Blasi, Damian E.
ICREA Research Professor at Universitat Pompeu Fabra (UPF).
Social & Behavioural Sciences
Short biography
Research interests
-
Worldwide cross-linguistic patterns. A fundamental facet of my research involves discovering and testing cross-linguistic (ir)regularities across diverse domains of language using large linguistic databases, with a special emphasis on those relevant for human cognition and culture. Recent examples include independent publications on (1) deep clausal embedding depths, (2) the efficiency of word encodings, (3) form-meaning mappings in basic vocabulary (4) grammatical gender, their modifiers, and semantic substrates, (5) surprisal effects on word choice, (6) syntactic dependency minimization effects.
-
Linguistic diversity through human history. I have researched different aspects of how the 6,500 languages in the world today came to be the way they are, covering processes and historical events ranging from the last hundred years to the onset of the Holocene. In particular, I have studied and published on (1) the complex relation between diet, behavior and speech sounds after the Neolithic revolution, (2) the rise of creole languages as the result of colonialism, and (3) the inter-related genetic, cultural, and linguistic histories of languages in North East Asia.
-
Development of open data in collaborative team.: I have participated in a number of projects aimed at digitizing and making publicly available different types of data relevant for the study of linguistic diversity. This includes independent databases of (1) grammatical information and (2) genetic panels, which are scheduled to be published this year.
-
More recently, I became interested in notions of linguistic fairness and low-resource languages. In this regard, I contributed to (1) a first attempt to quantify the degree of inequality on language technologies across the world’s languages, (2) a study on the lack of linguistic coverage in COVID-19 health guidelines, and (3) novel algorithms to assess the quality of specific computational linguistic resources when data is scarce.