Sunday 24 January 2010

On linguistic dreams

A correspondent writes to tell me of a linguistic dream he just had. As follows:

'I had a dream the other night with Significant Linguistic Content. It started out as the standard nightmare (mercifully infrequent these days) that I was teaching in a secondary school, as was the case long ago. But things improved and softened: the kiddies (11-year-olds, I’d say) were nice, and their music teacher, a fiftysomething German lady, came in and asked me would I mind if they practised the song that they were working on. I assented readily, and they sang a couple of verses of a song in German, about which I can remember nothing save that it was very sweetly done. I thought they deserved complimenting and encouraging, so I said to them “Ihr singet sehr gut auf Deutsch”.
I should explain that my German is far from good (I came to it very late), but they seemed to understand, and smiled and giggled pleasedly. I followed that up with “Wie hei├čt das Gesang?”.
Quick as a flash, the music teacher corrected me. “Der Gesang”, she said, entirely amicably; and I remember nothing after that.
Now, I was under the impression that German nouns that start with Ge- were uniformly neuter, so when I woke up I thought I’d better check it out in a dictionary; and, blow me, it is der Gesang. Now, what I can’t understand is how it’s possible to dream a person who knows the rules of a language better than one does oneself! Any thoughts?'

Not really - which is why a blog post might help. The only thing I can think of is that my correspondent must have encountered this correction before, and it impressed him at the time (as it did in his dream), and he's now forgotten all about it.

It's certainly possible to forget whole chunks of one's earlier linguistic life. I remember meeting an aphasic patient in his 70s who had lost his English very largely but who was able to produce some words in a foreign language. He himself denied he knew any of this language, and his wife had never heard him speak it, nor had they ever been to a country where it was spoken. But it eventually transpired he had lived in such a country for a while as a young child. And there was an interesting paper in the Journal of Child Language last year suggesting that some components of early childhood language memory can remain intact despite many years of disuse.

Sad to say, I don't recall ever having had a linguistically interesting dream. I can't be trying. But I'd like to hear of any others. I can't think of anything in the literature on this topic. Do phoneticians dream in accents? Do grammarians dream grammatically? Do lexicographers dream alphabetically? It's a whole new research domain: dreamlinguistics.

There's a thought. All those students who fell asleep in my lectures, over the years. Maybe they were doing research all the time.

And maybe that's what Chomsky's 'sleeping furiously' really meant.

Saturday 16 January 2010

On language and colic

A correspondent writes to ask whether singing to her baby will help when it's colicky.

The power of song over babies has long been recognized. The word lullaby has been in the English language since the Middle Ages - one of several, such as rockaby and hushaby, which show how generations of mothers and caretakers have helped their children fall asleep through music.

But music has greater power than this. As any adult knows, music has the power to engage all the emotions - from excitement to relaxation, from tears to laughter. Its power to calm is well recognized. The dramatist William Congreve summed it up in a famous quotation: 'Music hath charms to soothe a savage breast, to soften rocks, or bend a knotted oak'. But if the breast of adults, then why not the tummies of children?

Why does music have such a power over us? And why, in particular, does it have such an influence on tiny babies? A huge amount of research in infant perception over the past few decades has begun to reveal the answers. And it begins before birth, in the womb.

The womb is actually quite a noisy environment, and from around 30 weeks gestation the ears of the foetus are sufficiently well formed to enable it to hear what is going on. Indeed, the tiny little bones inside our ears, which transmit sound to our brain, are already fully formed by the time we are born. During some types of gynaecological examination, researchers have inserted a tiny microphone, called a hydrophone, into the uterus, enabling them to hear what the foetus can hear. And what the foetus hears is a great deal of background noise - the mother's heartbeat, the blood sloshing around the arteries, tummy and intestine rumbles, voices and loud noises from outside - and, above all, the mother's voice resonating through her tissues, bones, and fluids. The foetus is asleep a lot of the time, but when awake, its heart rate slows when the mother is speaking - the first evidence of a calming response.

It can't hear everything perfectly, of course. The effect is a bit like listening to a voice with cotton wool in our ears. The voice sounds distant and muffled. But there are certain things the foetus can hear very clearly. It can hear the intonation, or melody, of the mother's voice, and it can hear the loudness and rhythm of her speech. Sound and movement combine: when she laughs, the foetus can be seen to bounce around.

Once the baby is born, people have performed experiments which demonstrate just how much the foetus has heard. Researchers monitor the baby's heart-rate, the amount of cortisol in its saliva, the way it turns its head, the length of time it looks in a certain direction, or the rate at which it sucks on a special kind of nipple. The idea is that if a baby recognizes something, or is specially interested in something, then its heart-rate or saliva cortisol content will alter, or it will turn its head towards a stimulus, or it will look at a stimulus for longer, or it will suck faster.

And this is the sort of thing they've found. Newborn babies, even just a day old, prefer their mother's voice to that of a stranger. The show more interest when hearing their native language as opposed to a foreign language. And if the mother has told the foetus a particular story, during her pregnancy, they show a preference to that compared with an unfamiliar story.

The effect of music emerges too. One study played the same tune to a group of mothers every day throughout pregnancy; another group of mothers didn't hear the tune. When all the babies were born, the tune was played to them. The changes in heart-rate, movement, and general alertness of the 'musical' babies showed clearly that they recognized the tune. To check that it wasn't just a general response to music, the researchers played the babies a different tune, but they didn't react to it. Nor did they react to it when they heard the same tune played backwards!

There seems to be something special about the music of the voice. From the moment the baby is born, the mother - and other caretakers too - start talking to the baby in an unusual manner. That's what we call 'babytalk'. One of its most noticeable features is the way the voice ascends and descends throughout its whole pitch range - almost like singing in speech. And the exaggerated tones stay throughout the first year of life. The mother's voice is higher in pitch, and she speaks more slowly, when addressing the baby than when talking to others, and she is emotionally much more expressive. The effects can be clearly heard when playing simple games, such as peep-bo or round-and-round-the-garden. Games like this also draw attention to the importance of sensory reinforcement - sound, vision, touch, and movement all combine to create the maximum effect - 'and TICKLE him under there!!!'

Not surprisingly, then, the first features of the mother's language that the baby learns to reproduce are its intonation and rhythm. If we record babies' early vocalizations, at around a month or so of age, we cannot tell which language they are learning. Nor can we tell from their cooing or babbling. But at around 9 months the vocalizations start to sound 'shaped', and it's possible to distinguish babies who are learning English from those learning French from those learning Chinese, and so on. This is long before they learn any words, so what is it that we notice? The rhythm and intonation of the languages. The English baby is vocalizing with a 'tum-te-tum' rhythm. The French baby with a 'rat-a-tat-a-tat' rhythm. The Chinese baby with a sing-song rhythm. Why intonation and rhythm? It's no coincidence that these were the very features first perceived in the womb.

Melody, whether of speech or music, seems to be especially significant. Singing holds a special place in the emerging world of babies. They notice it. In one study, six-month-old infants were presented with pictures of their mother while she was singing and while she was speaking. They looked for longer at the singing one, and were less active while they did so, suggesting that they were paying more attention. Other studies show that infants can recognize melodies, even when they are presented at different pitch levels (sung higher or lower) or sung at different speeds. The melodic contour is the thing. Maternal singing, especially, is critical. It's different from normal singing, in that it is slower, higher in pitch, and emotionally more expressive - just like maternal speech.

Singing also simplifies vocal behaviour: words tend to be shorter, sounds are clearer and repeat more often, and they often rhyme. The value of repeated sounds and rhyme is well established, both for children's speech development and also later on in relation to reading. Repeated sounds are a major feature of babbling (babababa, dadadada); early words use them (mama, papa); and parents instinctively repeat syllables (doggie, bunny). Nursery rhymes work so well because they combine several effects - clear rhythm, repeated sounds, and rhyme. The effect of rhyme on babies can be measured from around 7 months. Primitive spontaneous singing can be observed in babies from around 9 to 12 months. When combined with vision, touch, and movement - as when telling a nursery rhyme while rocking a child in a cot or playing on a knee - the effect is extremely powerful. If the baby can see the singer, the effect is even more enhanced.

Maternal singing has a moderating effect on the emotional state of the baby. There are now many reports of music being used to calm sick babies. Music therapy is routine in many premature baby units. It seems to provide them with a safe and positive sensory experience. There is a bonus in that maternal singing seems to have a calming effect on the mothers too. Exploiting the melody and rhythm of song and speech would thus seem to be an ideal way of helping mothers to soothe their babies when need arises.

Sunday 10 January 2010

On the 800-word myth

A Sunday Times correspondent rang up last week to ask what I thought about the claim made by Jean Gross (described as the new UK 'communications czar') that 'the average teenager uses just 800 words in daily communication'. It was one of those waste-of-time interviews, where I spoke to the reporter for about 20 minutes, explaining how simplistic statements of that kind are rubbish, and what the linguistic realities are, and got one sentence in the report for my pains. Plus an ignoring of all the issues. The report was headed 'Youngsters are using just 800 words in everyday speech', as if this was a fact. I'm already receiving emails asking whether this is true, and I expect more as the week proceeds. So this post is to try to save a lot of time by summarizing the issues. In short: it isn't true, and I would call it the biggest load of chicken-droppings... except that I've already used that line in this blog [the post about the 'millionth word in English' claim - also, incidentally, listed as a fact in this Sunday Times report. Heigh ho... :( ]

Nobody has developed a satisfactory methodology for establishing the whole of someone's spoken lexicon. It isn't enough to take a sample of written material, such as counting words in blogs, as that is only going to be a partial reflection of speech. Few people have tried to record the whole of someone's spoken output in a day (e.g. by attaching a radio microphone to them and recording everything they say). When this was done - for example, in a study of young children in the Journal of Child Language some years ago - the word totals were in the thousands.

People always understimate vocabulary size. How many words do you know? Most people have no idea, or think it's just a few thousand. A few years ago, I read a report which said that the average size of a Sun reader's vocabulary was 500 words. I got hold of an electronic copy of the paper and counted the number of different words it contained (grouping inflectional variations, such as walk/walks/walking/walked, as a single item, or lexeme, as linguists call it). I found around 8000 - and that wasn't the complete total, as not everything in the paper was online. (Incidentally, how many different words are there in the King James Bible - excluding the names of people and places? Answer: about 8000.)

People know and use far more words than they (or communications czars) think they do. They forget about the whole year - about all the words to do with holidays, shopping, cars, animals, birthdays, Christmas... It's totally fallacious to think that the words you elicit from someone on a particular day or from a particular sample is an accurate index of all the words they know or use. The frequency stats on such networks as Google, Facebook, and Twitter show huge variations in the most popular words, day by day, depending on what's making the news. They're based on what people are talking about - or writing about. This week it's words about the weather. Next week...

It's even more difficult to make estimates when dealing with an unfamiliar world. Which is where teenagers come in. Few adults have any real idea about what teenagers talk about. When I've had the opportunity to listen in to such conversations, I've found it just as sophisticated as any adult area of vocabulary, for the topics which they find important. Listening to a group of kids rapping at each other recently, I heard them using a remarkable range of vocabulary, all with clever rhythms and rhymes. I certainly couldn't match their lexical range when they started to talk about pop music, clothes, favourite TV programmes, mobile phones and their applications - and I was aware I was only at the edge of their vocabulary when it came to some of the topics they would never dream of talking to adults like me about.

Moreover, in talking about these things, they weren't just using slang or text-messaging abbreviations or all the other things that adults imagine teenagers do. They were using a great deal of vocabulary of a general and useful kind, such as (to take a small extract from one list I compiled a few years ago): spoon, spot, spring, spy, squabble, square, squash, squeak, squeeze, squirrel, squirt, stain, stairs, stamp, stand, star, start, station, stay, steady... Find me a teenager that doesn't know any of these words.

I usually use a dictionary method when I want to get a rough idea of the size of a person's lexicon. I've reported it in several of my books over the years. Anybody can do it. You go through a sample of pages in a college-size dictionary (e.g. one containing around 100,000 entries) and mark the number of words you think you use (active vocabulary) or know (passive vocabulary), then scale up. I've rarely found estimates of adult active vocabulary falling below 40,000 words, and usually they are in excess of 50,000, with passive vocabulary being about a third larger again. In other words, most people expect at some point to have an opportunity to use about half the words in the dictionary. And when you look at the words they tick in such an exercise, like the ones above, it shouldn't be surprising. It doesn't take long to reach a total of several thousand. I've done this with mid-teenagers too, and I can't recall an estimate ever falling below 20,000. Some sixth-formers have a hugely impressive vocabulary, much higher than this.

So it's nonsense for anyone to suggest that teenagers have a lexicon of 800 words, which is how the media generally seem to have taken the comment. Nor is it very useful to say, as I've also seen in a couple of places, that 'the average daily vocabulary' of a teenager is 800 words. That's nonsense too, because (a) I don't know what an average lexical day is, either in content (weather today, something else tomorrow, Easter eggs soon...) or in quantity. I am writing this at the end of a day when I guess I have spoken for about an hour, in total. At the speed at which I speak, that means about 5000 words in all. The main subjects, insofar as I can recall them now, have been about the weather, the activities of my children, some points arising out of the papers, general mealtime chat, and holiday activities. I'd be very surprised if I used more than 800 different words today. (I read and heard tens of thousands of different words, of course, but that's a different matter.) Another day, involved in a discussion about linguistics or whatever, it would be very different. It's all a matter of subject-matter and motivation. Start an argument going amongst teenagers about the last episode of Dr Who, and you'll hear some fantastic vocabulary, from aliens to regeneration. Listen to one of the school debating competition contests, such as those organized by the English Speaking Union, and you'll be amazed at the lexical fluency of the participants.

And so, finally, on to the third strand in the report, that 800 words isn't enough 'to get a job', as Jean Gross put it. Well, we're agreed on that (for most jobs, anyway), but if most teenagers have a much larger vocabulary, then this argument is beside the point. So if she feels there is a problem here, the diagnosis must lie elsewhere. There are two possibilities. The teenagers she is thinking about may not have the right kind of vocabulary for the job she imagines they have in mind. Or they aren't able to use their vocabulary within the formal style of communication which is required in, say, an interview. I think there's some truth in both these observations. Acquiring the lexicon of areas outside your immediate situation is an important index of educational achievement. 'It pays to increase your word power', as Reader's Digest used to say, and this maxim is as true today as it ever was. And acquiring a mastery of diffent styles of English, so that one can switch confidently from everyday colloquial to formal discursive, is also critical. This is one of the aims of the National Curriculum in English, to develop that kind of language awareness in teenagers. And it was also the thrust of the research the Sunday Times reported by Professor McEnery, who has pointed out that teenagers are still developing their oral communication skills, and that oracy needs to be an educational focus alongside literacy. This is where Jean Gross needs to direct her energies: to supporting the need for more systematic language work in the school classroom. Promulgating myths about limited vocabulary and stereotyping youngsters as Vicky Pollards won't help.

It's such a shame. There was a great deal of importance in the interview with Jean Gross, drawing attention to the problems which affect early language learning in many children, the importance of speech and language therapy, and so on. These areas need the higher profile which she could help to provide. But all of this was overlooked because media attention focussed on one silly claim.

Saturday 2 January 2010

On OP 2010

One of the first enquiries of 2010 has been a request for information about Shakespearean 'original pronunciation', which was the subject of this blog some time ago. 'Is there anything in the pipeline by way of fresh productions?', my correspondent asks.

I know of one being planned, at the University of Kansas, where a production of A Midsummer Night/s Dream in OP is scheduled for towards the end of this year - an excellent choice, given the number of rhymes in that play which don't work in a modern English accent. I'll be involved in helping that event get off the ground. The director is Paul Meier, who many will know from his International Dialects of English Archive.

Judging by the number of requests I've had for the OP CDs, the interest is spreading, and there may well be some other activity going on that I don't know about. I don't know of anything that has happened in the UK other than the talks on the Sonnets in OP that I did during the year, and the associated write-up in Around the Globe 43, which will be online on my website soon.

One other thing has happened. The CD method was proving very awkward and time-consuming, so I've now replaced this by a direct downloading facility at the Pronouncing Shakespeare website, which is proving to be much more convenient for all.