One area where more and more language tests are being introduced is in the assessment of the language abilities of migrants. Before a migrant is let into a country, he or she could be asked to do a language test, in order to qualify for some government provided services, a migrant might have to do a language test, in order to become a citizen of his or her adopted country, a migrant can be asked to prove their language skills by doing a language test. You can see that a migrant might be asked to do a lot of tests. And it is these tests, and the language courses that are provided to migrants, that the Language Assessment for Migration and Integration (LAMI) special interest group within ALTE are interested in and I had the great pleasure to sit in on three related sessions on this subject led by the LAME group at the Krakow ALTE conference. Here is my interpretation of what was discussed.

A number of case studies looking at language policies, tests and courses associated with migrant workers in differnt countries were presented, all giving information about what laws and regualations surrounded languages for migrants, what levels of ability were demanded and what tests and courses were given. What surprised me was that there were such great differences. Some countries demanded level B2 to get citizenship, some required no proof of language ability at all. Some offered free courses to all migrants some offered nothing. Some had centralised control of language testing and learning, others just let individual teaching centres and teachers do what they wanted.

An extreme case of this is what happens in Belgium. In the Dutch speaking Flanders part, there are all sorts of interventions and demands. In the French speaking Waloon part, they don't have anything. In Flanders, you need to prove your langauge ability to get social housing, to get your children into school, to get unemployment benefit. In Waloon, you don't have to prove anything. And that is in one geographically small country. Some interesting questions come out of this. Why do I need formal proof of my language skills in one part of the country and not the other? Are formal proofs and the tests that are used to obtain them really necessary at all? And if they are, what is the level I need to get if B1 is stated in one part of the country and nothing in another?

A little research was done to answer some of these questions. Employers were asked if they demanded some sort of certificate to prove language ability. They said no, an interview would give them a good idea if someone's Dutch was good enough to do a job. And besides, there were some jobs that didn't really need Dutch language skills, such as being a cleaner, or for which there was such a shortage of skilled workers that they would take on the migrant skilled worker no matter what their language level was. They would work something out later. So, for many employers it wasn't necessary for the migrant to be B2. Other parts of the research showed similiar results in other areas of life. For example neighbours were happy to work around the limited language skills of migrants. It just simply wasn't necessary for many migrants to be at B2.

So why is B2 a requirement to get citizenship? There is a suspicion that these high demands for language skills are being used to exclude people, to keep them out. Not many people have such high level skills so demanding migrants have them and then giving them a language test to prove that they don't is a relatively cheap and easy way of excluding many of them. But, if high language skills aren't necessary to operate successfully in a country, then isn't it a little unfair to exclude people on that basis? Isn't that like saying we will exclude everyone who is shorter than 1m 90cm, even though we know being shorter than 1m 90cm won't affect your ability to live in the country successfully? If you don't need high level language skills, the state shouldn't demand you have them and tests shouldn't be devised to measure them.

Something has to be done to stop this hi-jacking of language tests for political purposes and LAMI are looking at just what that something can be. One suggestion is to go out into the real world and identify just exactly what language abilities and at what levels a migrant needs in order to operate successfully in it. Provide empirical proof of precisely what is required and then make sure state language requirements, and any courses and tests used with migrants, reflect them.

But migrants operate in many different worlds that require differnt language abilities at different levels. Think of labourer on a farm, a student at a university, a doctor in a hospital. So, there can't be just one performance level that can be used to determine whether a migrant enters or stays in a country, there have to be a number of differnt ones to fit the different situations migrants find themselve in. The labourer on the farm should be required to have the language abilities and levels required to be a farm labourer and likewise the doctor should be required to have the abilities and levels required to do his or her job. What is totally wrong is asking an aging parent who is coming over to join their son to be at B2 level when they're never ever going to need that level of language.

So, lets set language requirements, write tests and deliver courses that reflect the real world language demands placed on the migrant and do away with requirements and tests that are only their to provide a cheap and easy way of keeping people out.

On a personal note, I have to confess to being an interested party in all this. I've been a migrant for the last twenty years. Since I left the UK, I've lived and worked in many countries. And do you know, I've never ever been asked to prove my ability to speak the language of my host country? Not once. And do you want to know something else? I've managed to live quite successfully in many countries with hardly any knowledge of their language at all!!