David talks about his exploration of the changing role of English around the world and the impact that global change is having on how and when English is taught.
Countries worry where they rank but comparing levels of competence is futile
Tuesday 13 September 2011
Does China now speak English better than India? Will Spain soon challenge Scandinavia in English-language abilities?
A global league table of English proficiency recently published by the multinational language education provider English First (the EF English Proficiency Index) suggests the answer to the first question is “yes”, and to the second “no”.
The EF index both confirms common knowledge, but will raise eyebrows elsewhere: the Scandinavian countries, with the Netherlands, are top of the list in “a high proficiency” group, but Poland beats Switzerland and Hong Kong, and is just below Malaysia. China, meanwhile, trumps India, which in turn only marginally betters Brazil. But is India really in the same league as Brazil when it comes to English?
I’ve been monitoring the changing fortunes of English in different countries for some years. Indeed, the EF report draws indirectly on my own publications. But I have never attempted to produce a “country index” of this kind, and for good reasons.
The first is lack of data.
There is no standard way of describing language proficiency: in each country different kinds of tests are used, and results are often difficult to obtain. For example, the EF index claims China beats India but as I found out when I was researching my report for the British Council, English Next India, it remains a mystery exactly how many people in India speak English and to what level of proficiency.
Some countries regard the results of national exams as “state secrets“, and the detailed results of exams by independent organisations are often commercial secrets. Which independent exam provider would publish even the number of test-takers in each country?
EF, of course, have their own data – some simple averages based on four online tests, two of them placement tests taken by aspiring EF students. But there are serious problems with using data of this kind to construct “national scores”.
First, is the issue of representativeness.
The EF tests are a taken by a bunch of people who have elected to take them – they are not representative of the national population – or indeed, of any easily identifiable group. We can see this problem more clearly if we look at another international source of data – one that has annually produced average scores by country (and by first language of test-takers), but whose provider has never claimed any international comparative validity.
The Ielts test, created by Cambridge Esol, is taken by around 1 million people worldwide each year. (This compares to EF’s claim of 2 million test-takers over four years.) The country comparisons for Ielts data show a rather different picture from the EF scores. Germany tops the world in Ielts, though it languishes just above Malaysia in the EF index. Spain is above Hong Kong in Ielts, but is just beaten by Nigeria. No Scandinavian countries are reported, reflecting the low number of exam-takers from these countries.
Neither the EF nor Ielts test data represent adult national populations.
They simply reflect the different markets that each organisation has in each country.
For example, in some countries a significant group of Ielts test-takers may be students who cannot find university places in their home country but who have wealthy parents who are prepared to pay for overseas study. In Hong Kong, on the other hand, all second-year students are encouraged to take Ielts by the government, who will pay for their exam. In each country, the nature of the self-selection is very different, making an international comparison based on the data worthless.
Any global league table reduces a complex picture to a single figure.
But average scores, even if they were of a representative sample, are no longer of much use. We now need to know what proportion of a population fall into each broad proficiency band.
For example, using the Common European Framework of Reference (CEFR), which describes a proficiency scale from A1 (able to communicate in a basic way on familiar topics), to C2 (very near native-speaker competence), we can see two interesting but rather disturbing global trends.
The first is that the English proficiency in many countries is “bimodal” – there are some people who speak English but a lot who don’t.
Only the well-off in some countries have access to English, and English helps them become even wealthier.
The second trend has to do with the structure of economies – what kind of jobs are available, how much workers are paid for these jobs and whether they need English.
In many economies an increasing gap is emerging between the low-level service jobs (such as restaurants, hotels), which need no more than A1 or A2, and the higher level graduate, professional jobs, which require C1 or C2.
In research I carried out earlier this year, I found that as the structure of the Hong Kong economy has changed over the last 20 years – factories were moved to mainland China and Hong Kong was left as a services economy – there emerged two demands for English speakers in the workforce.
The largest is for graduate workers with at least C1 English skills, the other is for lower-level workers in hotels, retail and restaurants, with perhaps A2. The structure of the Hong Kong economy had least need for those with intermediate skills, such as B2.
Similar changes are now about to affect the whole of the Pearl River Delta, where those Hong Kong factories went in the 1990s. It is, of course, an economic success story: many of the low-cost labour factories that produce cheap clothes and electronic goods are now closing and shifting elsewhere – like west China and Vietnam.
Meanwhile, in cities such as Shenzhen, already bigger than Hong Kong, the factories are being replaced with office blocks and universities. I will be working during the coming year exploring what the impact of this economic transition will be on the demand for English at different levels. Averages will be of no help.
David Graddol is a writer, broadcaster, lecturer and consultant on issues related to global English.
Interview with David Graddol