Making equations with greek letters is easy -- for example, $2 \gamma = 1$. But is there any way to do it with arabic letters -- something like $2 \gim = 1$? I've run out of greek letters in my equations, so arabic is the logical next step...

Oh, and I have looked at this but it is... complicated... For example, I have no need for right-to-left support or the ability to make sentences. Just a single letter!

  • 3
    Welcome to TeX.SE. Wow, you have really used all the Greek letters, upper and lower case variants?? – Peter Grill Oct 10 '12 at 0:14
  • Hi Hooligan, you can use backticks (`) to mark inline code. That's what puts the fancy colored boxes around it :) – Scott H. Oct 10 '12 at 0:15
  • 3
    there are other alphabets. my guess is that a lot of people aren't as familiar with arabic as they are with cyrillic, so you might want to consider that. many arabic letters have the same basic shape, and differ only in the application of dots (above, below, or centered). since dots are also used to indicate differences in mathematical meaning when using greek or italic letters, this could be inherently confusing with arabic to someone who isn't familiar with that alphabet. – barbara beeton Oct 10 '12 at 12:31
  • 4
    I believe that Hebrew alphabet is a better choice. It letters are quite unique. As well, try to get the LaTeX Comprehensive Symbol List (google finds it easily), and there're some things like Runs, Smileys, (Web)dings, etc. You can use these as well. – yo' Oct 10 '12 at 23:24
  • I am a physicist, and we use a lot of different letters. The same letter can mean different things in different disciplines or even within the same discipline and topic. I have never ever heard of anybody running out of letters, when using good decorations, and sub-/sup-scripts. I have never heard of, or seen anyone using arabic letters. IT IS NOT A LOGICAL STEP. – hpekristiansen Mar 16 '13 at 1:21

Maybe this is only a comment, but I really mean it, so I post it as an answer.

If you've run out of all the latin, greek and \mathcal letters, then you're using so many variable names that the reader won't manage to remember them all! Seriously, please try and use less different letters.

I think one shouldn't use more than the 2*26 latin, some 24+7 greek and 26 \mathcal letters. (Probably it's too much already if you use them all.) In addition, you could use "speaking" indices, like a_{\textup{in}} and a_{\textup{out}}. Instead of \mathcal you can also use \mathfrak (which has additional 26 lower case letters), but I wouldn't combine them unless you only want to use a few of those. (Another thing: please don't use both \epsilon and \varepsilon, or both \phi and \varphi.)

Generally, when writing something, always think of the readers and make it as simple possible for them (but not simpler!). It does happen to me, too, that I think I'm out of letters; then I think the material through again and try to understand it better before writing it down.

| improve this answer | |
  • It's very typical in high energy physics to use \epsilon and \varepsilon (for instance one for a spinor, and the other for the totally antisymmetric tensor). – Bruno Le Floch Mar 15 '13 at 21:21
  • @Bruno: I was writing this from the point of view of a mathematician, where the notation is a lot more flexible. You're right, in some sciences there's some fixed notation that everybody adheres to, and then it's no problem to use the two different epsilons. (It is a huge problem if they're used in some ad hoc notation for something the reader hasn't seen before, and I've seen it happen!) – Hendrik Vogt Mar 15 '13 at 21:34
  • I agree that there is a big difference between notations that are "canonical" in a domain and those specific to the author. The problem of running out of Latin letters can happen when working on relations between two domains with fixed notations of their own. For instance in a recent paper we used all Latin and greek letters except o, u, x, y, E, G, L, O, P, T, X, zeta, iota, nu, omicron, rho, upsilon, chi, Delta, Theta, Xi, Phi, Psi (and we were using all the \var... ones). Trying to keep this consistent from one paper to the next while adding new objects is not easy. – Bruno Le Floch Mar 16 '13 at 4:26
  • @Bruno: I didn't say it's easy - and you had 23 letters left, after all :-) – Hendrik Vogt Mar 16 '13 at 8:00

I think I'd second barbara beeton's comment. If you still want arabic letters: I don't know anything about typesetting equations in LaTeX, but if you are using LuaLaTeX, you could try using the following as your preamble:


    [Script=Arabic,        % to get correct arabic shaping
     Scale=MatchLowercase] % adjusting the font size
    {Scheherazade}         % whatever Arabic font you like

\newcommand{\textarabic}[1] % Arabic inside LTR
           {\bgroup\luatextextdir TRT\arabicfont #1\egroup}

The you should be able to typeset arabic characters when ever you want by using the command\textarabic{...}:

2 + x = 4 - \textarabic{ع}

I don't know whether or not this would work in some equation or math environments.

| improve this answer | |

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.