CfP special issue of Computational Linguistics on "Computational approaches in historical linguistics"
Computational approaches play an increasingly important role in historical linguistics and typology. During the last two decades, scholars have made significant advances in automatizing and formalizing specific aspects of the workflow of the classic comparative method. But despite the growing interest in automated approaches to historical linguistics, it is obvious that automatic approaches are still far away from being able to replace human experts, and the majority of approaches still makes broad use of manually annotated datasets. More than two decades after the quantitative turn in historical linguistics, we think it is time to reconsider how computational approaches in historical linguistics can be further improved, and where their current limits can be found.
We invite submissions of papers which address one or several of the following questions:
1. How can we improve the automatic identification of cognates?
2. How can computational methods help to infer deeper genetic relations between the world’s language families?
3. How can big data approaches from computational linguistics help to improve classical approaches to historical linguistics?
4. What are the strengths and shortcomings of phylogenetic methods?
5. How does demography and geography influence the spread of languages through time and space?
6. Are there universal tendencies in the evolution of the world’s languages?
7. How to integrate typological features with lexical features for inferring language phylogenies and predicting typological features for ancestral languages?