Correct me if I'm wrong -- t + sh = ch. d + zh = j.

[Veris]

I've always thought that the ch and j sounds were similar in the same way that the f and v sounds or t and d sounds are. I just thought -- wait, isn't ch simply t and sh together? And then isn't j just d and zh together? By "together" I mean "both points of articulation being done at once."

So ch and j are voiced/voiceless counterparts, right? Well, I can see that much by checking the IPA -- indeed, they're called the voiced and voiceless postalveolar affricate. But is it true that ch is actually t + sh? Like, it's literally two sounds being made at once? Or do we just say that because it sounds like both being said at once?

[MrKrov]

No. A few langs do contrast affricates and clusters of, say, [t]+[ʃ] and [d]+[ʒ]. Affricates are also not two sounds made at once.

[Ossicone]

Compare 'cat shit' vs 'catch it.' Not the perfect example, but you get the gist between the difference in tʃ and t͡ʃ.

Affricates are not made by saying two phonemes at the same time. Two at once = coarticulation (more or less). In this case it's imposible, because when you close the airflow for the stop any fricativeness from escaping. According to the wiki, which is close to what I remember from class, the release of the stop is the fricative, which creates the affricate. They even used my example above: http://en.wikipedia.org/wiki/Affricate_consonant

Also, IPA or X-SAMPA helps!

[David McCann]

The phonetic difference between an affricate and the sequence of stop + fricative is largely a matter of timing: the affricate takes much the same time to pronounce as the fricative on its own.

In English, a good example is the contrast of "Wiltshire" /wɪlt·šɪə/ and "filcher" /fɪl·čə/. In this case the syllable boundary is important, as /tš/ cannot start a syllable. In Polish it can at the beginning of the word, so you can get a minimal pair like czy /čɪ/ "if" and trzy /tšɪ/ "three".

IPA is misleading (and inconvenient) in its lack of symbols for affricates, but one has to remember it goes back to the 1880s.

[Veris]

@MrKrov: Any example languages that do such?

@Ossicone: Yeah, crappy example -- since the stop between words is in a different place, it doesn't really help illustrate the difference between [t] then [ʃ] and [t͡ʃ].

I don't think IPA was necessary here, though (I don't know the alt codes and don't feel like trudging through Wikipedia to find the letters so I can copy-paste all the time) since I'm speaking in English, so you could assume any letters I was giving outside of [IPA brackets] were the English letters, rather than the IPA sounds. Thus j you could understand as the English 'jay' whereas [j] is the IPA sound.

@David McCann: It does sound like it's just timing.

And yeah, we sorta need a conference or something to create a new IPA. Having fifty tiny variations of the same symbol (e, hooked e, e with squiggle, e with accent, e with dot, e with two dots, e with accent facing the other way, e with hook facing the other way, e with line, e with a large coke and some fries) is annoying and difficult to get used to. Not to mention hellish to write for those of us with naturally small handwriting -- when writing in IPA, I have to make certain letters implausibly bigger than I normally write them to show the little hook or whatever I'm adding. It's annoying.

[MrKrov]

Klallam affricate /t͡s/ in k’ʷə́nc 'look at me' versus stop–fricative /ts/ in k’ʷə́nts 'he looks at it'.

I wish I could've found that article about IPA vs Americanist notation because there were a lot more.

[Ossicone]

Yeah, I suggested IPA because I didn't actually understand you question at first. :oops:

I assume you're using Windows because of you mention alt codes. If it annoys you enough, you can expidite the process with one of the methods described on this page:
http://ipa4linguists.pbworks.com/Windows-IPA:-Main-page (I haven't tested any of their methods, as I don't use windows much, but their advice looks sound.)

The problem with the example is not the the /t/ is in coda/word final position for cat, but rather that I have different vowel for cat and caʧ /cæt/ vs. /cɛt͡ʃ/ (and I feel that this effects the realization of the /t/ in cat). However, if you say the two words quickly it should give you a good impression of how /tʃ/ and /t͡ʃ/ are differentiated.

[sangi39]

AFAICT the difference between an affricate and a similar plosive+fricative cluster is somewhat similar to differences in VOT. The difference between an unaspirated voiceless plosive and an aspirated voiceless plosive is the difference in the time at which the vocal chords begin to vibrate in relation to the time at which the plosive was released. In unaspirated plosives the vocal chords begin to vibrate at the same time or just after the plosive is released while some time after (relatively) in aspirated ones.

Similarly, the difference between an affricate and a similar plosive+fricative cluster depends on the time it takes after the release for frication to appear. In an affricate the frication begins almost as soon as the stop is released, i.e. frication is the release, while in a cluster the plosive is fully released before frication begins.

[jal]

Any example languages that do such?

Slavonic languages, e.g. Polish (czy vs. trzy).


JAL

[Nortaneous]

And yeah, we sorta need a conference or something to create a new IPA. Having fifty tiny variations of the same symbol (e, hooked e, e with squiggle, e with accent, e with dot, e with two dots, e with accent facing the other way, e with hook facing the other way, e with line, e with a large coke and some fries) is annoying and difficult to get used to. Not to mention hellish to write for those of us with naturally small handwriting -- when writing in IPA, I have to make certain letters implausibly bigger than I normally write them to show the little hook or whatever I'm adding. It's annoying.

IPA is pretty much meant for typesetting, and it's a lot lighter on the diacritics than pretty much any other transcription system out there, besides canIPA, which is... well, to call it completely and totally insane would be more than a bit of an understatement.

If you don't like the diacritics, use tone letters instead; tones are the main thing that IPA uses diacritics for, besides maybe nasalization, and diacritics are a shit way of representing them anyway in anything but running transcription, where tone letters would get in the way too much for at least my preference. And borrow symbols from canIPA if you find yourself using the same diacritic+symbol combination a lot and you can't replace it with a simpler IPA symbol.

[zelos]

if you want an example of how they are treated differently think ont "catch it" and "cat shit"

[Khemehekis]

The problem with the example is not the the /t/ is in coda/word final position for cat, but rather that I have different vowel for cat and caʧ /cæt/ vs. /cɛt͡ʃ/ (and I feel that this effects the realization of the /t/ in cat).

Really? You use /c/ instead of /k/ for those words? I've never heard it before.

And yes, I say /kɛt͡ʃ/ instead of /kæt͡ʃ/.

[Ossicone]

The problem with the example is not the the /t/ is in coda/word final position for cat, but rather that I have different vowel for cat and caʧ /cæt/ vs. /cɛt͡ʃ/ (and I feel that this effects the realization of the /t/ in cat).

Really? You use /c/ instead of /k/ for those words? I've never heard it before.

And yes, I say /kɛt͡ʃ/ instead of /kæt͡ʃ/.

No, I don't.
I must have been tired when I wrote that.
I even spelled catch 'caʧ.'
Damn English orthography screwing me over.