It's not clear to me whether or not someone whose work aims to improve an NLP system may be called a "Computational Linguist" even when she/he doesn't modify the algorithm directly by coding.

Let's consider the following activities:

Annotation for Machine Learning: analysis of Morphology, Syntax, POS tagging Annotation, analysis, and annotation of entities (NER) and collocations; supporting content categorization; chunking; word sense disambiguation. Recording of technical issues of the annotation tool to improve its reliability. Recording of linguistic and logical particular rules adopted by the research team who develops the NLP algorithm to improve consistency between annotation and criteria previously adopted to train the NLP.

May be these activities considered "Computational Linguistics"? If not, which is their professional category and how should they be included in the resume in a word which synthesizes them?


(Disclosure: I am a PhD student and lecturer in Computational Linguistics)

It is true that annotation and debugging work with existing tools without modification can be considered Computational Linguistics.

And yet, most Computational Linguists program on a daily basis, since they actively develop tools. Just to give you some context, at major Computational Linguistics conferences such as ACL or EMNLP (the biggest ones), most authors did the coding themselves.

To say that coding is an unimportant side aspect of being a Computational Linguist, as claimed in the accepted answer, is a slight misrepresentation.

  • $\begingroup$ It depends. I studied Computational Linguistics, and currently work as a computational linguist. My day-to-day work does not involve much coding at all. So I'd say that programming on a daily basis is also a slight misrepresentation :) It also depends if you are more on the Computer Science side, or on the Linguistic side of it. CL is inherently interdisciplinary, so I guess we just have different experiences of it. $\endgroup$ Feb 3 '20 at 21:01
  • $\begingroup$ @OliverMason You are describing your individual experience, while I am trying to describe the research field as a whole. I still maintain that saying that programming is the "least important" aspect of CL is a misrepresentation. If the head of department at your university does not do any coding and never did, he is an outlier among heads of departments of CL. $\endgroup$ Feb 4 '20 at 8:33
  • $\begingroup$ I'm looking beyond academia; so I guess we just have to agree to disagree on this. $\endgroup$ Feb 4 '20 at 9:31

Yes. A computational linguist is someone who (among other things) uses computers to process/model/analyse/... natural language. Coding might be one aspect of it, but is about the least important: you can always get a non-linguist programmer to do coding for you.

I studied "Computational Linguistics" at university, and while programming was taught as part of the course, coding was only a minor aspect in the actual subject matter. The senior professor (and head of the department) wasn't able to do any coding himself; he came from the linguistics side of it.

Being able to program is useful, as it speeds things up and makes you more independent, but it is by no means an important part of being a computational linguist.

UPDATE: I have been accused of misrepresenting the field of CL. However, it is a broad, interdisciplinary field, and comprises many elements. Sure, on the academic/research side you might do more programming than in the applied/commercial side, but I maintain that you can easily work as a computational linguist without actually doing any programming. For most tasks, readily available software exists by now, so you don't actually need to program anything new.

  • $\begingroup$ Thank you for answering (I edited the title of the Q to make it more clear though). Therefore someone who does annotation for ML and all the activities described above is a computational linguist, right? $\endgroup$
    – franz1
    Jan 23 '20 at 21:09
  • 1
    $\begingroup$ @franz1 Annotation is part of the typical activities of a computational linguist these days, yes. $\endgroup$ Jan 24 '20 at 9:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.