Talk:Digraph (orthography)

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Writing systems (Rated Start-class, High-importance)
WikiProject icon This article falls within the scope of WikiProject Writing systems, a WikiProject interested in improving the encyclopaedic coverage and content of articles relating to writing systems on Wikipedia. If you would like to help out, you are welcome to drop by the project page and/or leave a query at the project’s talk page.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 High  This article has been rated as High-importance on the project's importance scale.

English examples[edit]

How about 'bough'? 'gh' isn't just /f/ or /g/. (talk) 09:27, 28 December 2007 (UTC)

From the article, under "Sequences": ". . . or is silent at the end of words . . . " (talk) 03:22, 18 January 2008 (UTC)

Please let me know some 'ch' words announcing /ʃ/.--Octra Bond (talk) 13:39, 7 July 2008 (UTC)

All of French origin: chaconne, chagrin, chauvinist, chef. −Woodstone (talk) 15:06, 7 July 2008 (UTC)


My data for ph, ff, f and ll was correlated from Welsh language, (when precisely listed, I added a {{ConvertIPA}}, maybe I should also create {{FullIPA}} for lack of IPA references?) and Omniglot. Unless you have other opposite references, I will restore all the IPA sounds references tomorrow. --Circeus 05:50, 31 Dec 2004 (UTC)


What about Maori language Wh and Ng? Can someone please research these, as I know they are listed in Māori language dictionaries as single letters. However I'm not a linguistics scholar and can't write anything meaningful about them for the purposes of this article.


There are more digraphs than shown here (Slovak 'dz' and 'dž' for example). Rmpfu89 19:35, August 19, 2005 (UTC)


Basque language indicates that 'tx' is not a digraph; this page states otherwise. --Ghewgill 19:47, 1 January 2006 (UTC)

Cyrillic and Devanagari[edit]

Why has neither Cyrillic nor Devanagari any digraphs? -- 12:57, 11 February 2006 (UTC)

Cyrillic has a lot of digraphs: all combinations of a consonant and a "hard" or "soft" sign can be considered a digraph. −Woodstone 14:14, 11 February 2006 (UTC)


If no one objects, I'm going to write a Quadragraph article.Cameron Nedland 02:25, 3 March 2006 (UTC)

Scandinavian digraphs?[edit]

Norwegian, Swedish, and Danish have one digraph (sj) in common; Norwegian and Swedish also share the kj and tj digraphs. Should I revise the article to give the relevant information? (Since all Scandinavian languages are Germanic, the ng digraph should also occur on that list...) -ISNorden 01:01, 14 July 2006 (UTC)

Go ahead and add your knowledge. No need to ask permission. −Woodstone 06:59, 14 July 2006 (UTC)

dr an english digraph?[edit]

hello, i was wondering if "dr" (as in drive, or drink) counts as a digraph or is it something else? thanks, Naufana:Talk 21:06, 19 September 2006 (UTC)

No, it's pronounced /d/ + /r/. FilipeS (talk) 20:44, 15 January 2008 (UTC)

Czech digraph dž[edit]

Czech digraph dž is also used in these words of czech origin:

- džbán (jug) 
- džber (ewer, pail, tub)

so it does not occur in words of foreign origin only

  An alternative (and correct) spelling is čbán and čber.

Should only digraphs that are considered letters be listed?[edit]

I've been leaning towards that option. The other cases can be exemplified in the specific articles for each language's alphabet or orthography. Any thoughts?... FilipeS 23:04, 3 January 2007 (UTC)

Well someone needs to sort out what this article is about, because it starts A digraph, bigraph or digram is a pair of letters but you are saying the article is only be about single-letter digraphs.
Joe Llywelyn Griffith Blakesley talk contrib 01:14, 2 June 2007 (UTC)

No, I am not saying that. I am (was) asking whether the article should be only about digraphs that are regarded as letters. As a matter of fact, though I have changed my mind since I wrote that comment. I think the article can discuss both types of digraph, the ones that are "one" letter, and the ones that are two. FilipeS 10:34, 2 June 2007 (UTC)

My point was that the lede should be altered to reflect the fact that digraphs may be single letters. I have now done so, although it still could still be significantly improved. —Joe Llywelyn Griffith Blakesley talk contrib 16:33, 2 June 2007 (UTC)

Thanks; that's even better now. However, I don't agree with calling them characters not graphemes: a character only refers to the representation in a computer and this is an orthography not a computing article, and, more importantly, multiple characters may be used to represent a single grapheme (even in Unicode). —Joe Llywelyn Griffith Blakesley talk contrib 18:27, 2 June 2007 (UTC)

You are thinking of the meaning of "character" in typography, but the word can also refer to "a mark or symbol used in a writing system". References here. FilipeS 18:39, 2 June 2007 (UTC)

? Please excuse my ignorance, but how can a digraph be "one" letter? (talk) 16:36, 15 January 2008 (UTC)

See Welsh alphabet, Hungarian alphabet, or Serbo-Croatian language, for instance. FilipeS (talk) 20:44, 15 January 2008 (UTC)
Of course (also Spanish [traditional], Schweizerdeutsch et al.). Thank you for your response. I was just having a little difficulty following the thread of this discussion because the lede question is somewhat unclear, and "single letter digraph" sounds nearly oxymoronic. Also, there are certain implicit ambiguities in words like "character", "letter", "mark" and "symbol". (talk) 03:02, 18 January 2008 (UTC)

Swahili ng'[edit]

Swahili ng' is considered a digraph or a trigraph? Why it isn't neither in digraph nor in trigraph articles? -- (talk) 22:25, 25 March 2008 (UTC)

To do list[edit]

Arabic: Kazakh vowels, Tatar ng, Kurdish uu = /y/. kwami (talk) 09:20, 6 August 2008 (UTC)

Yoruba: gb and sh. Bukky Olawoyin (talk) 16:29, 19 May 2015 (GMT)


What do we do w Irish? Are all V sequences digraphs etc., or do we need to place the vowels with the consonants when they're used to distinguish broad from slender? Or neither, as in Cyrillic? kwami (talk) 12:19, 11 August 2008 (UTC)

What about foetus and anaemic ?[edit]

So I come to this page trying to find out about words like foetus/fetus and anaemic/anemic ( British/American spellings ), and the way they are sometimes printed with the letters stuck together. Seems these are digraphs which should get a mention on this page somewhere.Eregli bob (talk) 05:09, 13 March 2010 (UTC)

I believe you're thinking of ligatures. — Ƶ§œš¹ [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi] 06:58, 13 March 2010 (UTC)
Well ligatures are a typography issue, which is not the whole story with anemic and anaemic. If it was just about ligatures, it would be a question of whether the "a" and "e" are stuck together, and not a question about why Americans leave out the second "a" in anaemic entirely. So really, there are three variant spellings, anemic, anaemic and an<ae>mic with the "a" and "e" stuck together. I think it is a digraph. Anyhow it is a very common source of confusion and should be mentioned somewhere. And why isn't the capital of Arizona Phenix ?Eregli bob (talk) 22:30, 19 March 2010 (UTC)

Hangul vowels as multigraphs[edit]

I wrote that most Hangul vowel glyphs are multigraphs, but somebody undid the revision, arguing that 'no, no more that "i" and "j" are multigraphs'. However, I stand by my claim.

First, "i" and "j" cannot be multigraphs, because they cannot be broken down into valid Latin letters (except perhaps Turkish dotless i). On the other hand Hangul vowels can be (at the time of its invention, before ㆍ was obsoleted).

Second, there is textual evidence that Hangul vowels are multigraphs. Section 制字解 (Character Design Explanation) of Hunminjeongeum says,

中聲凡十一字。 ㆍ舌縮而聲深,天開於子也。形之圓,象乎天也。 ㅡ舌小縮而聲 不深不淺,地闢於丑也。形之平,象乎地也。 ㅣ舌不縮而聲淺,人生於寅也。形之 立,象乎人也。 此下八聲,一闔一闢。

(Trans.: Middle sound [vowel] is comprised of eleven characters total. For ㆍ, the tongue recedes and the sound is deep, as Heaven opens from "子". Its shape is round like Heaven. For ㅡ, the tongue recedes a little and the sound is neither deep nor shallow, as Earth opens from "丑". Its shape is flat like Earth. For ㅣ, the tongue does not recede, and the sound is shallow, as Man comes from "寅". Its shape is standing, like Man.)


(ㅗ is like ㆍ, but mouth contracts; the shape is formed by combining ㆍ and ㅡ, taking the property of Heaven beginning to interact with Earth.)


(ㅏ is like ㆍ, but mouth lengthens; the shape is formed by combining ㅣ and ㆍ, taking the the property that the action of Heaven and Earth starts from objects, but to occur, waits for Man.)

And so on. Thus the rationale for Hangul states explicitly that the vowels are multigraphs.

In light of these facts, I feel justified to undo the deletion. —Preceding unsigned comment added by (talk) 11:29, 22 October 2010 (UTC)

Assuming that was actually the motivation for the design, rather than a post hoc rational, the fact remains that these are not sequences of letters today. At best they'd be ligatures, but even that would be dubious.
Also, please read WP:BOLD. If an edit is reverted, you should discuss it first, not edit war over it. — kwami (talk) 11:38, 22 October 2010 (UTC)
As far as I'm aware, it's an accepted mainstream view that these were the motivations for the vowel design. Some speculate that the Daoist metaphysical explanations could be post hoc, as these arguably don't flow with the rest of the text and as the extant texts are not first editions. However, there's no concrete evidence to support this nor is there an alternate explanation.
I'm not sure what you mean when you say that these are not sequences of letters today. Each individual part occurs on the same location it has always been, so time does not seem to be the issue.
Is your concern with the fact that the dot morphed into a short stroke and became conjoined with long strokes? The dot morphed during the evolution of the Palace script to accomodate ink brushes. If joining of glyphs is a problem then the Arabic and the Peace sign sections would also be invalid, and cursive English script would have no digraphs.
Is your concern with the fact that each vowel has a unique Unicode code points? If so, neither should "ㅚ" be considered a multigraph.
Is your concern with the possibility that these are not perceived as a sequence of letters by an average literate Korean person? If so, anyone who's texted in Hangul using a Samsung cell phone would disagree. The public school system teaches about the vowel designs, so it is at least possible that it is perceivable to Koreans that vowels are composites of letters.
Anyways, regardless of what you speculate about the current Korean orthography, I think the historical account of Hangul's vowel design as composition of glyphs is relevant in this article. —Preceding unsigned comment added by (talk) 01:02, 23 October 2010 (UTC)

Spanish digraphs[edit]

The section on Spanish digraphs seems to say the reverse of the Spanish textbook !Claro Que Si! 3rd Edition by Garner, Rusch, and Dominguez, Houghton Mifflin Company 1996. The textbook says that ch and ll were eliminated from the alphabet in 1994, but the article says "they are both considered part of the alphabet". The textbook also lists rr as a letter in its alphabet, but the article says that it was never a letter. Not really my field, but I remember learning rr as part of the alphabet in Spanish class decades ago, so I checked.--Wikimedes (talk) 08:37, 12 July 2011 (UTC)

What is the opposite of a digraph called?[edit]

I.e. single letters like "j" and "x" which actually represent two sounds.Adrigon (talk) 22:36, 4 February 2012 (UTC)

Some linguists I have asked suggest "diphone" (or "triphone"). However, unlike "digraph", a diphone can mean any combination of two phonemes, not just such a combination that is represented by one grapheme. Perhaps a descriptive term like "diphonemic grapheme" would work. Simon Abbott (talk) 12:05, 15 March 2014 (UTC)


"〈eu〉 represents /ɔ͡ʏ/ (open-mid back rounded vowel) followed by (near-close near-front rounded vowel)"

E.g. "Europa" (Europe) is "[ɔɪ̯ˈʀoːpa]" at de.wikti and "/ɔʏ̯ˈʁoːpa/" at en.wikti. So is it "<eu> = /ɔ͡ʏ/", "<eu> = /ɔʏ̯/" or "<eu> = /ɔɪ̯/"?

"〈ss〉 [..]. In German, this digraph was fused into the letter ß."

That's wrong or at least misleading. <ſz> (or sometimes <ſs>) became <ß>, thus the German name "Eszett" or "Esszett", which comes from "Ess" (<S>, <ſ> (long s) and <s>) and "Zett" (<Z> and <z>). -Yodonothav (talk) 12:52, 11 July 2014 (UTC)

English: ia, ou(r), (t)ch?[edit]

IPA letters accourding to

  1. E.g. Russia (/ˈɹʌ.ʃə/), Asia (/ˈeɪʒə/, /ˈeɪʃə/), Croatia (/ˌkɹəʊˈeɪ.ʃə/, /ˌkɹoʊˈeɪ.ʃə/), differential ([dɪfəˈɹənʃəɫ]), special (/ˈspɛ.ʃəl/).
    Is ⟨ia⟩ a digraph becoming /ə/ or is it like ⟨sia/tia/cia⟩ being trigraphs becoming /ʃə/?
  2. E.g. colour (/ˈkʌl.ə(ɹ)/, /ˈkʌl.ɚ/). Is ⟨ou⟩ a digraph becoming /ə/ or is it like ⟨our⟩ being a trigragraph becoming /ə(ɹ)/ɚ/?
  3. E.g. watch (/wɒtʃ/, /wɑːtʃ/). Is ⟨ch⟩ a digraph becoming /ʃ/ or is ⟨tch⟩ a trigragraph becoming //? On the one hand it's sometimes ⟨ch⟩ = // and often ⟨t⟩ = /t/ thus one could put it as ⟨wa[t]{ch} ⟩=/-[t]{ʃ} /; on the other hand it's watch but not /-ttʃ/.

-IP, 20:44, 12 July 2014 (UTC) & 22:47, 12 July 2014 (UTC)

ou and au for Dutch too. (talk) 14:04, 11 February 2016 (UTC)


The Hungarian ny and the Italian gn are for sure not the same sounds. Which of the two is written incorrectly here with the IPA? — Preceding unsigned comment added by (talk) 00:22, 8 March 2016 (UTC)

/θ/ (voiceless interdental fricative) in 'this'[edit]

"th" in "this" is not voiceless /θ/, is it not? -- (talk) 03:08, 12 February 2017 (UTC)