'Everyone is bilingual to some extent'

An interview with Enoch O. Aboh

1 March 2014

Enoch Aboh is professor of learnability at the University of Amsterdam. Before coming to the UvA, he worked at the Université de Genève. His research interests are: learnability of human language, theoretical syntax, comparative syntax, (including comparison between the West-African language family Kwa and Germanic, Romance, Sinitic, and Caribbean creoles), the discourse-syntax interface, language creation, and language change.

‘I started my university studies with economics in Benin, a small country on the west coast of Africa where I come from. I then moved to a neighbouring country, Togo, where I did my BA in English. The idea of the program was to train students to become foreign language teachers. But after my BA I was not so keen on becoming a teacher and I decided to study translation. To cut a long story short, I moved to Europe and found myself in Geneva, registered in the faculty of letters with linguistics as a major. I did my MA and my PhD in Geneva. After graduating in 1998, I worked at Université de Genève as a lecturer, for two years before I got a post-doc here in Amsterdam, in 2000. Looking back, I realise that I left Benin with no specific idea about having a career in academia.. It all happened as a consequence of different actions and events, which led me to make one choice or the other. But, of course, even if you do not plan such things ahead, there is always a point in time when you just feel like yes, this is what I want to do.’

The African Linguistics School

‘I think immigrants always consider going back to their home country. Immigrants of my age keep saying it, but it is the kind of thing you know will never happen. In my case, I made a conscious choice not to go back, because I think that the work I am doing here is more useful to people in Benin and other African countries, than the kind of work I could do in Benin. One of those things is providing literature on a language that before I started did not have any references. Since 2009 I am also involved in organizing the African Linguistics School, which is unique in the world. This school takes place every two years in an African country of our choice. We select about 70 to 80 students from various African countries, who are funded to attend the school. For two weeks, the students are exposed to top calibre teachers invited from Europe and the United States to teach introductory as well as advanced courses. The program revolves around four main pillars: formal syntax, phonetics and phonology, semantics, and language contact. It is very rewarding to see that the entire faculty accept to teach pro bono and many of them fund their own travel expenses to the venue. Clearly the school is a fantastic forum where the African students and the invited teachers contribute to the development and transfer of knowledge The school took place in 2009, 2011 and 2013, and in 2015 it will take place in Ivory Coast. Staying here thus does have the advantage of allowing me to contribute to such large scale initiatives and to be able to give back. Being in Benin would have had the advantage of being close to the field, but when I consider the social, cultural, and educational impact of these two options, I think that staying here weighs more.’

A pretty fine cognitive system

‘If you put a child in any kind of noise where people speak a language, the child will end up speaking that language. There seems to be a bias in human being, something about learning that helps the child to decide what is relevant and what she has to pay attention to. Once we agree to that, then we must also agree that there is something ‘in there’, a pretty fine cognitive system, that allows the child to make the right hypotheses about the rather challenging data she is exposed to. The question is whether this pretty fine system is language specific or not, but, actually, I don’t think the answer to that question will help us make any progress. This is the kind of question we want to ask once we really know how the system functions. Right now it is more important to figure out what kind of bias we are talking about, how it functions, and how children develop this bias.’

‘What is nice about this university, at least in the Linguistics department, is the fact that we have different backgrounds, but we can still function as a coherent research unit. It is true that generativists, like me, form a minority – though a strong minority – in this department, but that does not prevent us from working and interacting with other colleagues. For example, I collaborate closely with Kees Hengeveld, who is a functionalist and our views converge on many linguistic issues, though we may disagree on the exact implementation or analysis.. Actually, that is what I like here: you hear different voices, and it is up to you to decide how you want to integrate these different voices into your own perspective of what the human language capacity is.’

Learnability of Human Languages

‘The position that I have now is the Learnability of Human languages, which deals with the basic question of how children acquire the language that they are exposed to. One of the things that I want to do is to approach the notion of learnability from the perspective of bilinguals. If you look at the development of linguistics throughout the last 30-40 years, a lot of emphasis has been put on monolinguals. The idea is that you investigate what monolinguals do, and then you can infer what the linguistic system behind that is. I am trying to investigate this from the perspective of bilinguals, because I think that everyone is bilingual to some extent. Even speakers of a single language are considered to be bilinguals, in the sense that they master different varieties, registers or styles which sometimes may correspond to significant differences in grammars. The fact that you can speak standard Dutch as opposed to colloquial Dutch is a form of bilingualism, similar to the fact that I can speak Gungbe and French.’ The only difference is that in one case you are operating on closely related grammars while in my case I’m operating on typologically and genetically different grammars.

‘What I am interested in is how speakers draw on their knowledge of different grammars. We know that in normal contexts, speakers try to stick to one grammar or to one language at a time. But what we also know is that in multilingual settings speakers can easily switch from one language to another, and that they do this in a very systematic manner. The question is what the engine behind this is, that allows them to merge two or more grammars into one without causing communication to fail. And then the question arises how this relates to the learning process. My view is that learning boils down to developing an algorithm that allows the learner to put together pieces of information extracted from the input. If this is true, then what we know as ‘code switching’, moving from one language to another in an utterance, is actually a reflection of that algorithm. You take a piece from one language, and another one from another language, and then you put them together to form a new item. I am suggesting that even monolinguals do exactly this when they are acquiring a language. The child extracts a number of features from the input that she is exposed to, and then put those features back together, in order to form a grammar that allows her to communicate with other members of the community.’

‘If we take that line of thought, there are a number of interesting implications. One of them, suggested by studies on pathological code-mixing is that some bilingual speakers with brain damage may loose the inhibitory control system that allows selection and maintenance of one language during a conversation. So these speakers cannot stick to one language and, switch from one language to another without control. The interesting thing is that: the switch patterns that they produce cannot be formally distinguished from those made by speakers with no brain damage. This suggests that the entity, or module – or call it whatever you want because we do not know what it is – which generates these switches, is not damaged even in such patients. I suggest that this module, which is specialized in generating linguistic combinatorics is also what allows us to acquire language. This module is independent of the lexicon since it generates well-formed outputs made of linguistic features extracted from different lexicon. Since this module is not tied to any specific lexicon, what we apparently learn is how to control it so as to stick to some particular type of lexicon which itself relates to some particular type of grammar which we label as some particular language (e.g., Dutch, Gungbe, Michif, etc. If I am right, it seems to me reasonable to conclude that this module is language-specific. It’s only job is to recombine linguistic features into new linguistic items. My current research is to find out the properties of this module.’

by Gisela Govaart, March 2014

Published by  SMART Cognitive Science