F is for Functions

20 08 2017

7th function of languageThere are not many novels whose theme is linguistics but the book I took to read on vacation is one of them. It’s called The Seventh Function of Language, and is by the French writer Laurent Binet (2015; English translation 2017). It’s a sort of whacky thriller that plays with the idea that the death of the French semiotician, linguist and literary theorist Roland Barthes (he was run over by a laundry van only hours after lunching with François Mitterand in 1980) was not an accident. It appears that Barthes had stumbled upon an as yet unidentified function of language – one so powerful that, in the wrong hands, it might wreak havoc.

In order to enlighten the lay reader, Binet recaps the six functions of language as identified by one of Barthes’ most important influences, the Russian-born linguist Roman Jakobson, and spelled out in a lecture Jakobson gave when assuming the presidency of the Linguistic Society of America in 1956.

These six functions map neatly on to each of the six dimensions of any speech event – the context, the addresser and addressee, the physical and psychological channel (or contact) between them, the language (or code), and the message itself. They are

  1. the referential function, i.e. the way language refers to the context, whether local or global, real or imagined, in which it is used – e.g. ‘It’s 35 degrees in the shade.’
  2. the emotive, or expressive function, i.e. the way that addressers encode their attitude, or their degree of commitment, to the message, e.g. ‘It’s too darned hot!’
  3. the conative function, where the focus is on the addressee, e.g. in the form of a command: ‘Why don’t you turn on the fan?’
  4. the phatic function, where language is being used to lubricate the channel of contact, irrespective of its content, as when we make small talk: ‘Hot enough for you?’
  5. the metalinguistic function, where language itself is the focus, as in ‘How do you say heat-wave in Swedish?’ and
  6. the poetic function, where language draws attention to itself – its form, style, and aesthetics – as in the playful use of rhyme in the line ‘the rain in Spain falls mainly on the plain.’ Or, more sublimely, the cadences of Shakespeare’s song:

Fear no more the heat o’ the sun,
Nor the furious winter’s rages…

Jakobson himself noted that ‘although we distinguish six basic aspects of language, we could, however, hardly find verbal messages that would fulfill only one function’ (1990, p. 73).  That is to say, any one utterance can encode different functions, just as one function can be realized by various linguistic means.

Laurent BinetThe Binet novel is a useful reminder as to how seminal a figure Jakobson was: arguably the most influential linguist of the 20th century.

Born in Moscow in 1896, he studied philology but, even at a young age, he was frustrated by the failure of linguistics to see beyond the ‘scattered parts’ of language, thereby ignoring how it functions as a whole. In 1920 he moved to Prague and helped form the ‘Prague Circle’ where he was able to pursue his interest in the way that the parts of language – specifically its phonemes – form an interconnected system, whereby the parts can only be described in relation to other parts. Because of this concern for the inherent systematicity of language, Jakobson aligns with the structuralist tradition dating back to Saussure. But it would be wrong to think of Jakobson’s linguistics as purely formal (in the American tradition of Fries and Chomsky) and that he disregarded meaning: his interest in the functions of language – a line of enquiry he further elaborated after moving to the US in 1939 – attests to his ‘bi-focal’ view of language. Indeed, as Waugh and Monville-Burston note, in the introduction to their edition of Jakobson’s works (1990, p. 14):

For the Prague Circle, functionalism and structuralism were inseparable. Jakobson himself described his theory of language as one in which function (language as a tool for communication) and structure (language as a lawful governed whole) are combined…: language is structured so as to be suitable for communication.

The pedagogical implications of this two-pronged view of language continue to reverberate – and to challenge teachers and course designers alike. How do you reconcile the fact that language is a tool for communication while at the same time it is a rule-governed system (of considerable intricacy and complexity)? The pendulum seems to swing both ways without ever finding a point of equilibrium.

Thus, for structuralist-influenced approaches, such as audiolingualism, the syllabus was unapologetically structural and the major focus of instruction was pattern practice – although it would misrepresent audiolingualism to say that it ignored communication entirely. Indeed, a key document in the audiolingual canon observes that ‘probably the best way to practice a foreign language is to use it in communicating with others. Thus, teachers should provide time for meaning-oriented practice’ (Krohn 1971, p. viii).

JakobsonOn the other hand, the communicative approach, in seeking to redress the prevailing structural bias by substituting a syllabus of functions or tasks, may have erred in the opposite direction. Besides, as Brumfit was one of the first to point out, (1978, p. 41), a functional syllabus simply replaces one set of discrete-items with another: ‘No inventory of language items can itself capture the essence of communication.’

The reversion to a grammatical syllabus that now drives most general English programs, although notionally ‘communicative’ in their allegiance, seems to have sent the pendulum swinging back again.

It is testimony to the greatness of Jakobson that he was able to bestride these two poles with enormous intellectual, cultural and linguistic authority. It’s only a pity that he had no advice to give us language teachers.

Meanwhile – is there a seventh function of language? Well, you will have to read Binet to find out!

References

Binet, L. (2017)The 7th Function of Language (translated by S. Taylor). London: Harvill Secker.

Brumfit, C.J. (1978) “‘Communicative” language teaching: an assessment’, in P. Strevens (ed.) In Honour of A.S. Hornby. Oxford: Oxford University Press.

Krohn, R. (1971) English sentence structure. Ann Arbor: University of Michigan Press.

Jakobson, R. (1990) On Language. Edited by Waugh, L. R. & M. Monville-Burston. Cambridge, MA: Harvard University Press.


Actions

Information

28 responses

20 08 2017
Steve

Going to have to buy the book… Well, both books, Binet and Jakobson

21 08 2017
Scott Thornbury

Good luck, Steve – although see Philip’s caveat below!

20 08 2017
Justin Willoughby

Thanks very much for the post Scott. It came at a great time. I am currently trying to put together my DELTA LSA 3 on Speaking, specifically discussion. I was reading Penny Ur’s “Discussions that work” and her definition of a successful one was everyone participates, everyone is motivated, and finally a variety of language is used in terms of subject-matter and communicative function. I was just about to look into communicative functions and here they are. So I guess a good discussion would be full of the six functions above. In your own book “How to Teach Speaking” you seem to agree with Penny Ur saying that a good speaking activity must be “maximally language productive”. I wonder then if it would be useful to make advanced students aware of the 6 communicative functions prior to a discussion and then focus on them afterwards by an analysis of the recorded discussion.

21 08 2017
Scott Thornbury

Thanks, Justin, for your comment. It’s worth pointing out that when a writer like Penny Ur talks about a ‘communicative function’ she’s probably got in mind – not Jakobson’s list of six ‘macro-functions’ – but the more finely-tuned list of items that make up the kinds of functional syllabus that we associate with the early days of the communicative approach. For example van Ek, whose ‘threshold level’ syllabus of notions and functions, first developed in 1975, was a widely used reference for syllabus design, lists ‘six main categories of verbal communication’ (although not exactly Jakobson’s six) which are then subdivided into a long list of ‘micro-functions’.

For example, the first main category is ‘imparting and seeking factual information’ (roughly analogous to Jakobson’s referential function), which is then subdivided into: identifying, reporting (including describing and narrating), correcting and asking. Van Ek’s third function – expressing and finding out emotional attitudes – has 19 subdivisions. . (Van Ek points out that ‘the list of functions is far from exhaustive’ and speculates that it may be impossible to draw up a definitive list).

The principle underlying the functional syllabus was, of course, to teach learners ‘how to do things with language’, hence a communicative activity is one which targets one or more of these ‘micro-functions’, rather than simply practising language for its own sake.

Axiomatic to this functional view was the understanding that ‘there is no simple equation between linguistic forms and communicative functions’ (Widdowson 1972), hence teaching the meaning of linguistic forms without teaching their ‘value’ in actual communication would utlmately short-change the learner.

21 08 2017
Derek Keever

hi Scott,

could you give the Van Ek reference?

many thanks

21 08 2017
Justin Willoughby

Hi Derek,

I just found it by googling “van ek threshold level 1990”. If you click on the link, you get the PDF. It’s free. Language functions are on page 27.

21 08 2017
Scott Thornbury

Sure… van Ek, J. (1975/1980) Threshold Level English Oxford: Pergamon. There is also a substantial extract from it in Brumfit, C.J. and Johnson, K. (eds) (1979) The communicative approach to language teaching. Oxford: Oxford University Press.

21 08 2017
Justin Willoughby

Thank you Scott. I will look at Van Ek’s list of micro-functions and search for the ones that would be most useful for a group of advanced learners in an informal discussion on Spanish politics. How would one go about finding the most frequent linguistic forms for performing established communicative micro-functions using corpus data?

21 08 2017
Justin Willoughby

There are recommended exponents for the functions in Van Ek’s threshold, but they did not come from corpus data.

21 08 2017
Scott Thornbury

Off-hand, I would say that the functions most useful in a political discussion would be those related to the ones subsumed under van Ek’s category ‘Expressing and finding out intellectual attitudes’, which include expressing agreement and disagreement, expressing whether something is considered possible or impossible, expressing how certain/uncertain one is of something, etc, and also perhaps the general category Expressing and finding out emotional attitudes, including expressing hope, expressing satisfaction, expressing fear or worry, enquiring about preference, and so on.

In terms of using corpora to find ways that functions are typically expressed, it would be difficult, since I know of no corpus that is tagged for functions or speech acts. Most corpora are tagged for text type, language variety, mode (i.e. speaking or writing) and so on. You would need to find examples of people engaged in political discussion, and then comb through these looking for ways in which they typically put across their point of view. In fact the COCA corpus (i.e. Corpus of Contemporary American English) does include a lot of transcripts from broadcast talk shows. It might be interesting to make a sub-corpus of these texts and then do word searches (e.g. opinion, agree, I think, etc) to find associated functional exponents.

21 08 2017
Derek

Thanks for the link– found the pdf online. And got ‘On Language’ from the 2nd floor of the UOA library (a place where you may have spent some time?) Jakobson’s style of writing is interesting and compelling, and what a fascinating life story.

20 08 2017
Marisa Constantinides

Wonderful and highly fitting post paying tribute to a great linguist – was a pleasure to read.

21 08 2017
Scott Thornbury

Thanks, Marisa – I’m glad you enjoyed it. I enjoyed researching it!

20 08 2017
Bruno Andrade

I am just studying this topic for a professor position at the Federal University in Rio and some points came to my attention:

A functional grammar aims to match forms to function and meaning in context;

Functions are realized in a variety of ways according to the communicative effect desired;

Language can be seen in a discursive perspective in that each linguistic element is seen not in isolation but in relation to others, since it has potential to realise different functions;

The regular patterns of different kinds that can be distinguished reflect the uses which a language serves: the structural patterns known as ‘declarative’, ‘ interrogative’ and ‘imperative’ serve the purposes of expressing a multitude of types of social behavior.

Now I have to find this book! Thanks for sharing

21 08 2017
Scott Thornbury

Thanks, Bruno – and good luck with your career path!

All good points, and, to continue the quote from Widdowson (1972) I included in my response to Justin (above), he goes on to say ‘one linguistic form can fulfil a variety of communicative functions, and one function can be fulfilled by variety of linguistic forms.’

Of course, for teaching purposes, a functional approach has to make a selection of the most likely ways of fulfilling a particular function – but in the early days of the functional-notional syllabus, i.e. in the pre-corpus era – quite a lot of guesswork went into deciding what were the most typical or most useful exponents of different functions. I remember we used to spend a lot of time teaching the request form would you mind -ing? only to discover later on that it is relatively infrequent in spoken language (less than 1 occurrence every million words in the COCA corpus, for example).

20 08 2017
philipjkerr

As a novel, ‘The Seventh Function of Language’ is not, I think, terribly good … too clever-clever by half. A better novel, I think, is Binet’s earlier ‘HHhH’.

21 08 2017
Scott Thornbury

Thanks, Philip. Note that nowhere did I actually say I had enjoyed the novel 😉 But I may give HHhH a go, on your recommendation.

20 08 2017
shahram

An informative post!

The structural syllabus is similar to the functional syllabus in that both take a synthetic approach to teaching the grammatical and functional features of language. It is assumed that once the discrete functions are learned, learners can function in the language communicatively. This view of language learning is reductionist because it ignores the complex nature of communication.

In addition, the assumption that there is a direct relationship between forms and functions is a myth. Language learners are involved in an ongoing complex discourse process (including the physical context, the context created by the participants and many other variables) whereby they act upon discursive structures and negotiate for meaning and functions. Therefore, it would be wrong to think that there are a set of preselected functions mapping certain forms (Richards, 2015).

The aforementioned functions seem to almost match the ones proposed by Halliday, namely, instrumental, regulatory, interactional, personal, heuristic, imaginative, and informative.

21 08 2017
Scott Thornbury

Thanks, Shahram. With regard to the one-to-one relationship betwen form and function, see my earlier comments. And thanks for reminding me of Halliday’s ‘list’. As far as I can tell these were first formulated for his book Learning How to Mean (1975) as an attempt to describe early language development in children. As you say they sort of match Jakobson’s list, although Halliday splits the conative function into two (instrumental and regulatory) and adds an imaginative one, which only loosely matches Jakobson’s poetic function.

Interestingly, Halliday goes on to say ‘later on there is in fact a seventh [function] to be added to the list; but the initial hypothesis was that the seven function, although it is the one which is undoubtedly dominant in the adult’s use of language, and even more so in the adult image of what language is, is one which does not emerge in the life of a child until considerably after the others. This is the one that we can call the informative function of language, the “I’ve got something to tell you” function. The idea that language can be used as a means of communicating information to someone who does not already possess that information is a very sophisticated one which depends on the internalisation of a whole complex set of linguistic concepts that the young child does not possess’ (1975, p.21). This, of course, is what Jakobson calls the referential function.

22 08 2017
Luiz Otávio

Hi Shahram, yes, absolutely. But in fairness to the early proponents of communicative language teaching, most of them attempted to differentiate between content and process. Couldn’t we argue though that when students engage in semi-communicative work of a more functional/lexical nature, the ensuing interaction is perhaps more likely to call for the kinds of discourse strategies and negotiation of meaning that you’re referring to?

22 08 2017
shahram

Sure. I think that such concepts as process and content are used for discussion and differentiation. However, in a real classroom situation, things may turn out to be something different regardless of labels.

20 08 2017
Elka Todeva

Good morning, all.
Using Jakobson’s framework with my university students in Japan worked miracles. The phatic function was the one that made all the difference. I am so glad Scott wrote this post, a tribute to an intellectual titan — alas still little known to many teachers and … linguists. I hope people will go to Jakobson’s book and chap. 4 in particular (The Speech Event and the Functions of Language), as the phatic function is not exactly “small talk” as described above. It indicates engagement and interest in what the speaker is saying; it keeps the conversation going; it keeps the channel of communication open. The good news is that we do all this with very short, easy to learn and use phrases like “really?”, “when?”, “who with?”, “no way”, “are you kidding me?”. The phatic function adds a layer of fluency that listeners notice right away. All it took to get there with my students was to raise awareness around the importance of this function and some examples from their L1, e.g. hontoni? ah so desu-ka? It tuned out the students had hundreds of similar English phrases in their repertoire that had lain dormant until our class exploration, phrases that they had heard in all the movies they had been watching. Once the students got the idea of the phatic function, there was no stopping them. That dam wall broke (pardon the violent image) and they got a new type of fluency that surprised both them and their teachers. William Acton and I did some presentations on this experience in the 1990s. He has a publication in this area, I believe. Definitely worth checking with him. I talked a bit about Roman Jakobson and his phatic function at the BETA/IATEFLP/FIPLV Conference in June and I am considering doing an TESOL presentation on the topic. Such an easy function to master; so little pain with so much gain! Thanks, Scott for another great Sunday morning entry!
Thinking of all of you in Barcelona these last few days, ET

21 08 2017
Scott Thornbury

Thanks, Elka (I had a feeling you would like this post). And thanks for the corrective comment re the phatic function. It’s perhaps worth quoting in full what Jakobson had to say about this function:

‘There are messages primarily serving to establish, to prolong, or to discontinue communication, to check whether the channel works (“Hello, do you hear me?”), to attract the attention of the interlocutor or to confirm his continued attention (“Are you listening?” Or in Shakespearean diction, “Lend me your ears!” – and on the other end of the wire “Um-hum!”). This set for CONTACT, or in Malinowski’s terms phatic function (1953), may be displayed by a profuse exchange of ritualised formulas, by entire dialogues with the mere purport of prolonging communication… It is also the first verbal function acquired by infants; they are prone to communicate before being able to send or receive informative communication ‘ (p. 75).

(This last point echo’s Halliday’s claim that the referential or informative function is late-acquired).

By the way, do you realise that you and I have in common with Jakobson the fact that we all taught at The New School in New York?! Jakobson was one of the many European intellectuals fleeing Nazism (Hannah Arendt was another) who, in 1939, was offered a safe haven at the ‘University of Exile’ under the auspices of The New School for Social Research. Hard to imagine such generosity occuring uner the present political climate.

21 08 2017
Patrick

Ah, Roland Barthes…

brings back mixed and heady memories of the late 70’s lolling around in a Brixton “housing co-op” whiling away my gap year (half-decade) – a melange of counter culture, Barthes, Derrida, Foucault, and even Saussure got in there.

I think, on reflection, that a lot of it was/is deliberately obscure or just badly-written or -translated. But “semiology” was definitely “in”.

Fast forward to last month – Bob Hodge’s “Social Semiotics for a Complex World” (2017) was an entertaining distraction from the day job of writing courses and curricula!

21 08 2017
Scott Thornbury

Thanks, Patrick. Given your memories, you might enjoy the Binet novel, in which Derrida and Foucault have walk-on parts, as do Julia Kristeva, Gayatri Spivak, Paul de Man and even Noam Chomsky.

21 08 2017
patrick

Thanks Scott.

A while ago I was looking for material to help me relate Functional Grammar to the classroom, the 2 books that I found most useful are probably Graham Lock’s “Functional English Grammar: An Introduction for Second Language Teachers” and Jones & Lock “Functional Grammar in the ESL Classroom”.

21 08 2017
Scott Thornbury

Thanks Patrick. There are also some good books that came out of the Australian mainstream education sector, such as How Texts Work, by B. Derewianka, and English Grammar: A functional approach by J. Collerson, both publshed by the Primary English Teaching Association, NSW, in the 1990s. Note that these books, and the Lock one you mentioned, subscribe to Halliday’s systemic functional theory of grammar, which attempts to explain all language use in terms of three ‘mega functions’: referential meaning, interpersonal meaning and textual meaning.This approach seems to have widest applicability in the teaching of literacy.

22 08 2017
Patrick

Thanks for the tip Scott, and you’re absolutely right, “FG in the ESL Classroom” is geared to reading and writing, I’ve found it very useful, and Scott Thornbury gets namechecked in it too.

In my working context Young Learners have no exposure at all to spoken English outside of the classroom (including tv, internet etc), and often very little inside of the classroom, so texts and related activities really are crucial to their English language development.

I think this is an area that is often forgotten or marginalised in CLT (though of course there’s no reason in principle why it should be).

Re the discussion further up the thread about the Threshold Level. This is for my money the most useful document that a teacher can possess, along with the CEFR (I know some people will disagree violently!), and it’s fantastic that they are free.

Van Ek also produced “The Threshold Level for Modern Language in Schools” for the Council of Europe in 1976, not free, which might be useful for teachers of younger learners.

I’m sure the language examples are not corpus-based; I enjoy John Trim’s wry comment that a mature speaker has a larger corpus between their ears! But that’s just a matter of time.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s