Summary: HCOMET was created to be a more reliable system of measuring translation-quality. The Human COgnitive Metric for Evaluating Translation enables researchers to linguistically quantify improvements in translation quality for the development and tuning of machine translation algorithms.
By analyzing frequencies and collocates of character-divided subcorpora, this quantitative study describes the corpus-linguistic representation of gay men in Will & Grace. Results derived from these linguistic representations were used to provide evidence for or against proposed sociological theories on the subject.
Will & Grace provides, not a single-dimensional, positive, and socially progressive depiction of gay men, but a multifaceted portrayal that includes, whether intentionally or not, linguistic inequalities that depict homosexuality as not fully accepted or even acceptable. By placing these corpus-linguistic findings in conversation with the sociological studies mentioned, it becomes clearer that Will & Grace was hindered by the writers’ social perceptions about the audience it successfully attracted, in spite of being called a gay sitcom. (Download the Paper)
João’s Role: Author