[Lingtyp] query: 1st syllable deletion

Wed Nov 2 17:20:38 UTC 2022

Interesting discussion, everyone!

David, I've been thinking about this for a couple days now (but did not
reply because I wasn't aware of specific references to suggest). My initial
intuition was that this is not phonological, and not about syllables. But
since English pronouns (at least the basic ones) are all monosyllabic, it's
hard to test that. The messages from others have mostly fit with my
intuition, with some additional points to think about too.

A common (perhaps at this point strawman) functional explanation for
"pro-drop" is that rich agreement allows recovery of the subject identity
so the pronoun itself is optional. But rich agreement does not always
license pro-drop, such as in German or Russian. (And even in English, we
might assume it should be allowed in third-person singular present-tense.)
More importantly, there are also languages such as Japanese which allow
pronoun ellipsis but do not have rich agreement (or any agreement at all).

But in this case (which as pointed out by others is not "pro-drop" in a
strict sense), that kind of functional explanation does seem to be the
right approach: this kind of ellipsis is only found with contextually
salient forms.

The explanation cannot be purely phonological. No English speaker would
ever say:

*Guistics is fun!

That's too hard to understand. Omission of ling- is not possible, because
Linguistics is such a rare grammatical subject. Of course I can imagine
this could be primed, perhaps at the next ALT conference!

Therefore, this phenomenon must be frequency-based. And in fact that's
exactly what all of the examples are, such as Riccardo's in the previous
message. Almost all of these are set phrases or common pronouns in
contextually-relevant usage, such as questions with "you", or statements
with "I". I haven't done a corpus search (searching for ellipsis is
tricky!), but I assume we would find a substantial bias toward frequent
collocations, as well as some pronouns rather than others, given the type
of usage that is typical.

At the same time, it isn't really syntactic either, and it's not full
constituents ("...the hell", "...the fuck", "...cazzo"). I'd guess that
pronouns and the other partial-constituent expressions are distinct types,
but in general they seem to have similar properties, most importantly that
they are high-frequency and predictable.

What we're looking at then seems to be more a kind of *abbreviation* than
*ellipsis*. Sound change is regular except for idioms and other frequent
expressions (i.e. allegro speech). "God be with you" becomes "goodbye", but
only for that expression. So this is more about lexicalization and
frequency than a strictly phonological *or *strictly syntactic rule. The
pronoun type may need a more syntax-based explanation, but I think this is
still part of it.

The omitted syllable is probably just reduced to the point of being
unpronounced, because it is (extremely) unstressed, because it is
(extremely) predictable. In fact, I would bet that at least some of the
time the speaker actually mouths the first syllable but doesn't utter it
audibly. David, that matches some of what you suggested, but I'm
interpreting it differently, or at least in a restricted way. There is a
prosodic explanation, but most importantly within a frequency explanation.
And in a sense it's more about *morphology* than syntax or phonology.

If we tie all of this together, and here's where it gets interesting, this
seems to suggest that multi-word expressions allowing omission of the first
syllable (or more) are actually single, frequent lexical items. And
regarding pronoun subjects, they not only have characteristic prosodies,
which could in itself be considered something resembling lexicalization,
but arguably if we assume that phonologically null syntactic structure can
also be lexicalized, we might get, for example, "Q+you" for second-person
questions, which occurs so often that it can be omitted and still
recovered. Or just a lexicalized prosody explanation if you prefer.

Of course this can spread by analogy (but not a regular rule!) to new
expressions, but only when those expressions are also relatively frequent
and predictable.

Daniel

On Wed, Nov 2, 2022 at 9:36 AM Riccardo Giomi <rgiomi at campus.ul.pt> wrote:

> Dear David and all,
>
> For what it's worth, I tend to agree with Randy and Christian. One
> possible argument against a rule-based deletion/ellipsis account comes from
> central varieties of Italian (including my own), and concerns, once again,
> emphatic *wh-*questions -- suggesting a cross-linguistic pattern that, I
> suspect, may also be found in other languages.
>
> The "full" version of the type of question I have in mind makes use of the
> emphatic element *cazzo *(slang for 'penis', but functionally largely
> equivalent to English *fuck*), postponed to the clause-initial *wh-*word,
> e.g.
>
> - *Che cazzo fai? *'What the fuck are you doing?'
>
> I often hear (and probably produce myself) "reduced" versions of similar
> questions, including the following:
>
> - *Cazzo fai?*
> -* Azzo fai?*
> *- Zzo fai?*
>
> In the first case, it is only the first word and syllable *che *that
> would be deleted; in the second, it's this word/syllable plus the onset of
> the second syllable (which belongs to the second syntactic word); in the
> third the first word/syllable plus the whole second syllable (i.e. the
> first syllable of the second syntactic word). And I also recognize the
> forms *Zzo vai? *and *Azzo vai? *for *Dove cazzo vai? *('Where the fuck
> are you going?'), where the (by hypothesis) elided part consists of two and
> a half and three syllables, respectively, again cutting across syntactic
> domains.
>
> What this all suggests to me is that, if the correct explanation was
> indeed one in term of deletion or ellipsis, then either it is a composite
> rule that cuts across phonology and syntax, or it would be a merely
> phonological rule allowing deletion of phonological strings that do not
> necessarily coincide with syllables (which sounds kind of weird, at least
> for Italian), and may stretch over three entire syllables (and maybe more,
> but this is what I could think of for the moment). Rather than postulating
> such a complex rule, which for more I don't think is found in any other
> type of utterance in Italian, I find it much more economical and
> psycholinguistically plausible (but the latter is of course quite a shot in
> the dark) to assume that the point is precisely the omission of the
> predictable elements from this specific type of emphatic question. The
> constraint appears to be that only the most distinctive (and perceptually
> salient) element, i.e. the second syllable of *cazzo*, must be retained
> for the utterance-type in question to be easily identified.
>
> Sorry for my stream of consciousness!
>
> Best,
> Riccardo
>
> Randy J. LaPolla <randy.lapolla at gmail.com> escreveu no dia quarta,
> 2/11/2022 à(s) 13:35:
>
>> Hi David,
>> I don’t like the term “prodrop”, as it takes English, which is
>> typologically actually the odd man out, as the norm, and all of the many
>> languages that have not grammaticalised the grammatical mood constructions
>> that require pronouns to be retained in English are seen as aberrant, but
>> for languages that do not have such constructions, e.g. Chinese, the kind
>> of pattern we have been talking about is the norm.
>>
>> All the best,
>> Randy
>>
>>
>> > On 2 Nov 2022, at 7:29 PM, David Gil <gil at shh.mpg.de> wrote:
>> >
>> > Randy,
>> >
>> > Thanks for your comment.  The last couple of days I've received a
>> flurry of very helpful references and pointers concerning the phenomenon in
>> question, which seem to point to it not being "a simple pragmatic
>> phenomenon" of the kind you suggest.  Also, with the possible exception of
>> a reference to German, nobody so far has offered examples of similar
>> processes in other languages, and indeed, I can't think of anything like it
>> in the other languages I am familiar with.
>> >
>> > Best,
>> >
>> > David
>> >
>> >
>> > On 02/11/2022 12:33, Randy LaPolla wrote:
>> >> Good question, David!
>> >> Not a matter of phonetics or morphology, though.
>> >> Possibly a simple pragmatic phenomenon where predictable elements,
>> especially topics, can be left unspoken.
>> >> Common in many languages.
>> >> Not considered “grammatical” in English, but maybe English is changing.
>> >>
>> >> Randy
>> >>
>> > --
>> > David Gil
>> >
>> > Senior Scientist (Associate)
>> > Department of Linguistic and Cultural Evolution
>> > Max Planck Institute for Evolutionary Anthropology
>> > Deutscher Platz 6, Leipzig, 04103, Germany
>> >
>> > Email: gil at shh.mpg.de
>> > Mobile Phone (Israel): +972-526117713
>> > Mobile Phone (Indonesia): +62-082113720302
>> >
>>
>> _______________________________________________
>> Lingtyp mailing list
>> Lingtyp at listserv.linguistlist.org
>> https://listserv.linguistlist.org/cgi-bin/mailman/listinfo/lingtyp
>>
>
>
> --
> Riccardo Giomi, Ph.D.
> University of Liège
> Département de langues modernes : linguistique, littérature et traduction
> Research group *Linguistique contrastive et typologie des langues*
> F.R.S.-FNRS Postdoctoral fellow (CR - FC 43095)
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> https://listserv.linguistlist.org/cgi-bin/mailman/listinfo/lingtyp
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20221102/0a9ea9e5/attachment.htm>