ArticlePDF Available

A Sociolinguistically Based, Empirically Researched Pronunciation Syllabus for English as an International Language

Abstract and Figures

The starting point of this paper is the recent shift in the use of English, such that non-native speakers (NNSs) using English for international communication now outnumber its native speakers (Crystal 1997; Graddol 1997). This shift, it will be argued, has serious implications for ELT pedagogy. Principal among these is the need for empirically established phonological norms and classroom pronunciation models for English as an International Language (EIL), in which intelligibility for NNS rather than for native speaker (NS) receivers is the primary motivation. Three sets of data drawn from NNS-NNS interaction are provided in order to exemplify the kinds of empirical evidence that are necessary to enable us to make informed claims about phonological intelligibility in EIL, Then follows the author's proposal, based on such evidence, for a revised pronunciation syllabus for EIL, the Lingua Franca Core. This core approach, it is suggested, is better able to promote both intelligibility and regional appropriateness among EIL interlocutors, as well as being more teachable, than either of the two most commonly adopted classroom models, Received Pronunciation and General American. Finally, the importance of developing learners' accommodation skills as an integral part of pronunciation teaching for EIL is discussed.
Content may be subject to copyright.
A Sociolinguistically Based, Empirically
Researched Pronunciation Syllabus for
English as an International Language
King's College, London
The starting point of this paper is the recent shift in the use of English, such that
non-native speakers (NNSs) using English for international communication
now outnumber its native speakers (Crystal 1997; Graddol 1997). This shift, it
will be argued, has serious implications for ELT pedagogy. Principal among
these is the need for empirically established phonological norms and classroom
pronunciation models for English as an International Language (EIL), in which
intelligibility for NNS rather than for native speaker (NS) receivers is the
primary motivation. Three sets of data drawn from NNS±NNS interaction are
provided in order to exemplify the kinds of empirical evidence that are
necessary to enable us to make informed claims about phonological intellig-
ibility in EIL. Then follows the author's proposal, based on such evidence, for a
revised pronunciation syllabus for EIL, the Lingua Franca Core. This core
approach, it is suggested, is better able to promote both intelligibility and
regional appropriateness among EIL interlocutors, as well as being more
teachable, than either of the two most commonly adopted classroom models,
Received Pronunciation and General American. Finally, the importance of
developing learners' accommodation skills as an integral part of pronunciation
teaching for EIL is discussed.
At the start of the twenty-®rst century, most applied linguists are familiar with
the fact that English is now spoken by a considerably greater number of NNSs
than NSs. Several scholars have already begun discussing the implications for
English Language Teaching (see, e.g., Kachru 1992; Kachru and Nelson 1996;
Kasper 1998; Pennycook 1999, 2000; Seidlhofer 1999; Widdowson 1994,
1997). And with the conceptual leap implicit in Cook's recommendation that
`language teaching would bene®t by paying attention to the L2 user rather
than concentrating primarily on the native speaker' and should `apply an L2
user model' (1999: 185), second language acquisition researchers look set to
enter the debate. Nevertheless, while this paradigm shift is ®nally gaining
acceptance in theory, in practice it has so far had little impact on applied
linguistic research design and even less on English language teaching or
Applied Linguistics 23/1: 83±103 # Oxford University Press 2002
teaching materials: the NS remains a given, and the NS standard measure still
reigns supreme.
Within the ®eld of phonology, decisions as to what to include in
pronunciation syllabuses are, in the main, still grounded in NS intuitions,
despite the fact that evidence is emerging to demonstrate that phonological
intuitions may be inaccurate (see, e.g., Cauldwell 1996 on so-called `stress-
timing', and Levis 1999 on the pitch movement on the tonic syllable in yes/no
questions). And even where intuitions are correct, they are based on
intelligibility for NS receivers without any suggestion that intelligibility for
NNS receivers might make dierent demands. Meanwhile, corpora such as
Collins COBUILD and the British National Corpus have begun to provide
rather more reliable information than NS intuitions about English speech.
Once again, though, the focus is on what NSs do when they communicate
with other NSs, with the tacit assumptionÐwherever corpus ®ndings are
incorporated into ELT materialsÐthat communication between NNSs should
operate smoothly along the same lines. Learners are therefore encouraged to
adopt NS English assimilatory features of pronunciation such as elisions,
contractions, assimilation and weak forms, regardless of their often negative
eect on intelligibility for their NNS interlocutors, simply because these
features are facts of NS pronunciation. As Widdowson points out, however,
the descriptions provided by corpora `are ``factors'' to be considered . . . but
not facts to be uncritically incorporated into prescriptions' (1991: 20). In other
words, it is not sucient for pedagogical proposals (or `prescriptions') to be
based entirely on descriptions of (or, worse still, intuitions about) NS speech.
If we are to provide appropriate pedagogic proposals for EIL pronunciation,
then these must be linked directly to relevant descriptions of NNS speech, that
is, to the ®ndings of empirical research conducted in NNS±NNS speech
contexts in terms of (1) what constitutes optimum productive competence
and (2) what learners need to be able to comprehend. At present, though, I
am aware of only one spoken NNS±NNS corpus, Seidlhofer's Vienna-Oxford
ELF (English as a Lingua Franca) Corpus, which is being compiled at the time of
writing, and which will initially number approximately 500,000 words (see
Seidlhofer 2000, forthcoming).
In the second part of this paper I will present examples of my empirical
research into EIL phonology to demonstrate the sort of information that is
required to enable us to make informed decisions about EIL pedagogy and on
which I have based my proposals for an EIL pronunciation syllabus. Before
moving on to the empirical discussion, however, it will be helpful to clarify
my position as regards the appropriateness of regional accents in EIL and, in
this respect, to distinguish between EIL on the one hand and English as a
Foreign Language (EFL) along with other modern foreign languages on the
Received Pronunciation (RP), the prestige British accent, is thought to be
spoken now in its pure form by fewer than 3 per cent of British Speakers of
English (Crystal 1995: 365), while the majority of British people have either a
regionally modi®ed RP or a regional accent. The latter, unless overly broad
(and with the exception of the still-stigmatized accents of Liverpool,
Birmingham, and Glasgow), are fast gaining acceptance among the general
public. There is, then, a growing awareness in Britain of a fact that
sociolinguists have always recognized: that regional variation is the (accept-
able) rule rather than the (unacceptable) exception. Turning to the case of
EFL and other modern foreign languages, the purpose of learning is, by
de®nition, to speak the target language as a `foreigner' in order to facilitate
communication with NSs of the language. Here, it seems reasonable to argue
that the goal of pronunciation teaching should be the suciently close
approximation of an NS accent such that it can be understood by NSs of that
language. In this case, the choice of pedagogic model is (or should be) a matter
of selecting the NS accent which will have widest currency among the
learner's target (NS) community.
EIL is a dierent matter. Here, English is being learnt for international
communication rather than for communication with its NSs. Speakers of EIL
are not `foreign' speakers of the language, but `international' speakers. The
EIL target community is no longer an NS British (or any other NS) one: it is an
international community in which all participants have an equal claim to
membership. An intrinsic part of this claim, it seems to me, is the right for
speakers to express their (L1) regional group identity in English by means of
their accent, as long as the accent does not jeopardize international
intelligibility. This position ®nds support in Bourdieu's (1977) concept of
`legitimate discourse'. According to Bourdieu, in order to be considered an
example of `legitimate discourse', an utterance must meet four conditions, of
which the fourth is that it must employ `legitimate' phonology and syntax.
With respect to EFL, `legitimate' phonology entails speaking with an accent
that is intelligible and acceptable to the target NS English community.
However, when we shift our attention to EIL, `legitimate' in Bourdieu's terms,
implies that phonology must be intelligible and acceptable to the target
international, and therefore predominantly NNS, English-speaking commun-
ity. This will involve the making of adjustments by NSs as well as NNSs of
English, towards an agreed international (rather than NS) norm.
It is a current irony that L2 learners are nowadays the only English speakers
who are still encouraged to approximate an RP or General American (GA)
accent as closely as possible, while regionally accented (NS, but not NNS)
teachers are not discouraged from teaching in the accent of their birth if they
so desire. In this regard, ELT has not moved on from the `de®cit linguistics'
ideology referred to a decade ago by Quirk, a view of language in which any
item that deviates from idealized or attested NS English usage (in this case, RP
and GA) is an `error', and one which rejects arguments in favour of L2
regional English norms (see Quirk 1990 and Kachru's 1991 response). While
this view can to some extent be justi®ed for EFL and other modern foreign
languages, it is unjusti®able for EIL. Insofar as EIL is nowadays the rule and
EFL the exception in terms of the contexts of NNS use of English it is, then,
somewhat surprising that correctness continues to be judged in relation to NS
usage in ELT classrooms around the world.
However, this is not to make a claim that as far as NNS English accents are
concerned, anything is acceptable. The possibility of mutual unintelligibility is
currently a cause of much concern. In Trudgill's words, there is `a great fear . . .
that English is now used so widely around the world, and is in particular used
by so many non-native speakers, that if we are not careful, and very vigilant,
the language will quite rapidly break up into a series of increasingly mutually
unintelligible dialects, and eventually into dierent languages' (Trudgill 1998:
29). Trudgill himself considers this to be `a perfectly sensible point of view' for
a language that has more non-native than native speakers, and goes on to
predict that while English lexis is likely to undergo a process of `homogenisa-
tion' by means of `Americanisation', English phonology will take the opposite
route and undergo a process of disintegration.
In line with Trudgill's point, despite decades of teaching of RP and GA, the
NNS Englishes are thought to diverge from each other more in terms of
pronunciation than of the other linguistic levels (see, for example, Ioup 1984).
The links between accent and identity on the one hand and accent and
articulatory motor skills on the other are, it seems, so ingrained that
traditional English pronunciation teaching is destined to fail for all but a
small minority of L2 learners. Hence pronunciation already has a greater
potential to compromise mutual international unintelligibility than do the
other linguistic levels. And the worst-case scenario is, as Trudgill implies, that
the whole purpose of learning EILÐto engage in successful international
communicationÐwill be threatened by further phonological divergence. In
order to prevent the disintegration of international phonological intelligibility
there is, it follows, a strong case for pedagogic intervention of a new kind:
intervention that is no longer based on idealized NS models or NS corpora, but
that is both more relevant (in terms of EIL needs) and more realistic (in terms
of teachability). Given its primary concern to promote international
phonological intelligibility, as well as the broader purpose of developing a
research-based pedagogy, a new pedagogy for EIL must be based on evidence
drawn from EIL (NNS±NNS) interactions and, above all, from NNS listeners. It
is to such evidence that we now turn.
The purpose of this section is threefold: ®rst, to demonstrate the extent to
which intelligibility in NNS±NNS interaction can break down as a result of
problems at the phonological level; secondly, to identify which speci®c
phonological features are implicated in the breakdown; and thirdly, to
consider two other factors which contribute to (un)successful EIL commun-
ication: the processing of contextual cues and the use of accommodation
The section is structured around the analysis of three dierent types of data
collected in EIL contexts.
The ®rst set of data (®ve communication
breakdowns drawn from the author's ®eld data collected in a range of
classroom and social situations) establishes the general fact of communication
breakdown and provides evidence of speci®c phonological sources of mis- and
non-communication. The second set (two recorded information exchange
tasks) demonstrates that such phonologically derived breakdowns are likely to
be more dicult to resolve in EIL (NNS±NNS) than in NS±NS, or even NS±
NNS, interactions because EIL interlocutors below the level of bilingual
pro®ciency appear unable to process contextual cues to compensate for their
interlocutors' pronunciation errors. The third set (three further recorded
information exchanges and one recorded social exchange) demonstrates NNS
interlocutors' attempts to accommodate (converge) at the phonological level
(in other words, to adjust their pronunciation in order to bring it closer to one
another's) in speech situations where intelligibility for their interlocutor is
critical. The focus here is on which phonological features are selected for
convergence and how the speaker resolves the problem. The purpose is to
assess whether the features selected are those which have been identi®ed as
being likely to threaten EIL phonological intelligibility, and how satisfactorily
the problem is resolved.
In all three sets of data, the subjects are of upper-intermediate to low
advanced level as recognized by the University of Cambridge Local
Examinations Syndicate (UCLES) in that students hold the First Certi®cate
of English (FCE) or the Certi®cate of Advanced English (CAE) quali®cation. In
other words, these subjects have achieved a level of reasonable competence,
but would not be considered fully bilingual in English. Because this is
tantamount to saying that such speakers still have interlanguages, some
authors refer to their spoken interactions with each other as `interlanguage
talk' (see, e.g., Du 1986, Long and Porter 1985). My own use of the term
interlanguage talk (ILT) relates speci®cally to this type of EIL interaction.
Examples of pronunciation-based miscommunication in
interlanguage talk
We begin with ®ve examples drawn from the author's ®eld data. The majority
of this data was collected over a period of three years in classroom and social
settings, with the aim of establishing the extent to which miscommunication
in ILT is caused primarily by problems at the phonological level. Over this
period, I noted down every example of mis- and non-communication that
occurred in my presence and, wherever feasible, discussed the cause(s) with
the interlocutors involved. What emerged was a clear indication that although
pronunciation was by no means the sole cause of ILT communication
breakdown, it was by far the most frequent and the most dicult to resolve.
The following ®ve examples are typical of the errors which caused
intelligibility problems for ILT interlocutors in the data, and exemplify the
categories of phonological error which most often proved problematic:
consonant sounds, tonic (or nuclear) stress, vowel length, and non-
permissible (according to the rules of English syllable structure) simpli®cation
of consonant clusters.
[1] [aI `peIld] = `I failed'
In this example, an L1 Korean speaker of English had taken his driving test
that morning. He entered the classroom after the lunch break, and announced
that he had failed the test. His classmates (from a range of L1s), did not
understand the signi®cance of what he had said, and one followed the
announcement with the enquiry `Did you pass your test, Lee?' In this
example, the typical Korean sound substitution of /p/ for /f/ had rendered the
speaker's pronunciation unintelligible for his non-Korean receivers. There are
several other examples in the ®eld data of this speaker's /p/-/f/ sound
substitution causing problems for receivers. For example, `wife' pronounced
`wipe'; `®nish' pronounced `pinish'; `coee' pronounced `copy'; `father'
pronounced `pader'.
[2] I smoke more than you DO = `I smoke more than you do.'
In this example, the speaker, an L1 Taiwanese speaker of English, and his
Swiss-Italian interlocutor were discussing how many cigarettes a day they
each smoked. The Swiss-Italian subject had just told the Taiwanese subject
that she smoked around 20 a day. The latter replied that he smoked more
than she did, but instead of putting tonic (nuclear) stress on the word `you',
where it would have indicated intonationally the contrast he was making
lexically, he put it on the last item in the tone unit, that is `you'. After three
repetitions she still did not understand.
[3] Shakespeare's [bA:spleIs] = `Shakespeare's birthplace'
In this case, a Japanese speaker was giving a short presentation to a mixed-L1
class. He announced the title of his talk as what sounded most like
`Shakespeare's bathplace' (although the /T/ was pronounced /s/). During
the presentation he described Stratford-upon-Avon, talked about Shake-
speare's early years, his marriage to Anne Hathaway, his acting career, and his
plays. When the student had ®nished speaking and oered his audience the
opportunity to ask questions, the ®rst was an enquiry about the connection
between the content of the talk and a bath. Indeed, the majority of the group
admitted to having thought the talk had concerned a bath, although they had
not been able to identify a connection. The problem here was the substitution
of /Æ:/ with /A:/, a substitution which frequently caused intelligibility
problems in my data, and one which had previously been noted by others
such as Schwartz (1980). On the other hand, the replacement of /T/ with /s/
was not at all problematicÐand this was a phenomenon which recurred
regularly in the data.
[4] Have you got a blue VUN? = `Have you got a blue one?'
Here we have an example of a combination of phonological errors which
caused the most serious problems in my data: misplaced tonic (nuclear) stress
along with a consonant substitution within the wrongly stressed word. An L1
Hungarian student of English was talking with three other students,
respectively from Guatemala, (French-)Switzerland, and Brazil. They were
using coloured pens to make posters for the classroom wall. At one point the
Hungarian asked his fellow students if any of them had a blue pen. However,
he not only placed tonic stress on the ®nal item in the tone unit, but he also
substituted the /w/ of `one' with a /v/. The other students asked several times
`What is vun?' and only understood his meaning when the speaker located
the pen he needed, and held it up, saying `Blue vun like THIS'. It could be
argued that the problem in this example was also lexical: the listeners may
have assumed that `vun' was a word unfamiliar to them. However, the same
problem arises in my data when tonic stress is misplaced on words that are
both familiar to listeners and contain no segmental errors. It seems, then, that
any lexical diculties of the kind encountered in example 4 are compounded
by misplaced tonic stress.
[5] [dO
laIz@ fIz Áf 'skO] = `Don't rise the fees of school.'
In this ®nal example, another L1 Japanese speaker of English was giving a
short presentation to her mixed-L1 class. She had been discussing education
in the EC and, in particular, its high cost for international students such as
herself and concluded her talk with the words `don't rise the fees of school'.
However, only the other Japanese students in the group understood her
meaning, while the others remained completely baed even after she had
repeated the sentence four times. Eventually we `translated' it word for word
and the meaning was immediately clear despite the two lexicogrammatical
errors (`rise' for `raise'; `fees of school' for `school fees'). In this example, the
several pronunication errors all involve vowel length and consonant
substitution. Both these error types are frequent causes of unintelligible
pronunciation in my empirical data.
The problem of contextual cues
At this point, the reader may wish to argue that where contextual or cotextual
information is available, it is likely to clarify meaning where pronunciation
has failed to do so. However, in interlanguage talk this appears not to be the
case. In NS±NS interaction and in interaction where one or more participants
are ¯uent bilinguals, receivers regularly (though not always) make use of
contextual and cotextual information as aids to the clari®cation of meaning.
On the other hand, when the receiver and speaker are both NNSs, the
receiver tends to focus on the acoustic signal and direct his or her eort to
decoding what has been heard. Where this does not tally with visual and
other extralinguistic cues, or with the cotext then, time and again in my ILT
data, the receiver adjusts the context and/or cotext to bring them into line
with the acoustic information rather than vice versa.
The following two extracts demonstrate this process. In each case, the
listener has clear contextual information in the form of visual cues, and yet
opts to try to make sense of the inaccurate pronunciation that he hears. In the
®rst extract, a Japanese speaker (B) has just described one of six pictures to
her Swiss-German receiver (A), who had the same six pictures, but dierently
ordered. B had referred in her description to `three led cars'. With some
diculty, A had correctly identi®ed the picture being described. This recorded
conversation takes place immediately after the completion of the task:
A I didn't understand the let cars. What do you mean with this?
B Let [let] cars? Three red [Qed] cars (very slowly).
A Ah, red.
B Red.
A Now I understand. I understood car to hire, to let. Ah, red, yeah I
see. (Jenkins 2000: 81)
In other words, despite the fact that there was only one picture with cars, that
there were three cars, and that there was no picture showing cars for hire, B
chose to trust A's (mis)pronunciation rather than the visual information in
front of him, and to search for a non-existent picture containing hire cars.
The second extract comes from an interaction which took place four weeks
later. By now these two students are far more familiar with one another's
accents as they have spent 25 hours a week in the same classroom and
engaged extensively in pairwork together. Again, the Japanese subject, B, is
describing one of six pictures to her Swiss-German interlocutor, A, who has
the same six pictures in a dierent order. Again, his task is to identify the
correct picture that B is describing. At one point, the following exchange takes
B And second picture, the bottom of the bottom of the picture there's
mm gley [gleI] house,
A (frowns)
B (registers A's frown) grey [gòeI] and small house, it's very s-old?
A Yeah, there's a grey house, yeah.
B Mm, okay. (Jenkins 2000: 82)
This time, B quickly notices A's diculty, realises that she has substituted /r/
with /l/ and corrects her error using the approximant [ò] unlike her earlier
attempt at correction, which had initially produced the ¯ap [Q]. A's confusion
is solved and he is able to identify the right picture. From B's hesitation `mm'
just before she produces the word `grey', it appears that she cannot remember
the word and that her attention is focused on accessing the lexical item rather
than on her pronunciation. Even at relatively high levels of competence,
then, (these two speakers were about to sit the CAE examination) it seems
that processing overload can lead to serious pronunciation error.
In the follow-up discussion to this exchange, the Swiss-German receiver
commented that until his interlocutor had corrected her pronunciation, he
had been looking for a picture containing a `clay' house. This was despite the
fact that only one picture contained a house and that the house was grey, but
clearly not made of clay. He also pointed out that he had grown accustomed
over recent weeks to his interlocutor's confusion of /r/ and /l/, and that he
had been listening for the error. Nevertheless, he had not noticed it when it
occurred, nor had he trusted the visual information in front of him. Instead,
he had attempted to identify a picture which complied with the faulty
pronunciation he had heard. And this phenomenon was repeated regularly in
my data.
Given that NNSs of less than bilingual competence seem unable to make
successful use of contextual information, coupled with the fact that in
situations of processing overload NNSs are liable to make pronunciation
errors, it becomes crucial to identify which errors seriously threaten
phonological intelligibility in ILT. We can then focus on these items in the
classroom so that if speakers should make an error during an interaction, they
are better placed to recognize what has happened and to make the necessary
The accommodation data: same-L1 and different-L1 interactions
To sum up so far, the ILT data indicate that certain pronunciation deviations,
particularly in consonant sounds, vowel length and the placing of tonic stress,
render an NNS's pronunciation unintelligible to an NNS interlocutor; and that
when this happens, context and cotext do not provide much help in clarifying
meaning. As far as sounds are concerned, this position is corroborated by a
further set of data which investigate the extent of NNSs' attempts to
accommodate towards the pronunciation of their interlocutors. Here, the
sounds which speakers attempt to replace with something more targetlike
(certain sounds that are liable to be aected by L1 transfer), tend to be the
same sounds as those which appear in the ®eld data to be crucial for
intelligibility in ILT. In other words, it seems that, given the opportunity,
NNSs engaged in ILT work out for themselves which features of their
pronunciation are potentially unintelligible for their NNS interlocutors, and
endeavour to replace them.
Traditionally the accommodation literature discusses the accommodative
adjustments made by interlocutors in relation to an aective motivation (see,
e.g., Giles 1973). That is, speakers make their speech more similar to that of
their interlocutors because of their desire to be liked. But the early
accommodation theorists also identi®ed a motivation which they called
`communicative eciency' and which they found to be particularly relevant
to communication in a second language: the desire to be understood (Beebe
and Giles 1984). What seems to happen in ILT, however, is that instead of
converging on each other's pronunciation, when intelligibility is particularly
important, speakers converge on what they construe as a more targetlike
The following extracts demonstrate this phenomenon. The ®rst two extracts
compare the amount of L1 transfer when speakers are engaged in an
exchange ®rst with a partner from a dierent L1 and, secondly, with a partner
from the same L1. In each case, the task is an information exchange. In the
®rst, the Swiss-German speaker is describing a picture which his Japanese
interlocutor has to identify from a choice of six. In the second, the Swiss-
German speaker is describing a geometric pattern which his Swiss-German
interlocutor is endeavouring to draw. In both transcripts, only those words
containing non-targetlike sounds are transcribed phonetically.
Extract 1: SG1 (Swiss-German) in dierent-L1 pair
(J) indicates that the subject's Japanese interlocutor spoke brie¯y at this point.
1. Okay, it's a four storey house with two large balconies and one small balcon,
wIv bñlkÁnis ñnt
this the
2. small balcon is on topÐis the highest one (J) Balcon (J) I think it's the right word.
3. in front of the house are is a is a yes it's a road, and on this road is a a lorry.
@ ò@UtdIs
And and
4. in front of the house too there are is a parking a small parking space with let's say
@ wIv
5. one, two, three, four, ®ve, six, seven, eight parked cars and most of the cars are
6. covered with snow. And on the left side of the house there are there are four or ®ve
k@Uw@dd@ de
7. parked cars. Four are co-®ve are covered with snow and one is is, a red a red car is
8. not covered with snow. In the back of the hou-of the house you can see, on the right
k@Uv@dwIv d
9. side of the back of the house is you can see a mountain with er covered with with
10. trees and snow of course. And there are a f-few houses behind this main house I
ñnt bIhaInt
11. described to you
Extract 2: SG1 in same-L1 pair
(SG2) indicates that the Swiss-German interlocutor spoke brie¯y at this point
1. All I can see is one square, it's (unintelligible) ®rst with with two
2. I guess, this is the word, and now in every every corner of your square is er, is
wÆ:òt Áf skve
3. another er, the square is yeah, a small square in every corners of your big square is a
4. small one, and the length is about two, two-and-a-half, no three centimetres
d@ leÎgs sentImits
. . . (SG2).
5. Yeah. So you have four small squares in the big square. Then you have the er a
hñf skves d
en d
6. with the same size in the middle where the two diagonals diagonals crosses each
wIvd@ saIsd@ d@ daIñnñls
7. other, you have another square. (SG2) Same size as theother (unintelligible) (SG2)
hñf Ád
8. Yes, you have then (SG2) parallel to the the length of the big square . . .
hñf d@ leÎgs d
Okay, then you
9. have, if you have drawn this er small one in the middle er the four corners of this
hñf dò@UndIszmO:ld@ kO:ònsdIs
10. small square er hit the diagonals. (SG2) Then from there you draw a line to
skve daIñnñls den de dò@U
the middle
11. of the white, the length of the big square, so it gives you er (SG2) Four (SG2) Yeah,
Áfd@ leÎgsd@ skve gIfs
12. like arrows . . . They all have the same size . . . should have the same size.
ñò@Us hñf saIs
(Jenkins 2000: 59±60)
In the above two exchanges, the amount of subject SG1's transfer diers both
qualitatively and quantitatively. As regards the quantitative analysis, a chi
square test was carried out on the types of items which were aected (or not)
most frequently by transfer (for example, word-®nal consonant devoicing in
the case of subject SG1). The two conditions (same-L1 and dierent-L1 pairs)
were compared for the number of occurrences of transfer as a proportion of the
total number of candidates for transfer. Signi®cance was found at the level p <
0.01. In other words, the replacement of transfer with more targetlike sounds
in the dierent-L1 pairs was not due to chance. Qualitatively, we can make a
number of observations about the presence and absence of L1 transfer in these
exchanges. For example, in Extract 1, the dierent-L1 dyad, transfer tends not
to involve words which may be considered key to meaning (in the sense of
being important content as opposed to structural words), and mainly occurs on
the de®nite article, on the preposition `with', and on `and'. This is not at all the
case in Extract 2, the same-L1 pair, where there are many occurrences of L1
transfer on key items such as `diagonals', `square', and `arrows'.
In addition, many of the instances of transfer in Extract 1 involve only the
dental fricatives /T/ and //, which my ®eld data have consistently
demonstrated not to aect intelligibility for NNS listeners. Where other
transferred sounds are concerned, especially where these occur on key lexical
items, the speaker appears to make a considerable eort to replace the
transfers with a more targetlike sound. This is particularly noticeable in his
endeavour to pronounce the word `covered'. He has four attempts at this
word, each time approximating target production more closely (see Extract 1,
lines 6, 7, 8, and 9). However, he is aware of his problem and of his
interlocutor's diculty in interpreting this word. Because of this, he is careful
to repeat the word `snow' each timeÐeven adding it in line 10 (`and snow of
course') after he has attempted to convey its meaning by referring to the trees
on the mountainside as well.
In contrast, the two Swiss-German interlocutors were appalled at the extent
of their L1 phonological transfer when the recording of their exchange was
played back to them. Nevertheless, they pointed out that they had found one
another far easier to understand than their respective Japanese interlocutors.
In responses to a questionnaire, it also emerged that in the dierent-L1 pairs,
they had consciously been attempting to make adjustments in the direction of
more targetlike pronunciation (in other words, to converge phonologically)
for the sake of intelligibility for their interlocutors (the `communicative
eciency' motivation), by replacing L1 transfer with closer approximations to
target sounds where they judged this to be necessary. In these exchanges, the
adjustments occurred chie¯y on consonant sounds, so corroborating the
evidence of my ®eld data where consonant sounds proved to be the greatest
barrier to phonological intelligibility in ILT.
The accommodation data: evidence from two task types
A similar eect was found within the dierent-language pairs themselves, but
this time the comparison was across two task types: social interaction and
information exchange. Again, the dierence was identi®ed both quant-
itatively and qualitatively. For the Taiwanese subject of Extracts 3 and 4
below, for example, a chi square test compared actual occurrence of items
such as consonant deletion as a proportion of total candidates for such transfer
across the two task types, and yielded p < 0.001. In other words, when an
interlocutor's understanding was particularly important to a speaker, as in an
information exchange task with a measurable outcome, then the speaker
appeared to replace as much phonological L1 transfer as possible for those
pronunciations which she considered to have the greatest potential to
threaten this understanding.
In the following two extracts, even without the support of statistical
information, we would be able to identify a marked dierence in the amount
of L1 transfer leading to consonant deletion. It is also noticeable that almost
half of the transfers in the information exchange task involve substitutions of
/T/ and //, and the use of /U/for[à] (dark `l')Ðneither of which had caused
intelligibility problems in my ®eld data (and both of which occur regularly in
the pronunciation of many NS English speakers).
Extract 3: Taiwanese subject, social interaction task:
(K) indicates that the subject's Korean interlocutor spoke brie¯y at this point
1. Middle country, and here I I think I've been, I born there for a very long time,
mIdU bO
lO? taI?
2. I've never moved to the big city or the other place. Yes, but I've ®n-I've
3. just ®nished the senior high school and come to Britain .....London, when I
fInIS sInIO skU kö bòI
4. ®rst come here I don't-I didn't like London because ®rst I don't like the food,
kö dO
dId@ lOdO dO
5. yeah, it's quite terrible in here I think, you know in Taiwan (unintelligible).
kwAI? sk
6. (K) Then also I don't like the weather. (K) But now I'm used to. (K)What do,
ñz@U wez@ n@U jUz
what do
7. you think? (K) But I think in your country there are lot(s) of sunshine. In
skdeò lO?
your country
8. it's warm. (K) It's a dierent way.
Extract 4: Taiwanese subject, information exchange task
1. In my picture I think they're in a garden. The the house, be-er behind the
house, they
2. have the small garden. And there are one two three four ®ve six, six people in the
gA:d@ sIs
3. garden. And I think they er have er one man and with his wife and his
gA:d@ wIz
mother I think,
möd@ sk
4. and they've got er three children, two boy, one baby. And they are smiling, it
5. quite happy and . . . er, they're in the garden and (unintelligible) I don't know
what else
6. I can say, but the woman, ah she hold a baby, and . . . and, ah, the er old
Wm@nhOUd OUd
woman she sit
7. in the chair in the left my picture, left-hand, and the man sit on the right side.
òaI? saI?
And the
8. other people they are standing. (K) (Jenkins 2000: 64)
For EIL, the most interesting implications of this accommodation data are
that, given the opportunity to engage in ILT, learners become aware of the
features of their own pronunciation systems which are liable to be
unintelligible for interlocutors from other L1s, and endeavour to adjust
these features. Clearly, they will be unable to do so unless the adjustments
concerned are within their phonological and phonetic repertoires. Even then,
they may often be able to make these modi®cations only with considerable
diculty, and be unlikely to do so reliably unless interlocutor intelligibility is
crucial or interlocutor non-comprehension has been signalled.
Learners therefore need speci®c training to enable them to add to their
phonological repertoires those features which are most important for
intelligible pronunciation in EIL contexts. In addition, they need pedagogic
help in order to develop their accommodation skills, so that they become
more aware of the importance of making adjustments for speci®c interlocutors
and more able to identify the occasions when this is necessary. In the
following section we will consider what these two conclusions mean for
classroom pronunciation teaching.
The Lingua Franca Core
The empirical evidence above provided the basis for a phonological syllabus
for EIL learners: the Lingua Franca Core (LFC). This consists of those
phonological and phonetic features which, from an analysis of all the
miscommunication and accommodation data (of which the above are a
small selection), seem to be crucial as safeguards of mutual intelligibility in
ILT. Concentrating on these items is likely to be more eective than attending
to every detail in which an NNS's pronunciation diers from that of the
(standard) pronunciation of an NS. Additionally, it is also more relevant, since
the syllabus no longer attempts to address the comprehension needs of an NS
listener when, as we have already noted, in EIL the listener is more likely to
be an NNS.
The following is a summary of the main core items:
1. The consonant inventory with the following provisos:
some substitutions of /T/ and // are acceptable (because they are
intelligible in EIL);
rhotic `r' rather than non-rhotic varieties of `r';
British English /t/ between vowels in words such as `latter', `water' rather
than American English ¯apped [r];
allophonic variation within phonemes permissible as long as the pronunci-
ation does not overlap onto another phoneme, for example Spanish
pronunciation of /v/ as [b] leads in word-initial positions to its being
heard as /b/ (so `vowels' is heard as `bowels' etc.).
2. Additional phonetic requirements
aspiration following word-initial voiceless stops /p/ /t/ and /k/ e.g. in [p
(`pin') as compared with /spIn/ (`spin'), otherwise these stops sound like
their voiced counterparts /b/ /d/ and /g/;
shortening of vowel sounds before fortis (voiceless) consonants and
maintenance of length before lenis (voiced) consonants, for example the
shorter /ñ/ in `sat' as contrasted with the longer /ñ/ in `sad', or the /i:/in
`seat' as contrasted with that in `seed'.
3. Consonant clusters
no omission of sounds in word-initial clusters, e.g. in promise, string;
omission in middle and ®nal clusters only permissible according to L1
English rules of syllable structure, e.g. fa
ctsheet' can be pronounced
`facsheet' but not `fatsheet' or `facteet';
/nt/ between vowels as in British English `winter' pronounced /wInt@r/
rather than American English where, by deletion of /t/, it becomes /wIn@r/;
addition is acceptable, for example `product' pronounced [p@rdökUtO] was
intelligible to NNS interlocutors, whereas omission was not, for example
`product' pronounced /'pÁdök/.
4. Vowel sounds
maintenance of contrast between long and short vowels for example.
between `l
ive' and `leave';
L2 regional qualities acceptable if they are consistent, except substitutions
for the sound /Æ:/asin`b
ird', which regularly cause problems.
5. Production and placement of tonic(nuclear) stress
appropriate use of contrastive stress to signal meaning. For example the
dierence in meaning in the utterances `I came by TAXi' and `I CAME by
taxi' in which nuclear stress is shown in upper case. The former is a neutral
statement of fact, whereas the latter includes an additional meaning such as
`but I'm going home by bus'.
Non-core features
In eect, what I am claiming is that the items which are excluded from the
LFC are not crucial to intelligibility in EIL contexts, and that they can
therefore be considered as areas in which L1 transfer indicates not `error' but
(NNS) regional accent. In other words, what we have here is a rede®nition of
phonological and phonetic error for EIL: one which incorporates the
sociolinguistic facts of regional variation instead of regarding any deviation
from NS pronunciation as a potentially harmful error (the EFL perspective). It
should be acknowledged, nevertheless, that NNS comprehension has not been
widely researched in terms of responding to NS and NNS talk, and more
research would be crucial for this proposal to gain empirical validity.
In the following seven cases, the eschewing of an NS way of pronouncing in
favour of an NNS way (usually in¯uenced by L1 transfer) was not found in
my data to cause intelligibility problems for an NNS interlocutor. In addition,
as many pronunciation teachers are only too well aware, some of these
features seem to be unteachable. That is, however much classroom time is
spent on them, learners do not acquire them. It may be that the rules are not
suciently generalizable (as in the case of pitch movements) or are too
complex (as with word stress), or that the item is heavily marked, infrequent
in the world's languages and unnatural (as with /T/ and //). It may even be
that (false) NS intuitions have misleadingly lead us to believe in the existence
of certain aspects of NS pronunciation such as timing and the link between
pitch movement and grammar (see above).
Where the problem is one of overcomplexity or lack of generalizability, it
may be that the best teachers can do is to draw learners' attention receptively to
these items to prime learners for future acquisition outside the classroom,
should the possibility of extended exposure present itself. A further bene®t of
presenting these items receptively is that it might be helpful for learners to be
aware of these features in terms of both their own comprehension of NSs and
their understanding of the dierences between their own pronunciation and
that of NSs. Having said that, however, it is worth repeating that a lack of
these items did not emerge from my data as threatening intelligibility in EIL
interaction. So one could argue that in these cases, it is perhaps NSs who need
to make receptive adjustments rather than expecting NNSs to alter their
production in EIL contexts. The non-core areas are as follows:
1. The consonant sounds /T/, //, and the allophone [à].
2. Vowel quality, for example the dierence between /bös/ and /bUs/ as long
as quality is used consistently.
3. Weak forms, that is the use of schwa instead of the full vowel sound in
words such as `t
o', `from', `of', `was', `do'; in EIL the full vowel sounds
tend to help rather than hinder intelligibility.
4. Other features of connected speech, especially assimilation, for example
the assimilation of the sound /d/ at the end of one word to the sound
at the beginning of the next, so that /red peInt/ (`red paint') becomes
/reb peInt/.
5. The direction of pitch movements whether to signal attitude or
grammatical meaning.
6. The placement of word stress which, in any case, varies considerably
across dierent L1 varieties of English, so that there is a need for receptive
7. Stress-timed rhythm.
The relationship between the EIL and NS positions is summed up in Table 1,
whose purpose is both to demonstrate the degree (or otherwise) of overlap
between the two positions, and the fact that the EIL target is not simply a
subset of NS features. The NS position is based on those features most
commonly found in British EFL and American EFL/ESL pronunciation books.
The development of accommodation skills
Finally, in addition to training in the core items, learners need EIL practice to
enable them to develop their accommodation skills in relation to a wide
range of dierent-L1 interlocutor groups, and to be able to respond quickly
to phonological-based interlocutor incomprehension by adjusting their
pronunciation. Since the majority of NNSs who engage in EIL commun-
ication are not of bilingual pro®ciency (that is, they engage in ILT), we
cannot presume that they will produce the core items with 100 per cent
accuracy all the time. This is less important, however, than having an item
within their repertoire and being able to respond to a speci®c interlocutor's
needs as and when they arise. ILT pair and group work should prove highly
bene®cial in this regard.
Table 1: EIL and NS pronunciation targets
NS target EIL target
1. The consonantal
all sounds
all sounds except /T/,
// and [à]
RP non-rhotic /r/
GA rhotic /r/
rhotic /r/ only
RP intervocalic [t]
GA intervocalic [Q]
intervocalic [t] only
2. Phonetic require-
rarely speci®ed
aspiration after /p/
/t/ /k/
appropriate vowel
length before fortis/
lenis consonants
3. Consonant clusters
all word positions
word initially, word
4. Vowel quantity
long-short contrast
long-short contrast
5. Vowel quality
close to RP or GA
L2 (consistent)
regional qualities
6. Weak forms
unhelpful to intellig-
7. Features of con-
nected speech
inconsequential or
8. Stress-timed
does not exist
9. Word stress
reduce ¯exibility
10. Pitch movement
essential for indicat-
ing attitudes and
rectly linked to NS
11. Nuclear (tonic)
A basic problem for the development of EIL accommodation skills is the fact
that the vast majority of English teaching takes place in same-L1 classrooms in
the learners' own countries. As Table 2 below demonstrates (as did the same-
L1 exchange in Extract no. 2 above), when speakers come from the same L1
background, convergence will result in an increase of phonological transfer in
order to facilitate interlocutor intelligibility, as well as to signal shared group
identity. As Bygate (1988: 76±7) has pointed out, same-L1 group work `at
least allows and at worst encourages fossilization and the use of deviant L2
forms' (though it should be noted that the main thrust of Bygate's article is
that, in other respects, same-L1 group work is not necessarily disadvanta-
geous). On the other hand, interlanguage talk practice leads to precisely the
opposite scenario, since attempts to converge are more likely to result in
replacement of transfer if intelligibility for an interlocutor is potentially
threatened; that is, it will result in replacement of transfer in the core
phonological and phonetic areas.
In addition, whereas interlanguage talk practice is also helpful at the
receptive level by providing learners with exposure to a range of EIL accents
other than their own, same-L1 practice simply reinforces learners' familiarity
with their own accent. However, as things stand for the majority of learners,
the only accent they are likely to hear in the classroom other than their own
is that of RP- or GA-accented speakers on recorded materials.
The best way forward may therefore be for course materials producers to
give some thought to the development of video and audio tapes designed
speci®cally to provide exposure to a range of NNS accents, and for English
language teaching institutions to set up video conferencing activities with
Table 2: Model of predicted phonological outcomes in ILT and same-L1 talk
ILT ! convergence for
! replacement of L1
! interlocutor
and enhanced L2
exposure to other
EIL accents
! convergence for
eciency and
! non-replacement
of L1 transfer
! interlocutor
and fossilization
of IL phonology
lack of exposure
to other EIL
Adapted from Jenkins 2000: 192.
institutions in other L1 areas. This would seem to be a more realistic approach
than to hope for an increase in the availability of mixed-L1 classes around the
world. It would also solve a basic contradiction, which is that while mixed-L1
student groups are optimum for the development of accommodation skills,
the optimum teacher, as Seidlhofer (1999) argues, is often a bilingual English
speaker who shares her students' L1. This teacher will have acquired the core
pronunciation features but will also have clear traces of her regional accent.
She thus provides a more pedagogically realistic and sociolinguistically
reasonable model for her students.
Despite making sense, these proposals are, none the less, likely to prove
controversial. One issue which they will need to address is that of what
learners want to learn. While some learners are likely to respond positively to
the concept of L2 regional accents, others will no doubt share the view of
Andreasson, who argues that `it would ...befarfrom a compliment to tell a
Spanish person that his or her variety is Spanish English. It would imply that
his or her acquisition of the language left something to be desired'
(Andreasson 1994: 402). Similarly, research by Dalton-Puer et al. (e.g.
1997) into Austrian learners' attitudes to pedagogic models has found that RP
and GA are preferred to Austrian-English accents. The problem is com-
pounded by the fact that at present there is no academic course entitled
`English as an International Language'. The result is that the common-sense
philosophy underpinning the EIL pronunciation proposals, with their addition
of an intelligibility dimension to communicative competence (by specifying
core and non-core pronunciation features) and their promotion of accom-
modation skills, remains largely unrepresented and, therefore, uncompre-
hended. Eorts will need to be made to argue the EIL case more widely and
even then, it will be important not to patronize those learners who, having
heard the arguments, still wish to work towards the goal of a native speaker
accent, by telling them they have no need to do so.
It is nevertheless to be hoped that in the not too distant future, English
pronunciation research and teaching will come to terms with the changed
sociolinguistic environment in which they take place, and will generate more
empirical research of the kind described above, instead of relying on intuition,
laboratory experiments and corpora of NS speech.
(Revised version received June 2001)
A shorter version of this paper was presented at
the BAAL Annual Meeting, Cambridge, Septem-
ber 2000. I am most grateful to four anonymous
reviewers, to Martin Bygate and to Barbara
Seidlhofer for their perceptive comments and
criticisms of previous drafts of this paper.
1 The total database consists of approximately
30 hours of recorded interactions from both
classroom groupwork and paired conversa-
tions and information exchanges set up
outside the classroom; and of ®eld notes of
communication breakdowns made over a
four-year period in classroom and social
2 Upper case is used to indicate that the word
was pronounced with tonic stress.
Andreasson A.-M. 1994. `Norm as a pedago-
gical paradigm.' World Englishes 13/3: 395±
Beebe, L. and H. Giles. 1984. `Speech-accom-
modation theories: a discussion in terms of
second language acquisition' in H. Giles
(ed.): International Journal of the Sociology of
Language. The Dynamics of Speech Accommoda-
tion. Amsterdam: Mouton.
Bourdieu, P. 1977. `The economics of linguistic
exchanges.' Social Sciences Information 16/6:
Bygate, M. 1988. `Units of oral expression and
language learning in small group inter-
action.' Applied Linguistics 9/1: 59±82.
Cauldwell, R. 1996. `Stress-timing: observa-
tions, beliefs, and evidence.' Eger Journal of
English Studies 1: 33±48 (also available in
revised form on <
Cook, V. 1999. `Going Beyond the Native
Speaker in Language Teaching', TESOL
Quarterly 33/2: 185±209.
Crystal, D. 1995. The Cambridge Encyclopedia of
the English Language. Cambridge: Cambridge
University Press.
Crystal, D. 1997. English as a Global Language.
Cambridge: Cambridge University Press.
Dalton-Puffer, C., G. Kaltenboeck, and
U. Smit. 1997. `Learner attitudes and L2
pronunciation in Austria.' World Englishes
Duff, P. 1986. `Another look at interlanguage
talk: taking task to task' in R. Day (ed.):
Talking to Learn: Conversation in Second Lan-
guage Acquisition. Rowley, MA: Newbury
Giles, H. 1973. `Accent mobility: a model and
some data.' Anthropological Linguistics 15: 87±
Graddol, D. 1997. The Future of English? The
British Council.
Ioup, G. 1984. `Is there a structural foreign
accent? A comparison of syntactic and
phonological errors in second language
acquisition.' Language Learning 34/2: 1±17.
Jenkins, J. 2000. The Phonology of English as an
International Language. Oxford: Oxford Uni-
versity Press.
Kachru, B. B. 1991. `Liberation linguistics and
the Quirk concern.' English Today 25: 3±13.
Kachru, B. B. 1992. `Models for non-native
Englishes' in B. B. Kachru (ed.): The Other
Tongue. English Across Cultures. Urbana, IL:
University of Illinois Press.
Kachru, B. B. and C. L. Nelson. 1996. `World
Englishes' in S. L. McKay and N. H. Horn-
berger (eds): Sociolinguistics and Language
Teaching. Cambridge: Cambridge University
Kasper, G. 1998. `A bilingual perspective on
interlanguage pragmatics' in J. H. O'Mealy
and L. E. Lyon (eds): Language, Linguistics and
Leadership. Essays in Honor of Carol M.K.
Eastman. Honolulu, Hawai'i: University of
Levis, J. 1999. `The intonation and meaning of
normal yes/no questions.'World Englishes 18/
3: 373±80.
Long, M.H. and P. Porter 1985. `Group work,
interlanguage talk, and second language
acquisition. TESOL Quarterly 19/2: 207±28.
McArthur, T. 1998. The English Languages.
Cambridge: Cambridge University Press.
Pennycook, A. 1999. `Pedagogical implications
of different frameworks for understanding
the global spread of English' in
C. Gnutzmann (ed.): Teaching and Learning
English as a Global Language. Tubingen:
Stauffenburg Verlag.
Pennycook, A. 2000. `The social politics and
the cultural politics of language classrooms'
in J.K. Hall and W.G. Eggington (eds): The
Sociopolitics of English Language Teaching.
Clevedon: Multilingual Matters.
Quirk, R. 1990. `Language varieties and
standard language.' English Today 21: 3±10.
Schwartz, J. 1980. `The negotiation for mean-
ing: repair in conversations between second
language learners of English' in D. Larsen-
Freeman (ed.): Discourse Analysis in Second
Language Research. Rowley, MA: Newbury
Seidlhofer, B. 1999. `Double standards:
teacher education in the Expanding Circle.'
World Englishes 18/2: 233±45.
Seidlhofer, B. 2000. `Mind the gap: English as
a mother tongue vs. English as a lingua
franca.' Vienna English Working Papers 9/1:
Seidlhofer, B. 2001. `Closing a conceptual gap:
the case for a description of English as a
lingua franca.' International Journal of Applied
Linguistics 11/2: 133±58.
Trudgill, P. 1998. `World Englishes: Conver-
gence or divergence?' in H. Lindquist,
S. Klintberg, M. Levin, and M. Estling
(eds): The Major Varieties of English. Papers
from MAVEN 97. Vaxjo: Acta Wexionensia.
Widdowson, H. G. 1991. `The description and
prescription of language' in J.E. Alatis (ed.):
Georgetown University Round Table on Language
and Linguistics. Language, Communication and
Social Meaning. Washington DC: Georgetown
University Press.
Widdowson, H. G. 1994. `The ownership of
English.' TESOL Quarterly 28/2: 377±89.
Widdowson, H. G. 1997. `EIL, ESL, EFL: global
issues and local interests.' World Englishes 16/
1: 135±46.
... In this context, oral dimensions important for communication, such as speech intelligibility (as defined by Munro & Derwing, 1995), should be of interest to English pronunciation teachers and researchers worldwide. The lingua franca core (LFC) proposed by Jenkins (2000Jenkins ( , 2002Jenkins ( , 2007, a set of pronunciation features that allegedly preserve mutual intelligibility among NNSs, has generally (but far from unanimously, see e.g., Dauer, 2005;Szpyra-Kozłowska, 2008) met with enthusiastic reactions from teachers (Levis, 2016;Szpyra-Kozłowska, 2015;Tanner & Henrichsen, 2022). Besides the enthusiastic reaction to the LFC manifested by many pronunciation practitioners and in several circles (such as teachers' associations), the significance of this line of research is well-attested by the impact that Jenkins ' (2002) paper has had on the pronunciation literature. ...
... Besides the enthusiastic reaction to the LFC manifested by many pronunciation practitioners and in several circles (such as teachers' associations), the significance of this line of research is well-attested by the impact that Jenkins ' (2002) paper has had on the pronunciation literature. Thus, Demir and Kartal's (2022) bibliometric analysis showed that Jenkins (2002) is one of the most cited papers in the L2 pronunciation research literature. ...
... Osimk (2009) found that aspiration of voiceless plosives (one of the requisites of the LFC) facilitates intelligibility among NNSs, and that non-canonical realizations of interdental fricatives did not interfere with intelligibility, as predicted by the LFC. The role of rhoticity, which is to be preferred in an ELF context according to Jenkins (2000Jenkins ( , 2002, is less conclusive and demands more investigation. The L1s of the participants were all of European backgrounds, however. ...
Full-text available
Over twenty years ago, Jenkins (2000) put forth the Lingua Franca Core (LFC), a pronunciation syllabus for international intelligibility among non-native speakers (NNSs) of English. Although insufficient empirical research has been directed to validating the LFC proposals, the few studies that have tested this syllabus have produced mixed findings. One of the core features of the LFC is the use of British-based /t/, rather than General American (GA) flap [ɾ], which allegedly has a negative impact on English as lingua franca (ELF) intelligibility. There are, however, three additional types of flap in accents such as GA. In the current study, the intelligibility of the four types of flap typical of GA were tested experimentally, in the context of learners at an intermediate level with four European language backgrounds (n = 78). Using a matched-guise technique, learners were presented with the flapped and non-flapped versions of words including the four types of flap. The results of two experiments indicate that three of the four types of flap proved detrimental for ELF intelligibility to a large extent. More expectedly, word frequency and experience with GA were moderately associated with flap intelligibility.
... English originated in Britain, but now it has become a global or an international language and has been used worldwide as the means of international communication (Jenkins, 2002). Many countries have accepted this language as a foreign language and allowed their students to study this language at schools among the other school subjects (Horwitz, 1988). ...
Full-text available
English has become the world's language, and it has been used by many people in different countries as a means of international communication. Motivation to learn English has been seen as the most important factor contributing to the success of students' learning. The current study employed a quantitative design and was conducted to find out the motivation levels of students studying English as a foreign language at a private high school in Phnom Penh. 60 students participated in a survey after we sought the agreement from school principals and teachers of English there. The findings show that the students in the studied context were more extrinsically motivated to learn English rather than intrinsically. The sum of intrinsic and extrinsic motivations also revealed students' moderate motivation to learn English. Due to the small scale of the survey study, a future study should be conducted with a larger sample size and at other educational institutions across Cambodia. Qualitative and mixed-methods designs are also highly recommended.
... Tanner and Landon 2009;, whereas some others state the opposite (e.g. Jenkins 2002). Zielinski (2015), however, approaches this discussion differently by challenging the premise that segmental and suprasegmental features are independent entities. ...
Full-text available
In EFL classrooms today, teachers are facing the challenge of helping their students improve their intelligibility in international communication in English. Feedback is one of the main ways through which teachers are performing the task. However, feedback on students’ English pronunciation has not received commensurate attention in EFL contexts. Aiming to shed some light on this essential but neglected aspect of English language teaching (ETL), this study examines the beliefs and practices of ten experienced teachers of English at a non-English-major university in Vietnam. Findings derived from semi-structured interviews and classroom observations show that teachers largely shared a strong belief in the importance of feedback in learners’ pronunciation development. Teachers were found to mainly focus on instant feedback rather than delayed feedback, and on individual errors rather than collective ones. Findings also indicated that teachers’ approaches to feedback provision diverged greatly, which were confined by their varying expectations in pronunciation pedagogy, different knowledge, and beliefs regarding how to correct, what to correct, and who would benefit from error correction. On the basis of the findings, implications for classroom practices and teacher training were discussed.
The chapter takes the reader from the concrete phonetic descriptions of sounds, found in Chapters 11 and 12, to the use of these sounds in English. As in every language, sounds are influenced by their context. A large part of phonological description of a language is an effort to describe how the “same” sound is pronounced differently in different contexts, both phonetic and morphological. The chapter provides the phonemes of English, which are the distinctive units of sound, and examples of how they vary in context. It also illustrates the variation of English morphemes in context, by providing examples of allomorphy. Some implications of variation in context for teaching English are discussed.
The current trend of speakers of English as a second or an additional language (English language learners) outnumbering speakers of English as a first language (native English speakers) has shifted the focus of English language teaching from the nativeness principle to the intelligibility principle. Following the intelligibility principle, this review examined the effectiveness of interventions on the intelligibility of English language learners (ELLs) and the comprehensibility of their speech by native speakers in two related but independent meta-analytic studies. Study 1 focused on intervention studies for ELLs to improve intelligibility in their speech. Robust variance estimation (RVE) generated significant effect sizes of 0.62 ( p = 0.00) from 33 effect sizes in 18 independent studies. Study 2 focused on interventions for L1 native English speakers to improve their comprehensibility of ELL speech. RVE generated a statistically significant effect size of 0.24 ( p = 0.04) with 20 effect sizes in 10 independent studies. Moderating analyses revealed that the measures of intelligibility and comprehensibility, and the speech task type, were significant factors explaining the effect size variations between the included studies in both analytic reviews. However, the length of scales did not significantly differentiate the effectiveness of interventions. The effects of pronunciation instruction are not sensitive to research setting, and interventions aiming to change native English speakers’ (NESs’) attitudes towards ELLs’ accented English rather than those aiming to improve NESs’ familiarity with ELLs’ accented English have statistically significant effects. The current analysis generated unique and important implications for future educational practice and research on intelligible and comprehensible communication between L1 speakers and language learners beyond English.
The present paper stems from an awareness that English has become the most widely used means of intercultural communication on a global scale. Therefore, intercultural communication is more likely to occur through English used as a lingua franca than in any other language used as a lingua franca. English has transcended boundaries and has allowed people from distant cultures to come closer and find common grounds. If, on the one hand, the rise of English has been criticized as a threat to minority languages and cultures, on the other, English has been the means by which people are connected across national and international borders. European Universities and University Language Centres are known to be multicultural environments that provide students with opportunities to familiarise with diverse cultural backgrounds and experience non-native English speech. If therefore, university staff and professionals engage regularly with a multilingual population, they have to be prepared to deal with and respond to their different needs. Within this framework, University degree programs need to be able to cope with a changing cultural and linguistic environment where multilingual speakers increasingly interact in English with other non-native English speakers. In the light of these considerations, this small case study intends to raise awareness of the need to integrate academic degree programs with courses which specifically address Intercultural Communication and English as a Lingua Franca. A sample of Italian university websites has been analysed with a view to identifying the extent to which the aforementioned issues are incorporated within the course programs observed. Preliminary results will be described and considerations suggested.
This chapter takes up the issue of authenticity in language pedagogy. Traditional views of authenticity take the native speaker to be the primary authority for linguistic norms. Written standard language is especially highly valued here. It is argued herein that TELL environments are equally valid as learning environments, and that students can use the freedom they provide to develop their own locally negotiated cultural and linguistic norms. Evidence is provided that students on a net-based MA program develop their own norms for reducing language, and use them and other means to mark membership of a local TELL community. Thus, TELL is a rich and authentic environment for learners of English to become what is referred to as “language practitioners.”
This mixed-methods study investigated Chinese tertiary-level students’ perceptions of different English accents existing in mainland China. Also examined were their perception changes and factors that contributed to their perception changes. An electronic questionnaire survey was employed to investigate the general trend and the possible correlations and differences in the participants’ perceptions of different English accents. Follow-up interviews and diary studies were conducted to examine the participants’ perception changes and the contributing factors. The main findings show that the participants were inclined to accord privileged status to Inner Circle English accents, whereas their perceptions of Outer Circle and Expanding Circle English accents tended to remain unfavourable. Their perceptions were found to be under the substantial influence of the ongoing exonormative English education, English language assessment system and employment requirements. We also found that the explicit instruction of the development of English and the frequent cross-cultural communication were likely to engender changes in the participants’ language beliefs and perceptions of different English accent varieties.
The use of group work in classroom second language learning has long been supported by sound pedagogical arguments. Recently, however, a psycholinguistic rationale for group work has emerged from second language acquisition research on conversation between non-native speakers, or interlanguage talk. Provided careful attention is paid to the structure of tasks students work on together, the negotiation work possible in group activity makes it an attractive alternative to the teacher-led, “lockstep” mode and a viable classroom substitute for individual conversations with native speakers.
This article argues that language teaching would benefit by paying attention to the L2 user rather than concentrating primarily on the native speaker. It suggests ways in which language teaching can apply an L2 user model and exploit the students' L1. Because L2 users differ from monolingual native speakers in their knowledge of their L2s and L1s and in some of their cognitive processes, they should be considered as speakers in their own right, not as approximations to monolingual native speakers. In the classroom, teachers can recognise this status by incorporating goals based on L2 users in the outside world, bringing L2 user situations and roles into the classroom, deliberately using the students' L1 in teaching activities, and looking to descriptions of L2 users or L2 learners rather than descriptions of native speakers as a source of information. The main benefits of recognising that L2 users are speakers in the own right, however, will come from students' and teachers' having a positive image of L2 users rather than seeing them as failed native speakers.