Pandey, Pramod. (To appear). Indian English Prosody. In Gerhard Leitner, Azirah Hashim and Hans-Georg Wolf (Eds.

)
Communicating with Asia. Cambridge University Press.

Indian English prosody

Abstract

This present paper reviews studies on Indian English prosody, in the main, syllable structure,

word-stress, speech rhythm, intonation and information structure. The review is supplemented

with a critical discussion of the investigations of the prosodic features, especially those that have

a bearing on related issues, such as word stress, and that are a major factor in the difference

between native varieties of English (NE, for short) and Indian varieties of English (IndE, for

short). The topics are presented with a view to approaching the problem of difficulties in the

communicative import of IndE speech.

1 Introduction

No aspect of Indian English pronunciation is as complex as its prosodic phenomena, namely,

stress, rhythm and intonation. The main factors that lend them complexity are the following.

One, the pedagogy of English in India since the earliest times excludes teaching of prosodic

phenomena. These are included today in the curricula of graduate studies of English, which

means a large section of the educated public has no training in them. Their pronunciation of

English is expected to have a predominance of the features of their first languages (L1). Two, the

rise of the media has led to an exposure of the Indian public to native English (NE) and standard

non-native (NNE) English speech. There is thus a situation today where the standards of prosody

in the speech of the speakers of Indian English have begun to change in favour of a move away

1
from a very close affinity to the first language. Yet, by and large, the prosodic features of IndE

have their own characteristic features.

A factor that has lent complexity to the substantive features of prosody in Indian English

is the variation in the prosodic patterns in the source language of Indian English speakers. Each

source language has its own prosodic features, which are further differentiated on account of

their belonging to different language groups, such as Dravidian, Indo-Aryan and Tibeto-Burman,

among others. The variation involves differences in syllable-structures, word-stress patterns,

phonological phrasing and intonation contours. For a difference in the intonational contours

between the Indic languages, see, for example, Harnsberger (1994) for Hindi and Keane (2007)

for Tamil. Wiltshire (2005) is, thus, right in pointing out that Indian English should not be seen

as a single variety. Across all regions, two main varieties are especially noticeable- one, a

basilectal variety, containing strong source language features, and an acrolectal variety which is

the ideal variety, also proposed as General Indian English (GIE) (e.g. Masica and Dave 1972)

Notwithstanding the complexity inherent in a discussion of the prosodic phenomena in

IndE, an attempt is made here to give an account of those aspects of IndE prosody that have been

well-studied and that give us a clear idea of the unique features of IndE prosody as well as of the

common features that it shares with Native English (NE) varieties, especially British English

(BrE) and American English (AmE). The variety of IndE that is at the basis of the discussion

here is the variety that is between the basilectal, „Babu‟ or Butler English variety (e.g. Hosali and

Aicthison 1986) and the acrolectal, highly educated variety. This is the variety that has both L1

transfer features as well as independent features and is the most widely distributed across all

regions and economic and educational classes in India. The reason for this selection is that there

2
are detailed investigations of this variety, with hardly anything for the acrolectal variety to

compare with. The discussion focuses on the description of the prosodic phenomena as well as

the phonetic evidence for them.

2 Syllable Structure and syllable division

2.1. Syllable structure

This is an area that has lent Indian English its complexity on account of L1 interference in both

syllable structure types and syllable division in words. As IndE speakers are distributed in all

regions of India and cumulatively give it its character, it is reasonable to see why this should be

so. The most significant reason for the complexity here is the variety of canonical syllable

structures in the L1 of IndE speakers. Pandey (2014) lists 50 types of canonical syllable

structures found in Indic languages. A lot of them are fairly close to NE, presented below (see

Roach 2013 [1993].

The canonical syllable structure of NE is(C)(C)(C)V/C (C)(C)(C)(C), that is, a vowel

or syllabic consonant as the Nucleus and a minimum of nil and a maximum of three onset

clusters and four coda clusters. The consonant clusters have restrictions, stated as follows (C= a

consonant, L= a liquid /r l/, G= a glide /j w/:

(1)

2 Consonant Clusters: Initial: (i) C+L/G; (ii) s+C.

Final : general
Media : general.

3 Consonant Clusters: Initial: /sprstrsmr/

Final: C+C+s/z/t/d/θ
Medial: general.

4 consonant Clusters: Initial: Nil

3
 Medial: general
Final: lfθs, -mpts, -kst/θs

It is interesting to observe that some varieties of IndE share the syllable structure of NE, except

for the absence of syllabic consonants, as, for example, Hindi English (Pandey 1985). But a large

number differ in Onset and Coda clusters. Pandey (2014: 241-51) presents a detailed discussion

of the syllable structure types in Indic languages. To what extent and how exactly L1 interferes

in the organization of syllable structures in Indian English is a topic of much research interest.

While there are cases of syllable structures in some sub-varieties such as Hindi English that are

close to NE, there are others, like Bengali English, where there are differences between L1 and

NE (Bengali has the canonical structure ((C) (C) (C) V (C)), but there is no interference in the

sub-varieties; Bengali English has the complex onsets and codas of NE syllables. There are also

cases where there is strong L1 interference in a sub-variety, as in Khasi English (Raghavan 2014)

and Chakma English (Bardhan 1999). Some examples of the pronunciation forms in Khasi

English and Kokborok English are presented below:

(2) Khasi English Chakma English

Surface forms Surface forms

present [prezn] adopt [ædɒp]

matched [mæʧ] touched [ta:ʧ]
object [ɔbʤek] task [ta:ʧ] or
cold [kol] bold [bol]
stamp [stæm] last [laʧ] or [la:t]
scent [ʃen]
texts [teʧ] or [tek]

Chakma English has two allophones of /s/- [ʃ], occurring word-initially or in the onset position in

a syllable, and [ʧ], occurring word-finally or in the coda position in a syllable. As can be seen in

4
(2), the English consonant clusters are simplified in the two varieties with the elision of final

consonants. In Chakma English, in some cases, the penultimate consonant may be deleted,

instead of the final.

In the case of complex onsets, in the varieties of Indian English whose source languages

do not permit them, the words surface with vowel epenthesis or prothesis. For instance, the word

school is pronounced as [is.ku:l] in many areas in the northern and the north-eastern region, such

as the Hindi-speaking areas, Assamese English, and Chakma English. The word surfaces as

[səku:l] in Punjabi or Kashmiri speaking areas (e.g. Bhatt 2000). Notice that the consonant

clusters in the English words with word-initial sC clusters, such as school and state, apart from

differing from L1-specific constraints, also violate (as noted by Bhatt 1995) the universal

Sonority Sequencing Constraint (Goldsmith 1990). These could well be seen as instances of the

emergence of the unmarked features of syllable structure.

2.2 Syllable division

As extensively discussed in Chomsky & Halle (1968), the grouping of word-internal

clusters in NE (leading in effect to syllable division), does not allow a division between CL/G

and sC clusters, as in the following words:

(3) (a) be.tray, be.tween, a.pply, re.ply, se.cure

(b) mi.stake, mi.ni.ster, mi.ni.scule

A consequence of there being no division between the Consonant and a glide or liquid (i.e. w j r

l) in the forms in (a) is that the voiceless plosive is stressed and, consequently, aspirated-

[bɪˈʈʰreɪ], [əˈpʰlaɪ] etc., *[bɪʈ.ˈreɪ], *[əp.ˈlaɪ] etc. Had the consonant been grouped with the

preceding vowel, then it would not be a candidate for aspiration. Again, a consequence for there

5
being no division within an sC sequence is that the second syllable from the right does not

qualify for stress, it being a „weak cluster‟ in Chomsky & Halle‟s (1968) description of a light

syllable- [ˈmi. ni. stə]. Had there been a division within sC, the second syllable from the right

would have been stressed, because then that would have been a ‘strong cluster’ or a heavy

syllable- *[mɪˈnɪstə].

In most varieties of Indian English, there is a syllable division between the two

consonants in two consonant sequences. CG/L and sC sequences are no exception. Thus even

though the canonical syllable structure in Hindi English is very close to the same as in NE, on

account of syllable division, stress patterns are different- [miˈnisʈər].

A syllable division within the two types of clusters is responsible for another process, a

transfer from L1 of a process of gemination taking place following a general process in many

Indic languages, whereby a CL/G cluster is realized as CCL/G cluster, as for example,

be[ʈʈr]ay, be[ʈʈʋ]een, ə[ppl]y, re[ppl]y, se[kkjo]re.

3 Word-stress

3.1 Word-stress pattern

Word-stress in NE plays a significant role in its phonological organization. As is generally

recognized, aspiration of voiceless plosives takes place in the initial position in stressed

syllables, stressed vowels are fully pronounced and unstressed vowels tend to be reduced as can

be seen in the realizations of the vowels in alternating forms such as ˈph[əʊ]to, ˈphot[ə]ˌgr[a:]ph,

6
ph[ə]ˈt[ɒ]gr[ə]pher, ˌphotoˈgr[æ]phic. Most of these word-stress dependent phenomena are not

found in the common varieties of IndE. Thus, in all the forms of the words containing photo, the

pronunciation remains constant- [fo:ʈo:].

Structural descriptive accounts of NE word-stress (see e.g. Roach 2007[1993]) show that

it is sensitive to the morphological and syntactic structures of words: simple and complex nouns

have stress assigned up to the antepenult syllable (e.g. Aˈmerica, diˈvinity, etc.), while

adjectives and verbs have stress assigned up to the penult syllable (e.g. speˈcific, soˈlicit,

deˈvelop, etc.). Almost all accounts of IndE varieties, for example, Babu (1971), Dhamija

(1976), Vijaykrishnan (1978), Gokhale (1982), Pandey (1980), Chaudhary (1984), and Wiltshire

(2005), show that stress assignment is not sensitive to the grammatical category and

morphological structure of words. The following patterns of word-stress in Hindi English present

in (4) are a case in point.

(4) Verb
eˈmerge suˈrrender ˈdiminish
eˈlect ˈdiffer *ˈdevelop
aˈdopt *ˈsolicit
Adjective
seˈcure *siˈnister *ˈterrific
suˈpreme *seˈmester *ˈprolific
Noun
aˈlarm uˈtensil ˈbenefit
saˈloon aˈsylum Aˈmerica

The description of a stress foot in IndE within the framework of Metrical Theory of stress

(Hayes 1995) varies in its sub-varieties because of the filter of the L1 of the speakers. In Hindi

7
English and Punjabi English it is quantity-sensitive(e.g. Pandey 1985, Sethi 1980), as also in

some of the Tibeto-Burman varieties in which stress is placed on syllables depending on whether

they are heavy or light (e.g. Wiltshire 2005:295). It is quantity-insensitive in Tamil English (e.g.

Vijaykrishnan 1978), in which the first syllable is stressed, irrespective of the weight of the

internal syllables. This difference can be more intricate. The quantity-sensitivity of stress

systems may depend only on vowel length but not on the presence of codas in syllables, as in

Malayalee English (e.g. Sadanandan1981)or it may depend only on codas but not on vowel-

length as in Assamese English (e.g. Mahanta 2001). Among quantity-insensitive stress systems,

we have word-initial systems, such as Tamil, and word-final systems which characterize a large

number of Tibeto-Burman languages of north-east India, including the languages of Arunachal

Pradesh and Manipur.

In general, the following features characterize word-stress features of IndE:

a. Word-stress in most cases is predictable. In the sub-varieties such as Hindi English or

Punjabi English, final syllables are stressed only if they are superheavy (i.e. V:C or VCC type)

such as exˈplore, maˈchine etc. Words ending in V:/ VC syllables are in general not stressed, e.g.

*ˈcanoe,*ˈdegree,*ˈtotal, etc. Words ending in diphthongs are stressed finally not because they

function as long vowels, but because they function as sequences of two vowels (e.g. Pandey

1981 [1992]), e.g. a:u or au, a:i or ai, i:ə or iə, etc. as in aˈllow, deˈfy, etc.

In the sub-varieties with quantity-insensitive systems as L1, such as Tamil English (e.g.

Vijaykrishnan 1978), they are usually stressed on the first syllable, ˈexplore, ˈmachine, ˈallow,

etc. In languages with quantity-sensitive systems as L1, such as Punjabi English (e.g. Sethi

8
1980), stress is placed on the consideration of syllable-quantiy and syllable position from the

right, as in NE. In both, stress patterns are derived (i.e. predictable).

The resultant effect of predictability and regularity in word-stress in IndE is that the

regular and lexical stresses in NE have undergone levelling. Stress patterns in NE that must be

seen as lexical as, for example, in caˈsette, caˈnoe, and deˈgree (e.g. Hayes 1982), are regular in

IndE: *ˈcassette, *ˈcanoe, *ˈdegree, etc.

b. The segmental structure of many words in IndE has undergone restructuring, with

changes in vowel length, involving both shortening of long vowels, as in medieval *[ˈmedivəl] as

well as lengthening of short vowels, as in event *[ˈi:vent].

d. Compound words, with stress on the two or the three words, have a fixed pattern as

shown in (5): the first member of the compound has primary stress in most sub-varieties of IndE.

(5) (i) *ˈcollege ˌcanteen ˈWhite ˌhosue

*ˈnorth-ˌeast *ˈtwo-ˌwheeler
ˈsuit ˌcase ˈtea-ˌcup

(ii) *ˈwaste ˌpaper ˌbasket *ˈkitchen-ˌtowel ˌrack

3.2 Phonetic cues of word-stress in IndE

A general observation about the phonetic cues of word-stress in IndE (e.g. Mohanan

1986, Pickering 1999, Pickering and Wiltshire 2000, Wiltshire and Moon 2003, Puri 2012, 2013)

is that stress is realized in terms of a fall in pitch on the stressed syllable. Wiltshire and Moon

(2003), in an investigation of the phonetic correlates of word-stress in Hindi and a comparison

between American English (AmE) and IndE arrive at the conclusion that increase in amplitude

9
and duration are also associated with stressed syllables in addition to fall in F0, although the

former are not reliable cues of word-stress. They follow Beckman (1996) in using the term „pitch

accent language‟ to characterize IndE, like Japanese, in which a fall in F0 is the main phonetic

correlate. A recent study by Féry, Kenntner and Pandey (2014) offers a different interpretation of

the facts. What has been generally observed as a fall in F0 is often accompanied by a

concomitant rise, giving the shape of a „hammock‟, in addition to amplitude and duration as not

so significant correlates. This is especially prominent in focused words, as discussed in section 6

below.

3. 3 Stress clash resolution

When adjacent words in a phonological phrase (e.g. Nespor and Vogel 1986) have

adjacent stressed syllables, the phenomenon is called „stress clash‟, as, for example, in thirˈteen

ˈmen. Languages differ with regard to whether they allow stress clash to persist or to be resolved

by means such as deletion of stress or movement of stress to a different syllable. English is

known (e.g. Hayes 1984) to resolve stress clash by shifting stress to a non-adjacent syllable, as in

ˈthirteen ˈmenˈthirteen ˈmen. Puri (2012), shows that in the speech of IndE speakers from Delhi

with 1-2 years‟s stay in the U.S. stress clash in constructions such as routine check-up, sixteen

cakes, Japanese dialects is resolved by a change involving a simplification of the pitch contour

with a loss of pitch accent on the following syllable. The pitch contour L*H L*H on adjacent

stresses changes to L*HL.

4 Speech rhythm

10
English speech rhythm is known (e.g. Adams 1979) to be „stress-timed‟, that is, the duration

between stresses in NE spoken utterances is roughly equal. Thus, the duration of the stretches

between ˈ ma- and ˈ here in the sentences Theˈ manager is ˈ here and Theˈ man is ˈ here are

roughly equal in the speech of a NE speaker. Stress-timed rhythm is characterized by the

process of weakening of the unstressed syllables manifested by one of the following processes:

(a) the reduction of the unstressed vowel to schwa, (b) the deletion of the unstressed vowel,

especially in fast speech, (c) contraction of function words, as exemplified below:

(6) (a) ˈphoto, ˈphotoˌgraph, phoˈtographer

(b) oˈccasionally, poˈtato [ptʰɪetəʊ]

(c) we‟ve, they‟re

The underlined unstressed vowels in (a) and (b) are reduced and deleted respectively. The

contracted forms involve mainly function words, such as have, had, are etc. These words are

unstressed in the neutral context and occur in their „weak‟ forms, for instance, [həz]/ [əz] for has,

[kən] for can. When stressed, in given contexts, they occur in their „strong‟ forms, for instance,

[hæz] for has, [kæn] for can. Phoneticians (e.g. Ladefoged1993, Wells 2000, Roach 2001) have

stressed the need for the pronunciation of weak forms in English in order to avoid

miscommunication and sounding very different from NE.

In contrast with speech rhythm in NE, speech rhythm in IndE is generally labeled as „syllable-

timed‟ (e.g. Bansal 1969, Wells 1982, Kachru 1983, Gargesh 2004), that is, the number of

syllables in it determines the duration of spoken units. The classification of speech rhythm in

IndE as syllable-timed has come to be questioned on close examination. As pointed out by

Sailaja (2009), the „syllable-timed‟ nature of rhythm in IndE has not been verified. Besides, the

11
definition of „syllable-timed‟ speech rhythm itself has come to be seen as controversial (Roach

1982) and to be inapplicable to languages such as Tamil (Balasubramanian 1980) and Telugu

(Babu 1971), the substrate languages of the IndE varieties Tamil English and Telugu English.

However, some of the features of stress-timed rhythm are certainly not found in IndE varieties.

For instance, related forms like ˈphoto, ˈphoto graph and phoˈtographer tend to have full vowels

once they are stressed. Besides, function words such as has and can tend to have the vowels in

them pronounced in full, giving the impression of pronouncing them in strong forms. There is

some ground thus for distinguishing between the two types of speech rhythm (see e.g. Ohala and

Gilbert 1979, Auer 1993), although the terms need to be redefined. Is there acoustic evidence for

syllable-timed rhythm in IndE? Fuchs (2012), on the basis of a comparative study of BrE with

educated or acrolectal variety of IndE with Hindi, Bengali, Telugu or Malayalm as their L1,

found that there was some evidence for the syllable-timed speech rhythm in IndE, due to less

variability in the durations of vocalic intervals in IndE than in BrE.

In a study of the pronunciation of function words in IndE by a group of 20 students of the

age group 20-23 years, Madhavi (2009) reports the finding that out of a reading test of 36 words,

consisting of 26 function words and 8 contracted forms, only 10 of the 34 forms were

pronounced in their weak forms or contracted forms by a meagre 5% subjects and only 2 forms

by 10% subjects. All the subjects pronounced 22 of the 34 forms in their strong forms. The

forms were placed in the context in which they are pronounced in their weak forms in NE (as in

the underlined words in, for example, I asked her to take them away).

Among other aspects of rhythm, Pre-Boundary lengthening (PBL) has been reported to

characterize many languages. PBL is manifested in the lengthening effects on the duration of

vowels before prosodic boundary, as a cue to the prosodic unit. Puri (2013) presents a detailed

12
discussion of PBL effects in languages. I her experimental study, she finds evidence for PBL in

intonational Phrases. Thus stressed vowels at the end of an Intonational Phrase (IP) are longer

than the stressed vowels in the middle of an IP or at the end of a Phonological Phrase (PP) in

both Hindi and IndE. Examples of sentences in (7) and (8) are for Hindi and IndE.

(7) a. unke na:m hai ̃ sǝndi:p aur 'pu:tna:

Their names are Karan and Puutna
Their names are Karan and Puutna

b. unke na:m hai ̃ 'pu:tna: aur sǝndi:p

Their names are Putna and Karan
Their names are Putna and Karan.

(8) a. She lived in Jaˈpan and China.

b. She lived in China and Jaˈpan.

The vowels in pu:tna: and Japan are longer in (7b) and (8b) than in (7a) and (8a). It

was found that Hindi and IndE both shared this feature. This similarity could be

attributed in our view to L1 transfer or a universal. In this aspect, IndE was seen to

share the feature of PBL with NE (e.g. Beckman and Edwards 1990).

5 Intonation

5.1 Intonational unit

Most intonational studies on IndE have been impressionistic, on the one hand, and of specific

sub-varieties, on the other. This statement is true of Latha (1978) and Nair (2004) for Malayalee

English, Joseph (1984) for Telugu English, and Gokhale (1980, 1984) for Marathi English,

among others. There have been some studies of intonation of GIE, in the main, Babu (1971),

13
Khan (1974) and Shekhar (1993). The latter two examine the intonation of All-Indo Radio

announcers of IndE. One of the claims consistently made in these studies is that IndE is

characterized by multiple stresses in an intonational unit. The impression is so strong, in the case

of both Indic languages as well as IndE, especially to foreign ears, that Féry (2011) proposes a

new type of intonational language that properly describes these languages, namely, Phrase-

Accent language. There have been some recent experimental studies of IndE varieties that

corroborate the observation regarding the placement of multiple stresses in an intonational unit.

These are, in the main, Verma (2005), Puri (2013), Féry, Pandey and Kenntner (2014) for Hindi-

English, and Wiltshire and Harnsberger (2006) for Tamil English and Gujarati English.

5.2 Types of tone

In studies of the intonation of Marathi English, Gokhale (1978, 1980) describes three

simple tones- Fall, Low-Rise, and High Rise, and one complex tone- Fall-Rise. Studies on

rhythm and intonation in IndE (e.g. Babu 1971, Khan 1974, Joseph 1984, Shekhar 1993, Shuja

2005, Sailaja 2009) are in general in agreement with Gokhale‟s observation regarding the lack of

a crucial difference between IndE varieties and NE with regard to the use of the types of tone in

them. The tones in IndE varieties form a subset of the tones used in BrE. One exception here is

Latha (1978). She posits the following tones in Malayalee English: One simple tone- Fall, and

four complex tones- Rise-Fall, Fall-Rise, Drop-Rise and Drop-Rise-Fall. The types of tone in

Malayalee English are obviously quite different from BrE in the (lack of) employment of some

of the tones. Notice that a simple Rise tone is missing in Malayalee English, in which a Rise

begins with a drop.

The use of the nuclear tone in IndE varieties is not predictable, as in NE, where the

nuclear accent is placed on the last important word in the neutral context. There is no such

14
reliable rule in IndE (see Bansal 1976:27). We look further at this phenomenon in the Section 6

on information structure below. The only reliable generalization that holds for the placement of

the nuclear tone is the penultimate word in the following IP constructions (in the framework of

Basic Linguistic Theory (see Dixon 2010)): (a) those ending in an NP with a Modifier + Head

structure, such as Mary wore a blue dress, (b) those ending in clauses with a simple VP + NP

or VP + AdvP constructions, such as John found the book or John went home or John is

standing there. The nuclear tone in these intonational units is placed not on the final Head noun,

but on the penultimate word. This appears to be a general pattern in Indian English, as also in

Indian languages as diverse as Hindi (Bansal 1976, Puri 2013), Malayalam (Mahesh 2014) and

Meitei (Meiraba 2014).

5.3 Discourse features: Tag questions

Tag questions in NE are short questions at the end of statements, as in (9) below:

(9) a. He has been there, hasn‟t he?

b. You will be coming with the papers, won‟t you?

A tag question is in the negative, when the matrix sentence is a positive statement, as in

(9), and is in the positive when the matrix sentence is a negative statement, as in (10):

(10) a. He hasn‟t been there, has he?

b. You won‟t be coming with the papers, will you?

Tag questions are normally used to confirm the truth of a statement or to encourage the

interlocutor to continue a conversation. In terms of the prosodic structure, tag questions are like a

„tail‟ in an IP, following the nuclear pitch. They can also function like an independent question

and thus be an independent IP.

15
 The common, regional varieties of IndE are known to use an undifferentiated tag question

„Isn‟t it?‟ added in any sentence with the matrix sentence in the positive or the negative, and the

Subject being always „it‟ (see e.g. Trudgill and Hannah 2002, Mesthrie and Bhatt 2008, Sailaja

2009).

In regional varieties of IndE, questions are also formed with an invariant „no‟ as a tag

(see e.g. Sailaja 2009: 59). (Pandey 2011) carried out a brief study of the undifferentiated tag

questions „Isn‟t it?‟ and „no?‟ in IndE, with 6 speakers, 3 male and 3 female, all graduate

students at Jawaharlal Nehru University. They were given 6 sentences each for two types of

questions, one containing „no‟ and the other containing „Isn‟t it?‟ at the end of matrix sentences.

There were thus 36 sentences in all, equally divided between the two types of questions. It was

found that while the invariant „no‟ showed features of a sentence final question word, the

undifferentiated „Isn‟t it?‟ showed features of an independent question, as can be seen in Figure 1

and Figure 2 below.

16
 0.4824

500
-0.4752
0.09612 1.164
400 Time (s)

300
Pitch (Hz)

200

75
you have been to Agra No

.8s .2s

0.09612 1.164
Time (s)
Figure 1: You have been to Agra, no?

17
 0.5598

500
-0.6178
0.01644 2.049
400 Time (s)

300
Pitch (Hz)

200

75
the train was on time today isn‟t it?

1.55 sec1.61s .084s .35s

0.01644 2.049
Time (s)

Figure 2: The train was on time today, isn’t it?

A comparison between Figure 1 and Figure 2 shows that whereas there is a continuity of

the pitch contour in the „no‟ question, there is a break in the pitch contour before the „isn‟t it‟ tag.

The duration is .084 seconds, the average being .12 seconds. The interval duration between the

matrix sentence and the tag question was found to be close to .3 seconds in 2 of the 18 „isn‟t it‟

tag, as in Figure 3 below:

18
 0.6936

500
-0.6449
0.08267 1.588
400 Time (s)

300
Pitch (Hz)

200

75
you know Remesh Isn‟t it?

.83s .29s .37s

0.08267 1.588
Time (s)

Figure 3: You know Ramesh, isn’t it?

The difference between the prosodies of the tag question „Isn‟t it‟ and the „no‟ question in IndE

speech needs an explanation. The one that comes readily to mind is that the undifferentiated tag

is an independent question about the truth of the statement in the matrix sentence. What

differentiates the undifferentiated tag question in IndE from such a question in NE is that the tag

question in IndE is undifferentiated and that it functions always as an independent question,

unlike in NE where it can optionally function as an independent question when so intended. It

should be noted, however, that the use of the undifferentiated tag question and the „no‟ question

is found more commonly in the basilectal varieties of IndE than in the acrolectal, idealized

model. Their preponderance, however, in IndE speech is common.

19
6 Information Structure: Focus

The characteristic features of rhythm and intonation in IndE have been found to cause

considerable problems of communication, noted by many authors (e.g. Bansal 1972, 1983, Wells

1982). The basic issue here is of the difference in the information structure conveyed by the

intonational phrase. On account of multiple pitch accents and unpredictable placement of the

tonic, the intonational pattern of a clause does not match the information structure in NE. There

have been some recent studies of the phonetics of information structure in IndE that are pertinent

to the discussion here.

Information structure in a language is commonly studied in relation to the packaging of

information in a sentence in terms of the notions of Focus, Givenness and Topic and their

complementary notions of newness, background and comment. All other aspects of information

such as style, speaker‟s identity or encyclopaedic knowledge are left out of consideration. The

two most important notions are Focus, that gives new information and Given, that contains the

background information.

Verma (2005) in a study of the realization of Focus showed that Hindi bilingual speakers

of Indian English transferred the pitch melody of Hindi to IndE. Hindi is known to have L*H

melody (Dyrud 2001) for stress, while NE (e.g. Gussenhoven 2004, Pierrehumbert 1980,

Steedman 2000, among others) has the H*L melody for Focus realization. While NE does not

have a pitch melody on the Pre-focus elements, IndE speakers showed considerable instability in

switching between the L*H melody of stress to the H*L melody of Focus. No pitch melody was

found in the Post-Focus elements.

Féry, Pandey and Kenntner (2014) is an experimental study of the prosodic correlates of

focus in Hindi and Hindi bilingual speakers of IndE. The data for the study contained both

20
focused and given words, collected in a recorded form as response to a task of QUIS

questionnaire (Skopeteas et al. 2006), known as „Anima‟. The focused words included three

main types, namely, Correction (the informant was asked to correct information), Information

(the informant was asked to give new information), and Selection (the informant was asked to

choose between alternatives). The focused words belonged to two constituents- Subject and

Object. The data on IndE showed that Object Focus showed significant results more than the

Subject focus, realized as higher F0 with L*H melody on non-final elements and H*L melody on

final elements, a „hammock‟ structure realized as H*LH melody, higher amplitude and greater

duration. The last two are not significant. The „hammock‟ structure is found also for Hindi in this

study. It has been reported to characterize question intonation in Dutch (see e.g. Hahn et al.

1997:98). However, as a melody for the realization of focus, it has been found to characterize

Indic languages mainly (see also Mahesh 2014 for Malyalam).

The following illustration from the data collected for the study in Féry and Pandey (2012)

shows two aspects of the prosody of focus in IE. The sentence The man is pulling a woman with

Object focus, shows the HL melody on the Object, with H at the beginning of the Phonological

Phrase „woman‟ and L at the end. The Subject and the Verb group show the LH melody. Apart

from the melody, Focus was found to be aligned to the left of a phonological phrase. The

sentence „The woman is hitting the plant‟, with Object focus, shows the results of alignment on

the phrasing being different from native English, where the phonological phrases show different

groupings among words. There are three phonological groups- the man is, pulling a, and woman.

The separation of the article from the Noun in the Object shows encliticization as well as glottal

stop insertion to be features common for Focus realization in both Hindi and IndE, as (discussed

21
in greater detail in Féry et al. 2014). The symbols Φ and ι stand for the phonological phrase and

the intonational phrase.

0.5378

400
-0.2738
2.563 4.313
Time (s)

300
Pitch (Hz)

200

75
the man is pulling a woman

2.563 4.313
Time (s)

LΦHΦ HΦLι

((The man is)Φ (( pulling a) womanΦ))ι

Puri (2013) is a detailed investigation of three aspects of prosody of late and

simultaneous bilinguals of Indian English from Delhi- pre-boundary lengthening, pitch accent

and focus. Her experiments show that both Hindi and IndE have pre-boundary lengthening like

many languages, including American English. The study shows that there is no difference in the

two groups in the domain of pre-boundary lengthening: both groups indicate the end of IP

through pre-boundary lengthening effects (a) on the final as well as the penultimate syllable

22
before the IP, (b) more on the rhyme than on the vowel, and (c) of stress on the final or

penultimate syllables. With regards to pitch accent, the study gives evidence for the use of the

Hindi LH pitch accent in IndE as well. However, the simultaneous bilinguals use both the Hindi

LH and to a limited extent (17% of the time) use the British English H*/H*L pitch accent. Thus,

while late bilinguals show static interference in the use of the Hindi LH pitch accent, the

simultaneous bilinguals show a fusion of NE and L1. The results on Focus show that both

simultaneous and late bilinguals use a bigger F0 excursion in narrow focus, and both groups have

post focal compression due to lower duration, RMS amplitude and F0 range in Hindi and IndE.

Summary and conclusion

In the above discussion of intonational prosody in IndE we have tried to show that it is

much affected by the prosody of the source language. The amount of affection may vary from a

full to a small degree. The more basilectal the speech, the greater is the rate of transfer of the

prosodic features of the source language. The case from the Tibeto-Burman varieties of IndE

offers a different picture as discussed in Wiltshire (2005). Wiltshire quotes from Lalrindki (1989:

29) that I reproduce below:

… to a lay person, a Mizo speaker speaking English sounds rather „flat‟ or „toneless‟.
This seems to suggest that Mizo speakers consider the pitch system ofMizo and English
to be so different that the former should not be allowed to influence the latter at all and
hence the tonelessness of the latter.

Wiltshire adds (p. 296):

From the recordings of both the readings and the interviews, this observation seems to be
true of all TB-IE speakers; while GIE uses more pitch accents than American or British
English, the TB variety of English uses fewer.

23
We can stop here with the remark that although there is considerable amount of uniformity in the

varieties of IndE prosody, lending it its „natural‟ character, regional variation in the basilectal

varieties should be as naturally expected.

7 Conclusion
The present paper has taken a close look at IndE prosody in relation to the phenomena of stress,

rhythm and intonation. These phenomena were discussed in relation to the prosodic categories

proposed in the theory of Prosodic Phonology, namely, the Syllable, the Phonological Word, the

Phonological Phrase and the Intonational Phrase. The emphasis in the paper has been on those

aspects of IndE prosody that have been well-studied and that give us a clear idea of the basis of

the unique features of IndE speech as well as of the common features that it shares with NE

varieties, especially British English and American English.

Acknowledgements

The data on Indian English from the project from Féry and Pandey (2012) are drawn from a joint

NSA-DFG project on focus in Hindi and Indian English with cooperation between Caroline Féry

(Goethe University, Frankfurt) and Pramod Pandey (JNU Delhi) and partly funded by DFG. I‟m

grateful to M. Mahesh for the recording and preparation of material for a brief study by the

author on Tag Questions in Indian English.

24
References

Adams, C. (1979). English Speech Rhythm and the Foreign Learner. The Hague: Mouton.

Auer, P. (1993). Is a rhythm-based typology possible? A study of the role of prosody in

phonological typology. KontRI Working Paper 21: Universität Hamburg.

Babu, P. (1971). Prosodic Features in Indian English: Stress, rhythm and intonation. Research

diploma dissertation. Hyderabad: CIE&FL

Balasubramanian, T. (1980). Timing in Tamil. Journal of Phonetics 8. 449-68.

Bansal, R. K. (1976 [1969]).The Intelligibility of Indian English. 2nd edn. Monograph 4.

Hyderabad: CIE&FL.

Bansal, R. K. (1983).Studies in Phonetics and Spoken English, Monograph 10. Hyderabad:

CIE&FL.

Bardhan, S. K. (1999). A phonological study of Chakma and English. M. Phil. dissertation.

Hyderabad: CIE&FL.

Beckman, M. E. & Edwards, J. (1990). Lengthenings and shortenings and the nature of prosodic

constituency. Kingston, J. & Beckman, M. E. (eds.), Papers in Laboratory Phonology I:

Between the Grammar and the Physics of Speech. Cambridge: Cambridge University

Press. 152–178.

Beckman, M. (1996). Stress and Non-stress Accent. Dordrecht: Foris.

Bhatt, R. M. (2000). Optimal expressions in Indian English. English Language and Linguistics 4

(1): 69–95.

25
Chaudhury, S. (1984). Some aspects of the phonology of Indian English. Ph. D. Dissertation.

Hyderabad: CIE&FL

Chomsky, N. and Halle, M. (1968). Sound pattern of English. New York: Harper & Row.

Coelho, G. (1997). Anglo-Indian English: a nativized variety of Indian English. Language and

Society 26. 561–89.

Dauer, R. (1983). Stress-timing and syllable-timing reanalyzed. Journal of Phonetics, 11,51-62.

Dauer, R. (1987). Phonetic and phonological components of language rhythm. Paper presented

at the XIth International Congress of Phonetic Sciences, Tallinn.

Dhamija, P. V. (1976). A phonological description of Rajasthani English. M. Litt. Dissertation.

Hyderabad: CIE&FL.

Dixon, R. M. W. (2010). Basic linguistic theory, Volume 2. Oxford University Press.

Dyrud, L. (2001.) Hindi-Urdu: stress-accent or non-stress accent? MA thesis, University of

North Dakota, Grand Forks.

Féry, C. (2010). Indian languages as intonational „phrase languages‟. Hasnain, I. & Chaudhury,

S. [Eds.], Festschrift to honour Ramakant Agnihotri. New Delhi: Aakar Publisher. 292-

316.

Féry, C. & Pandey, P. (2012). Focus in Hindi and in Indian English. Collaborative research project

at the Institute of Linguistics, Johann Wilhelm Goethe University, Frankfurt-am Mein and

Department of Linguistics, Jawaharlal Nehru Univeresity, New Delhi.

Féry, C., Pandey, P. & Kenntner, G. (2014). The prosody of focus and givenness in Hindi and

Indian English. Under submission.

Fuchs, R. (2012). A duration-based account of speech rhythm in Indian English. Poster

presentation at Laboratory Phonology (LabPhon) 13, Stuttgart.

26
Gargesh, R. (2004). Indian English: Phonology. Edgar, W. Et al. (eds.), A Handbook of Varieties

of English, Vol 1:Phonology. Berlin: Mouton de Gruyter. 993–1002.

Gokhale, S. B. (1982). The intonational system of Marathi and Marathi English. Ph. D. thesis.

Hyderabad: CIE&FL.

Goldsmitth, J. (1990). Autosegmental and metrical phonology: A new synthesis. Oxford:

Blackwell.

Gussenhoven, C. (2004). The phonology of tone and intonation. Cambridge University Press.

Haan, J., van Heuven, V. J. , Pacilly, J. & van Bezooijen, R. (1997). An anatomy of Dutch

intonation. Coerts, J. A. & de Hoop, H. (eds.), Linguistics in the Netherlands 1997. Pp.

97–108.

Harnsberger, J. (1994). Towards an intonational phonology of Hindi. Unpublished dissertation.

Madison: University of Wisconsin.

Harnsberger, James D. (1999).The role of metrical structure in Hindi intonation. Paper presented

at the South Asian Language Analysis Roundtable, University of Illinois, Champaign-

Urbana.

Hayes, B. (1982). Extrametricality and English stress. Linguistic Inquiry 13, 227-276.

Hayes, B. (2004). The phonology of rhythm in English. Linguistic Inquiry 15, 33-74.

Hosali, P. & Aitchinson, J. (1986). Butler English: A minimal pidgin? Journal of Pidgin and

Creole Languages 1, 11: 51–79.

Joseph, K. G. (1984). Intonation Patterns of Telugu English. M.Litt. dissertation, CIE&FL ,

Hyderabad.

Kachru, B. (1983). The Indianization of English: The English language in India. Oxford: Oxford

University Press,

27
Keane, E. (2005). Prominence in Tamil. Journal of the International Phonetic Association, 36: 1-

20.

Keane, E. (2007). Investigating the intonational phonology of Tamil. ICPhS 2007 Satellite

Workshop.http://www.linguistics.ucla.edu/people/jun/Workshop2007ICPhS/Papers/Kean

e_handout.pdf

Khan, N. (1974). A Phonological Analysis of English Spoken by All India Radio Newsreaders.

Post-Graduate Research Diploma dissertation. Hyderabad: CIE&FL,.

Ladefoged, P. (1993). A Course in Phonetics. Fort Worth: Hartcourt Brace College Publishers.

Lalrindiki, T. F. (1989). Some aspects of the autosegmental phonology of English and Mizo. M.

Litt. Thesis. Hyderabad: CIE&FL.

Latha, P. (1978). Intonation of Malayalam and Malayalee English: a study of comparison and

contrast. M. Litt. Dissertation. Hyderabad: CIE&FL.

Madhavi, V. (2009). A Phonetic Study of Strong Forms and Weak Forms in Telugu English. M.

Litt. Dissertation. Hydrerabad: The English and Foreign Languages University.

Mahanta, S. (2001). Some aspects of prominence in Assamese and Assamese English.M.Phil.

dissertation. Hyderabad: CIE&FL.

Mahesh, M. (2014). Rhythm and Intonation in Malayalam.Ph. D. thesis. New Delhi: Jawaharlal

Nehru University.

Masica, C. & Dave, R. B. (1972).The Sound System of Indian English. Monograph 4. Hyderabad:

CIE&FL.

Meiraba, T. (2014). Prosodic phonology of Meite-lol. Ph. D. thesis, New Delhi: Jawaharlal

Nehru University.

28
Mesthrie R. & Bhatt, R. M. (2008). World Englishes: The study of new linguistic varieties.

Cambridge: University Press.

Mohanan, K. P. (1986). The theory of lexical phonology. Amsterdam: D. Reidel & Co.

Nair, N. G. (1996). Indian English phonology: A case study of Malayalee English. New Delhi:

Prestige Books.

Nespor, M. & Vogel, I. (1986). Prosodic phonology. Dordrecht: Foris.

Ohala, J. J. & Gilbert, J. B. (1979). Listeners' ability to identify languages by their prosody.

Léon, P. & Rossi, M. (Eds.) Problèmes de Prosodie,Vol. II. Ottawa: Didier. 123-131.

Pandey, P. (2011). Tag and „no‟ questions in Indian English. Ms. Centre for Linguistics,

Jawaharlal Nehru University, New Delhi.

Pandey, P. (2014). Sounds and their patterns in Indic languages, Volume 1. New Delhi: CUP

India.

Pandey, P. (1980). Stress in Hindustani English: A Generative Phonological Study.Hyderabad:

CIE&FL.

Pandey, P. 1981 [1994].On a description of the phonology of Indian English.

CIE&FLBulletin17:11-19. [Also in R. K. Agnihotri and A. L. Khanna (eds), Second

Language Acquisition: Socio-cultural and Linguistic Aspects of English in India, New

Delhi: Sage Publications, 198-207.]

Pandey, P. (1985). Word Accentuation in Hindustani English in Relation to Hindustani and

English.Ph. D. dissertation. University of Poona, Pune.

Pandey, P. (1989). Word accentuation in Hindi. Lingua 77:37-73.

Pierrehumbert, J. B. (1980). The phonetics and phonology of English intonation. PhD thesis,

MIT. Bloomington: Indiana University Linguistics Club.

29
Pickering, L. (1999). An Analysis of Prosodic Systems in the Classroom Discourse of Native

Speaker and NonnativeSpeakerTAs.PhD thesis, University of Florida, Gainesville.

Pickering, L. & Wiltshire, C. (2000). Pitch accent in Indian-English teaching discourse.World

Englishes19 (2): 173-183.

Puri, V. (2012). Stress Clash Resolution by Hindi-English Simultaneous Bilinguals. Ms.

University of Illinois at Urbana-Champaign.

Puri, V. (2013). Intonation in Indian English and Hindi late and simultaneous bilinguals.

Doctoral dissertation, University of Illinois at Urbana-Champaign.

Raghavan, R. G. (2012). Second language phonology: Investigating Tamil English. Saarbrücken:

LAP Lambert Publishing Company.

Roach, P. (1981). Phonetics. Oxford: University Press.

Roach, P. (1982). “On the distinction between stress-timed and syllable-timed languages. In

Linguistic Controversies, edited by David Crystal, 73-79. London: Edward Arnold.

Roach, P. (2013 [1993]). English phonetics and phonology: A practical course. Cambridge:

CUP. 4th edition.

Sadanandan, S. (1981). Stress in Malayalee English: A generative phonological approach. M.

Litt. Dissertation. Hyderabad: CIE &FL.

Sailaja, P. (2009). Indian English. Edinburgh: Edinburg University Press.

Schiering, R. Bickel, B. & Kristine, H. A. (2012). Stress-timed = word-based? Testing a

hypothesis in prosodic typology. STUF - Language Typology and Universals 65(2): 157-

168.

Sethi, J. (1980). Word accent in educated Punjabi speakers‟ English. Bulletin of the Central

Institute of English 16. 35–48.

30
Shekhar, G. (1993). A Study of Accent and Intonation in All India Radio Newsreaders Speech.

M. Phil. Dissertation. Hyderabad: CIE&FL.

Shuja, Asif. (1995). Urdu-English Phonetics and Phonology. New Delhi: Bahri Publications.

Skopeteas, S., Fiedler, I., Hellmuth, S., Schwarz, A., Stoel,R., Fanselow, G. ,Féry, C. and Krifka,

M. (2006). Working Papers of the SFB632, Interdisciplinary Studies on Information

Structure (ISIS) 4. Potsdam: Universitäts verlag Potsdam.

Steedman, M. (2000). Information structure and the syntax-phonology interface. Linguistic

Inquiry 31. 649-689.

Trudgill, P. and Hannah, J. (2002). International English: A guide to the varieties of Standard

English. London: Arnold.

Upendran, S. (1980). The intelligibility of English spoken by Tamilians.M.Phil. dissertation,

CIE&FL, Hyderabad.

Verma, D. (2005). Some Aspects of the Prosodic Phonology of Narrow and Contrastive focus in

Hindi and Hindi-English. M. Phil. dissertation, CIE&FL, Hyderabad.

Vijaykrishnan, K. G. (1978). Stress in Tamilian English: A study within the framework of

generative phonology. M. Litt. Dissertation. Hyderabad: CIE&FL

Wells, J. C. (1982). Accents of English. Cambridge: Cambridge University Press.

Wells, J. C. (2000). Longman Pronouncing Dictionary. London: Longman.

Wiltshire, C. R. (2005). The “Indian English” of Tibeto-Burman language speakers. English

World-Wide 26: 275–300.

Wiltshire, C. R. & Harnsberger, J. D. (2006). The influence of Gujarati and Tamil L1s on Indian

English: A preliminary study. World Englishes 25. 91–104.

31
Wiltshire, C. R. & Moon, R. (2003). Phonetic stress in Indian English versus American English.

World Englishes 22(3). 291-303.

Wiltshire, C. R. (To appear). Emergence of the unmarked in Indian Englishes with different

substrates. Filppula, M., Klemola, J. & Sharma, D. (Eds.) The Oxford Handbook of

World Englishes. Oxford University Press.

32