I know that diaeresis is used to show that two adjacent vowels are not a diphthong but should be pronounced separately, as in naïve or Zoë. Is there an equivalent mark or format in current or historical use that shows that a pair of consonants that usually form a digraph (e.g. "sh" or "th") should be read separately?

Cases where a word is is made up of identifiable parts are easy to deal with. One can do nothing and rely on the reader's understanding of the separate morphemes (e.g. knighthood) or with true compounds one can put in a hyphen, e.g. pot-hook.

However there is more of a problem when transcribing a word or personal name that comes from an unfamiliar foreign language (so the reader is unlikely to know its spelling conventions), is not a compound, and yet contains a syllable ending with "s" or "t" immediately followed by a syllable beginning with "h", or or another easily misread combination.

Right now I can't think of any words either from English or a from a foreign language which present this problem, but among all the vast multitude of proper names and languages in the world that sometimes need to be written in English it must sometimes occur. It also would come up in transcribing fictional constructed languages so as to sound "alien" yet still be easily readable. In fact my question here was inspired by this question on Writers' Stack Exchange , in which it was asked how to represent words from a fictional language that would be likely to be mispronounced in English.

Inserting a hyphen into a word that is a single unit of meaning seems wrong. Inserting an apostrophe might be better, but an apostrophe suggests either a glottal stop or the marking of omitted letters, neither of which might be present. I seem to recall once seeing a full stop placed between letters to show this but that might have been a quirk of an individual writer. Use of the International Phonetic Alphabet, explanations in brackets, or asterisks all disrupt the flow of reading.

Is there an existing convention or a better solution?

Added later: Some real life examples of words whose pronunciation would be clearer with a consonantal diaeresis: - posthumous - shorthand - Mathias (German proper name) - Kuthumi (name of a nineteenth century Indian mystic) - methemoglobin / methaemoglobin / methæmoglobin (medical term, in which the prefix "met" means "change in") - Ishak (Arabic proper name). In practice with the exception of the occasional hyphen these words seem to have no orthographic device to mark the correct pronunciation, thus answering my question in the negative, unless there are counter-examples I haven't yet met.

  • At first sight your final paragraph seems to exclude all possibilities! Please can you show an example of a text (perhaps with 'knighthood' ) with some indication of how you would mark it with this word without a comment in brackets. Commented Sep 1, 2015 at 11:22
  • Giving the IPA rendering of a word that few English readers will have encountered is surely wise? Transcriptions are often not easy. And 'words from fictional languages' are, per se, off-topic here. Tolkien provided a complete language guide along with his inventions. Commented Sep 1, 2015 at 11:22
  • @chaslyfromUK, alas, I don't know of any such marking. I was hoping that one exists of which I was ignorant, or perhaps a style that has fallen into disuse that I could revive. Commented Sep 1, 2015 at 11:29
  • @EdwinAshworth, I mentioned the question about constructed languages mostly to give credit to the person who got me interested in this issue. As I said, it surely must come up in real life sometimes. Also, I think the topic of English usage can reasonably include questions about whether there is a standard English usage to deal with a potential problem. In practice I'd probably use the IPA or an imitated pronunciation in brackets, but I'm just interested to know if there is a "smoother" way. Commented Sep 1, 2015 at 11:39
  • I'm voting to close this question as off-topic because it doesn't fall within the purview of this site. Commented Sep 2, 2015 at 20:43

3 Answers 3


Maybe you are asking for a diacritic.

How about this for example?


Other symbols that may be useful:

You can find their specifications at http://graphemica.com

If you are writing a scholarly article, all you have to do is define what it means before the main text.

Definition of diacritic in English: noun

A sign, such as an accent or cedilla, which when written above or below a letter indicates a difference in pronunciation from the same letter when unmarked or differently marked.

Oxford Dictionaries

  • 1
    Just out of interest, is this your own invention or something you have seen used elsewhere? Commented Sep 1, 2015 at 11:33
  • What, you mean the umlaut over the h? It is a diacritic but I haven't seen it used in English. Whether it has been I don't know. Probably not! That is why I suggested giving a definition before using it. Commented Sep 1, 2015 at 11:49
  • 1
    Do you have a link that shows how to write this in html? Commented Sep 1, 2015 at 11:54
  • 1
    As Wikipedia says, the "h" in English "sh" and "th" are already "in-line diacritics" anyway, because they modify the sound of the letter preceding them. So antḧill, for example, would imply change /t/ to a dental fricative, then change that to something else. Since the dental fricative could be either voiced /ð/ (as in this) or unvoiced /θ/ (as in thing) it could get a bit tricky deciding which change was intended. A typographically simpler approach might be to just use a capital letter - antHill. Commented Sep 1, 2015 at 12:04
  • 1
    @Lostinfrance - I'm sure you have noticed but in that page I linked to, you can scroll sideways to find other such symbols. I've added a couple to the answer.Some are less intrusive then the umlaut. Commented Sep 1, 2015 at 12:34

No, there is no special dieresis mark for consonants.

In your question, it seems to me that you have already excluded all of the most likely ways of marking that a sequence of "consonant letters" is not a digraph. I would recommend reconsidering your rejection of them (or just not worrying about ambiguity), because I don't know of any better or more established conventions.

Hyphens are not just used in compound words. They are also used to split words across line breaks. And English speakers may be familiar with the use of hyphens as syllable-boundary markers in special contexts, such as dictionary pronunciation guides.

The apostrophe would likely only suggest "a glottal stop" to linguistically sophisticated readers. I think most English speakers don't even know what a glottal stop is, let alone that there could be a contrast between a CC sequence and a CʔC sequence. So in practice, I doubt there is much risk of speakers pronouncing unintended glottal stops here. It's true that the apostrophe may suggest omitted letters.

The full stop (period) is the symbol used in the IPA to mark syllable boundaries. This is a specialist convention that the average English speaker would not find familiar, although the role of the period as a separator at the sentence level might make its use as a syllable divider somewhat intuitive.

Using a transcription or explanation might take up more space, but it's the only way to clearly convey this kind of information about the pronunciation of a word or phrase.

Discussion of specific examples.

The "h" in posthumous is non-etymological, and not normally pronounced at all. This is not a situation where a dieresis would be appropriate. If you are concerned about the possibility of it being pronounced with /θ/, you could use the spelling postumous; this is archaic and would be eccentric today, but I think no more eccentric than spellings that resort to diacritics (like "posthḧumous", "postḥumous" or "postḫumous").

Shorthand is a compound of "short" and "hand", so your objection to the use of the hyphen doesn't seem to apply here. The Oxford English Dictionary indicates that hyphenated spellings were used in the past.

Mathias is an example of the digraph "th" (used in Latin as an equivalent to the Greek letter θ). Since "θ" is a single letter (representing a single Greek phoneme), and the name is pronounced in German and in English with the single phoneme /t/ (not a sequence /t.h/), I can't see how a dieresis would be appropriate here.

Kuthumi seems to have often been spelled "Koot Hoomi", with a space, in the past. If you think it is necessary to use a spelling that more clearly conveys the pronunciation of this name, you could switch to using this one. The Wikipedia article on him even indicates that the spelling "Koot Hoomi" is used in the signatures of "K.H.'s early letters to Sinnett".

Methemoglobin doesn't really need a dieresis. When "t" and "h" were brought together in Ancient Greek, they combined to form the sound "th" (θ), as in cathode < Greek kathodos (κάθοδος) < kata- + hodos. The OED records both pronunciations with /t.h/ and pronunciations with /θ/ for methemoglobin. While analogy with hemoglobin and awareness of the presence of the met(a)- prefix probably will cause most English speakers to naturally pronounce /t.h/ in this word, it shouldn't be seen as an error to use /θ/.

Ishak is an Arabic name, so English speakers who don't know Arabic will mispronounce it no matter what you do, and English speakers who do know Arabic will probably know how to pronounce the "ambiguous" sh based on familiarity with the name. If despite this, you do want to give an accurate indication of the pronunciation, it would be essential to use some widely accepted system for transcribing Arabic; in the ALA-LC scheme, the name would be written Isḥāq. If you use the number-based transcription of Arabic consonants that is often seen on the Internet, you could write the middle consonant cluster as "s7". Although English speakers won't know how to pronounce the 7, at least they will probably know that they don't know how to pronounce the 7, as opposed to not knowing that they don't know how to pronounced the "sh" in "Ishak".

  • 1
    The Arabic name is also one case where an apostrophe would make some sense. If a lack of knowledge of Arabic is presumed, a lack of knowledge of Arabic transcription schemes may also be presumed. If you just want an easy way to make the Anglicised spelling Ishak less ambiguous, you could go for Is’hak, which most readers would presumably recognise as a foreignish name with an apostrophe they don’t know what means – but not a name with the sound /ʃ/ in it. Commented Jan 4, 2019 at 10:49

Distinguishing between unionized (being in a union) and unïonized (not being ionized) is an example where a consonant/vowel diaeresis mark seems to be genuinely useful. I can't think of a non-contrived consonant/consonant equivalent though.

  • 1
    From personal experience, I know that haphazard is not necessarily etymologically obvious if you first encounter it in writing. As a child, I long thought it was pronounced /ˈhafəzərd/, and I think I was a teenager before I realised that it was actually the spelling of the word I knew from spoken language as /hapˈhazərd/. It’s not a minimal pair, but it is a case where actual ambiguity and error can be introduced. Anthony (variously pronounced /ˈantəni/ and /ˈanθəni/) is another example of actual ambiguity. Commented Jan 4, 2019 at 10:52

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.