New Blog Post in the How-To Series
Already last week, my first blog post in our "How to do X in linguistics" series appered, concentrating on How to write an initial review for a journal. Here is the abstract, more can be found in the actual blog post.
Writing reviews for a journal is one of those things which most scientists never actively learn. For laypeople, this may be surprising, given how often the scientific method with its rigorous peer review procedure is being mentioned in the news nowadays. How can it be, one may ask oneself, that this procedure that is usually presented as the core principle of scientific reasoning, is never really actively taught? If the review by experts is the core of the scientific method and what decides about the acceptance of an article, how can it be that scientists do never take a course on article reviewing, and how can it be that reviewers are (as I have previously discussed in a German blogpost) themselves never reviewed or graded?
On Monday, we (John Miller, Tiago Tresoldi, Roberto Zariquiey, César Beltrán, and Natalia Morozowa), submitted a new preprint on automated borrowing detection in monolingual wordlists, which can be accessed here. The abstract is given below.
Native speakers are often assumed to be efficient in identifying whether a word in their language has been borrowed, even when they do not have direct knowledge of the donor language from which it was taken. To detect borrowings, speakers make use of various strategies, often in combination, relying on clues such as semantics of the words in question, phonology and phonotactics. Computationally, phonology and phonotactics can be modeled with support of Markov n-gram models or – as a more recent technique– recurrent neural network models. Based on a substantially revised dataset in which lexical borrowings have been thoroughly annotated for 41 different languages of a large typological diversity, we use these models to conduct a series of experiments to investigate their performance in borrowing detection using only information from monolingual wordlists. Their performance is in many cases unsatisfying, but becomes more promising for strata where there is a significant ratio of borrowings and when most borrowings originate from a dominant donor language. The recurrent neural network performs marginally better overall in both realistic studies and artificial experiments,and holds out the most promise for continued improvement and innovation in lexical borrowing detection. Phonology and phonotactics, as operationalized in our lexical language models, are only a part of the multiple clues speakers use to detect borrowings. While improving our current methods will result in better borrowing detection, what is needed are more integrated approaches that also take into account multilingual and cross-linguistic information for a proper automated borrowing detection.