Pret-ty trick-y

« previous post | next post »

Today's Zits:

The idiom "You've got to get up pretty early in the [morning] to ___" anchors the context, and helps us interpret the spelling "pret-ty" as implying the division of pretty into two separated syllables, with associated aspiration (rather than flapping and voicing) of the /t/. (And maybe a glottalized copy of the /t/ at the end of the first syllable.)

Other examples of the same thing are out there. COCA has one example of the spelling "pret-ty", from a short story by Alice Munro published in the 3/22/2004 issue of the New Yorker:

A Google search for "pret-ty" turns up what seem to be some more examples of the same prosodic idiom, e.g.

this fire dog stuff can get pret-ty boring    [link]
Pret-ty snea-ky, Senator: GOP tries to conceal weapons in credit-card reform bill   [link]
Yep, this could get pret-ty, pret-ty, pret-ty bad. Or at least it would if, once again, it were true.   [link]

When I hear these five examples in my mind, they're performed with a downstepping intonation on the "pret-ty __" part.

In these examples, the whole package — hyper-separated syllables and downstepping intonation — seems to be associated with ironic and somewhat jocular emphasis, though as usual with such meanings, it's easier to feel the vibe than to give it a precise definition. And the various pieces of the package can certain be used separately or in combination with other performance techniques, though (again as usual) it's hard to tell whether the combinations are fully compositional in meaning.

Question for readers: Is this a purely North American pattern? Or does it exist in England, Scotland, etc., as well?


  1. Pflaumbaum said,

    October 23, 2010 @ 7:43 am

    It certainly exists in England. It is used for ironic/jocular emphasis, but also for other kinds of emphasis. It's sometimes equivalent to that kind of emphatic whistle people sometimes make.

    Repeating it three times to modify 'good' or 'bad' is often a reference to Curb Your Enthusiasm, as it is in your link above. In the show it's a kind of catchphrase in that form.

    It's interesting that when AmE speakers use it, they seem to realise the /t/ as an unvoiced plosive, as many British speakers do, rather than the normal flapped [ɾ] or the vocalic one of 'purdy'. Is that because it's a true doubled consonant, as in Italian, or because the long syllable break is equivalent to starting a new word, or what?

    [(myl) I assume that the voiceless aspirate release of the /t/ is because the separation of syllables disrupts the pattern associated with the flapping-and-voicing allophone (though this doesn't necessarily make the second syllable prosodically equivalent to a separate word). But there's also typically (or at least sometimes) a glottalized closure on the first syllable, which is certainly like what you'd see (in these varieties of English) if there were really two underlying consonants, as in flattop or hat trick.

    One theory about flapping-and-voicing of /t/ (and other sorts of consontant lenition as well) is that it happens when the consonant is "ambisyllabic", or so to speak caught ambivalently between two adjacent syllables. (I believe that this suggestion was first made by Dan Kahn in his 1976 thesis "Syllable-based generalizations in English phonology".) You could view the apparently replicated consonants in examples like these as the result of separating the syllables at a relatively superficial or phonetic level. ]

  2. Pflaumbaum said,

    October 23, 2010 @ 7:51 am

    Sorry, that was a bit confused. The flapped [ɾ] becoming a plosive and the vocalic [ɝ] losing its r-colouring are obviously different things.

  3. Willie said,

    October 23, 2010 @ 8:01 am

    I feel this discussion is incomplete without reference to Larry David and his recurring use of "Pret-ty, pret-ty, pret-ty good" on Curb Your Enthusiasm. (see here: He of course didn't start this usage, but I do feel that he brought it to a new level of cultural recognition.

    [(myl) Thanks to you and Pflaumbaum for pointing out this connection. Curb Your Enthusiasm is one of a number of areas where my cultural formation is sadly lacking.

    Larry David's pronunciation (in the YouTube collection cited) involves a voiceless unaspirated /t/ with an enormously elongated final vowel, which is different from what I imgined and described (which was a glottalized closure with an elongated stop gap and an aspirated release). He's got the same downstepping intonation, though.]

  4. Shawn Weil said,

    October 23, 2010 @ 8:31 am

    The most salient example I can recall was the commercial for the children's board game "Connect Four." Here is is: – it occurs around 0:19.

  5. Pflaumbaum said,

    October 23, 2010 @ 8:43 am

    @ Prof. Liberman – thanks for the explanation and the link, looking forward to reading that.

    One other query. In heavily stressed "WHAT-ever!", I think I'm right in saying most AmE speakers tend to opt for lengthening the first vowel and retaining the flap, rather than the glottal closure + voiceless aspirate you find in 'pret-ty'. Is there an obvious reason for the difference?

    In the many British accents that realise these instances of /t/ as a glottal stop, both options (i.e. retaining the glottal stop with lengthened preceding vowel, or 'doubling' the consonant and realising the second part of it a [t]) seem to be available.

  6. Ellen K. said,

    October 23, 2010 @ 9:17 am

    @Pflaumbaum. I think the difference is that in "whatever", the t goes with the first syllable, if they are broken up. I'd pronounce the t in that heavily stressed WHAT-ever the same as I would in "what" by itself. Which I think is not so much a tap but an unreleased stop, though I'm not sure.

  7. George said,

    October 23, 2010 @ 9:20 am

    Does anyone know what the origin of the orthographically doubled consonant is? Does it have any germinate history in English or German?

  8. language hat said,

    October 23, 2010 @ 10:13 am

    The OED cites Dutch prettig 'pleasant, nice, agreeable, comfortable,' German regional prettig 'sportive, funny, humorous,' and Old Norwegian prettugr 'tricky, deceitful,' so there's definitely a history of gemination, but why I couldn't tell you.

  9. mollymooly said,

    October 23, 2010 @ 10:36 am

    Given the irregular sound-spelling of "pretty", I am surprised how few ghits there are for "pritt y good", "prit tee good" etc compared to "pret ty good". Whereas "pritt y good" makes the original word less clear, IMO it makes the target pronunciation clearer.

  10. John said,

    October 23, 2010 @ 10:37 am

    Shawn beat me to it! Pret-ty sneaky, Shawn!

    (And "Here, diagonally!" is also recognizable as a pre-internet meme, I dare say.)

  11. Dan Milton said,

    October 23, 2010 @ 10:39 am

    Last night I heard ScienceComedian Brian Malow explaining some violation of the Commandments "I'm only Jew-ish".

  12. Pflaumbaum said,

    October 23, 2010 @ 10:54 am

    @ Ellen – true, but I think we're talking about two different outcomes (aside from the pret-ty one that MYL's talking about). One is to more or less completely separate them (What-, -ever!), as you're saying. The other is to considerably elongate the /ɑ/ but then run on as normal into the [ɾ].

    By the way, looking on youtube I found this slightly disturbing clip of AmE speakers teaching a child to say 'Whatever!' with a glottal stop for the /t/:

    I didn't realise AmE speakers had this allophone, though I know it's used for /d/ in AAVE pronunciations like " di'n't ".

    @ George – sometimes a double consonant does derive from a real geminate, e.g. full < PGmc. *full-az.

  13. A. Linguist said,

    October 23, 2010 @ 12:29 pm

    I've been wondering on and off about this for some time, but have never really got around to looking into it in detail. In my experience, the articulation is not simply /t/ with aspiration, but obligatory gemination of the /t/. For me, that is contrastive with the non-geminated form. What I'm curious about is whether this occurs with a very narrow set of lexical items: perhaps "pretty" only. Is "This is a hap-py day!" or "He's a pic-ky eater!" possible for anyone? How productive is this?

  14. John said,

    October 23, 2010 @ 12:41 pm

    I seem to recall the 'pret-ty _____' phrasing from the 1960s US TV show, 'Get Smart'. The bumbling hero, Maxwell Smart, would often offer his appraisal of some nefarious plot as 'pret-ty clever' or 'pret-ty tricky'. There was no irony in these appraisals (beyond the total send-up in the show's premises), but rather approval of the deviousness.

  15. Pflaumbaum said,

    October 23, 2010 @ 1:16 pm

    @ A. Linguist – 'hap-py' is definitely heard, as in "Hap-py New Year!" As far as I can tell that has exactly the equivalent pronunciation to 'pret-ty' as MYL described it: a glottalized closure on the first syllable followed by a voiceless aspirate release.

  16. George said,

    October 23, 2010 @ 1:53 pm

    Maybe my question (9:20am) about doubling the consonants in writing was not clear.

    To my knowledge there, is no phonemic gemination in English today – in speech, 'predy (flapped),' 'prety, pretty, pret-ty, etc. all have the same meaning (the differences would be dialect or emphasis). So why the orthographic doubling of consonants? Does this suggest that at one time in English, or its ancestors, gemination did distinguish meaning?

    language hat demonstrates above (10:13am) that the orthography is found in related Germanic languages. Is/was gemination a (speech) feature of those languages.

  17. Faldone said,

    October 23, 2010 @ 2:02 pm

    A. Linguist: Is "This is a hap-py day!" or "He's a pic-ky eater!" possible for anyone?

    Dunno about the "pic-ky" but I've been known to wish any and all a hap-py New Year.

  18. Kym said,

    October 23, 2010 @ 2:03 pm

    I'm reminded of Looney Tunes cartoons, where Daffy Duck is prone to saying "You've got to get up pret-ty early in the morning to outsmart THIS little black duck," though because the cartoons have been remixed for television so many times over the years, I'm not sure whether it only appeared in one cartoon and was reused, or if it was something of a catchphrase.

  19. Dw said,

    October 23, 2010 @ 2:08 pm


    Yes: English used to have phonemic gemination of consonants, corresponding pretty much to orthographic doubling. I think it was lost over the seventeenth and eighteenth centuries, though I could be out be a century or two.


    I'm interested that you perceive "pret-ty" to indicate an _aspirated_ medial [t]. As a British native, I have a plain unaspirated [t] in that position. Would you say that the [p] in "happy" is also aspirated?

  20. Jarek Weckwerth said,

    October 23, 2010 @ 2:37 pm

    @ George: Yes, gemination is still there in Swedish.

  21. Pflaumbaum said,

    October 23, 2010 @ 2:41 pm

    @DW – Are you sure you don't aspirate it when you extend the syllable break in pret-ty? This BrE speaker certainly does.

    I think the aspiration is due to the separation of the syllables. Normally (I assume for you too) /t/ would be aspirated word-initially, as in 'tea', and before a stressed vowel, as in the first /t/ of 'pretentious'. Whether in the case of pret-ty the /t/ is re-analysed as word-initial, or the second syllable receives more stress, or something else, MYL may be able to tell us. Though in his reply to me above, he's cautious about concluding that the second syllable is prosodically equivalent to a second word.

  22. LQ said,

    October 23, 2010 @ 2:49 pm

    @Dan Milton Yes, a friend of mine does that too (Jew-ish, stress on "ish" and an accompanying hand-waggling motion).

  23. Mr Punch said,

    October 23, 2010 @ 4:26 pm

    IIRC, about 50 years ago, in Beyond the Fringe, Jonathan insisted, "I'm not 'a Jew' — I'm Jew-ish."

  24. Robert Furber said,

    October 23, 2010 @ 7:55 pm

    I think English has a little bit of gemination. In particular, the word 'betweenness' (from graph theory) seems to have a double n. However I don't know enough phonetics to know if that's the real thing or not.

  25. Mark F. said,

    October 24, 2010 @ 1:19 am

    Pflaumbaum — Not only were they using a glottal stop, they were non-rhotic. That's not a normal pronunciation in any American accent I've ever heard. I think they were affecting some kind of London accent, but I'm not sure why.

  26. Pflaumbaum said,

    October 24, 2010 @ 4:03 am

    @ Robert Furber – lots of compound words, eg MYL's flattop and hat-trick examples above.

    @ Mark F. – ah, that'll be it.

  27. George said,

    October 24, 2010 @ 6:06 am

    @DW and Jarek: Thanks for the information. Can you point to a minimal pair with and without gemination?

    If not, maybe English just doubled stops in certain environments and this could have motivated the orthographic convention. As an example, I can think of no orthographically doubled consonants word initially.

  28. George said,

    October 24, 2010 @ 6:12 am

    @DW and Jarek: Thanks for the information. Can you point to a minimal pair with and without gemination? In some languages, gemination distinguishes meaning.

    If not, maybe English, at one time, just doubled stops in certain environments and this could have motivated the orthographic convention. As an example, I can think of no words with orthographically doubled consonants word initially.

  29. Marion Crane said,

    October 24, 2010 @ 6:35 am

    @language hat: "The OED cites Dutch prettig 'pleasant, nice, agreeable, comfortable,'"

    Yes, but Dutch doesn't have the word 'pretig' that would mean something else because there's only one t, which is what geminaton means, right? The number of consonants determines the pronounciation of the vowel, but not meaning.
    We do have the word 'pret', fun, but whether the one stems from the other I don't know.

  30. George said,

    October 24, 2010 @ 7:55 am

    @Marion: "there's only one t, which is what geminaton means, right?"

    Actually it is the other way. Gemination involves doubling of a consonant. An example from Arabic: 'darasa' (to study) vs. 'darrasa' (to teach). The difference is pronounced and since the meaning is different, it is phonemic.

  31. Mark F. said,

    October 24, 2010 @ 10:03 am

    George — Marion's full quote should have been (brackets added):

    [Dutch doesn't have the word 'pretig' that would mean something else because there's only one t], which is what geminaton means, right?

    Basically the same idea you were getting at with 'darasa' vs 'darrasa'.

    George — I can't think of any true minimal pairs of words, but I can think of a place where it distinguishes meaning at word boundaries: 'black cops' vs 'black ops'.

  32. CS said,

    October 24, 2010 @ 12:09 pm

    Responding to the commentary on the youtube video, prefaced by saying that I am a midwestern-all-my-life college student not studying linguistics (just interested in language): That pronunciation of "whatever!" sounded pretty American to me, if a little '90s. Earlier in this thread, people have discussed pronouncing "whatever" with an emphasis on the first syllable, which I can sort of imagine, but among people of my acquaintance it's more frequently said with an emphasis on the second syllable in a valley girl style, "wha-EVa" or "wha-EVer"; I personally would either drop the "t" entirely or pronounce it with a distinct glottal stop.

  33. Jerry Friedman said,

    October 24, 2010 @ 11:49 pm

    @George: If I remember correctly, unaimed and unnamed are a minimal pair for gemination. (Am I the only one who wants to spell that word with two ms?

    @Pflaumbaum: I'm pretty sure I say rotten, mountain, etc., with a glottal stop for the /t/. The following syllable is a syllabic n. I think this is pretty common in America. Some younger Hispanic people here in northern New Mexico have what sounds to me like a glottal stop and an un-nasal vowel in rotten: [ˈrɑʔən] or even [ˈrɑʔɪn]. (I am not a linguist.)

  34. Pflaumbaum said,

    October 25, 2010 @ 5:01 am

    @ Jerry – yes, you're absolutely right now I think about it. Does it happen with syllabic /m/ or /l/? My instinct says not but maybe your Hispanic speakers do?

    Also, that's the same environment as the glottal stop for /d/ in AAVE di'n't. I don't know distribution for that allophone at all though. Would you hear su'enly for 'suddenly', for instance?

  35. GeorgeW said,

    October 25, 2010 @ 7:41 am

    @Jerry Friedman: "If I remember correctly, unaimed and unnamed are a minimal pair for gemination."

    Thanks. I think it occurs in English only at word and morpheme boundaries.

  36. Xmun said,

    October 25, 2010 @ 10:35 am

    Yes. My father, who learned English as a third language (after mother-tongue Maltese and Italian), and who generalized too broadly the rule that double consonants in English are not geminated as they are in Italian, always used to pronounce "penknife", for example, as [pɛnaɪ̯f].

  37. GeorgeW said,

    October 25, 2010 @ 2:24 pm

    @Xmum: Interesting. Is gemination phonemic in Italian?

  38. Xmun said,

    October 25, 2010 @ 7:25 pm

    I don't know Italian, but yes gemination is phonemic in that language. One example is enough. "Dama", for example, means "lady". "Damma", on the other hand, means "fallow-deer".

  39. Aaron Davies said,

    October 26, 2010 @ 5:06 pm

    for me there (at least) two emphases possible for "whatever"–the "long A" one, which has the standard american flapped t (from "latter", etc.) and the "syllable break" one, which doesn't. i'm not sure what consonant it does have, but it's not just the glottal stop of (sterotypical) AAVE "wha'eva", and it's not exactly the same as in "what" said in isolation. (but it may be the same as "what" said at speed in the middle of a sentence, e.g. "i don't know what you're doing".)

  40. Aaron Davies said,

    October 26, 2010 @ 5:08 pm

    gemination seems to a be a major feature of japanese

  41. Pflaumbaum said,

    October 27, 2010 @ 8:52 am

    @ Aaron – is it possible that your "syllable break" what-, -ever has something close to a geminate glottal stop? i.e. the /t/ is realised as a pre-glottalised, unreleased [ʔt] followed not by [tʰ], as in MYL's pret-ty, but by a glottal stop. Something like [(h)wɑʔt.ʔɛvɚ]

    Mind you, this is based on my own pronunciation of "what-, -ever" in my best AmE accent (which is probably something like that of a native Canadian brought up in New Jersey by Nebraskan wolves).

  42. Epik said,

    November 10, 2012 @ 3:43 am


    "You've got to get up PRET-ty early in the morning to outsmart THIS little black duck!"

    Did you ever find out what episode that's from?

RSS feed for comments on this post