Universiteit Leiden

nl en

‘The linguist’s work is by no means done’

Brain research and statistics are advancing our understanding of language and language acquisition. Linguists are still essential, however, says Professor of Dutch Linguistics Sjef Barbiers. Inaugural lecture on 8 December.

Believe it or not, but linguistics was amazingly sexy and popular in the second half of the last century, says Professor Sjef Barbiers. ‘Cognitive sciences and computer science, two important sources of inspiration for modern linguistics, were young and promising. The syntax of natural language was more or less a terra incognita in which linguists could venture forth.’

Discipline in crisis?

Fast forward to 2017, and many a linguist thinks this approach to linguistics has lost its sex appeal. The discipline is in a crisis, they say. If you want to study language as a cognitive phenomenon, you can do brain research and immediately see which regions of the brain show activity when a person is speaking or learning a language. And you can also study languages perfectly well with big data and algorithms. Why would you still need standard linguistics?

In his inaugural lecture Barbiers shows that theoretical linguistics is still in the prime of its life, whatever its critics may say. The linguist’s work is by no means done. Because if you want to use neurocognitive research and data science to advance linguistics, you still need linguistic analysis to interpret the results properly. ‘Language technology still can’t do essential things that people can. Google Translate still doesn’t understand a thing about language.’ According to the Head of Google Research Europe, Emmanuel Mogenet, artificial intelligence is currently approaching the intelligence level of a snake. To put it mildly, there is much work to be done.

General syntax principles

Barbiers gives the example of sentences in which two words that belong together do not appear together, such as the waar and mee in the sentence ‘Waar denk je dat hij de auto mee repareert?’, a common phenomenon in human language. Such constructions are a problem for translation machines. Young children, however, tend to separate words that can’t be separated, such as the verb noem (to call) and the past tense suffix -de in the sentence Dan noem-ik-de jou Sinterklaas, a sentence that is incorrect in adult Dutch.

Barbiers: ‘This phenomenon starts to become interesting in light of the hypothesis that, at the abstract level, syntax is the same in all languages in the world.’ Then it becomes clear that this child’s language sentence makes creative use of the verb positions that are available in very many languages. Children don’t just learn their mother tongue from imitation, because they don’t hear adults using this kind of sentence. Barbiers believes that you can only discover these general principles of syntax by comparing as many languages and dialects as possible. He sees this as a job worthy of linguists.

Netherlands in strong position

Barbiers believes that Dutch linguists are in a strong position compared with their counterparts abroad. ‘Nowhere in the world are there such extensive, varied and accessible digital data collections, descriptions and atlases of a language and its dialects. The way that linguistics in the Netherlands is brought together in the National Graduate School of Linguistics is also unique in the international perspective, ideal conditions therefore for the Dutch language and Dutch linguists to play a key role in the future of linguistics.’ Who knows, perhaps he’ll make linguistics as sexy as it once was?