How a new theory is revolutionising our understanding of language

Group photo
(from left) Prof. Dr. Thomas Herbst, Prof. Dr. Stefan Evert, and Prof. Dr. Ewa Dąbrowska carry out research together at FAU’s Linguistics Lab (Image: FAU/Klotz/Dąbrowska)

At FAU’s Linguistics Lab, linguists are conducting research into how humans learn language.

How do we learn language? The theory of innate language ability has existed for a very long time. It is now being questioned by a new theory. In the newly-founded Linguistics Lab at FAU, Prof. Ewa Dąbrowska and Prof. Thomas Herbst, Chair of English Linguistics, are conducting research into usage-based construction grammar. In an interview with their colleague, Prof. Stefan Evert, they explain what impact the new theory could have on language teaching and why the Linguistics Lab was set up.

You’re working on a new language theory called usage-based construction grammar. What does this theory involve?

Prof. Herbst: Construction grammar, which has mainly been developed in the USA, assumes that knowledge of language in humans is stored in the brain in the form of a multi-dimensional network of pairings of forms with meanings, or constructions. Such constructions can be single words, or abstract grammatical structures.

Prof. Dąbrowska: For example, in German, you could use the construction subject-verb-object-adjective to say that an action leads to a specific result. Since we are familiar with this construction and its meaning, in German we can form sentences like ‘Schmeck dich fit!’ or ‘Schmeck dich prickelnd!’, as used in an advert for sweets, and understand them, although the word ‘schmecken’ (taste) isn’t usually used in this construction.

What has the theory that has been dominant up to now assumed?

Prof. Evert: Noam Chomsky’s theories are based on the assumption of a specific genetic disposition that enables people to acquire language or languages. To put it more simply – Chomsky imagines that every person is born with the knowledge of which structures can arise in languages. In his model, sentences are transferred from a fundamental deep structure via rules to a so-called surface structure.

Prof. Herbst: But there’s no convincing evidence for this. We’re therefore following a what-you-see-is-what-you-get approach with construction grammar, which means we assume that the meaning of a sentence is generated directly from the constructions that form it.

Prof. Dąbrowska: Several experiments have also shown that children are able to learn such form-meaning combinations. If we pursue the construction grammar approach, children learn grammar by abstracting a large number of constructional patterns from the language they hear and connecting these patterns into a network. This can be explained via general cognitive processes, without the need for positing an especially linguistic genetic disposition.

Construction grammar questions the significance of traditional grammar rules. How? And can you give some examples?

Prof. Herbst: If you translate a German sentence such as Wir zwei waren ein halbes Jahr in England into English with We two were in England for half a year, you won’t break a single grammar rule. The sentence still sounds hopelessly German however. The analysis of corpora, which are collections of written or spoken texts that today comprise several hundred million words, enables us to determine that compared with the two of us, we two occurs extremely rarely in the English language. The same applies to half a year compared with six months. The two of us spent six months in Britain sounds much more idiomatic and we can also say why.

Prof. Dąbrowska: Sentences are not formed by creating a structure using grammar rules in which you put some words. Rather, we use pre-fabricated stored combinations of words to a great extent.

Prof. Evert: Incidentally, this is one research approach that has a very long tradition in foreign language linguistics in Erlangen.

How does this affect foreign language teaching?

Prof. Herbst: It would be nice if this research approach did have an effect on foreign language teaching. The curricula should finally be changed to match the latest research in this field. Currently, schoolchildren in Bavaria following the curriculum for English have to learn such absurd rules such as the difference between the participle singing and a gerund singing, even though the most important scientific grammar texts of the English language have considered this differentiation unjustifiable for several decades now. It would be feasible, for example, to stop the separation of grammar and vocabulary in textbooks and to increase the introduction of words with the constructions in which they are found. From a construction grammar perspective, repetitive combinations of words are very important, but I must say that some school books are already excellent in this respect.

You founded the FAU Linguistics Lab. What research do you carry out there? What’s special about the lab?

Prof. Evert: Several subjects at FAU carry out research into language. The Linguistics Lab was set up as a coordination centre for all research projects on the subject of language and it increases the exchange of expertise in these different fields that range from neuroscience to machine data processing. One focus of our work is researching associations that exist between words and constructions. For example, we’re interested in finding out to what extent differences in the knowledge of language of individuals correlate with their cognitive skills in other areas. We will be carrying out psychological tests as well as analysing corpora. In this context, we’re also looking at the development of computer-aided methods for automatically generating environment profiles of words.

Prof. Herbst: The latter is being conducted with the prospect of creating a whole new reference work for the English language, a ‘Konstruktikon’ that covers the entire bandwidth of linguistic constructions – lexicon and grammar rolled into one.

Prof. Dąbrowska: The Linguistics Lab has also been set up to link linguistics research at FAU with international discussion on the topic as was the case during The Constructionist Challenge conference held at FAU for the opening of the Linguistics Lab in October 2019, which was attended by leading researchers from Germany and abroad. It is our objective to turn FAU into a centre of linguistics research with the Linguistics Lab that is well known internationally for its research.

Further information:

Prof. Dr. Ewa Dąbrowska
Chair of Language and Cognition
Phone: + 49 9131 85-29263

Prof. Dr. Thomas Herbst
Chair of English Linguistics
Phone: + 49 9131 85-22936

Prof. Dr. Stefan Evert
Chair of Computational Corpus Linguistics
Phone: + 49 9131 85-22426