Language desk
< November 16	<< Oct \| November \| Dec >>	November 18 >

Welcome to the Wikipedia Language Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

November 17

Is the t in match silent??

We often label the t in the word match as a silent letter. However, listen to the ch sound carefully. It has a t at the beginning; without the t sound it becomes the sh sound. If the t in match were objectively silent, it would be pronounced exactly like mash. Anything I am confusing?? Georgia guy (talk) 02:52, 17 November 2023 (UTC)[reply]

We do? I certainly wouldn't call this t a "silent letter". I suppose some people might, if they already think of the ch digraph as representing /tʃ/, but that's sort of an odd take here, given that "mach" is pronounced -- depends on context, but /mɑːk/ or /mɑːx/ seem like the only real choices.

Honestly the notion of a "silent letter" is mostly for teaching kids; it's not that useful a concept when you look at English spelling and phonology in detail. Is the w in "bow" (as in bow and arrow) silent? The word is pronounced with a diphthong (/boʊ/ or /bɛʊ/ depending on which side of the pond you're on) and it's not completely unreasonable to say that the w stands for /ʊ/, but it's also not completely clear, given that you'd pronounce "bo" the same way. --Trovatore (talk) 03:00, 17 November 2023 (UTC)[reply]

Yes, I'm "some people", and I say you have it right in your "I suppose". "Ch" represents the two sounds "tsh", so the T in "match" is silent. "Mach" is irrelevant because it's not pronounced as it's spelled. The spelling corresponding to "mash" is "mash". --142.112.221.156 (talk) 04:10, 17 November 2023 (UTC)[reply]

To be clear, I think that's wrong, or at least a less than useful account. The ch digraph sometimes represents the /tʃ/ consonant cluster, say in the word "chip". But I don't see any justification for claiming that it represents that cluster in the word "match". Where in English does ch following a vowel represent /tʃ/? Let's try a few: Bach, Dachshund, Mach — not coming up with any after "a". How about after "e"? "Tech" is the first one I can think of. No "tsh" sounds yet.

After "i"? OK, maybe "rich"; so this is English and you're going to get a few exceptions, but it doesn't seem to be common for ch after a vowel to represent /tʃ/.

Can't think of any after "o". After "u" there's "much" and "such". So so far I have three words where ch represents /tʃ/ after a vowel, and many more possibilities where it doesn't. Go ahead, play along; see what you can come up with. --Trovatore (talk) 04:23, 17 November 2023 (UTC)[reply]

Interesting though — it does seem to depend on the vowel. The ones taught as "long vowels" (for example /iː/ and /uː/) seem to be much more prone to being followed by a /tʃ/ cluster spelled as ch. Reach, leach, mooch, hooch, etc. Or maybe it's not the vowel quality but the fact that the vowel is written as two letters? That would explain "touch". In any case I don't see that this has much bearing on "match". --Trovatore (talk) 04:39, 17 November 2023 (UTC)[reply]

There's definitely a T sound in there. Looking at etymonline, it seems that the roots of a number of these -atch words were originally spelled with a double-C.[1] If the ch in match is considered to already include a T sound, then maybe you could argue that the explicit T is "redundant" rather than "silent". ←Baseball Bugs ^{What's up, Doc?} carrots→ 05:24, 17 November 2023 (UTC)[reply]

I think there are two or possibly three reasonable readings, none of which are "silent t":

You could take the three letters "tch" as a trigraph pronounced /tʃ/, without worrying too much about which component letter makes which sound. Personally I think this is the most reasonable option. I can't think of any exceptions in English to "tch" being pronounced /tʃ/.
You could analyze it as t+ch, where the ch digraph takes the value /ʃ/, as in "chiffon" or "chanty". That's nice because it's intuitive as to how it treats the t, but it's a little suspicious because of the rarity of rendering ch as /ʃ/ in genuine English words.
You could analyze it as t+ch, coming out as /ttʃ/, more or less as Bugs was suggesting. I don't quite buy it because English (almost) doesn't have phonemic gemination, though if you want to stretch a point you can find some minimal pairs, like "night train" vs "night rain.

But considering the t to be silent? No. That's silly. --Trovatore (talk) 07:09, 17 November 2023 (UTC)[reply]

As an aside, aren't there some phonologists who consider /tʃ/ to be a separate phoneme in English? It even has an IPA symbol, I think, but I haven't managed to find it. --Trovatore (talk) 07:18, 17 November 2023 (UTC) [reply]

I'm actually suprised that some consider /tʃ/ a consonant cluster rather than a single affricate consonant. But then, my native language, Polish, makes a very strong distinction between /t͡ʂ/ (a single phoneme) and /tʂ/ (a cluster). The word czy /t͡ʂɨ/ 'whether', mentioned two threads above, makes a minimal pair with trzy /tʂɨ/ 'three'. In English this distinction is not phonemic, so I guess it's a moot point. — Kpalion^(talk) 09:22, 17 November 2023 (UTC)[reply]

@Kpalion: thanks for pointing me at the affricate article — interesting stuff. The minimal pair worst shin versus worse chin is actually reasonably convincing. (I wish I'd known about night rain versus night train when I was learning Italian; I think that would have helped a lot with understanding gemination.)

I still can't find a single IPA for the ch affricate and now don't think there is one, but Americanist phonetic notation seems to have many different variants of it, all some sort of c with various excrescences. --Trovatore (talk) 21:25, 21 November 2023 (UTC)[reply]

IPA doesn't use single glyphs for affricates, by design... 惑乱 Wakuran (talk) 00:58, 22 November 2023 (UTC)[reply]

Historically, (Middle) English consonants were doubled in spelling to indicate a short vowel; cf. hater vs. hatter. For the two affricates, normally spelled ⟨ch⟩ and ⟨g⟩, the "double spelling" came to be ⟨tch⟩ and ⟨dg⟩, instead of the more straightforward *⟨chch⟩ and *⟨gg⟩. A very similar spelling rule applies in German, where the affricate normally spelled ⟨z⟩ is "doubled" to ⟨tz⟩, instead of the more straightforward *⟨zz⟩, when indicating a short vowel.

This also explains why Trovatore can find many instances of ⟨-ch⟩ after historically long vowels, but very few after the short ones. A nice minimal pair is beach vs. bitch. 147.234.66.93 (talk) 12:50, 17 November 2023 (UTC)[reply]

This makes a lot of sense. I suppose the same goes for ⟨k⟩ vs ⟨ck⟩ (e.g., liked vs licked). — Kpalion^(talk) 13:45, 17 November 2023 (UTC)[reply]

I think we have a winner! That seems to make Bugs right after all; it's "redundant" rather than "silent". You wouldn't say that either of the t's in "hatter" is silent. --Trovatore (talk) 17:54, 17 November 2023 (UTC)[reply]

The trigraph -tch- is quite useful, but is never afaik found at the start of a word. Except for some rusted-on transliterations of a few Russian surnames that start with /tʃ/, such as Tchaikovsky and Tcherepnin. But Chekhov is no longer Tchekhov since wiser heads prevailed. -- Jack of Oz ^{[pleasantries]} 21:02, 17 November 2023 (UTC)[reply]

In the 19th and early 20th century the spelling Tschechow,^[2] copied straight from the standard German transliteration, was more common. --Lambiam 00:33, 18 November 2023 (UTC)[reply]

The obvious one that comes to my mind is tchotchke, though whether that counts as an English word is debatable, I suppose. --Trovatore (talk) 04:09, 18 November 2023 (UTC)[reply]

Also Tchad Blake and Tchad, seen in atlases where native spellings are preferred. 2A02:C7B:301:3D00:81DD:C2C4:7AA0:EE73 (talk) 15:17, 18 November 2023 (UTC)[reply]

You are right: English words don't start with tch- for the same reason why they don't start with double consonants, dg- or ck-. 147.234.72.96 (talk) 20:52, 19 November 2023 (UTC)[reply]

I take your general point. (Except for "I am bleakly proud of the crowds of slimy street-sweepers claiming their free glucose-stuffed platitudes and truculent drop-bears streaming through their flashy phantasmagoria.") -- Jack of Oz ^{[pleasantries]} 22:00, 19 November 2023 (UTC)[reply]

Djack, you mmust have ckompletely mmissed the ppoint. — Kpalion^(talk) 23:22, 19 November 2023 (UTC)[reply]

Mmoi? Pperish ththe ththought. -- Jack of Oz ^{[pleasantries]} 07:47, 20 November 2023 (UTC) [reply]

Tchaikovsky of course comes through French, where ch means only /ʃ/. —Tamfang (talk) 20:30, 22 November 2023 (UTC)[reply]

More than half

When the moon is between half and full I have always referred to it as "gibbous" (with a hard "g"). Holly Willoughby does likewise [3]. The Concise Oxford Dictionary confirms this. However, Webster's New Collegiate Dictionary (London, 1975) gives "jib-...gib". Do some Americans pronounce this word with a soft "g"? 2A00:23D0:C78:B501:79E1:2FE5:1008:424B (talk) 09:57, 17 November 2023 (UTC)[reply]

As someone who lives in North America I've only heard it start with a /g/ instead of /d͡ʒ/. It appears out of the 54 instances noted on YouGlish, only 1 pronounces it as the latter, and that's from a UK field reporter. —Tenryuu 🐲 ( 💬 • 📝 ) 11:20, 17 November 2023 (UTC)[reply]

Wiktionary only gives the "hard g" pronunciation, which is what I always assumed. Most words with the spelling "gi" or "ge" and a similar etymological path into English are pronounced with a "soft g" (dzh), however... AnonMoos (talk) 12:05, 17 November 2023 (UTC)[reply]

Softly, softly. No one mention GIF. --Trovatore (talk) 19:06, 17 November 2023 (UTC) [reply]

A very controversial topic. 惑乱 Wakuran (talk) 10:46, 18 November 2023 (UTC) [reply]

The words gear, geat, geck, gecko, geek, geezer, gehenna, geld, gelding, gerenuk, get, gibbon, Gibbs, Gibson, gift and – at least in the UK – words prefixed with giga- are pronounced with a hard g. --Lambiam 11:40, 18 November 2023 (UTC)[reply]

Lambiam -- Of the words you mentioned, only "gehenna" and "giga" have a Latin etymological connection, as far as I know, and they're somewhat special cases -- "gehenna" is of Hebrew origin and could have been reshaped by an attempt to go back to the original Hebrew, while "giga-" belongs to the realm of unit prefixes, where they feel quite free to make stuff up ("yotta-" etc). AnonMoos (talk) 22:36, 18 November 2023 (UTC)[reply]

I think "giga" may be temporal rather than spatial. As recently as, I don't know, maybe the 1970s, the soft-g pronunciation was fairly common, maybe even the standard. But not too many people outside of engineering and science had much use for it. I suspect the hard g became standard as some sort of side effect of its exploding usage in computing. --Trovatore (talk) 20:27, 18 November 2023 (UTC)[reply]

OED lists both hard and soft for words starting with giga. My 1983 Chambers also lists both. DuncanHill (talk) 20:31, 18 November 2023 (UTC)[reply]

Some may recall that in Back to the Future (1985) Doc Brown uses the soft-g pronunciation of gigawatts. Deor (talk) 22:50, 18 November 2023 (UTC)[reply]

The online OED gives /ˈɡɪbəs/ (GIB-uhss) for British English, and /ˈdʒɪbəs/ (JIB-uhss) and /ˈɡɪbəs/ (GIB-uhss) for American English. DuncanHill (talk) 15:14, 18 November 2023 (UTC)[reply]

As a counterexample, the US-origin word "Gerrymander" is usually pronounced with a 'soft g', although it derives from a politician surnamed Gerry pronounced with a 'hard g'. I wonder why that came about. {The poster formerly known as 87.81.230.195} 51.241.161.192 (talk) 01:30, 18 November 2023 (UTC)[reply]

Perhaps people intuitively think of Gerald and its related names? Bazza (talk) 09:29, 18 November 2023 (UTC)[reply]

The online OED gives a soft g for both British and American pronunciations of gerrymander, but notes "N.E.D. (1899) gives the pronunciation as (gerimæ·ndəɹ) /ɡɛrɪˈmændə(r)/". DuncanHill (talk)

I've always pronounced the name Geraint with a soft "g", but it seems I've been doing it wrong. 2A02:C7B:301:3D00:81DD:C2C4:7AA0:EE73 (talk) 15:24, 18 November 2023 (UTC)[reply]

You have been doing it wrong, there is no "J" sound in Welsh apart from some modern loan words like garej and ffrij which you can work out for yourself. Alansplodge (talk) 19:14, 18 November 2023 (UTC)[reply]

I remember thinking Geraint had a soft G when I was about 8 or 9 and had never met one. Of course if it were an English name one would expect it to be soft given the e, and it does tend to get confused or conflated with Gerontius, which I rather think is more commonly pronounced with a soft G, despite the best efforts of the etymologistical wing of the musicologists. DuncanHill (talk) 22:11, 23 November 2023 (UTC)[reply]