Crosspost from Latin SE.

Oxia (Q on English keyboard) to the left, tonos (semicolon on English keyboard) to the right.

When polytonic Greek was abandoned, they for some reason chose to design the tonos (modern Greek, right in the image) different from the oxia (Ancient Greek, left in the image). In most fonts, there is verily no difference at all, but in some (I would assume those with a longer tradition, such as Garamond), the difference is quite pronounced; as demonstrated in ‘What is the difference between the accent on q and the accent on semicolon?’ (Latin SE), the Ancient Greek accents mirror each other, whilst the tonos has a steeper angle than the ancient oxia.

We have had a conversation in comments and chat on Latin SE, and although modern Greek is off-topic (on Latin SE), unless in relation to Ancient Greek, this question was deemed on-topic due to it concerning the transition from Ancient to Modern Greek. This question was therefore originally posted on Latin SE, and from the conversations in chat, I was recommended to crosspost this to a different SE that might be better qualified to answer my question. Though being on-topic on Latin SE, it is considered unlikely to get an answer, due to the question’s topic being on the side of the expertise found there.

Thus, my question is focused on the development from Ancient to modern Greek. I am especially considering the transition from Ancient to modern Greek grammar and typography, leading to the questions: Why are these symbols different from each other? How did the oxia develop into the tonos?

  • 2
    Not sure this has anything to do with graphic design to be honest. It's a pretty specialist question you'd probably be better asking an expert in ancient Greek about - someone who is also into paleography perhaps.
    – Billy Kerr
    May 16, 2021 at 23:25
  • Seems to be some info here: en.wiktionary.org/wiki/oxia#English - I've no idea if it's correct or not. Sorry
    – Billy Kerr
    May 16, 2021 at 23:31

2 Answers 2


This is pure speculation, only informed by my experience of typeface design:

In polytonic Greek, you want to ensure that an oxia is not confused with other diacritics, in particular the varia (βαρεῖα, grave accent). For this purpose, it is better when the two accents are not very steep. By contrast, in monotonic Greek, there is not much to confuse a tonos with. If you have a diacritic over a vowel, it can only be a tonos or diaeresis and those are pretty easy to distinguish. Thus you have more liberty in designing the accent and can make it steeper if you think that befits the rest of your typeface.

To give an analogue from other languages: If I design a typeface exclusively with the German language¹ in mind, it doesn’t matter if my ö looks like this: ő. Almost every German reader will parse an ő as an ö without further thought, because the language has only one type of diacritic and even loanwords do not feature any other double diacritic. There are indeed some typefaces (in particular when emulating handwriting) whose ö looks rather like an ő. On the other hand, if I design a typeface for the Hungarian language¹, I must ensure that ö and ő are visually different, as this distinction matters for that language.

Another idea was that a steep oxia is somehow more difficult to combine with other diacritics, but it rather seems the other way round.

¹ or more precisely: its standard orthography
  • As a comment to what you say about the design of the German ö: In the cursive I was taught, ø i rendered as o̍ (and being that it is written with a slant, the vertical line is parallell to the slant). In fact, Norwegian uses all three accents: ò, ó and ô, and that needs to be rendered in handwriting. You never get an ø with an accent, but an o̍ and an ó can look quite similar.
    – Canned Man
    May 20, 2021 at 8:34
  • That sounds like a horrible convention, in particular as I don’t see any difficulties to render an ø in cursive. Does anybody actually use this outside primary school?
    – Wrzlprmft
    May 20, 2021 at 8:42
  • No problem at all; as you yourself state in your answer, there is no reasonable argument to make that a German would misread ő for ö. Similarly, in Swedish typesetting, e.g. in cartoons, it is common to write ä and ö as ā and ō; nobody would mistake them for a/o with macron. Similarly, in the generations before me, going at least as far back as Gothic cursive, it was common to write u as either ū (Latin cursive) or ŭ (Gothic cursive). So although o̍ and ó can look similar, context dictates interpretation, so there really is no reason to expect misinterpretations. Also, diacritics stand out.
    – Canned Man
    May 20, 2021 at 10:30
  • @Wrzlprmft Virtually no one uses o̍ (or o͗, as was often the case) in primary school anymore. It was the standard way of writing ø up until around the 1940s or 1950s, but it’s very rare nowadays in the younger generations – at least in Danish, and I assume also in Norwegian (Danish can occasionally use ó, but not ò except in loan words). Also compare Icelandic, where the accent is usually very vertical (a̍ e̍, etc.), to Irish, where it’s usually quite horizontal (a᷄ e᷄, etc.) – and then Baltic languages where it’s completely horizontal (ā ē, etc.). May 21, 2021 at 10:05
  • I began primary school in 1987, and we were taught that way of writing (o with a stroke) from the very first year. Som quick googling on Norwegian newspapers (‘skjønnskrift’ or ‘løkkeskrift’), revealed that it is still taught in many schools at least as late as 2012.
    – Canned Man
    May 21, 2021 at 21:54

Nick Nicholas's website is a good source for information on Greek typography and Unicode.

Nicholas says that in the monotonic system, which became official in 1982 and was increasingly used in the '70s, the accent initially was often written as a dot or wedge. However, "the Greek government decreed in 1986 that the tonos shall be the acute." Based on this, in modern Unicode there is no canonical distinction between the precomposed letters with tonos and oxia in their names (and both are equivalent to the base letter and a combining acute accent).

I don't know the history of how the distinction came to be found in typefaces such as Garamond, but it sounds like it would date to less than a century ago, and probably is not representative of any very long typographic tradition.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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