Role of Computational Linguistics in Translation
Thread poster: ljs1ect

Local time: 21:52
Apr 21, 2007

Hello all. This is my first post on this forum, and I am basically interested in the area of overlap between computational linguistics and translation, assuming such exists to some significant extent.

As I currently understand things, the most obvious link is probably machine translation, although speech-recognition is another area of interest. What other major areas exist? Also, is the field (computational linguists) best approached from the computational side or the linguistics side? I ask because I am currently studying a computing course at university, and I wonder what formal training in linguistics is required.

I hope this is in the right forum; I eagerly await your responses!


Peter Linton  Identity Verified
Local time: 21:52
Member (2002)
Swedish to English
+ ...
Answered yesterday Apr 23, 2007

Your question was answered in some detail yesterday at the 21st Anniversary Conference in London of the Institute of Translation and Interpreting (which I attended).

A noted expert on Computational Linguistics and anaphora resolution. Prof Ruslan Mitkov of Wolverhampton University, gave a talk entitled "When computers should be used for translation (and when not)?"

To answer your question briefly and amateurishly, the main link is naturally machine translation (MT). The aim is to make MT smarter by going beyond the current rather limited ability for syntact translation by aiming for semantic translation -- in simple terms producing software that will not only translate the words but will understand the context and the meaning of words. For example, English word like "run" or "set" or "fit" have an amazing number of different meanings. Syntactic translation usually stumbles over this. The hope is that semantic translation will understand which of the many meanings is appropriate. But that immediately illustrates the fundamental problem -- computers are brilliant at handling numbers, pretty useless at handling words and languages.

Speech recognition, as I understand it, is in a totally different category. Unlike MT, speech recognition does not understand language -- nor does it need to. SR is merely a statistical process of comparing your digitised voice with a library of phonemes, and then selecting the statistically most probable matches. SR therefore has no need to parse a sentence to understand the subject, verb, object, or to distinguished between nouns and verbs and other parts of speech, as done in computational linguistics.

I am getting out of my depth here, and there is quite a lot of info about these topics on the Web. You might start at this very long URL:


Local time: 21:52
Thank You Apr 28, 2007

Thanks for the reply.


To report site rules violations or get help, contact a site moderator:

You can also contact site staff by submitting a support request »

Role of Computational Linguistics in Translation

Advanced search

SDL MultiTerm 2017
Guarantee a unified, consistent and high-quality translation with terminology software by the industry leaders.

SDL MultiTerm 2017 allows translators to create one central location to store and manage multilingual terminology, and with SDL MultiTerm Extract 2017 you can automatically create term lists from your existing documentation to save time.

More info »
CafeTran Espresso
You've never met a CAT tool this clever!

Translate faster & easier, using a sophisticated CAT tool built by a translator / developer. Accept jobs from clients who use SDL Trados, MemoQ, Wordfast & major CAT tools. Download and start using CafeTran Espresso -- for free

More info »

  • All of
  • Term search
  • Jobs
  • Forums
  • Multiple search