While answering a question on computational complexity of parsing, I attempted to give a mathematical formula with TeX, but it did not work.

The same formula did work on other SE sites such as physics.

Is there a reason why it is not possible to do that on SE-linguistics?

I realize that most contributions, but not all, tend to discuss purely linguistics phenomena, and are not concerned much the mathematisation of linguistics and the development of computational linguistics.

The unavailability (unless I misused the thing) of mathematical notation can only reinforce this trend, Is it intended ?

This seems regretable. For one thing, people are clearly concerned with issues that only have a mathematical statement and answer, such as complexity issues. Another point is that some pragmatic problems that have appeared are better addressed when properly formalized in a mathematical way (such as the problem of representing ambiguity - sorry for mentionning only topics I looked at).

Also, my impression is that the site is pretty quiet. Making available the tools that would help developping issues in the mathematics of linguistics and in computational linguistics might attract more people, and give a broader scope to the site for the benefit of all.

I would also like to point out that some views and terminology differ between people doing only linguistic theory and people more oriented towards mathematical analysis of linguistic problems.

Typically, that struck me with jlawler's answer to the question What's the difference between syntax and grammar? > Though I am not disputing it, it is certainly not in agreement with the way I used these words in my academic work, which was more computational linguistics.

Being confronted with an increased diversity is always educational, especially for the downvote trigger happy.

  • If I remember correctly, most sites do not support TeX because the javascript library needed to parse and generate the symbols is very heavy. So, as a trade-off, only the sites that cannot live without it (such as math and physics) import such library. Dec 14, 2013 at 17:11
  • @OtavioMacedo Thanks for the reply. I thought the computational cost was on the user side, but I never really checked. However, my comments about the scope of the site and the users activity still stand.
    – babou
    Dec 14, 2013 at 18:25
  • @babou it is on the user side (in latency, bandwidth and load time). I did a writeup on this a while ago: meta.stackexchange.com/questions/203977/… Dec 15, 2013 at 3:24
  • This may not be the answer you were looking for: why not sent CS questions over to Computer Science? Your defined scope seems not to include formal language theory, anyway, and we have lots of domain experts -- and MathJax. (cc @OtavioMacedo)
    – Raphael
    Jan 26, 2014 at 14:26
  • @Raphael If you look at my page in SE-Linguistics, you may see that I never asked a question. Hence my problem was with giving answers, for which the SE site is not for me to choose: I have to stay with the question. I am not sure what you mean by "Your defined scope seems not to include formal language theory". I am concerned with that, among other things. But I was struck by the fact that answers in SE linguistics can be mathematically too naive (not that I consider myself a mathematician). Unavailability of Latex will not help improve this.
    – babou
    Feb 4, 2014 at 10:38
  • By "you" I meant the community as a whole. Everybody (with >= 15 rep) can flag a post for migration, after all. Regarding scope, I was referring to the list in the help center which does not seem to contain formal languages. (Maybe they are implicitly included in one of the items listed?)
    – Raphael
    Feb 4, 2014 at 11:35
  • @Raphael Sorry for misunderstanding. However, I am not sure your suggestion answers the issue. Understanding and use of formal approaches has to be integrated to natural linguistics, not seggregated from it. Also, formal structures used in linguistics have often been ignored by the formal language crowd. Another problem is that, in my experience, the two communities have linguistic and cultural differences. Given how little forgiving (not to say how irresponsible) SE users can be when commenting or voting contributions, your suggestion for migration might only lead to more misunderstanding.
    – babou
    Feb 5, 2014 at 20:36
  • @babou: I appreciate the goal of incorporating formal models into traditionally less formal disciplines, but that does not mean a question may not be better served somewhere else. I suggest you handle it similarly to Computer Science "vs" Mathematics: if you can and will handle a CS question, fine. If you think it does not get the attention/care/expertise it deserves, consider migrating it over.
    – Raphael
    Feb 5, 2014 at 21:56
  • possible duplicate of Use of LaTeX commands Mar 28, 2014 at 18:22


You must log in to answer this question.

Browse other questions tagged .