Phonology problems

[zapp4975]

I have said that for this problem /ə/ → [a] / when part of [+stressed] and/or [+long] $
 however my friend told me that you can't use [+stressed] as a feature. Is this true? If so, I can't find a way that that [ə] and [a] are in complementary distribution. Does this mean that they form two separate phonemes?

C.   KAROK (California)

1   ʔátra:x   arm           5   ʔəʧna:t   rat
2   tátəʧ   mama           6   ʔáxək   two
3   sárə   bread             7   ʔásər   wet
4   ʧəná:k   mosquito     8   tənukjá:nər   shovel

Are [a ə] allophones of one phoneme? State your evidence.

Phonetic notes:
   [  ́ ] in this problem represents a combination of strong stress and high pitch.
   Unmarked vowels have low pitch.

My solution:

[a] occurs only in syllables which are stressed and/or lengthened
[ə] occurs only in syllables which are neither stressed nor lengthened

Thus, [a] and [ə] are in complementary distribution and are two allophones of one phoneme /ə/.

Distribution:

/ə/ → [a] / as part of [+stressed] and/or [+long] $
     → [ə] / elsewhere

/ə/ → [a] / when part of [+stressed] and/or [+long] $

Where $ = syllable

For this problem - I have found that they are not in complementary distribution, but am I missing out on an essential feature?

B.   VENDA (South Africa)

1   ha n̪ u
   at your place               7   ene   he
2   l̪ i n̪ o   tooth               8   hana   childhood
3   m u n̪ e   master          9   khouno   there
4   n̪ a r i   buffalo             10   βatanu   five
5   pf ʰ e n̪ e   baboon         11   βonani   see!
6   β a n̪ a   four                 12   zʷino   now

Are [n̪   n] allophones of one phoneme? State your evidence.

Phonetic notes:
   [n̪] is a dental, [n] is an alveolar, voiced nasal.
   [l̪] is a dental lateral approximant.
   [β] is a voiced bilabial fricative.

My solution:

[n̪ ] and [n] both occur between:
 
[a]_
_

[e]_[e]
[a]_[a]

They thus share four identical contexts, and are therefore not in complementary dis-tribution, forming two separate phonemes.

[Daniel]

We're not going to do homework for you (the point of homework is that it's practice for you to do!), but if you can ask a very specific question we may be able to help with that. The assignment should not be so difficult that you can't do it yourself, possibly with help from your instructor at office hours.

[zapp4975]

I know, of course I don't expect that - that's why I have already completed the assignment myself. I just wanted some help with questions 2 and 3 - i have changed the post to make it more specific

[panini]

If your instructor said "You can't use "+stressed as a feature", then that's the rule. If your friend said that, get a better friend. It is true that we have a different way of talking of stressed stuff that doen't involve a feature, but for the level you're dealing with, it's a perfectly legitimate feature. It is harder to say "in a stressed syllable" using conventional rule formalism. As a general rule, if you have a simple and true generalization and can't find a way to formalize it in a rule, the problem probably lies in your understanding of rule formalism or in the theory of formalism itself, and not in the solution.

Does your instructor require you to write one rule for each allophone, or can you write one rule for the variant in context X and no rule is necessary for "elsewhere"? If you can get away with one rule, you should explore the consequences of different assumptions about the underlying value of the vowel in question.

My objection to your Venda solution is that you don't identify any examples where [n̪] and [n] both occur in comparable contexts. Extra points for clarity and explicitness.

[zapp4975]

Ok thanks so much!!

I'm just a bit unsure about what you mean by the underlying value of the vowels?

[Daniel]

The underlying value is whatever the starting value is, in the representation of the word. Then the phonological rules apply to generate all variations (allophones) from that.

So an underlying form might be /e/ which then becomes [a] or [e] depending on context. Or it might be /a/ that becomes [a] or [e] depending on context. These assumptions can change the analysis, of course, because they require different rules to generate the surface forms.

[zapp4975]

Oh right ok of course -  thank you again!!

[zapp4975]

oh i just realised that my full solution for venda didn't really show properly, I say:

[n.] and [n] both occur between a_u, i_o, e_e and a_a.

Is this providing examples of where they occur in comparable contexts?

i had to take off the brackets because they were causing formatting problems

[panini]

Is this providing examples of where they occur in comparable contexts?

"Examples" are actual words which constitute instances of one fact vs. the other. You described abstract generalizations, now I (and your instructor) want the words that the generalizations refer to. Understanding the proper relationship between data and generalization constitutes the keys to the kingdom in linguistics.

[Muikkunen]

Thus, [a] and [ə] are in complementary distribution and are two allophones of one phoneme /ə/.

Distribution:

/ə/ → [a] / as part of [+stressed] and/or [+long] $
     → [ə] / elsewhere

/ə/ → [a] / when part of [+stressed] and/or [+long] $

Where $ = syllable

Isn't the undelying vowel /a/ rather than /ə/? I think it might be an example of vowel reduction, when /a/ is centralized to /ə/ in unstressed and short syllables.

[panini]

Isn't the undelying vowel /a/ rather than /ə/? I think it might be an example of vowel reduction, when /a/ is centralized to /ə/ in unstressed and short syllables.

Your term "short syllable" is mildly problematic. Vowels can be short or long, but syllables are light ~ open ~ heavy ~ closed. If the syllable has a final C, it is closed, and if it ends with a vowel it is open. Whether or not a syllable of the kind CVC is heavy or light depends on the language, and these data don't indicate that "heavy" vs. "light" is relevant in the language. If you make the correct assumption about the direction of the change, then a simple conjunction of two properties of the vowel will give you the rule (both must be true: you wording could be interpreted as "...either, or").

[Muikkunen]

Your term "short syllable" is mildly problematic.

Sorry, I meant short vowels.
Thank you for the reply. I will think more about this task.