Features Of Connected Speech In English English Language Essay

Blurred boundaries, sounds and even words that are completely swallowed up, chewed and mangled words that force you to listen with all your acquired knowledge and a general sense of hopelessness that you are never going to manage this ‘foreign tongue’ are but the most common frustrations expressed by the ESL learners I have taught. L2 users educated in a system that prides in teaching the most precise and appropriate pronunciation leaves them bewildered when they hear English (even snippets of conversation) as spoken by L1 users. They suffer a ‘devastating diminution of phonetic information at the segmental level when they encounter normal speech’ (Brown, G.1990, p.60)

Connected speech is not a familiar feature to even fluent speakers of English in India and so they tend to fully form the words even in informal situations, giving the impression to L1 speakers and other L2 speakers that, ‘he’s so arrogant about it all’ (Crystal and Davy 1975,p.8)

In this essay I would analyse the features of connected speech in English, the problems they cause to L2 learners in India. In the first section I will examine the features of connected speech and move onto the challenges they cause to L2 speakers especially Indian ESL learners. I would also like to analyse my own and the prevailing attitude and practice towards learning and teaching connected speech. Finally I would like to identify the criteria that I would apply in deciding the different aspects of connected speech suitable to my learning/ teaching environment.

Connected speech

Underhill A (1994) defines connected speech as a “flow of sounds which are modified by a system of simplifications through which phonemes are connected, grouped and modified” This simplification is an unconscious process and most of the L1 speakers are often unaware of this as they practise it. (Collins B, and I. Mees 2008). The goal of connected speech is to reduce the articulatory effort required to send the message. Even though whole chunks of phonetic details are left out by the L1 speaker to maximise the ease of communication, native listener decodes these messages using the different types of support knowledge in his repertoire.

This reduction often takes place within words or words in a ‘stream of speech’ (Dalton C. and B. Seidlhofer, 1994), where word boundaries get blurred. Words flow seamlessly, allowing the L1 speaker to make use of the stress system to emphasise the content of the message. In this flow adjacent sounds are modified to resemble each other- Assimilation-; sounds are completely left out of words – Elision-; and certain sounds maybe inserted between the words to fuse the words together at boundaries- Linking-.

Features of connected speech

Assimilation

L1 speakers speak at a pace of 350 syllables per minute in formal speeches and 400- 450 words in informal speech. (Crystal.D 1992; 1999) When the tongue has to move from one articulating position to another at this pace, only an approximation of the phoneme can be produced. Thus all phonemes occurring within a word or in a stream of speech influence one another and ‘adjust with the phonemes of their neighbours’ (Brown, G.1990). Underhill (1994) summarises assimilation as “the natural result of the various speech organs ‘cutting corners’ as they perform their complex sequence of movements…”

Though any sound can influence any other sound theoretically assimilation is limited to a few phonemes (Dalton C. and B. Seidlhofer, 1994) “Alveolar consonants /t/, /d/ and /n/ at the end of a word often assimilate to the place of articulation of the consonant at the beginning of the next word” (Underhill, 2005, p.60)

Eɡː

Great Britain

/ greɪtbritən /

/ greɪʔpbrɪtən /

Won’t come

/ wəʊntkʌm /

/ wəʊnʔkʌm /

Down by law

/ daʊnbaɪlɔ: /

/ daʊmbaɪlɔ: /

Good girl

/ gʊdgɜ:l /

/ gʊggɜ:l /

Elision

When the speech effort is reduced for maximum efficiency, the articulation of individual phonemes gets weakened. When these phonemes are minimised markedly they are dropped from connected speech. Elision is the process of dropping a sound (a vowel or a consonant), from a word when it is uttered as part of connected speech. RP sounds short and clipped to L2 speakers and speakers of other varieties of English due to elision of the schwa (/ə/) sound. The most commonly elided sounds in English are /t/, /d/ and /ə/. The consonants that are elided almost as frequently as these are /v/ and /ð/

/t/

First three

fɜstθri

fɜsθri

Last year

lɑstjiə

lasjiÉ™

 /d/

ground pressure

ɡraʊndpreʃə

ɡraʊmpreʃə

Banned for life

BændfÉ™laɪf

BæmfÉ™laɪf

/É™/

Chocolate

tʃɔ kə lət

tʃɔk lət

Vegetable

vɛʤə tə bl̩

vɛʤ tə bl̩

/v/

We’ve been considering

Wivbinkənsideriŋ

Wɪbɪnkənsideriŋ

Needs of the

Nidzəvðə

Nidzəðə

/ð/

I think that was

aɪθɪŋkðətwəz

ʌθɪŋkətwəz

Went the way of the

Wentðəweɪəvðə

Wentəweɪːðə

Vowel reduction

Unstressed vowels in the stream of speech are shortened and are often centralized to a schwa (/É™/) sound. An unaccented diphthong in a similar setting can lose the length of the vowel glide or could even be reduced to a monophthong.

Must

mÊŒst

məs

There

ðeə

ðʒː

Strong and weak forms

Function class words in unaccented positions are reduced to their weak forms in connected speech. The degree of reduction depends on the rate of speech delivery; the faster the speech, the greater the reduction of vowel sounds.

Stronɡ

Weak

And

ænd

ən

Of

É’v

əv

To

tʋ

tÉ™

Linking

Dalton C. and B. Seidlhofer (1994) describe linking as the consequence of two vowel sounds meeting at a vowel boundary. In such situations an extra sound is inserted to mark the transition between the two vowels.

Linking / r/

In RP /r/ is not realised in pronunciation except when it is followed by a vowel. In connected speech /r/ is articulated when the following word starts with a vowel.

Sooner

suːnə

sooner or later

suːnərɔːleɪtə

Sure

ʃɔː

sure enough

ʃɔːrɪnʌf

Intrusive /r/

Some speakers insert a /r/ even when there is no /r/ in the spelling. Intrusive /r/ carries a certain social stigma and educated L1 speakers often deny having an intrusive /r/ in their connected speech. The presence of intrusive in connected speech is quite frequent among non-rhotic L1 speakers.

Idea

aɪdɪə

idea of it

aɪdɪərɒvɪt

Intrusive /w/ and /j/

When a word ending in a vowel is followed by another vowel, L1 speakers insert a consonant sound that is nearer to the sound of the first vowel.

É¡o out

ɡəʋwaʋt

I am

aɪjam

She is

ʃiːjis

Catenation

In connected speech, L1 speakers tend to connect the last consonant sound in a word with the first vowel sound of the next word.

Keep out

kiːp aʋt

Key pout

kiː paʋt

A name

ə neɪm

An aim

ən eɪm

Connected speech- Teach Reception and or Production?

Challenges to L2 learners and teachers

An awareness of connected speech features is essential to facilitate listening comprehension in an ESL learner. A lack of these features does not affect intelligibility drastically and the decision of how much (the degree) and when to convert the declarative knowledge to procedural knowledge should be left to the L2 learners. Jennifer Jenkins (2000) argues that “the assimilatory process will be acquired naturally if learners’ progress in their knowledge and control of the language is sufficient to enable them to speed up the rate of their speech”. Thus we understand that a learner’s acquisition of knowledge in one area would positively affect his performance in another area, and that even though most if not all of this is eminently learnable with exposure to an L1 atmosphere, it is not always teachable. The criteria for teaching connected speech would be arrived at by observing the teachability- learnability scale. What the L2 learners need at this point could be a simulation of the maternal speech to children “… mothers alternate between clarified and ‘distorted’ forms and this seems to enable the children to develop rules of correspondence between the model and distorted forms they will normally encounter” (Ratner 1984). Exposing the L2 learners to both ‘careful colloquial speech’ and ‘rapid colloquial speech’ could help them to understand these varieties at a declarative level and thereby position them for a procedural development at their own pace.

Indian learners of English (ILE) are unfamiliar with the connected speech feature assimilation. In my anecdotal experience I have noticed that the educated speakers in India (for that matter even the uneducated speakers) utter each word in a sentence almost as though they are articulated in isolation. This syllable timed articulation is carried forward even when they speak a stress timed language like English. This trait becomes problematic at the reception level as ILE listeners expect the same vocal and phonemic clarity even when they are listening to English. Gillian Brown’s (1990) observation that an L2 listener suffers a ‘devastating diminution of phonetic information’, in such situations aptly summarises the confusion and panic they feel. So I believe that L2 learners of English should be made familiar with the assimilation features, so that they can understand the content of the discussion/ interaction without having to decode language features each and every time. In my teaching practices I think I have been unconsciously following the suggestions that Gillian Brown (1990) proposes: “I have already suggested that I do not approve of teaching students to produce ‘assimilated forms’ and elided forms. Sophisticated students who have been taught to be aware of these forms will introduce them into their own speech in a natural context when they feel able to control them”

Assimilation becomes a problem at the production level only when the L2 speakers interact with the L1 listeners, but then L1 listeners are mostly aware of this difficulty of the L2 speakers and make necessary adjustments to their linguistic and emotional responses. As an L2 user and as a teacher I believe that a few words uttered painfully slow to achieve connected speech would create more confusion to the L1 speaker than a total absence of assimilatory/ connected speech features. “… slower speech made to incorporate features of connected speech is reduced to gibberish” (Crystal D. 1992;1999). One of the major arguments that exist against this stand is that an absence of assimilatory features would prevent the L2 speakers from using intonation patterns and rhythm and would lead to loss of fluency. (Underhill A. 1994) and he suggests exposing ESL learners to rapid colloquial speech as a useful activity for improving listening comprehension.

Even though elision is a feature that is present in most Indian languages, the Indian ESL learner may struggle to understand and participate in a conversation with an L1 speaker employing rapid colloquial speech. This is because elision is not a feature that is taught in Indian classrooms. Moreover, elision carries a certain social stigma among educated Indians and so they might actively resist adopting elision in their speech practices as well. Dropping sounds is equated with a lower social class and could also be interpreted as a sign of poor education.

As with assimilation, elision can cause confusion at the reception level for the L2 listener, if he has always been exposed to ‘correct’ and fully articulated speech. An L2 speaker trying to include elision in a slow speech would confuse the L1 listener. “As demonstrated by Crystal in his experiment the sentence’ I wouldn’t have been able to’ in fluent speech may become /É™ wÊŠbmpɪneɪblÉ™/. Spoken more slowly by a learner who is being encouraged to use the various features of connected speech, it could come across like this: /É™ wÊŠbÉ™m pɪ neɪ blÉ™/,with each ‘syllable’ being carefully articulated. In this case it would at best sound ridiculous and at worst would be rendered completely incomprehensible.” (Jennifer Jenkins, 2000)

Another contentious area is the teaching of weak forms. “As pointed out by David Brazil (1994) the contradiction of focusing in the classroom on a feature whose quality is precisely the result of speakers not focusing on it. This pedagogic focus may then, paradoxically, impede the later acquisition of weak forms in learning outside the classroom…” (Jennifer Jenkins, 2000). Jennifer Jenkins challenges the notion of the need to weaken an unimportant sound to highlight an important one and points out that if the important sounds are stressed then the meaning would become clear to the listener. Reading about this approach and thinking further on it has convinced me that this is the way to go forward. The hours of drilling that I have been forcing on my students might have been counterproductive. I feel that this is another area where declarative knowledge should be allowed to mature to a procedural knowledge at the students’ own pace.

Listening comprehension and thereby communication is enhanced by dissimilatory practices rather than assimilatory processes, as they give primary importance to the hearers needs by subordinating the speech strategies of a speaker.

I would still like to believe that the assimilatory features are what give the English language its identity. But for learners at a transactional level with L1 speakers or with other L2 speakers conquering these features could be daunting and unnecessary for everyday communication. As David Crystal (1999) suggests The possibility of L2 learners becoming competent in both syllable and stress based speech and being able to shift effortlessly from one to other, depending on the need of the circumstance, either to proclaim a national identity or to improve international intelligibility, is not an improbability and that is what I would like to strive for and encourage my learners to aim for.