Background Two Examples of Contour Tone Distribution

Background

1. Two Examples of Contour Tone Distribution

The term ‘tone language’ usually refers to languages in which the pitch of a syllable serves lexical or grammatical functions. In some tone languages, the contrastive functions of pitch are sometimes played by pitch changes within a syllable. Pitch changes of this kind are called contour tones. The distribution of contour tones in a language, i.e., under which phonological contexts are contour tones more readily realized, has been of much theoretical interest, as it sheds light on both the representation of tone (Woo 1969, Leben 1973, Goldsmith 1976, Bao 1990, Duanmu 1990, 1994a, Yip 1989, 1995) and the relation between phonetics and phonology (Duanmu 1994b, Gordon 1998, Zhang 1998). This dissertation is an in-depth investigation of the distribution of contour tones.

1. Contour Tones on Long Vowels Only

By way of an example, let us consider languages which have both contrastive vowel length and contour tones. In these languages, it is often the case that contour tones are restricted to phonemic long vowels; e.g., Somali (Saeed 1982, 1993), Navajo (Hoijer 1974, Kari 1976, Young and Morgan 1987, 1992), and Ju|'hoasi (Snyman 1975, Dickens 1994, Miller-Ockhuizen 1998) all display this pattern.

The ubiquity of this type of contour-tone restriction prompts analysts to posit the following principles regarding tonal representation: first, the mora is both the contrastive segmental length unit and the tone-bearing unit (TBU); second, a contour tone is structurally composed of two level tones; and third, each mora can only be associated with one tone (Trubetzkoy 1939, McCawley 1968, Newman 1972, Hyman 1985, McCarthy and Prince 1986, Zec 1988, Hayes 1989, Duanmu 1990, 1994a, Odden 1995, among others). Working together, these principles ensure that a contour tone can occur on a phonemic long vowel, which has two moras, but not on a phonemic short vowel, which has only one mora.

In the Optimality-theoretic framework (Prince and Smolensky 1993), the above principles can be translated into the markedness constraint in (0), which bans many-to-one mappings between tones and moras. Here, we assume that a ‘tone’ means a ‘pitch target’.

hf : two tones cannot be mapped onto one mora.


If we assume that the relevant tonal faithfulness constraint here is Max[Tone], as defined in (0), then by ranking the markedness constraint in (0) over the faithfulness constraint in (0), as shown in (0), we can capture the restriction of contour tones to phonemic long vowels. The tableaux in (0) show that, under this ranking, when two tones are associated with a short vowel underlyingly, only one tone will survive on the surface—(0a); but when they are associated with a long vowel, both tones can survive—(0b).
(0) Max[Tone]: if tone T is in the input, then it must also be in the output.
(0) *T₁ T₂

hf » Max[Tone]

hf | | | | | |

| | | hf hf

V V V V V

T1 T2 hf  | V	*T1 T2 hf 	Max [Tone]		T1 T2 | |  hf V	*T1 T2 hf 	Max [Tone]
T1 T2 hf  | V	*!			T1 T2 | |   hf V
T1 |   | V		*		T1 fh  hf V		*!
T2 |   | V		*		T2 fh  hf V		*!

Instead of explaining this contour tone restriction representationally as shown above, we may opt to provide a positional faithfulness (Alderete 1995, Steriade 1995, Beckman 1998) account in Optimality Theory. Generally speaking, this approach singles out the faithfulness constraint specific to a prominent position from the context-free faithfulness constraint and ranks positional faithfulness over context-free faithfulness. Then when a relevant markedness constraint is ranked between these two constraints, the marked value will be able to surface in the prominent position, but not elsewhere.

To show this account schematically, let us posit the constraints as in (0) (McCarthy and Prince 1995, Beckman 1998).

b. *[+F]: no [+F] is allowed in the output.

c. Ident-P₁[F]: let  be a segment in position P₁ in the output, and  be any correspondent of  in the input; if  is [F], then  is [F].
Constraint (0a) requires the faithful realization of F from the input to the output; constraint (0b) bans [+F] in the output; and crucially, constraint (0c) requires the faithful realization of F provided that its carrier segment is in position P₁ in the output. Then with the constraint ranking in (0), we generate the pattern in which the marked value [+F] is only allowed in the prominent position P₁, as illustrated in the tableaux in (0). When [+F] occurs in position P₁ in the input, it will be faithfully realized as [+F], since the candidate only violates *[+F], while its rival with an unfaithful rendition [-F] violates the most highly ranked positional faithfulness constraint Ident-P₁(F). This is shown in (0a). When [+F] occurs elsewhere however, it will be realized as [-F], since changing the feature specification in this situation only involves the violation of the low-ranked general faithfulness constraint, while its rival with a faithful realization [+F] violates the more highly ranked *[+F]. This is shown in (0b). Of course, [-F] in the input will always be realized as [-F], since there is no markedness constraint against [-F]. Therefore, we generate the pattern in which F is contrastive in position P₁, but neutralized elsewhere.
(0) Constraint ranking: Ident-P₁[F] » *[+F] » Ident[F]
(0) a. [+F] is faithfully realized in P₁:

[+F] in P1	Ident-P1[F]	*[+F]	Ident[F]
 [+F]		*
[-F]	*!		*

b. [+F] is realized as [-F] elsewhere:

[+F] in P1	Ident-P1[F]	*[+F]	Ident[F]
[+F]		*!
 [-F]			*

Specifically to the contour tone case in question, the key idea is to acknowledge that a phonemic long vowel has a longer sonorous duration, hence provides better perceptual cues for the identification of the tonal contour and better opportunities to fully realize the tonal contour articulatorily. The positional faithfulness constraint is then Ident-Long(Tone), as defined in (0a). The context-free faithfulness constraint Ident(Tone) and the relevant markedness constraint *Contour are defined in (0b) and (0c) respectively. In these definitions, a contour tone is not considered a concatenation of level tones, but a tonal unit whose pitch changes during its time course.

b. Ident[Tone]: let  be a vowel in the input, and  be any correspondent of  in the output; if  is has tone T, then  has tone T.

c. *Contour: no contour tone is allowed on a vowel.
To account for the restriction of contour tones to long vowels, we employ the ranking in (0). The first ranking in (0) ensures that a contour tone on a long vowel will be faithfully realized in the output, while the second ranking in (0) ensures that no contour tone will surface on a short vowel.
(0) Ident-Long[Tone] » *Contour » Ident[Tone]
The third possible account for this restriction is to refer to the phonetic properties of long vowels directly, and I will term this as the ‘direct approach’. Like the positional faithfulness approach discussed above, it also recognizes that phonemic long vowels are better contour tone bearers because they have a long sonorous duration, which is the crucial phonetic dimension on which the realization of contour tones rely, and it also uses a positional faithfulness schema. But unlike traditional positional faithfulness, which only refers to the phonological feature that distinguishes a phonemic long vowel from a phonemic short vowel, namely, [+long], it directly refers to the phonetic properties that are crucial to contour tone realization—duration and sonority. Let us assume for now that the contour tone bearing ability of a syllable is proportional to an index C_CONTOUR, which is a weighted sum of duration and sonority.¹ Then the positional faithfulness constraint under this approach is Ident-C_CONTOUR(Long)[Tone], as defined in (0).
(0) Ident-C_CONTOUR(Long)[Tone]: let  be a vowel whose C_CONTOUR is greater than or equal to that of a [+long] vowel in the output, and  be any correspondent of  in the input; if  has tone T, then  has tone T.
With the same constraints Ident[Tone] and *Contour as in (0b) and (0c) and the ranking as in (0), this approach also accounts for the restriction of contour tones to long vowels, as the previous two approaches.
(0) Ident-C_CONTOUR(Long)[Tone] » *Contour » Ident[Tone]

2. Contour Tones on Stressed Syllables Only

Another commonly attested restriction on contour tone distribution is that they are only allowed on stressed syllables. For instance, in the penultimate-stress language Xhosa (Lanham 1958, 1963, Jordan 1966, Claughton 1983), contour tones are generally restricted to the penultimate syllable of a word. In Jemez (Bell 1993), the initial syllable carries the word stress, and it is the only position in which a contour tone is allowed.

This contour tone restriction can again be captured in three different ways.

First, we may assume that stressed syllables are bimoraic while unstressed syllables are monomoraic. This assumption does not stem from the mora as the contrastive segmental length unit. Rather, we are taking the mora as a unit of weight here, and assigning two such units to a stressed syllable, which we assume to be heavy, and one such unit to an unstressed syllable, which we assume to be light. Further assuming that contour tones are concatenations of level tones and each level tone needs a mora to be realized, we can see that the restriction of contour tones to stressed syllables is explained just as the restriction of contour tones to phonemic long vowels.

Second, in a positional faithfulness approach, ‘stress’ can be justifiably singled out from the context-free faithfulness constraint. This is because, as a prosodically prominent position, stress typically induces longer duration and higher amplitude, both of which facilitate the identification of phonetic features. Therefore, the positional faithfulness constraint is Ident-Stress[Tone], which is defined in (0). The constraint ranking that captures this contour tone restriction is shown in (0).
(0) Ident-Stress[Tone]: let  be a vowel that is [+stress] in the output, and  be any correspondent of  in the input; if  has tone T, then  has tone T.
(0) Ident-Stress[Tone] » *Contour » Ident[Tone]
Third, we can also appeal to the ‘direct approach’ and refer to the index C_CONTOUR for stressed and unstressed syllables in the account. The positional faithfulness constraint is Ident-C_CONTOUR(Stress)[Tone] as defined in (0). It requires the faithful realization of a tone provided it surfaces on a syllable whose C_CONTOUR is no less than the C_CONTOUR of a stressed syllable. Then with this constraint outranking *Contour, which in turn outranks the context-free Ident[Tone], as shown in (0), the restriction of contour tones to stressed syllables can likewise be captured.
(0) Ident-C_CONTOUR(Stress)[Tone]: let  be a syllable whose C_CONTOUR is greater than or equal to that of a stressed syllable in the output, and  be any correspondent of  in the input; if  has tone T, then  has tone T.
(0) Ident-C_CONTOUR(Stress)[Tone] » *Contour » Ident[Tone]

2. 
   
   Directory: files
   files -> Answer True False 2 points Question 2
   files -> Integer programming and game theory
   files -> 4 Integer Programming
   files -> Bpa vehicle Window Repair Scenario #1 task: Procure vehicle window relacement. Objective
   files -> Pop Warner History
   files -> North Carolina Inclusion Initiative Mapping Where Children with ieps are Being Served Purpose
   files -> Fall 2013 Spring 2014 Program Data: Standard 1 Exhibit 4d
   files -> Hanban – asia society confucius classrooms network 2010 request for proposal
   files -> Northern England’s set-jetting locations
   
   
   
   Download 3.01 Mb.
   
   Share with your friends: