View this page "Challenge of the Week!"

[Mike]

Hi folks! Try this Challenge of the Week!

Click on http://groups.google.com/group/kusaal/web/challenge-of-the-week
- or copy & paste it into your browser's address bar if that doesn't
work.

[Ryan....@gmail.com]

I'm on it. I had a long stretch of overtime here at work that's
finally coming to a close. I'll be back to work from now. Let's get
past all the minimal pairs, hopefully some interesting things will
turn up along the way. I'd like to take a stab at some phenomenon
using Optimality Theory. Get back to you soon.

[Ryan....@gmail.com]

Would the pair of [põbe?] and [võõme?] be the best example since the
enviroments are the immediate environments are the most similiar? In
this case [b] and [m] are both preceded by [õ] and followed by [e?].
Since [p] and [v] are somewhat similiar in production, bilabial and
labiodental respectively, possible influence is less likely from these
segments.

That makes this pair better than [kpibit] and [nimit] despite idential
immediate environments, given the lack of commone features between
[kp] and [n] and better than the first pair [kwiibok] and [lEmok]
because of different prededing environments, namely [ii] verus [E].

Please let me know if my comments are on target or missing anything.
Despite my degree in Linguistics, I only had on course in Phonology/
Phonetics from my university's small Linguistics department three
years ago. I should progress as I review my previous texts.

[Ryan....@gmail.com]

Also, all pairs have the same word class. Thanks for teaching me that
this is an important requirement.

[Ryan....@gmail.com]

word final
[t] [d]
[wakat] 'moment' (adv) [da?ad] 'day
before yesterday (adv)
[albahat] 'onion'
[ja?at] 'cheek'

The above environments are similiar enough to provide evidence of
contrast word finally as they show both [t] and [d] follow [a] and in
the case of [albahat], [ja?at], and [da?ad], they both follow [a] when
[a] is preceeded by a glottal like [?] or, for [t], [h]. Also, we
know already the pair [p] and [b], differing only in voicing just as
[t] and [d] do, are contrastive word finally and there is no
differences in the above environments that could account for a voicing
difference.

[Ryan....@gmail.com]

word medially
[t] [d]
[lotom] 'calf of leg' [jããdowa?] 'eggplant'
[nintɛŋ] 'sunshine [dindeok] 'chameleon'
[tigita?] 'tax' (n) [jidɪm] 'family'

As evident above, both [t] and [d] are found before [o], after [n] and
before an open-mid, front vowel as in [nde] and [ntɛ], and following
[i]. Nothing about these analgous environments can account for a
difference in voicing.

[Ryan....@gmail.com]

Believe it or not, the analgous environment for proof of contrast is a
little new to me, or I forgot about it. Let me know if my examples
aren't quite good enough and, if so, why.

[Ryan....@gmail.com]

word medial: The pairs below show [t] and [d] in analgous
environments, specifically preceding [o], following [n] while
preceeding either of the open-mid, front vowels [e] or [ɛ], and
preceeding either of the closed, front vowels [i] or [ɪ]. There is no
evidence to account for a difference in voicing

[Mike]

I agree with you 100% on this one, Ryan. Although [põbe?] and
[võõme?] are not ideal (see below), they are the best we have here.

Your reasoning is good. The primary difference between [b] and [m] is
nasalization, so we need to determine if nasalization, or the lack of
it, in the surrounding environment (particularly immediately preceding
the phones in question) is causing one or the other to be used. An
initial suspicion could be that [m] occurs following nasal vowels,
something which [võõme?] seems to support. However [põbe?] is a good
contrast because it shows that [b] can also follow a nasal vowel. The
only objection that might be raised is that in [võõme?], the [m] may
be the result of a long nasal vowel (with the longer articulation of
nasalization on the vowel causing that feature to be carried over to
the following consonant).

Actually, [kpibit] and [nimit] aren't bad examples either since the
immediate environments are identical and the syllable structure is the
same. The main potential objection I can see with them, however, is
again related to the influence of nasalization. Notice that [m]
occurs in the word that begins with a nasal consonant [n], and [b]
occurs in a word that begins with a non-nasal consonant (the [kp]
cluster).

[Mike]

Ideally, yes, all pairs should be of the same word class. If,
however, no minimal pairs can be found in the same word class (in
either identical or analogous environments), a minimal pair in
different word classes may be the best we can do. But someone could
legitimately raise the point that the difference in phoneme use (e.g.
using [p] instead of [b]) could be due to the difference in word
classes, with one phone being used in nouns and the other in verbs,
for example. This is something that occurs in some languages. In
that case, further investigation would have to be done.

For instance, if in our minimal pair the [p] occurs in a noun and the
[b] in a verb, we would check to see if [p] ever occurs in a verb
(preferably in the same position in the word) and if [b] ever occurs
in a noun. This would help discount the suspicion or theory that [p]
only occurs in nouns and [b] only in verbs.

[Mike]

Of the examples you gave, I think I like [ja?at] and [da?ad] the best,
not only because of identical immediate environments, but also because
they share identical syllable structures. In addition to being of the
same word class, having identical or very similar syllable structures
is also highly desirable for good contrast.

You did well to use the fact that [p] and [b] contrast word-finally as
support for the claim that [t] and [d] should too. This is in
accordance with a basic principle of phonological analysis, namely
that "sound systems tend to be symmetrical". In other words, what is
true of one set of phones in a particular class (stops, nasals,
fricatives, etc.) is likely to be true of the other sets of phones in
that class. Therefore, for instance, we would expect the pairs of [p]
[b], [t][d], and [k][g] to all behave in the same way.

[Mike]

Boy, this is a tough one. My main hesitation with these pairs is that
they all have different syllable structures. However, as you so well
point out, this cannot account for the difference in voicing that
characterizes [t] and [d].

How about [tutuuk] 'bush' and [kudook] 'trousers'? They have
identical syllable structures and very similar phonetic environments.
Someone may say that [t] (a voiceless consonant) occurs in words
beginning with voiceless consonants (also a [t] in the first word of
the above pair). But the "voiceless consonant influence theory" is
discounted by the second word of the pair, in which [d] also occurs in
a word beginning with a voiceless consonant (in this case a [k]). The
difference in place of articulation of [t] and [k] occuring at the
beginning of a word cannot explain the difference in voicing between
[t] and [d] (the specific phones in question) word-medially.

Does this make sense?

On Dec 2, 4:13 am, "Ryan.Lo...@gmail.com" <Ryan.Lo...@gmail.com>
wrote:

[Ryan....@gmail.com]

Yeah, it makes perfect sense to me. I don't know why I didn't think
to account for syllable structure before. It has been three years
since I took a phonology course since the last two years of my study
really focused on teaching (which I kind of feel I'm not really
interested in). I'm reviewing my text like I said and should get my
brain back up to speed shortly (I hope). Since the syllable structure
are already contained in Phonolgy assistant, it wouldn't be too much
to code them like: CV.kV. when looking at [k] and make lists of words
in this format. I then have a program that could find minimal pairs
this way (it doesn't work with the SIL fonts). I'll give it a try.

> > difference in voicing.- Hide quoted text -
>
> - Show quoted text -

[Mike]

Just a little slow on the uptake here due to the holidays, but I'll
post progress and a new challenge shortly!