چكيده لاتين :
The morpheme /-h@/ is the most productive plural marker in Persian. Besides
the formal style, its full form i.e., [h@] is pronounced in two environments in
the spoken style; when it is attached to words ending in vowels /e/ and /@/,
and when it emphasizes the notion “so many” in few idiosyncratic words.
However, its /h/ is deleted when it is attached to words ending in a consonant
or one of the three vowels /i/, /u/ and /o/. In the latter case, /h/ is replaced by
an intervocalic glide to resolve hiatus. This research aimed at analyzing
various environments and phonological processes which affect the
pronunciation of the plural morpheme /-h@/ within Optimality Theory
(Prince and Smolensky, 1993/2004). Idiosyncrasies which fail /h/ deletion,
even when its conditioning environment is met, are analyzed using Lexically
Specific Constraint theory (Pater, 2006, 2008) and Lexical-access
Constraints theory (Borsma, 2001) as OT’s subtheories.
1. Introduction
The morpheme /-h@/ is the most productive plural marker in Persian
which can attach to every noun. Its pronunciation is determined by different
phonological and semantic factors. Apart from the formal style, its full form
i.e., [h@] is pronounced in two environments in the spoken style; when it is
attached to words ending in vowels /e/ and /@/, and when it emphasizes the
notion “so many” in few idiosyncratic words. However, its /h/ is deleted
when it is attached to words ending in a consonant or one of the three vowels
/i/, /u/ and /o/. This research aimed at coming up with answers to the
following questions: 1. What constraints and rankings affect the pronunciation of the plural
marker /-h@/? 2. How does Optimality Theory
analyze idiosyncrasies which fail /h/ deletion, even when its conditioning
environment is met?
2. Literature Review
Studies like Sadeghi (1969), Sadeghi & Arjang (1979), Lazard (1992)
and Darzi & Ghadiri (2011) have discussed the morphological, syntactic and
semantic properties of the plural morpheme /-h@/. However, the present
research is the first study to discuss its phonological properties. Although
classic Optimality Theory (Prince and Smolensky, 1993/2004) is capable of
analyzing various phonological environments and phonological processes
which affect the pronunciation of the plural marker /-h@/, it is incapable of
analyzing idiosyncrasies in which /h/ deletion fails to apply, even when its
conditioning environment is met. According to Gouskova (2012) as
exceptions follow a pattern which is inconsistent with the rest of the
grammar, OT has difficulty establishing a uniform constraint ranking
without additional mechanisms. Pater (2004) claims that an adequate theory
of exceptions should be capable of expressing the distinction between
regular and exceptional forms as well as between exceptional and
ungrammatical forms. Thus, in order to deal with this challenge Pater (2006,
2008) proposes Lexically Specific Constraint Theory. In addition, there are
cases in which the pronunciation of the plural morpheme /-h@/ is dependent
on the interaction between phonology and semantics. Since classic
Optimality Theory (Prince and Smolensky, 1993/2004) has no mechanism to
analyze phonology-semantics interface, Borsma (2001) proposes Lexicalaccess
Constraints theory. Both Lexically Specific Constraint Theory and
Lexical-access Constraints theory are OT’s subtheories.
3. Methodology
This research applies classic Optimality Theory (Prince and Smolensky,
1993/2004) to analyze various phonological environments and phonological
processes which affect the pronunciation of plural marker /-h@/. However,
idiosyncrasies which fail /h/ deletion, even when its conditioning
environment is met, are analyzed using Lexically Specific Constraint theory
(Pater, 2006, 2008). In addition, the cases in which the pronunciation of the plural morpheme /-h@/ depends on the interface between phonology and
semantics are analyzed using Lexical-access Constraints theory (Borsma,
2001).
4. Results
In the informal style the /h/ in the plural marker deletes in most of the
environments, i.e., when it is attached to words ending in a consonant or in
one of the three vowels /i/, /u/ and /o/. Deletion of /h/ results in hiatus which
is disallowed in Persian. Therefore, in order to resolve hiatus, intervocalic
consonants [j] and [w] are inserted after words ending in /i/ and round
vowels /u/ and/o/ respectively. In fact, these intervocalic glides replace /h/
because the sequence of one of the three vowels /i/, /u/ and /o/ and a glide
which are all articulated in the mouth is easier to produce than the sequence
of one of these three vowels and the glottal /h/ which is articulated in the
glottis rather than the mouth.
When the plural morpheme /-h@/ is attached to words that end in /@h/, the
deletion of /h/ in the stem bleeds the deletion of /h/ in the plural marker, and
vice versa. Because the deletion of both occurrences of /h/ in the stem and in
the plural marker results in hiatus which as mentioned earlier is disallowed
in Persian. Spencer (1996: 168) calls this situation mutual bleeding.
Apart from the formal style, the full form of the plural morpheme /-h@/ is
pronounced when it attaches words ending in /e/ and /@/. In this environment
no glide replaces the /h/ in the plural marker because these two vowels and
the glides [j] and [w] do not agree in the feature [+high] or in the feature
[+round]. Moreover, there is no way other Persian intervocalic consonants
replace /h/ as they would make the pronunciation more difficult.
5. Discussion
This research aimed at coming up with answers to the two research
questions mentioned earlier.The answer to the first research question is that
the following ranking is capable of explaining the pronunciation of the plural
morpheme /-h@/:
ONSET >> *C.hPL
1
>> AGREE[round] >> AGREE[height] >> MAX,
DEP Furthermore, the full form of the plural morpheme /-h@/ is pronounced
when it emphatically expresses the sense of “so many” in some exceptional
words. In these words the /h/ in the plural marker fails to delete although the
phonological environment is ready for its deletion.
The answer to the second research question is that Optimality
Theory applies Lexically Specific Constraint theory (Pater, 2006, 2008) to
analyze idiosyncratic cases in which /h/ fails to delete, even when its
conditioning environment is met. In addition, the cases in which the
pronunciation of the plural morpheme /-h@/ depended on the interface
between phonology and semantics were analyzed using Lexical-access
Constraints theory (Borsma, 2001). This indicates that OT is the only theory
capable of explaining phonology–semantics interface.
6. Conclusion
This research is the first study to discuss the pronunciation of the plural
morpheme /-h@/ in Persian. Moreover, this paper was an attempt to explain
idiosyncratic cases in which /h/ fails to delete, even when its conditioning
environment is met. These idiosyncrasies included exceptionalities as well as
the cases in which the pronunciation of the plural morpheme /-h@/ depended
on the interaction between phonology and semantics.
