8

Apologies in advance for the slightly blog-like nature of this question.

The Background

Some of the comments in relation to this question here:

... made astute observations about the difference between words with intervocalic /t/ in Standard British and American Englishes (RP and General American).

It is well know that voiceless consonants cause pre-fortis clipping (a term coined by John Wells 1982) on preceding vowels. So the vowels in beat and bead are significantly different in length. Because /t/ is a fortis (unvoiced) consonant, it causes the vowel in beat to be clipped or shortened in the word beat. The word bead on the other hand ends in lenis (voiced) /d/, so there is no shortening of the vowel in this context.

Now if the /d/ at the end of bead is devoiced for some reason, we will still be able to tell these words apart. The lack of shortening of the vowel will tell our language brains that this consonant is the lenis /d/, and not the fortis /t/. This is very good news, because in actual fact, consonants which are normally voiced such as /b, d, g, v, ð, z, ʒ/ and /dʒ/ will become partially or fully devoiced when not surrounded on both sides by voiced sounds, so for example at the end of a word said in isolation. So in actual fact, the preceding vowel, not the consonant itself provides our brains with the crucial information to identify the consonant at the end of beat or bead as /d/ or /t/.

The Peter Shor point:

So the comment from Peter Shor, which precipitated this question was as follows:

The vowel length certainly plays a large part in decoding the phoneme. But I would think the voiced/unvoiced quality must also play some role in our decoding the phoneme. Otherwise, Americans wouldn't have any more trouble that Brits distinguishing bidder and bitter (and we do).

So, Peter's point here is that in Southern Standard British English bitter and bidder have distinctly different consonants. The first is fortis (unvoiced) and the second is lenis (voiced). And, unsurprisingly British English speakers can clearly distinguish between words like bitter and bidder.

However, in most examples of General American, the /t/ in bitter is realised by a voiced alveolar tap. This variety of English has an allophone (alternative pronunciation) of /t/ when it occurs between two vowels. So /t/ is normally realised by an unvoiced sound, but when it occurs between a vowel and another unstressed vowel, it becomes a voiced tap. Now the thing is that /d/ in American English also becomes a voiced tap in the same environment.

Now, in many, if not most instantiations of General American, the words bitter and bidder are homophones when they have a voiced tap (unlike the words bit and bid). In other words if listeners hear these words in isolation, they will not be able to tell them apart. So speakers of American English occasionally have some difficulty distinguishing these words.

Peter's point, then, is that if listeners weren't able to hear whether an actual /t/ or /d/ were voiced or not, then it seems at first blush that speakers of British and American English would have the same problem with pairs like bitter & bidder and writer & rider. But they don't.

Now just to make things more complicated here, most speakers of standard Canadian English also have a voiced tap for intervocalic /t/ and /d/ in these words too. But in Canadian English these words are not homophones. In other words speakers of Canadian English can easily tell these words apart even if said in isolation!

The question:

So the question is:

  • What is it that causes writer and rider to be distinguishable for speakers of standard British and Canadian Englishes?

  • Why are writer and rider indistinguishable in General American?

Note:

Of course, there are many varieties of American, British and Canadian English. Within these there is quite a lot of variation. I tried to put lots of careful hedges in the question (most, often and so forth). Of course many of the features described in American English may apply to speakers of Canadian and sometimes British Englishes and vice-versa. What I'm most interested in is what features help speakers distinguish these words and how this works in each case.

  • 1
    Possibly, they are not completely indistinguishable. I've read some interesting things about neutralizations that suggests that in some cases, even when speakers can't hear a difference, phonetic analysis still shows some differences on average between words with different underlying "phonemes." I forget where I read about this w.r.t English t-voicing in particular, but I read one paper that discussed it in the context of German final obstruent devoicing. – sumelic Nov 23 '15 at 11:24
  • 2
    Southeastern Pennsylvania near Philadelphia, me. The words rite and ride are not indistinguishable for me and my homeys. The vowels in write and ride are distinguishable not only by the longer vowel in ride (or the shorter vowel in rite) but by their place of origin in the throat. In my dialect and other AmE dialects, the vowel in ride originates much farther back in the throat than the vowel in rite; the latter vowel has a nasal quality. Unless there's ambient noise, we can pretty well hear the difference between the head/nasal tone and the throat/chest tone. – TRomano Nov 23 '15 at 11:42
  • 1
    @TimRomano Yes, but those don't have a word internal intervocalic /t/. What we want to know is whether writer and rider are the same or not - and why? Let us know! – Araucaria Nov 23 '15 at 11:54
  • 1
    @Araucaria: The same is true with writer and rider. The jaw drops on the latter. – TRomano Nov 23 '15 at 11:55
  • @TimRomano Saying that the jaw drops on rider may just be the other side of the coin from saying that writer has raising. – tchrist Nov 23 '15 at 12:29
8

It turns out that writer and rider are not “indistinguishable” in much of the United States. The difference is that although both rider and writer have an alveolar flap in their middles, writer is [ˈɹʌɪɾɚ] with a raised and somewhat shortened diphthong, whereas rider is simply [ˈɹaɪɾɚ].

The original /t/ of write was enough to trigger so-called Canadian raising in the diphthong, since write has it but ride does not. This distinction is preserved in the longer versions ending in ‹r›.

Understand that although the phenomenon is called Canadian, it is by no means limited to that country, but extends to much of the United States as well.

We don’t think of the raising as being phonemic, but it is enough to disambiguate what would otherwise be homophones. In the referenced Wikipedia article, they also point out that this is what allows us to distinguish high school (the one after junior high) from a high school (one that is high).

  • 2
    Isn't it the stress that disambiguates high school compound noun from high school adjective plus noun? – Araucaria Nov 23 '15 at 12:01
  • 1
    Though [ˈhʌɪskul] and [ˌhaɪ ˈskul] have differing primary stress under normal (non-emphatic) prosody, the raised diphthong in the first is a give-away, length and stress notwithstanding. The lack of voicing in the final consonant of tight triggers the raising of that word’s diphthong compared with the one in tide, where the voiced final consonant blocks it. This distinction remains even in environments where the final voicing is lost or the consonant reduces to a flap [ɾ] or glottal [ʔ]. This compensatory phonologic effect is critical for understanding, esp. in fast, connected speech. – tchrist Nov 23 '15 at 12:14
  • 1
    Got it. Right, so what's happening is that the /s/ in the compound is part of the first syllable in /ˈhaɪskul/ triggering /aɪ/-raising and giving us [ˈhʌɪskul]. In contrast the /s/ in /'haɪ 'skul/ occurs at the beginning of the second word and syllable, and for this reason does not trigger /aɪ/-raising. So there we get ['haɪ'skul]. I wish I'd remembered about /ai/-raising before I asked this question!! I'd have chosen a different pair of words... – Araucaria Nov 23 '15 at 13:52
5

In his classic 1962 article "Phonology in Generative Grammar" (Word 18, 54-72), Morris Halle gave the dialect difference between the rider/writer distinguishers and those who don't distinguish as an argument in support of rule ordering as a part of human phonological systems. If rules modifying vowel quality differently before voiced versus voiceless consonants and the flapping rule which neutralizes the difference between t and d apply in different orders, we get different results. We do actually find dialects of English of both sorts, which we can distinguish in a natural way by formulating phonological systems for the dialects with exactly the same rules, but applying in a different order. This shows the linguistic reality of rule ordering in phonology.

This is a rough paraphrase of the argument, according to my recollection, since I couldn't find a copy on line. The article was reprinted in the 1964 collection of essays The Structure of Language.

In response to the question, no, in my own native dialect, I don't distinguish the pronunciations of "rider" and "writer". Do I hear a difference in others' speech? Sure, if there is one, since I'm a linguist.

  • This is exactly what I was angling for. I'd be very grateful if you could unpack it a bit more for readers though, please? – Araucaria Nov 25 '15 at 8:01
  • @Araucaria, I'll refer you to Halle's article for details. Other answers have information about how the t/d distinction affected the preceding vowel: the shorter vowel before t is raised from [a] to caret, in Canada. For a time in phonology there was much interest in order interactions, which has subsided. Paul Kiparsky proposed at one point that opaque order interactions should be expected to be eliminated by historical change, because they are harder to learn. An opaque order is one which results in a rule applying in a derivation when its conditions are not met in the surface (cont. ..) – Greg Lee Nov 25 '15 at 8:22
  • @Araucaria (cont. ...) pronunciation. If the raising to caret rule applies for "writer" because of the voiceless t, then subsequently t is flapped and made voiced, this is an opaque interaction, since there is no surface motivation for the raising of [a] to caret. – Greg Lee Nov 25 '15 at 8:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.