Transcription, lenition and allophonic variation
« previous post | next post »
I doubt that many native speakers of American English will recognize this word:
But with a little more context, more people will get the message:
And if we play the whole pause group, it becomes obvious:
The clip comes from Terry Gross's 10/22/2015 interview with Sarah Silverman, at about 12:13 of the download, and of course the transcript is
gee he didn't even wait to get his braces off you know because
The OED gives /ˈdɪdn(t)/ as the US pronunciation of the word "didn't", with this online sound clip:
That pronunciation actually has six clear phonetic segments in it (more if we divide stop closures and releases):
And the IPA form of Sarah Silverman's pronunciation is [dɪn], not [dɪdn̩t] or [dɪdn̩] (supposing that the OED's 'n' is meant to be syllabic):
Sarah Silverman is of course not the only person ever to reduce "didn't" to [dɪn] — here's Carrie Brownstein, at about 11:24 of her 10/27/2015 Fresh Air interview:
not true — didn't even mention Buffy is Fluffy in the book
On the other hand, "didn't" isn't always reduced to that point — here's Terry Gross at about 9:51 of that same interview:
and so growing up in- in the house that you did
with a father who didn't acknowledge to
<
As indicated in the image above, an IPA transcription of that rendition of didn't would be something like
[dɪdn̩ʔ]:
… or maybe [dɪɾn̩ʔ], though in this case the medial /d/ is arguably a weak voiced stop rather than a tap.
This is the commonest pronunciation of the word "didn't" in even rather formal American English — the second-syllable reduced vowel and the final /t/ are hardly ever seen.
But that division into five phonetic segments obscures important aspects of what's going on, as we can see in a schematic articulatory score:
In the second syllable, the tip of the tongue closes off the oral tract for the [d], and this closure remains in place through the following nasal murmur and glottal stop; the velum opens to create the release of the [d] into the syllabic nasal [n̩]. and then remains open through (much of the) glottal stop; the glottis constricts to close off the nasal murmur — and then remains constricted as the velum closes again in preparation for the start of the initial vowel of acknowledge.
It's that articulatory spreading that makes the syllabic nasal a more natural outcome than the vowel+nasal combination. And it's the tendency to weaken non-pre-stress /t/ and /d/ to the point of losing the oral closure that makes it natural to turn the final /t/ into nothing more than a weak glottal stricture, and to weaken the medial /d/ to the point of extinction.
The OED is not the only dictionary to get the pronunciation of this word wrong.
If we look up didn't in the online version of cmudict, an open source pronouncing dictionary widely used in the speech technology field, we get
D IH D AH N T
which is the arpabet version of IPA [dɪdənt].
The full cmudict lists four variants (IPA equivalent added):
DIDN'T 1 D IH D AH N T = [dɪdənt]
DIDN'T 2 D IH D N T = [dɪdnt]
DIDN'T 3 D IH D AH N = [dɪdən]
DIDN'T 4 D IH N T = [dɪnt]
All of these are wrong in one way or another — either they postulate a second-syllable schwa segment, or they have a final [t], or both.
The online Merriam-Webster gives this pronunciation field
ˈdi-dənt, -dən, dial also ˈdit-ən(t) or ˈdint
which (eliding stress and syllabification) I think is meant to correspond to the five IPA variants
[dɪdənt], [dɪdən], [dɪtənt], [dɪtən], [dɪnt]
with this audio:
And the Wiktionary gives these pronunciations:
IPA: /ˈdɪd(ə)nt/
(General American): [ˈdɪɾn̩(t)], [dɪʔn̩(t)]
with the audio
What's the point of this little example?
Well, the first thing is not news: dictionary pronunciations don't give a very good account of how people actually talk.
And the second, related point is not news either: IPA segment sequences — or other ways of segmenting and alphabetizing pronunciation — are not a scientifically very satisfactory representation of speech.
There are at least three reasons for this failure:
- Articulatory gestures interact in terms of a multi-layered "score" rather than a single segmental sequence;
- The lexical representation of word pronunciation is digital, but there's a complex, context-dependent, and variable analog process of "phonetic realization" between symbols and sounds;
- The result of that process is often acoustically similar or identical to the output expected from a different input, e.g. where a gesture is weakened to the point of changing its nature (say from stop to fricative to approximant to nothing), or where two gestures merge to create what might have been originally just one.
As a result , there are three plausible but fundamentally different accounts for a given observation: modification of the lexical pronunciation in the process of phonetic realization; adoption of a different lexical pronunciation; symbolic modification of the lexical pronunciation between the lexicon and the process of phonetic realization.
There's more to say about this, but that's enough for now.
Thorin said,
January 8, 2017 @ 9:54 pm
I don't know if it's because I'm from Michigan or what, but I still say "I didn't even" like "I'd'n even".
FM said,
January 8, 2017 @ 10:40 pm
I keep hearing a little nasal release in all the examples, I think they're saying [dɪdⁿ]. And I heard the initial isolated example (before reading anything else) as "did". But maybe I'm hearing ghosts.
(And Thorin, "I'd'n even" is definitely possible for me and I've never lived in Michigan.)
Elonkareon said,
January 8, 2017 @ 10:51 pm
I likewise heard the initial example as "did" before reading anything besides the opening sentence. Of course this was after several replays as the soundbyte began and ended faster than I could process it.
ella said,
January 9, 2017 @ 4:10 am
the first clip didn't play for me at all
Bob Ladd said,
January 9, 2017 @ 7:50 am
It may not be news that "IPA segment sequences … are not a scientifically very satisfactory representation of speech", but there are still an awful lot of people who either haven't heard the news or, perhaps more accurately, don't want to hear it. Practically all phonological grammars simply assume a segmented symbolic "output", and IPA phoneticians who certainly know about the kinds of phenomena MYL discusses in the post resist the idea that segmented representations of speech are therefore a problem (see e.g. Klaus Kohler in Phonetica 69:254-273 (2012), DOI:10.1159/000351218) (probably not free to view, unfortunately).
Cervantes said,
January 9, 2017 @ 8:08 am
Bob, Kohler raised three questions about your chapter. Do you feel he addressed any of them reasonably?
Andrew Usher said,
January 9, 2017 @ 8:49 am
I got it immediately after one listening to the first clip. I don't see how one couldn't, simply by process of elimination (what else could it be?). I would transcribe it, with the best precision standard IPA allows, as [dɪ̃ʔ]. The glottal closure may not be complete, but the 't' is not completely vanished; it's certainly not homophonous with 'din' as this article's transcription would have us believe. I think this is true in general.
[(myl) There's no evidence for a "glottal closure", incomplete or otherwise, at the end of the realization of the /n/ in those two clips. In particular, there's no local drop in amplitude and no change in the sequence of pitch periods of the sort that would indicate glottalization. If you're hearing it, I submit that it can only be via the "phoneme restoration effect", i.e. a perception created by top-down expectations.]
The dictionary transcriptions are not 'wrong'; they're supported by audio of native speakers, and can you believe they're all intentionally mispronouncing? The OED's example (though I'd put schwa, not KIT, in the second syllable, at least phonemically) is what I'd say in the most careful speech, without specifically trying. And if dictionaries record only one pronunciation it should be the careful one, not some approximation of running-speech reduction processes that can't be adequately represented in IPA anyhow.
k_over_hbarc at yahoo.com
Adrian said,
January 9, 2017 @ 8:51 am
I know I'm used to this sort of thing, but I heard "didn't" and "didn't even" without prompting. In fact I think quite a lot of people hearing the "din" of the first clip would think it had to be "didn't", but maybe I'm being too generous.
Mark Meckes said,
January 9, 2017 @ 9:28 am
I didn't listen to the clips (someone else working in the same room who I didn't want to disturb), but I happened to guess exactly what the word in question would be from the first three sentences of the post.
I'm not positive, but I think that when speaking quickly I sometimes pronounce something like /dɪʔn̩/ (I don't know the IPA; I'm making my best guess based on this post) — i.e., not reducing quite as far as Silverman does, but with some sort of noticeable stop in place of the second d. I'm trying to figure out whether I would ever pronounce a final [t]. I suspect that if I were ever trying to enunciate that precisely, I'd be much more likely to say "did not".
Vulcan With a Mullet said,
January 9, 2017 @ 10:29 am
First clip didn't play for me either… but with the 2nd clip I had the gist.
Bob Ladd said,
January 9, 2017 @ 11:15 am
@ Cervantes: I'm tempted to reply "no comment". I considered writing a reply when Kohler's article came out, but decided against it.
I'm handicapped in answering your more specific question right now, because I had to move all my books and papers home when I retired, and I can't actually find my copy of the article in question. No doubt a further search would locate the correct box, but "no comment" remains an attractive option.
Joe said,
January 9, 2017 @ 1:34 pm
Reminds me of this alternative pronunciation in "Oh No You Didn't" meme replicated in this Family Guy snippet. I guess the "tongue-tip closure" of the second "d" is removed from the second syllable – is that correct?
Bloix said,
January 9, 2017 @ 2:08 pm
I am trying very hard to say din for didn't in a sentence ("he didn't even wait") and I can't quite manage it. I have a slight cessation or restriction of breath between the I and the N as my tongue flaps up to the palate. So "di'n." For din as in "it's time for dinner" or "the din was deafening" I have no such cessation.
I am a mid-Atlantic speaker with a trace of New York. To me, "He din even wait" would sound deep south or Texan, along the lines of bidness for business and foo'ba' for football.
Cervantes said,
January 9, 2017 @ 2:43 pm
I had a fourth grade teacher who pointed out that "jeet jet?" means "did you eat yet?" Many people — including those actually running for the office — call it Prezneh Unigh Stay. Many consonants become glottal stops or disappear completely, and of course all unstressed vowels are schwas. I don't know how we manage to communicate.
Bloix said,
January 9, 2017 @ 2:44 pm
PS- unfortunately I am not able to listen to the embedded links (I never can on this blog). I went to the Terry Gross interview and I hear Silverman saying "di'n." This may because of an illusory "phoneme restoration effect", or it may that I and others are hearing something that doesn't show up in the graph.
David Eddyshaw said,
January 9, 2017 @ 4:32 pm
I got it on the first listen of the first clip, though I suspect that the form of the post may have tipped me off before I played it.
Might there be a Brit versus Yank difference in this? (Surely all Adrians are British?)
Coby Lubliner said,
January 9, 2017 @ 5:22 pm
"IPA segment sequences … are not a scientifically very satisfactory representation of speech."
As much as the IPA has been tinkered with in order to achieve that goal, it hasn't really gone that far beyond its original one, which was to help French students approximate the sounds of foreign languages.
speedwell said,
January 10, 2017 @ 11:04 am
After mentioning this to my Irish husband (I'm an American resident in Ireland who previously in Texas), he observed that my "didn't" resembles "dnn", with a prolonged n sound and a nearly full reduction of the "i". He says other Texans do the same thing… this being Houston, specifically.
Bob Ladd said,
January 10, 2017 @ 11:06 am
@Coby Lubliner: Actually, if I understand the history right, the original purpose of the IPA was to help French students deal with the vagaries of English orthography, though the idea of a "universal alphabet" lurked in the background right from the start.
Jerry Friedman said,
January 10, 2017 @ 3:18 pm
David Eddyshaw: (Surely all Adrians are British?)
I've known two or three American Adrians.
Alec said,
January 10, 2017 @ 5:46 pm
The problem with this, to my basically British ear, is that I often find it almost impossible to tell if Americans are saying "can" or "can't". Obviously this can/can't be embarrassing.
Bloix said,
January 11, 2017 @ 12:50 pm
Alec – in my mid-Atlantic US accent, the difference between "I can do it" and "I can't do it" is a cessation of breath between can't and do, and no such cessation between can and do. My tongue never leaves my palate between the end of can't and the beginning of do. Also, ordinarily I reduce the a to a schwa in can but not in can't (unless I'm stressing for emphasis – "your wrong, I CAN do it").
Rose Eneri said,
January 12, 2017 @ 2:07 pm
I'm from Philadelphia. I find it odd when people articulate the middle d in didn't. It sounds to me like an overcorrection. My second syllable is all swallowed up in my throat as if I'm imitating a frog.
To Alex and Bloix: When I first moved to the South (Baltimore) I could never tell whether people were saying can or can't. In Philly, we used different vowels for these 2 words. The vowel in can't is tense, as in trash can. The vowel in can is much more open, like the a in apple. So, I don't have to articulate the t at the end of can't to be understood. The different vowel makes the meaning clear.
Phonetics Weekly 2/2017. – "Phonetics & Phonology" said,
January 15, 2017 @ 4:39 pm
[…] Transcription, lenition and allophonic variation | Mark Liberman | Language Log […]