[Lingtyp] Tonal inventories: High vs Extra-High

Wed Jan 27 17:34:20 UTC 2021

Hi Laura,

I don't know how many readers are interested in tones, but to be more
concrete, if I now understand what you are asking, the following systems
have been proposed:

/H, L, Ø/ = common (see below)
/H, M, Ø/ = less common (Mary Paster proposed this for Leggbo, the only
case I know): Ø is realized L.
/M, L, Ø/ = unattested? (I don't know of any language that has been
analyzed that way)
/S, H, Ø/ = you are proposing? I don't see why this couldn't exist
/S, L, Ø/ = ?
etc.

Now, the question is how Ø is realized in /H, L, Ø/ systems. There are /H,
L, Ø/ analyses where Ø has a phonological behavior distinct from /H/ and
/L/, but ultimately is assigned [H] or [L] vs. those where Ø is realized on
a distinct, third phonetic pitch height, e.g. [M]. I think you are only
interested in these latter. For other /S, H, Ø/ systems, I suspect Punu
(Bantu; Gabon), but again, a lot depends on interpretation, e.g. whether
/L/ is needed instead of /Ø/.

One other thing I forgot to say; The interpretation often depends on
frequency. We have a tendency to first use /H/ and /L/ and then other
tones, such that if the highest pitch were rare in the system (e.g. derived
only in a specific environment), it would likely be called "superhigh". But
this wasn't the issue you were interested in.

On Wed, Jan 27, 2021 at 9:18 AM ARNOLD Laura <Laura.Arnold at ed.ac.uk> wrote:

> Thank you all for the helpful discussion!
>
> I realise I phrased the original question poorly. I’m not so much
> interested in the labels – we can call the two tones whatever we want – I’m
> more interested in the relation of the two tones with reference to
> non-tonal units.
>
> So, in languages that contrast two level tones (let’s call them X and Y)
> and toneless units, I’m interested in cases where X and Y are
> phonologically specified to be realised with a higher F0 than toneless
> units (i.e., toneless < X < Y). For that matter, I’d also be interested to
> hear about cases where the two tones are lower than toneless units (i.e., X
> < Y < toneless). I may be wrong, but it’s my impression that these two
> distributions are less common than the more frequent X < toneless < Y.
>
> Many thanks again,
> Laura
> ------------------------------
> *From:* Larry M. HYMAN <hyman at berkeley.edu>
> *Sent:* 27 January 2021 16:47
> *To:* ARNOLD Laura <Laura.Arnold at ed.ac.uk>
> *Cc:* lingtyp at listserv.linguistlist.org <lingtyp at listserv.linguistlist.org
> >
> *Subject:* Re: [Lingtyp] Tonal inventories: High vs Extra-High
>
> This email was sent to you by someone outside the University.
> You should only click on links or attachments if you are certain that the
> email is genuine and the content is safe.
> There are at least three problems in this discussion.
>
> First, there is the problem of whether we are talking phonetics or
> phonology.
>
> Second, there is the problem of whether the labeling of tones can be
> uniquely derived from the phonetics, continuous F0 alone (as we can
> identify vowels from the acoustic vowel space).
> As Mark Donohue once put it to me, "We don't hear tones. We hear pitch."
> (and anything co-varying with the pitch: duration, other laryngeal activity
> etc.)
>
> Third, there's the sometimes arbitrary labeling of tones.
>
> I agree that in some languages the pitch of the highest tone sounds higher
> than in others (and sometimes also individuals speaking the same language).
> It also is the case that in some languages the lowest tone is at rock
> bottom (trailing off especially before pause), while in others, the lowest
> tone has a level "middish" F0.
>
> In many cases, the languages where terms like "superhigh" or "extrahigh"
> are invoked not only have some very high pitch realizations, but often ones
> that are derived. For example, in Elaine Thomas' analysis of Engenni
> (Edoid, Nigeria), a H tone is raised to what she calls "top" tone
> (=superhigh) when followed by a L tone (and in certain morphological
> contexts). Since the L sometimes is not realized, this creates a system of
> with a "superhigh", "high", and "low" pitch contrast which then has to be
> interpreted. (I have proposed that the "high" tone is really phonologically
> Ø, such that the system starts out as a /L/ vs. Ø one, but develops into a
> three-height system, S, H and L.) There are other languages where the
> highest tone is derived from the simplification of a H + downstepped H
> sequence (or contour on the same vowel/syllable), and others reported where
> H tone is realized as "superhigh" on high vowels.
>
> In my dissertation on Fe'fe' dialect of Bamileke (Grassfields Bantu,
> Cameroon), the general non-low tone was called M, even though there are
> rules raising it (and even L) to H, e.g. in most noun classes, where a M +
> M noun + noun in a genitive relation (N1 of N2) is realized H + M. I
> certainly could have called the M a high tone, and called the H tone as
> superhigh tone. The reason I didn't do this is that I didn't feel that the
> H was any higher than I would expect in a H, M, L system like Yoruba, Nupe
> etc.
>
> Mid tone languages are also different from each other. In some the /M/ is
> there underlyingly, in others it is interpreted as the realization of Ø,
> and in still other languages it is derived. (I'm now remembering that I
> once gave a minicourse just on the mid tone--in Leipzig in 2006 :-)!
>
> I don't think labeling is as crucial as the characterization of the tonal
> system itself, the relation between the tones, the phonetic implementation
> etc. Maybe someone will come up with an objective way to go from pitch to
> tone category, but so far it isn't automatic (or agreed upon by different
> researchers).
>
> On Wed, Jan 27, 2021 at 7:06 AM ARNOLD Laura <Laura.Arnold at ed.ac.uk>
> wrote:
>
> Dear colleagues,
>
> Does anyone know how frequent two-tone inventories contrasting only High
> and Extra-High are? I’m working with data from a language which has an
> inventory that can possibly be analysed this way (the two tones also
> contrast with toneless syllables). I suspect this is quite an unusual
> inventory, cross-linguistically – it would be helpful to confirm this. I
> would also be interested to hear about similar examples elsewhere in the
> world.
>
> Many thanks,
> Laura
>
> ~~~
> Laura Arnold – British Academy Postdoctoral Fellow
> laura-arnold.org <https://www.laura-arnold.org/>
>
> Room 1.13, Dugald Stewart Building
> School of Philosophy, Psychology & Language Sciences
> University of Edinburgh
>
> The University of Edinburgh is a charitable body, registered in Scotland,
> with registration number SC005336.
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> http://listserv.linguistlist.org/mailman/listinfo/lingtyp
>
>
>
> --
> Larry M. Hyman, Professor of Linguistics & Executive Director,
> France-Berkeley Fund
> Department of Linguistics, University of California, Berkeley
> https://linguistics.berkeley.edu/~hyman
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> http://listserv.linguistlist.org/mailman/listinfo/lingtyp
>

-- 
Larry M. Hyman, Professor of Linguistics & Executive Director,
France-Berkeley Fund
Department of Linguistics, University of California, Berkeley
https://linguistics.berkeley.edu/~hyman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20210127/ca11b832/attachment.htm>