Talk:Dutch alphabet

name of letters

I'll need know how man calls the y, simply as [i] or as [ej].
and ÿ is call lange ij?
name for letter u is [u] or [y]?
and q, man calls it as [ku] or [ky].

I think y is called [ej], u is [y], q is [ku], am I right? However I can't imagine the spelling for ÿ.

The letters b, c, d, g, h, j, k, p, q, t, v, w are called like [be:], [ce:], etc., with a large vowel? I think they are pronounced more large than the letters f, l, m, n, r, s, x, z.

It'll be interesting to put on this remarks in this article. Really I am developing another article in Catalan Wikipedia and I'll want to know these data. THANKS
—Ludor 19:14, 21 August 2005 (UTC)

In Dutch language many diphthong vowel-combinations are used (for example: ie, oe, ui, ei, ij, but also oo, aa, uu and ee are different from pronouncing the single vowle twice).

Two of them sound like Y. The ij is called "lange ij" (long y) and the ei is called "korte ij" (short y). The character Y is called "Griekse ij" (Greek Y), "i-grec" or "ypsilon".

ÿ is not used in Dutch, it has no special name (but sometimes in handwriting the combination of ij will look like ÿ)

The y indeed usually appears in words borrowed from other languages. The y/ij confusion appears, but as far as I know officially the y is the 25th letter of the alphabet and the ij is seen as a combination like ui, oe, ie and ei.

I am not used to phonetics, so I find it hard to answer your other questions.

- Wouter, 30 October 2006

"Y" (as a separate letter) is most commonly pronounced as "Griekse ij" (Greek y), "i-grec" (from French) or informally simply "ij". Ÿ does not exist, "lange ij" is used to distinguish the diphthong "ij" from "ei" (which carry the same pronunciation but a different spelling). U is called IPA y, not u (which is spelled "oe" in Dutch, as in "voet"). Q is called IPA ky, not ku (koe, which would mean "cow", haha). Salaskan 18:34, 26 May 2007 (UTC)

IJ and Ÿ

In French Wikipedia man wrote:

The cluster ij (graph : IJ/ij) is considered as a whole letter and it is collated after the letter i. Into the past, in certain regions, it was wrote with or without trema (graph : Ÿ/ÿ)

Here it is not write as that. Where is it right?
—Ludor 01:27, 22 August 2005 (UTC)

ÿ has never been an acceptable variant for ij. However in old versions of Dutch ij was indeed written as y or i. The Afrikaans language still uses y where Dutch uses ij.

As for collation, this all depends on where you look. When ij is treated as a single letter, it is either collated on its own after x, or combined with y (after x). When treated as i+j, it is collated under i. Jordi·✆ 05:44, 26 September 2005 (UTC)

Clarity

IJ is sometimes treated as a single letter indeed (e.g. with capitalisation and word games), but is never given a place in the alphabet. The Dutch alphabet is the standard Latin alphabet, and children are taught at school that the alphabet consists of 26 letters. The current article is quite confusing, as it says that the Dutch alphabet has 27 letters according to some, but it is always considered to have 26 letters by native speakers. I think native speakers create the language... Anyway, I think it should be noted in the article that IJ is generally not treated as a separate character, but rather as a special kind of digraph (which is capitalised together, filled in one square in word puzzles and letter separations, etc). Salaskan 18:55, 26 May 2007 (UTC)

I disagree... in elementary school I was taught the alpabet ends with x, ij, z. The y was introduced later as being a letter used in foreign words like 'baby'. And every Dutchman or -woman who recites the alphabet ends with x, ij, z and never ever with x, i-grec, z. That the alphabet consists of 26 letters, that's a thing I do agree with. Richard 12:55, 30 May 2007 (UTC)

When pronouncing the alphabet (in order), we call it ij (iks ij zet), as "Griekse ij" would be a little too long. But have you ever seen it written with "x ij z"? It's always something like this: [1][2][3] SalaSkan 19:39, 8 June 2007 (UTC)

I don't think any of these images depicts a accurate representation of a standard, (hand-) written, modern-age Dutch alphabet. The image on the page IJ (letter), though crude, is more truthful. But since you seem to be on a crusade against the letter 'ij' I remove myself from this endless discussion. Richard 09:59, 11 June 2007 (UTC)

Of course I am not, the article IJ (digraph) describes the status of IJ pretty well. I just remove the info about IJ from this article as it should be included in IJ (digraph) and not here. Also, you're quite right that this image doesn't represent modern Dutch handwriting, that's why it is called "Dutch alphabet (1560)". I'll get a better image sometime soon. SalaSkan 14:55, 23 June 2007 (UTC)

In that case your remark: It's always something like this is invalid. And by the way - in my alphabet I ALWAYS write x, ij, z and never x, y, z. Richard 13:56, 25 June 2007 (UTC)

The 'letterkaart' with the alphabet how it's taught in elementary school: [4] [5]. Scoub 21:32, 27 June 2007 (UTC)

That's how "y" is generally written in handwriting. SalaSkan 16:19, 28 June 2007 (UTC)

On 24 June, Ruud Koot added a similar image to the article. Richard 09:21, 28 June 2007 (UTC)

Image caption

Look, the image clearly shows either a 'ij' or a 'ÿ', I don't really care either way, but it equally clearly shows a "Y" and not a "IJ". The conclusion that because it shows "ij", it has to be "IJ", can just as easy be reversed: since it clearly shows a "Y", it has to be a "y" (or ÿ") lowercase. Either we describe what it actually shows instead of what we want it to show, or we remove it altogether for being not really enlightening (obviously). Fram 18:52, 3 October 2007 (UTC)

A capital IJ and capital Y look the same in (this form of) Dutch handwriting. (And because it cannot be a 'ÿ' (which does not occur in Dutch language) the argument can't be reversed as you claim.) —Ruud 23:09, 3 October 2007 (UTC)

Final devoicing

Woodstone, I'm having a good time trying to figure out the proper wording of the devoicing note with you, but it seems that now's about a good time to bring it to the talk page. It seems, according to this edit, that you're saying that the final devoicing is something other than a phonological process and that the voiceless consonants that emerge are phonemic. This can't be the case. The word for horse, paard, for example, is pronounced [paːrt] and adding the pluralizing suffix -en makes it pronounced [paːrdən]. This alternation upon suffixation is practically the definition of allophony. Do you need greater proof? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 03:28, 19 December 2009 (UTC)

Hi Aeusoes, in my view it is not correct to make a phonemic distinction in final position of /-d/ and /-t/. They are completely equal and there are no minimal pairs distinguishing them. I would look at the example above the other way: the word [paːrdən] when singularised, becomes [paːrt]. The example I put in the article shows this particularly well because it creates homonyms: both [laːdən] and [laːtən] give rise to a fully identical [laːt] in present singular. At most you could argue for /d/ and /t/ as archephonemes, but I always found those a bit shady. −Woodstone (talk) 14:34, 19 December 2009 (UTC)

I've seen that same sort of evidence used for other languages that feature final devoicing (German, Russian) in arguing that the surface realization of a voiceless consonant can actually be a voiced phoneme. In fact, I could get such an example pretty quickly from an introductory phonology textbook, though using that as justification for Dutch would be WP:SYNTH. I could probably scrounge around and find something specifically about Dutch. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 19:06, 19 December 2009 (UTC)

Here we go Final Devoicing and Voicing Assimilation in Dutch Derivation and Cliticization. Clear as a bell on the second page. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:00, 19 December 2009 (UTC)

The examples shown in the document are correct as written, but not according to the definition of what a phoneme is: (WP) "a phoneme is the smallest segmental unit of sound employed to form meaningful contrasts between utterances". The referred article does not note down a "unit of sound", but the "underlying (voicing) specification". So /pɒd/ is just a theoretical construct, not a phonemic representation, which should be /pɒt/, just as there is /lɒt/. The fact that plurals are /pɒdən/ and /lɒtən/ has nothing to do with the sound of the singular. This is most obvious in homonyms. The word pronounced [laːt] can either be derived from [laːdən] or [laːtən]. Listening to the sound alone, there is no way to distinguish the meanings. −Woodstone (talk) 09:24, 20 December 2009 (UTC)

As I said, the interpretation of the data that I've laid out is very common crosslinguistically. A quick search has brought a relevent resource backing it up. Do you have any sources to support your claims? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 05:15, 21 December 2009 (UTC)

The easiest place to look for is the IPA handbook. It states "/b, d/ are fully voiced" and "voiced obstruents and /ɦ/ do not occur in the coda". −Woodstone (talk) 14:23, 22 December 2009 (UTC)

That's a bit vague, don't you think? I mean, by comparison to the very explicit source I presented. Here are some more sources:

• Warner, Natasha; Jongman, Allard; Cutler, Anne; Mücke, Doris (2001), "The Phonological Status of Dutch Epenthetic Schwa", Phonology, 18 (3): 387–420
  • Transcribes (p 398) schillenmand, and bollenveld as /ˈsxiləˌmɑnd/ and /ˈbɔləˌvɛld/, respectively, while indicating final devoicing phonetically on pages 417 and 418.
• Grijzenhout, Janet (2003), "untitled review of The Morphology of Dutch by Geert Booij", Journal of Linguistics, 39 (2): 419–421
  • "Consider in this respect the fact that even though the verbal stem red 'save' ends in the voiced sound /d/, this form never surfaces in isolation. When no morpheme is attached to the stem, a phonological rule applies that devoices final obstruents and the result is the surface form [rɛt]. If a vowel-initial morpheme is added to the stem (e.g. -er, to form a noun that refers to the subject of the verb), the phonological rule mentioned above does not apply and the voiced sound appears (as in redder [rɛdər] 'one who saves, saviour')." (p 419)
• Gussenhoven, Carlos (1993), "The Dutch Foot and the Chanted Call", Journal of Linguistics, 29 (1): 37–63
  • Gussenhoven, the author of the HIPA Dutch article, calls refers to final devoicing as a "postcyclic lexical rule", that is one that applies after morphology.
• Booij, Geert; Rubach, Jerzy (1987), "Postcyclic versus Postlexical Rules in Lexical Phonology", Linguistic Inquiry, 18 (1): 1–44
  • In their discussion of final devoicing (which they call syllable-final devoicing), the authors transcribe, for example, held as /hɛld/ and say specifically that the rule is both postcyclic and lexical, analogous to Catalan's cluster simplification.

Do you have anything more explicit? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:17, 22 December 2009 (UTC)

I fail too see what is vague about "voiced obstruents do not occur in the coda". That is a very explicit statement that Dutch just never has voiced obstruents in final position. Of course it is possible to analyse "underlying" concepts and several scientists do as you have shown, and as is reflected in Dutch spelling rules. But it remains contradictory to the definition of the phoneme as a distinguishing unit of sound. If the sound is not different it cannot distinguish and therefore not be a phoneme. So before the postulated final /-d/ makes sense, the definition of a phoneme would need to be modified (which I would not cherish). Are you aware that the most common spelling mistake in Dutch verbs is exchanging final d's and t's? Native speakers do not feel any difference. −Woodstone (talk) 20:48, 22 December 2009 (UTC)

You're free to publish your theories regarding the underlying nature of final obstruents in Dutch, but until you produce sources with the thoroughness that I have, I'll have to assume that you're providing an OR approach. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 07:04, 23 December 2009 (UTC)

You are confused. The term "underlying" is not mine, but consistently used in the reference you intially gave to describe the nature of the final voiced obstruents. The IPA handbook is very clearly ruling out any use of /b, d/ as finals. There is no question of OR. −Woodstone (talk) 09:39, 23 December 2009 (UTC)

We all know that {[b] and [d] do not appear syllable-finally and this is what Gussenhoven is most likely saying in his HIPA article. Because it's one sentence, because he doesn't use brackets or slashes for what he's saying, it's not clear if he means that it's just the phones that don't appear or the actual phonemes. Let's say, though, that your reading is correct. Two things are still true: 1) it's not explicit enough to use as a source since Gussenhoven neither explains the alternations nor articulates if he means phones or phonemes 2) the body of literature on Dutch overwhelmingly endorses the view I've presented. This isn't just "some scientists", it's everybody else. You have one line from an article giving a quick illustration, I have five articles. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:36, 23 December 2009 (UTC)

Inflection is essential to the whole concept of devoicing. Without any inflected form where the voiced obstruent is heard, talking about devoicing is meaningless. The obstruent in a word that has has no inflected forms where it is voiced can by definition not be devoiced, it is simply unvoiced. −Woodstone (talk) 21:10, 23 December 2009 (UTC)

I was thinking about that after my edit. My problem was more with the wording because it could be read that voiced consonants could appear at the end of a syllable if inflection wasn't part of its appearance. Perhaps it's better to mention inflection with the example, something like "For example, the word paard ('horse') is pronounced [paːrt]; inflectional affixes show that this final [t] has been devoiced, so that paarden ('horses') is pronounced [paːrdən], with no devoicing." — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 21:21, 23 December 2009 (UTC)

I see what you mean. There are in fact two aspects to this:

1. final obstruents in Dutch are never voiced
2. consequently if inflection makes a (voiced) obstruent final, it will be devoiced

So far the contentious footnote is talking only about the second part, but the wording does not address the first. −Woodstone (talk) 13:05, 24 December 2009 (UTC)

I think one of the sources listed above mentions an example or two of words that don't take part in inflection but are still spelled with a final voiced consonant (so it's not clear at all if they have underlying voiced consonants). Should we mention that as a contrast or is that the level of detail that should go at Dutch phonology/final obstruent devoicing? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:52, 25 December 2009 (UTC)

If there are words spelled with a voiced final obstruent, there is problably an etymological reason, but spelling is not essential to the subject. The way you now rephrased brings back the contradiction: "Dutch language does not have voiced obstruents in the syllable coda" vs "whenever such a consonant does occur at the end of a syllable ...". In that case it undergoes final obstruent devoicing, so the voiced occurrence is hypothetical. That is why bringing in the inflection brings clarity, stating that "the voiced obstruent would occur (by inflection)". −Woodstone (talk) 20:03, 25 December 2009 (UTC)

Does my change of "do" to "would" remedy your concern? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:06, 25 December 2009 (UTC)

It's better, but leaves the reader in the dark how this "would occur" comes about. I still maintain that the only way to know that an obstruent has been devoiced is to show a voiced inflected variant. (P.S. what were those words ending in "d" without stemming from inflection? Just curious.) −Woodstone (talk) 22:17, 25 December 2009 (UTC)

My bad, the examples were in German: weg ('away') and sind ('they are'). Dutch weg (the adjective, not the noun) seems to be parallel.

Asking how a voiced obstruent could occur in the syllable coda in Dutch is like asking how initial /t/ could occur or how /bl/ could occur. The reasons are etymological. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:10, 26 December 2009 (UTC)

Yes, but in the case of /-d/ the answer is: it does not occur. So a statement as to what would happen if it occurred is purely hypothetical. But enough said, I can live with the current formulation if you insist. Dutch does not have /g/, except in cultured pronunciation of loan words. Instead the <g> is pronounced /x/ or /χ/, and only in some dialects as /ɣ/. So devoicing is progressed much further. −Woodstone (talk) 09:23, 27 December 2009 (UTC)

Oh yes, I was fooled by the spelling. Well, if wiktionary is any guide, some potential examples: boord, dood, inderdaad, overeind, reeds, steeds, and iemand.

I'm glad we got this squared away. I should be able to incorporate some of the information I found at Dutch phonology. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 00:54, 28 December 2009 (UTC)

There are the related forms (with English cognate, all with d) boorden (boards of ship), dode (dead), daden (deeds, actions), einde (end), bereiden (to make ready) and gestadig (steady). Rarely one can hear the plural iemanden (someone, no English cognate found).

Interesting is the situation in Thai, which is an isolating language (no inflections at all), with no voiced final obstruents. It still has many words spelled with final letters that are voiced when initial. This is purely etymological. For example ดาว /daːw/ (star) and วัด /wat/ (temple), spelled with the same ด. It must be a complete nightmare for children (and definitely is for me) to learn which final to write. −Woodstone (talk) 10:04, 28 December 2009 (UTC)