional duties, and to assess them against standards which in addition may not serve their clients either.

I can give one example. Many years ago, I attended a French summer course in Pau, in the French Pyrénées-Atlantiques. There I met a few other Portuguese students, who spoke a different Portuguese dialect from mine. In particular, we pronounced our so-called “rolled-r” differently, as in the ‘rr’ spelling of my surname. Mine is a uvular articulation, at the back of the soft palate, theirs was an alveolar one, at the upper gum ridge, and we used our respective r’s in our French too, as we had done ever since we first learned to speak French. We soon found out, however, that by doing so I was being a good student, whereas they weren’t “putting in the required effort”. I wondered what kind of “effort” I was giving evidence of, since all of us were doing exactly the same thing, speaking French as we always had. The issue was that my French ‘r’ happened to match the standard Parisian one which was required as proof of “good” command of the language. The irony of it all was that their ‘r’ matched the mainstream Béarnais French accent, which was the one we heard around us.

Being required to learn a standard variety of a language is not an issue in itself: whatever the variety or varieties of our language(s) that we use outside of official circles, we all need to learn to navigate (some) standard of those languages. But it wouldn’t hurt to also learn that languages come in many standards, and that what people sometimes call “the” standard is just one of them.

I can’t remember whether our teachers at the French course spoke in their own accents or in the “good” one with one another and with us students, outside of the classroom – probably because everyone understood everyone else, when we were using the language to talk rather than to demonstrate classroom-bound linguistic skills, an issue I’ll come back to some other day. But I was constantly reminded of this episode in my later language teaching career, when, as a beginner teacher, I took it as my duty to comply with unwieldy textbooks and assessment materials on offer, and equally unwieldy students who, because they’d been brainwashed about “good” uses of language being “the” language, were persuaded that, say, Standard Lisbonese (or whatever you choose to call it) and Parisian French were in use, or should be, in places like Luanda and Liège, respectively, to where they were relocating after completing their language courses with me. 

The issue has nothing to do with murky concepts like “nativeness”. Béarnais accents are as native as Parisian ones, which makes one wonder what (certain) people might have in mind when they say that “competent” language learning means emulation of “native” proficiency. The issue has nothing to do with linguistic competence or intelligibility either – unless we wish to argue that native Béarnais French speakers should also put in some “effort” in order to sound “good”. The issue is, to me, a non-issue, because it stems, yet again, from an ingrained confusion about what terms like good, standard, competent, native, intelligible, and so on, might mean. The next couple of posts deal with these matters.

© MCF 2012

Next post: “Good”, “standard”, and other intriguing language qualifiers. Wednesday 29^th August 2012.