Laurence Horn laurence.horn at YALE.EDU
Mon Apr 5 17:38:39 UTC 2010

At 4:23 PM +0100 4/5/10, Damien Hall wrote:
>Larry and Paul commented on the non-pronunciation of the /s/ in
>'Fudgesicle'. Possibly it is dialectal, but isn't the simplest explanation
>phonetic / phonological? It seems to me very likely that in a sequence of
>affricate /J/ + fricative /s/
>- which could be simplified to
>(/d/ +) voiced fricative /Z/ + fricative /s/
>one of the two will get elided, at least in fast speech. If you believe in
>the Obligatory Contour Principle (not saying I don't, but not everyone
>does!), that would be one way of describing it.

Phonology/phonetics is relevant to be sure, but I'd never pronounce
John Hodgson's name [haJ at n] rather than [haJs at n], and I'm sure it's
not just frequency (in that this would be the case even I were
Hodgson's cousin, or Jon Stewart, or the other guy in those Mac vs.
PC commercials).  As you point out, the case for the phonological
explanation can be rendered more compelling than just [Js] > [J], but
I really do believe that my underlying form was "Fudgicle" and not
"Fudgesicle".  I suspect that the first time I heard the /s/
pronounced therein I assumed the speaker was blending "Fudgicle" with


>That's the phonology, but in my view English phonetics make it almost
>inevitable that one of these two phonemes will be elided in fast speech.
>It's clear that English final voiced consonants are often / usually not
>really voiced if you look at them on a spectrogram: usually the only
>difference between 'sinch' and 'singe' is in the length of the vowel, which
>is longer before the notionally- / phonologically-voiced consonant in
>'singe', while both final consonants are voiceless. So for 'fudge' (though
>I haven't done the spectrogram to check). I would expect that the final
>consonant of 'fudge' would be [tS], the same as the final consonant of
>'futch'.* If this is so, then we have
>[S] + [s]
>in the middle of 'fudgesicle', and I think that, if you say it fast, even
>if you are saying both consonants, it's very difficult for the listener to
>tell. It therefore sounds like 'fudgicle', and it's a short step from it
>sounding like that to people actually making that their target
>pronunciation and spelling it accordingly.
>* I know that there's a little, ahem, fudging of the argument here, since
>the final consonant of 'fudge' isn't word-final in 'fudgesicle'; but isn't
>it likely that people analyse it as if it _were_ word-final, and therefore
>final devoicing applies (turning 'fudge' to 'futch'), since the 'fudge'
>syllable is only there in 'fudgesicle' because it is a word in its own
>Date:    Sun, 4 Apr 2010 17:26:44 -0400
>From:    Paul Johnston <paul.johnston at WMICH.EDU>
>Subject: Re: Popsicle
>I (and my parents)said that too!  Maybe Fudgicle is a Northeastern
>Paul Johnston
>On Apr 4, 2010, at 11:09 AM, Laurence Horn wrote:
>>---------------------- Information from the mail header
>>Sender:       American Dialect Society <ADS-L at LISTSERV.UGA.EDU>
>>Poster:       Laurence Horn <laurence.horn at YALE.EDU>
>>Subject:      Re: Fwd: Re: Popsicle
>>At 1:17 AM -0400 4/4/10, victor steinbok wrote:
>>>I am still trying to figure out why there are so many references to
>>>Epperson coining Creamsicle, Fudgesicle, etc.
>>Was it just me and people I knew who never pronounced the "s" in the
>>latter?  It was always "Fudgicle", and I was surprised the first time
>>I read the package to see that it was technically "Fudgesicle".  No
>>such problems with the other (cream, corpse) varieties.  I guess if
>>we start from "icicle" the process is similar to that for the
>>-((c)o)holic family.
>>The American Dialect Society - http://www.americandialect.org
>Damien Hall
>University of York
>Department of Language and Linguistic Science
>YO10 5DD
>Tel. (office) +44 (0)1904 432665
>     (mobile) +44 (0)771 853 5634
>Fax  +44 (0)1904 432673
>The American Dialect Society - http://www.americandialect.org

The American Dialect Society - http://www.americandialect.org

More information about the Ads-l mailing list