This new commentary article is a response to a target article from Laura Becker and Matías Guzmán Naranjo titled “Replication and methodological robustness in quantitative typology”.
In this paper, Becker and Guzmán Naranjo explore what happens when typological work is reproduced by another team, and what the differences in results mean for the field of linguistic typology.
I wrote this article alongside my former co-chairs of Linguistic Data Interest Group (LDIG) of the Research Data Alliance, reflecting on the work needed to ensure that we are all doing the kind of research that is useful to a more transparent way of doing work in linguistic typology.
From the commentary:
We thank B&GN for their work, but we also thank Dryer (2018), Seržant (2021), Shcherbakova et al. (2023) and Berg (2020), for doing work that could be reproduced (a benchmark much scholarship falls short on) and subsequently having their work scrutinised and evaluated in this way. As B&GN note, “[t]here is no specific reason for choosing these papers other than the fact that the authors made their datasets available”. To work in an open and transparent manner is to open yourself to critical evaluation. Linguistic typology advances because of researchers who have created accessible data as part of their work. This includes individual researchers working on specific languages, whose data is the basis of typological work, as well as those typologists who share the databases of their work. Open ways of working require open mindedness from the whole research community.
Reference
Gawne, L., H.N. Andreassen, L. Ferrara, A.L. Berez-Kroeker (2025). Open research requires open mindedness: commentary on “Replication and methodological robustness in quantitative typology” by Becker and Guzmán Naranjo. Linguistic Typology. doi: 10.1515/lingty-2025-0018
Related links