Keynote Speakers

Linguists have clear notions about diachronic changes in languages, about diatopic (dialectal) variation in language as well as potential relations between them. In order to trace diatopic changes over time, one needs a long series of locally representative data for a sufficient number of geographical locations. The Frisian language material from the current province of Fryslân seems to be one of the rare cases where such series are available for multiple centuries in a row, sometimes up to seven hundred years. This offers a spectacular insight in the geographical dynamics of linguistic changes.

This talk will showcase some of the most interesting findings regarding syntactic theory starting from the study of heritage languages.

We will explore verb movement and its relation with domain delimitation, and the syntactic role of clitic doubling and differential object marking.

In May 1993, the world’s first academic meeting about mixed languages took place at the University of Leiden, co-organized with the University of Amsterdam. This culminated in the book Mixed Languages. 15 cases in Language Intertwining (Bakker & Mous eds.). Before that event, there had been a consensus among linguists that mixed languages do not exist, and that all languages have one primary genealogical affiliation. 

Since then, most linguists accept the existence of mixed languages, as having two genealogical affiliations, none of them primary. Several monographic descriptions and discussions of mixed languages have been published in the meantime, and several anthologies have been published with discussions about the more theoretical aspects.

There are languages with a lexicon from one language and grammatical system from another. There are languages with verbs from one language and nouns from another. There are languages with a dichotomy between the most conservative and unborrowable parts (like verbal paradigmsand personal pronouns) being from one language, but everything else from another language. There are also languages that have massively copied patterns or fabric from other languages, thus having changed their typological properties radically. There are languages that have borrowed verbs or nouns together with their inflectional paradigms (“parallel systems borrowing”). 

There are only a few dozen known mixed languages in the world. Yet, the existence of mixed languages is of high interest not only for specialists in language contact, but also for psycholinguists, historical linguists and theoretical linguists of all persuasions. 

In my talk I will give an overview of mixed languages and the types, and discuss controversies, and whether mixed languages differ categorically from other language contact phenomena such as borrowing and codeswitching. I will also briefly say why pidgins and creoles in general are not mixed languages.

In the last few decades, the top-level structure of the Indo-European family tree has taken centre stage in Indo-European linguistics, due mostly to the impact of the archaic Anatolian branch, and to the application of tree-producing computer models to Indo-European. The majority of scholars would now agree that Anatolian was the first branch to have split off from the family, and Tocharian the second. In this lecture, I will share my current thoughts about the position of these two branches based on the relevant findings of my PhD about Anatolian and my postdoc about Tocharian. While I found more evidence which may plausibly be interpreted as corroborating the hypothesized position of Anatolian, Tocharian has given me reason to take a view that is radically different from the emerging consensus.

Linguistic typology has a fundamental tension between the level of detail obtained about a linguistic phenomenon and the desire to be able to make meaningful comparisons across languages, i.e. the comparability problem (Evans 2020:418). As linguists, we are accustomed to using labels like “SVO”, “ergative-absolutive”, and “7-vowel system” to describe and categorise languages. However, reducing languages’ properties to these kinds of labels can give a misleading picture of homogeneity and can also lead us to miss out on more insightful characterisations of the language’s grammar. In this talk, I illustrate these methodological points with two case studies from my own work on word order variation, namely (i) the use of the label ‘S-Aux-O-V-X’ for the syntax of various languages from West/Central Africa, and (ii) the use of the label ‘SVO’ for the syntax of the Bantu subgroup of the Niger-Congo family. I show that detailed analysis of the syntax of these languages raises issues with the validity of these two labels, and discuss how being willing to give up familiar labels can lead us to more accurate description and comparison.