Derived from french.stackexchange's one.

As we have already one question dealing with typography and formatting issues and as there will likely be many of these, I think it would be fitting to set down a typography guide, e.g. like Wikipedia's one. Even more since we will likely have many specific or technical content to write e.g.

  • Syntactic trees
  • Features matrices à la Jakobson, HPSG attributes matrices...
  • IPA transcriptions
  • Glosses
  • ...

Said guide should be a community wiki answer to this question.

Note that I don't suggest a foolish consistency — which is the hobgoblin of little minds — enforcement sacred book but rather a guide to help potentially confused users (me included) keep a legible and consistent style.

| |

These are the most popular propositions so far. Note that these are merely guidelines, and not intended as strict policy rules. This is a community wiki; feel free to improve it.

Words and phrases as a topic of dicussion

When a particular word or affix is the topic of discussion it should be written in italics, which is formatted with asterisks. A short phrase should also be presented in italics, with a translation in quotes afterwards:

(Spanish) feliz cumpleaños 'happy birthday'


IPA transcriptions can be presented inline as in [aɪ.pi.eɪ tɹænˈskɹɪp.ʃəns]. The normal linguistic covention is to use square brackets for phonetic transcriptions (either broad or narrow) and slashes (//) for phonemic transcriptions.

In cases where the phonetic vs. phonemic distinction is irrelevant or tangential to the question at hand, it may be acceptable to use backticks (``) to render the IPA symbols as monospaced text (often called "code formatting"). For exampleː "should I use the single glyph ʧ (U+0287) or the sequence (U+0074 U+0283)?" Beware however that the monospaced fonts may do a poorer job of handling IPA than the main text font (especially when using diacritics), so it is probably best to use code formatting only when it makes the question or answer substantially clearer or removes ambiguity.


Longer or more detailed examples should be presented as three or four line glosses, with the original in the first line, a word by word, or morpheme by morpheme gloss in the second, and a free translation in the third.

Unfortunately there is no good way to present these in the formatting supported by Stack Exchange. For simple examples you can present them using a quote box, with the first line in italics:

Gwenn eo ar c'haz
White is the cat
“The cat is white”

For more complex examples it helps if the words can be aligned. To do this, write the example in monospaced type by indenting each line with four spaces:

Gwenn  eo  ar   c'haz  
White  is  the  cat  
“The cat is white”

Slashes and dashes may be used for compound words:

Bayerische Seenschifffahrt
Bavarian / lakes-shipping
“Bavarian lakes shipping”

Diagrams and trees

You can use EL&U's tools for diagramming

  • Phonetic/phonemic transcripts should not be presented using code formatting. They should not be presented in monospace - that does not match existing linguistic journal/book conventions. Instead, just present them in normal running text with brackets ([]) or slashes (//). – curiousdannii Apr 26 '15 at 13:33
  • before starting an edit war, I'll ask again here (as I did here) why @curiousdannii believes that linguistics.SE ought to look just like a linguistics journal. I'll also mention that the linguistics.SE help page does not call it "code formatting" but rather "preformatted text", and the example is a recipe for cocoa. So I'm not sure why the opinion that "monospace is for code only" is so strongly held. – drammock Apr 27 '15 at 3:50
  • @drammock If it's not going to look like a linguistics journal then what will it look like?? Linguistics style sheets are pretty consistent, and I don't see any reason not to emulate them where possible. – curiousdannii Apr 27 '15 at 14:00
  • I agree that users should use brackets or slashes where appropriate, if they know what they mean and the distinction is relevant to their question. But I also assert that SE is a website, not a print journal, and formatting standards are not the same for print and web. Moreover, SE comes with a built in, easy to use formatting tool that is visually salient and doesn't see much use on linguistics.SE. why not use it to aid clarity? – drammock Apr 27 '15 at 23:32
  • 2
    The EDIT part of this response in English Language and Usage Stack Exchange explains how I use the graphic and textual resources afforded here at *.SE for comment on linguistic matters. – jlawler Apr 30 '15 at 0:09
  • @drammock Because I've seen no evidence that it's helpful. How would it aid clarity? Why do transcriptions need visual saliency? I think perhaps we should make a poll question here on Meta to get the community's opinion about using monospaced text. – curiousdannii Apr 30 '15 at 6:59
  • 1
    The existing conventions for journals already provide visual salience. Proposing a convention that is orthogonal to standard practice in the field cannot improve clarity, since the probability is low that a nominal convention that contradicts everyday linguist's life will actually rise to the level of being an observed convention. I do grant, though, that "preformatted" is useful in overcoming the lack of other good way to present aligned text-and-annoation. – user6726 Apr 30 '15 at 16:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .