Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Disclaimer: The Parliamentary Library does not warrant or accept liability for the accuracy or usefulness of the transcripts. These are copied directly from the broadcaster's website.
Genetic anomaly could explain severe difficul -

View in ParlViewView other Segments

Most Australian Aboriginal languages are deficient in words for numbers. Brian Butterworth sought
out children who were monolingual in such a language. He set them tasks involving numbers and
counting involving spatial strategy. The children completed the tasks as well as children who grew
up with English. This finding is important as there is debate as to whether you can enumerate when
there are no counting words in your language. The suggestion is that numbers are wired into the
human brain and perhaps a genetic anomaly explains why some children have severe difficulties with

Robyn Williams: Could you count the numbers of apples in a bowl or coins in your hand if you didn't
have names for numbers? Well, Aboriginal kids can somehow. Here's Professor Brian Butterworth.

Brian Butterworth: Well, if you read books of Australian linguistics like, for example, the great
book by Bob Dixon, then you find out that probably all Australian languages are very deficient in
number word vocabulary, they don't have counting words. So you try and find a group that's large
enough to have a decent number of kids still, kids who've been brought up just in that language,
monolingual in one of the Australian languages.

Robyn Williams: So you went to which particular groups to do your studies?

Brian Butterworth: We went to one group who speak a language called Walbiri in a place called
Willowra which is north-west of Alice Springs out in the desert, and another group that speak a
language called Anindilyakwa on Groote Eylandt in the Gulf of Carpentaria. We had a number of
tasks, only some of which we've reported so far. First of all we wanted to find children who were
brought up mono-lingually in one of those two languages, and of course we can't ask them questions
like, 'How many dolls are there on the mat?' because they don't have a word for 'how many' and they
don't have a word for the answer if there's more than two. So we had to ask them different sorts of
question, and so it took some ingenuity to develop a proper battery of tasks.

One of the ways in which we did it was like this; we would put out some objects on a mat in front
of us and the child would have a mat and we'd say, 'Make your mat the same as our mat,' and we'd
make it clear that the exact arrangement of the objects on the mat didn't really matter. What
mattered was how many things there were on the mat. So that was one way in which we did it, and we
had a number of variations on that.

A further way in which we did it would be I put some things out on my mat or our local assistant
because we had to speak to these kids in their own language by a native speaker in that language,
so she would put out a mat with some things on it, then she'd cover it up. We would then ask the
child to remember what was on the mat and put some things out on his or her mat. Then the assistant
would add something under the cover so the number of things on the mat would change and then the
child would have to put out a number of things on his or her mat which corresponded to the number
of things on the assistant's mat.

But of course those tasks can be done just by remembering what the objects were and more or less
where they were on the mat. In fact we did an analysis in which we looked at this kind of spatial
strategy for doing the task and comparing our Northern Territory kids and our comparison group of
kids in Melbourne. The Melbourne kids never used that, they used presumably some kind of verbal
strategy, they would go 'one, two, three, four, five' and then put out five things. But because the
Northern Territory kids couldn't do that, they used a spatial strategy.

But we had some further tests which couldn't rely on that strategy and this is a task we called
cross-modal matching. So we take some sticks and we'd bang them together a certain number of times,
and the child had to put out on his or her mat a counter for each time we hit the sticks together.
Here you have rather abstract matching; you're matching a sequence of auditory events with a set of
objects on a mat. So it's really quite an abstract representation of the number of both the sounds
and the objects on the mat. Again, we found that the children in the Northern Territory who didn't
grow up with counting practices or with counting words did just as well as the kids in Melbourne.

Robyn Williams: What did the kids think of the experiment? Did they think this was boring nonsense
or tremendous fun? Did they muck about?

Brian Butterworth: Particularly in the central desert where they seemed to not attend all our
sessions, we called them 'special games' and I think the kids quite enjoyed it. It made a change
from what they did normally, so I think they had quite a good time doing it.

Robyn Williams: And presumably the Melbourne kids were Aboriginal as well.

Brian Butterworth: Well, certainly had some Aboriginal antecedence, so there's a special school in
Melbourne for children of indigenous parents and we tested those as well.

Robyn Williams: So you're saying that the kids responded as well, even though they may not have had
words for 'one, two, three, four, five, six, seven' but they could clearly count, nonetheless.

Brian Butterworth: They could certainly enumerate. They weren't counting using words, they were
counting in some other way. This actually is an important finding because there's a fundamental
theoretical debate that's raging in the literature between those who think that you can enumerate
exactly without having counting words and those who think you can't. There were a couple of studies
published four years ago with Amazonian groups who also spoke languages which didn't have counting
words, and the findings from there were that these people didn't do as well on exact number tasks
as...well, where there were controls, they were French people. So they didn't do as well as French

But we think that there were some problems with that study. For example, in probably the better of
the two studies they used computerised tasks, and I think maybe that creates an extra problem,
whereas we tried to use tasks that were culturally appropriate and also developmentally
appropriate, because our kids were four to seven years old, so you have to have something that kids
of four to seven can do and would actually want to do.

Robyn Williams: Were the findings consistent in high numbers? Because I can imagine an awful lot
of...well, even animals can work out 'three, four, five' but once you get to double figures you
tend to have a problem.

Brian Butterworth: The research that we've reported so far only goes up to nine. The theoretical
boundary is between sets that you can enumerate without counting in a glance and the limit there is
about four rather than five. The question is was there a discontinuity between performance up to
three or four and performance from five to nine. And we didn't find that discontinuity. So we think
that whatever it was the children were doing, they were doing more or less the same thing for all
the numbers between one and nine. Of course it gets more difficult as you get larger numbers and so
there were more errors for lager numbers, but basically we think they were doing the same thing.

Robyn Williams: Did you ever ask the kids how they were doing it? Did you say, 'How did you know?
What were you doing in your head to work out what the numbers were?'

Brian Butterworth: We didn't do that exactly but what we did was we recorded the way in which, for
example, they would lay out the counters in the task which required them to do that. So we could
see whether they were using some kind of spatial strategy without actually asking them. You know,
you ask a four-year-old 'What strategy did you use to do this task?' and you probably wouldn't get
a terribly useful answer. But we could observe. For example, if we laid out four in a perfect
square and the child laid out four in a perfect square we'd say, well, this child is using spatial
strategy. If the child, in Melbourne for example, is going 'one, two, three, four' we say they're
using a verbal strategy. So we did do careful observation of what the children were doing so we
could figure out what kinds of mental processes seemed to underlie their performance.

Robyn Williams: Of course the significance of this is to say that numbers are somehow in our heads
in the beginning, irrespective of a whole number of things, and now including language. So if the
numbers are wired in, you can tell which groups by some sort of mishap have missed out as a result
of some sort of genetic anomaly. Is that more or less the story?

Brian Butterworth: That's certainly our story, and we think that the kids who do have the severe
problems learning arithmetic, these kids we call dyscalculic children or dyscalculic learners, and
we think that they do indeed have a genetic anomaly which means that their brains have developed in
an atypical way and the particular circuits in the brain needed to do just these simple number
tasks on which the rest of arithmetic is built, that seems to be abnormal and therefore doesn't
provide the basis for learning arithmetic. That kind of genetic anomaly seems to be fairly
prevalent. In studies that we've done and other people have done, it's about 6% or 7% of the
population. That's a lot of people, and as far as we can tell it persists into adulthood. So we've
been testing quite a lot of very high functioning adults who nevertheless have great difficulty
even enumerating quite small sets of objects.

Robyn Williams: And it's irrespective of language, as you've just shown.

Brian Butterworth: It seems to be, yes.

Robyn Williams: Professor Brian Butterworth is at the Institute of Cognitive Neuroscience in
London, and his colleague at the University of Melbourne is Professor Robert Reeve.