URDU CONSONANTAL AND VOCALIC SOUNDS - ABDUL MANNAN SALEEM, HASAN KABIR, MUHAMMAD KHURRAM RIAZ, MUHAMMAD MUSTAFA RAFIQUE, NAUMAN KHALID, SYED RAZA ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
1
URDU CONSONANTAL AND VOCALIC SOUNDS
ABDUL MANNAN SALEEM, HASAN KABIR, MUHAMMAD KHURRAM RIAZ,
MUHAMMAD MUSTAFA RAFIQUE, NAUMAN KHALID, SYED RAZA SHAHID
ABSTRACT 2. LITERATURE REVIEW
It is a fact that no scientific speech processing According to Kachru (1990), there are seven
research has been done so far, that can be long oral vowels, and three short oral vowels,
the basis for improved applications and further and according to Bokhari (1985; 1991) there
research in Pakistan. One of the primary are seven long oral vowels, but seven short
reasons is the absence of any core material oral vowels. Bokhari (1985; 1991) contains
related to the phonetic inventory of Urdu. This many allophones of the corresponding long
paper addresses this matter by attempting to vowels as discussed by Kachru (1990).
provide a listing of all possible sounds that are Kachru (1990) maintains that the front low
present in Urdu language, and also attempts cardinal vowel [æ] exists as front middle low
at justifying their presence. vowel [] in Urdu. As a result the back low
cardinal vowel [] is shifted to the low center,
1. INTRODUCTION making it [a]. Khan (1997) also agrees with the
Urdu, the national language of Pakistan, is long and short vowel distribution of Kachru
partially spoken and more than partially (1990).
understood in South Asian countries like India, Bokhari (1985; 1991) and Khan (1997) list ten
Bangladesh and UAE. Hindi seems to be nasalized vowels including five short and five
phonetically similar to Urdu, but it differs in its long nasalized vowels. Kachru (1990), on the
script and historical characteristics. The other hand, has not listed any nasalized
pronunciation of Urdu varies from region to vowel, but mentions in the text that oral and
region due to different characteristics of their nasal vowels contrast, and that nasalization is
regions. distinctive (Kachru, 1991, p. 55).
The word Urdu has a Turkish origin, meaning Collectively Kachru (1990), Bokhari (1985;
‘camp or army with its follower’. It is popularly 1991), Khan (1997), and Hussain (1997) have
regarded as an offspring of Persian. It borrows listed forty-three (43) consonantal sounds of
words from different languages to expand its Urdu (see TABLE 1), out of which twenty-eight
vocabulary. Major languages participating in (28) sounds are agreed upon by all the above
the camp of Urdu are: Persian, Arabic, authors. Kachru (1990) lists 37 consonants
Portuguese and English (Saksena). and has missed [, h, r, n, m, l]. Hussain
It is a fact that no scientific speech processing (1997) lists 36 consonants and has missed
research has been done so far, that can be [, , r, n, m, l, q]. Bokhari (1985; 1991)
the basis for improved applications and further lists 36 consonants and he has missed
research in Pakistan. One of the primary [f, , , z, q, x, r]. Bokhari misses interestingly
reasons is the absence of any core material many basic sounds, which are listed by
related to the phonetic inventory of Urdu. This Kachru and Hussain. Khan (1997) lists, most
paper addresses this matter by attempting to of all, 42 consonants and has missed only one
provide a listing (inventory) of all possible consonantal sound [].
sounds that are present in Urdu language, and
also attempts at justifying their presence. Overall, the controversial consonantal sounds
Some comparison with other published lists is are [, , r, n, m, l, ]. This paper will try to
also done as a literature review. This listing discuss these cases in more detail.
(inventory) is the basis for any research to be
done on Urdu language speech processing.2 Center for Research in Urdu Language Processing
3. METHODOLOGY of a high fidelity (Hi-Fi) microphone, a Teac
integrated stereo amplifier and two high quality
speakers with 8-ohm impedance.
3.1 Subjects
For the purpose of finding Urdu consonantal 3.3 Experimental Conditions
and vocalic sounds, a group of eight native All subjects were required to speak a list of
Urdu speakers were surveyed, and consulted. Urdu words collectively containing all the
Furthermore, to confirm data, some relevant vocalic and consonantal sounds. Due
established and well-known dictionaries (see to the unavailability of a sound proof room, the
section 0 below) and books of Urdu language datum thus collected was vigorously screened
were studied. Many published texts were for errors.
consulted which are listed in the references
section.
4. RESULTS
3.2 Data Recording and Processing
All acoustic analysis of the speakers was 4.1 Consonants
carried out on a speech processing software
Xwaves® 5.3 by Entropic®, and Speech The inventory of all consonants is given in
Analyzer® by SIL®. The equipment consisted TABLE 1.
Sound
Type Place Manner Minimal Pairs
Symbol
Stops Bilabial Voiceless p pp, bp, bp
Voiced b pln, pln
Aspirated Voiceless p pn, mn
Aspirated Voiced b bnd, pnd
Nasalized Voiced m br, mr
p, b
pln, mln
br, mr
Aspirated Nasalized Voiced m tme
Dental Voiceless t tl, tl, dl
Voiced d tli, tli
Aspirated Voiceless t tr , dr
Aspirated Voiced d tn, tn, dn, nan
Nasalized Voiced n d, d
dm, nm
tn, dn
Aspirated Nasalized Voiced n
Alveolar Voiceless ln, ln
Voiced ,
Aspirated Voiceless l, l
Aspirated Voiced n, n
l, dal, l
Velar Voiceless k kli, li
Voiced kn, kn
Aspirated Voiceless k kai, i3
Aspirated Voiced n, kn
Nasalized Voiced ti, ti
k,
s, sn
Uvular Voiceless q qlm, lm
Glottal Voiceless lim, qlim
Fricatives Labio-Dental Voiceless f vrk, frk
Voiced w (v) f, l
Alveolar Voiceless s
ser, zer
Voiced z
Palatal Voiceless d, sd
Voiced l, l
Velar Voiced rib, qrib
Uvular Voiceless x xn, n
Glottal Breathy Voiceless h hm, nm
Affricates Alveolar Voiceless t
tl, dl
Voiced d
tl, tl
Aspirated Voiceless t
d, d
Aspirated Voiced d
Trills Palatal Voiced r mqrr, mqrb
Aspirated Voiced r
Flap Palatal Voiced b, bl
Aspirated Voiced
Approximants Back Central j ld, jd
Middle Lateral l j, n
Aspirated Lateral l
TABLE 1 Urdu Consonantal Sounds, tabulated by survey of native speakers, and some established dictionaries [2,3,4]
4.2 Vowels
The vowels present in Urdu language are
tabulated in TABLE 2
Position Sound
Type Minimal Pairs
Front/Back High/Low Symbol
Long Front High i
din, dn, dn
Middle e
bæl, bel
Low æ
Back High u
sun, sn, son
High-Middle o
bn, bon
Low-Middle
kn, kn
Low
Short Front High bn, bn4 Center for Research in Urdu Language Processing
Middle shr, shr
Back High sn, sn
Middle Middle kl, kl
Nasalized Front High i phni, phni
Long
Middle e khe, khe
Low æ hæ, hæ
Back High u hu, hu
High-Middle o dno, dno
Low kh, kh
TABLE 2 Urdu Vocalic Sounds, tabulated by survey of native speakers, and some established dictionaries [2,3,4]
4.3 Counts
Stops 23 Fricatives 9 Vowels 17
Bilabial 6 Labio-Dental 2 Long 7
Dental 6 Alveolar 2 Short 4
Alveolar 4 Palatal 2 Nasalized Long 6
Velar 5 Velar 1 Front 8
Uvular 1 Uvular 1 Back 8
Glottal 1 Glottal 1 Middle 1
Affricates 4 Trill 2 VOWELS 17
Alveolar 4 Palatal 2
Approximants 3 Flaps 2
CONSONANTS 43
Back 1 Palatal 2
Middle 2
TABLE 3 Summary of the number of different sounds
suggest that being given freedom to
pronounce anything from [v] to [w], the
5. DISCUSSIONS speakers generally tend towards a more loose
pronunciation, somewhere between the
Distinction Between [v] and [w]
Kachru explains that the phoneme [v] was approximant [w], the fricative [v], and the
originally part of Urdu (called Highly vowel [u].
Persianised Urdu by him, and Khari Boli by The discussion in (Khalid, 2002) also reveals
some others). At that stage Urdu was part of that the phoneme [w] is a more appropriate
Old Indo-Aryan languages. During the member of the consonantal inventory for Urdu.
transition of Urdu from Old Indo-Aryan to New
Indo-Aryan languages, the membership of [v] Aspirated Nasalized Stops
became questionable, and many linguists
claimed of [w] being its replacement. A
confusion whether [v] or [w] is now present in A study conducted on nasal aspirants in Urdu
Urdu was created and is present even now revealed that [n] and [m] do not occur in
(Khalid, 2002). word initial and word final positions (Aziz,
2002).
Khalid (2002) discusses the existence and
usage of [v] and [w] in much detail in his A word medial aspirated [m] is seen to occur
paper. The experiment and its analysis when its following segment is a non-vowel.
revealed that the native speakers of Urdu do One thing is certain: [m] occurs word
not distinguish between the use of [v] or [w] in
medially. However, no case for existence of
words. The message is conveyed equally well
in both cases. Also enough data is present to [n] could be found (Aziz, 2002).5
Although this suggests that [m] is more often 6. REFERENCES
broken down in two separate phonemes [m], Aziz, Omar 2002. Nasal Aspirates in Urdu.
and [h], than uttered as [m], but there are Center for Research in Urdu Language
some cases described by Aziz (2002) that Processing (CRULP), National University
confirms the existence of [m]. Also the of Computer and Emerging Sciences,
existence is highly speaker dependent (Aziz, Lahore, Pakistan.
2002).
Bokhari, Sohail 1985. Phonology of Urdu
Possible Vowel (Front, Low-Middle) in Urdu Language.
Bokhari, Sohail 1991. Urdu Zubaan ka
Soti Nizaam.
Farhang-e-Talffuz by Shaan-ul-Haq Haqqi,
Another possible vowel was encountered in Muqtadra Qaumi Zubaan. ISBN 969-474-
the study. This vowel [] was found in words, 153-X.
such as [khr]. The spectrogram of this word
is shown in FIGURE 1, and it is quite clear that Feroz-ul-Lughat Urdu, ISBN 969-000-514-
the [] vowel is spoken while pronouncing 6
these words of Urdu language.
Hussain, Sarmad 1997. Phonetic
Correlates of Lexical Stress in Urdu.
Jadeed Urdu Lughat—Tulba ke Liye, ISBN
969-474-051-7.
Kachru, Yamuna 1990. Hindi-Urdu in The
Major Languages of South Asia, The
Middle East and Africa, edited by Bernard
Comrie.
Khan, Mahboob Alam 1997. Urdu ka Soti
Nizaam.
Khalid, Nauman 2002. Existence of [v]
and [w] in Urdu Language. Center for
Research in Urdu Language Processing
(CRULP), National University of Computer
FIGURE 1 Spectrogram Analysis of [khr] and Emerging Sciences, Lahore, Pakistan.
Saksena, Ram Babu A History of Urdu.
The above example for the existence of []
Shah, Mustafa 2002. Urdu Nasal
gives a high degree of chance that [æ] and [] Consonants and Their Phonological
are allophones of each other. However, this Behavior. Center for Research in Urdu
controversy has to be examined in greater Language Processing (CRULP), National
depth, and is not discussed in this paper. University of Computer and Emerging
Other controversies of [l], [r], and [] are Sciences, Lahore, Pakistan.
also not discussed here, but they should be
explored and corrected.You can also read
