I phoned my Spanish internet provider the other day and tried to explain a problem I was experiencing. Clearly, I was unintelligible because the operator immediately switched me to an English-speaking operator. Even then, I had trouble getting my message across, because I didn’t know how to say ‘tráfico de datos’ in English. Was I again being unintelligible, or simply incomprehensible?
This reminded me that, in a session on my MA TESOL course last summer, during a discussion on the goals of pronunciation teaching, one student mentioned the fact that she’d heard that there was a distinction between intelligibility and comprehensibility, and she asked me to explain the difference.
I volunteered an off-the-cuff explanation (as one does!), suggesting that intelligibility is a function of speakers (and particularly of their pronunciation), while comprehensibility (invoking Krashen) is a function of texts. Or, put another way, output is (to a greater or lesser extent) intelligible, while input is (to a greater or lesser extent) comprehensible.
Even as I said this, I could see there were problems. Communication is by definition reciprocal, so is it possible to gauge either intelligibility or comprehensibility without reference to interlocutors – either listeners or readers? Moreover, whether listening to a speaker or reading a text, your degree of understanding is going to be experienced in a similar way: ‘I understand it a bit, a lot, or not at all.’
Since that awkward day (sorry, Autumn, my bad!), I’ve had a chance to research the difference. For example, Munro, Derwing and Morton (2006, p. 112), referencing earlier work by the first two authors, define intelligibility ‘as the extent to which a speaker’s utterance is actually understood’, whereas comprehensibility ‘refers to the listener’s estimation of difficulty in understanding an utterance’. (So I wasn’t entirely wrong, perhaps). They further distinguish both from accentedness, i.e. ‘ the degree to which the pronunciation of an utterance sounds different from an expected production pattern.’ And they add: ‘Although comprehensibility and accentedness are related to intelligibility, they are partially independent dimensions of L2 speech. An utterance that is rated by a listener as “heavily accented,” for instance, might still be understood perfectly by the same listener. Furthermore, two utterances that are fully intelligible might entail perceptibly distinct degrees of processing difficulty, such that they are rated differently for comprehensibility.’
On the other hand, and markedly differently, Nelson (2011), referencing papers by Smith (1992) and Smith and Nelson (1985), defines intelligibility as ‘word and/or utterance recognition, involving the sound system’, and comprehensibility as ‘word/utterance meaning, or locutionary force’. To further complicate matters, they introduce the term interpretability, i.e. ‘the meaning behind the word/utterance, or illocutionary force’.
MacKay (2002, p. 52) helpfully (?) unpacks these distinctions with an example:
If a listener recognises that the word salt is an English word rather than a Spanish word, English is then intelligible to him or her. If the listener in addition knows the meaning of the word, it is comprehensible, and if he or she understands that the phrase, ‘Do you have any salt?’, is intended to be a request for salt, then he or she is said to be able to interpret the language.
Put another way, if you’re having trouble understanding someone, it may be a case of not recognizing what they’re saying (likely their fault), or not knowing what they mean (probably your fault), or not knowing what their intention is (could be anyone’s fault). Going back to my exchange on the phone, I can sort of apply these distinctions, but I’m also wondering if accentedness was the reason why I was switched to the English-speaking operator, since the first operator made no attempt even to negotiate some sort of understanding. (Mercifully, in a subsequent conversation with yet another operator, I was actually congratulated on my Spanish – probably because, although heavily accented, I was intelligible. Or do I mean comprehensible?)
This raises another issue related to intelligibility: that it is highly subjective. As Rajagopolan (2010, p. 467) argues, ‘No matter how one tries to define intelligibility from a neutral standpoint, the question that cries out for an answer is: “intelligible for who?”’ Why was I intelligible to one of my interlocutors but not to another? Was it, indeed, nothing to do with accent at all, but more to do with attitude? After all, it is not accents that are intelligible, it is people. I never tire of quoting Bamgbose (1998) on the subject: ‘Preoccupation with intelligibility has often taken an abstract form characterized by decontextualised comparison of varieties. The point is often missed that it is people, not language codes, that understand one another” (quoted in Jenkins, 2007, p. 84). Thus, intelligibility may have as much to do with our overall impression of a speaker as it has to do with the intrusiveness of their accent (or lack thereof) – not dissimilar to the notion of ‘comfortable intelligibility’ (Kenworthy 1987) or ‘perceived fluency’ (Lennon 2000, cited in Götz 2013).
Either way, this doesn’t provide a lot of solace to those who have to assess a learner’s pronunciation, as in the kinds of oral tests favoured by many public exams nowadays, using descriptors such as these:
- is easy to understand throughout; L1 accent has minimal effect on intelligibility
- can generally be understood throughout, though mispronunciation of individual words or sounds reduces clarity at times
What are the chances that any two raters will agree?
References
Götz, S. (2013) Fluency in native and nonnative English speech. Amsterdam: John Benjamins.
Jenkins, J. (2007) English as a Lingua Franca: Attitude and Identity. Oxford: Oxford University Press.
Kenworthy J (1987) Teaching English Pronunciation. Harlow: Longman.
McKay, S. (2002) Teaching English as an International Language. Oxford: Oxford University Press.
Munro, M.J., Derwing, T.M., & Morton, S.L. (2006) ‘The mutual intelligibility of L2 speech.’ Studies in Second language Acquisition, 28.
Nelson, C. L. (2011). Intelligibility in World Englishes: Theory and Application. New York: Routledge.
Rajagopolan, K. (2010) ‘The soft ideological underbelly of the notion of intelligibility in discussions about “World Englishes”.’ Applied Linguistics, 31/3.
Smith, L. E. (1992). Spread of English and issues of intelligibility. In B. B. Kachru (ed.) The Other Tongue: English across Cultures (Second Edition). Urbana, IL: University of Illinois Press.
Smith, L. E. and Nelson, C. L. (1985). ‘International intelligibility of English: Directions and resources.’ World Englishes, 4(3).
Illustrations by Quentin Blake from Success with English, by Geoffrey Broughton, Penguin Education, 1968.
An A-Z of ELT 
Thank you Scott for such a practical yet tricky blogpost. Tricky in terms of the effort needed in the process of achieving intelligibillity and comprehension within an L2 interaction. In the real-life example with the operator in Spain, Rost 2001 states that “to become successful participants in target-language conversation, listeners need to employ a great deal of ˋinteractional workˋ(including using clarification strategies) in addition to using linguistic processing.” It seems like the operator used no such ‘interactional work’ as you were switched immediately to an English-speaking one. So …. perhaps you could have made yourself more intelligible if you had looked into Spanish adjacency pairs and functional exponents for this kind of instrumental interaction. You could have also predicted and looked up key lexis like ‘trafico de datos’ (I don’t what that is, maybe mobile internet connection or something like that) and then actually taken it a step further and practiced the interaction with a friend, roommate, classmate, etc back-to-back in the kitchen prior to making the call. Vandergrift 1996 advocates the use of metacognitive strategies such as planning and Underwood 1989 also supports the use of pre-activities such as pre-learning of lexis and Nunan 1995c promotes predicting prior to the interaction. So next time, you could prepare a little before making the call or just dive-in and take a more heuristic approach. I think it depends on how long you want to spend speaking to people in South American call centres.
Thanks for your thoughts, Justin. Yes, your point that intelligibility is negotiated is crucial – and that it requires strategizing as well. Perhaps we should be aiming to help our learners develop communication strategies, such as paralinguistics, that compensate for having accents that might otherwise impede full intelligibility. Of course, such strategies are likely to work better in face-to-face interaction: the disembodied nature of phone communication makes it particularly susceptible to breakdown – and, again, I wonder if we do enough to prepare our learners for it.
Thanks again Scott for your feedback. I am curious about paralinguistics. In your opinion, which ones should we focus on in class and should it be looked at for use or just for interpretation? I was also wondering if paralinguistics would be something that transfers naturally from L1 to L2 and hence would not need to be explicitly learned.
Justin, you’re right that paralinguistics (e.g. use of gesture) are probably not conducive to explicit teaching, although we might encourage learners to reinforce their speech with more overt gestures – see G for Gesture in this blog.
The last part of your post articulates a problem I’ve had trouble discussing with colleagues for some time. When using speaking band descriptors with terms like “generally understood throughout”, it places the burden of understanding on the speaker. So, if two different listeners had two very different interpretations of an utterance, the speaker would somehow be responsible for these differences in processing. Or if one examiner understands you, then you pass but if you have a different examiner who doesn’t then you don’t. It just seems like there must be a better way of wording them.
Thanks for making this point, Richard.I wonder if any research has been done on the role of the interlocutor/examiner in tests of speaking such as IELTS. Would a candidate interacting with an interlocutor who is mute and/or expressionless (or even behind a screen) be judged more negatively than with one who is providing signals, both verbal and non-verbal, on the speaker’s intelligibility? If, as I commented to Justin above intelligibility is negotiated, do examiners get instruction as to the extent they are allowed to negotiate? (In the dynamic assessment approach advocated by proponents of sociocultural learning theory, a candidate’s capacity to respond to feedback while communicating is considered to be a significant indicator of competence.)
Dear Scott,
Congratulations for taking up your A-Z again which is always interesting to read.. – I wrote to you earlier on the subject of L1 in FLT. May I let you know that Butzkamm & Caldwell can now be downloaded from ResearchGate? Guy Cook wrote in 2010:
“Such a comprehensive vision of bilingual language teaching is provided by Butzkamm and Caldwell’s landmark book The Bilingual Reform: A Paradigm Shift in Foreign Language Teaching (2009). Impressively grounded in a scholarly understanding of the history and theory of the debate between monolingualists and bilingualistsâ¦â
All the very best to you,
Wolfgang Butzkamm
Thanks, Wolfgang – I’m glad that this important book is more freely available.
Interesting. I’ve always used ‘comprehensible’ reflexively i.e. the speaker/writer aims for comprehensibility to their listener/reader; if they are not, they need to work on something to improve it, be it pronunciation, grammar, or clarity. It’s often come out of my mouth that comprehensibility is the goal before accuracy. But this post has made me question its meaning when compared with intelligibility–a word I use more rarely. Hmmm.
Thanks, Tyson. The term intelligibility has come to the fore in the light of discussions about ELF (English as a lingua franca), where scholars are keen to find a substitute for a native-speaker-like accent as a goal for pronunciation teaching. Jennifer Jenkins was one of the first to explore speech features affecting the mutual intelligibility of speakers using English as a shared means oif communication, out of which she developed the notion of a ‘phonological core’. In her 2000 book (The Phonology of English as an International Language, OUP), she spends a lot of time struggling with various definitions of intelligibility, and opts for a use of the term that ‘concerns the production and recognition of the formal properties of words and utterances and, in particular, the ability to produce and receive phonological form’ (p. 78), but avoids the question as to whose standards determine ‘the formal properties of words and utterances’. So, if you say ‘toMAHto’ and I say ‘toMAYto’ we are still probably mutually intelligible. But if you stress anything but the second syllable (‘TOmato’, or ‘tomaTO’), you may well not be – the pronunciation of the word has departed too far from the standard – whoever decides what that standard is. (Perhaps it is simply familiarity , I am simply not used to hearing the word pronounced like that?).
She also acknowledges that intelligibility is negotiated, ‘with a two-way process involving both speaker and listener at every stage of the interaction’.
In the end, the distinction between the terms intelligibility and comprehensibility seems to be a scalar one: intelligibility applies to discrete features of pronunciation, while comprehensibility is more holistic, taking into account other verbal and non-verbal features of communication.
All makes perfect sense–moving away from what is most familiar that decreases intelligibility. Still, that familiarity is flexible, eh?
Incidentally, I’ve now heard that Tracey Derwing and Murray Munro are both speaking at TESL Canada Conference Jun 8-10, so I’ll considering attending their talks.
Thanks for the interesting differentiation of closely related terms while analysing the quality of communication of the learners of a language. I have found that with increased exposure to varied speakers, learners find their comprehensibility increasing during listening, just as the more the speaker communicates with varied listeners, the intelligibility of speakers too is enhanced for the learners. Examples are common when a subject teacher communicates with students using the target language as a medium of instruction for science/ humanities / commerce etc. The success of interpretation depends on both during communication, though only the learners’ weakness is usually highlighted !
Thanks for this comment, and your point about ‘increased exposure’ partly echoes my comment to Tyson that maybe familiarity is key to intelligibility: if we pronounce the target language in ways that are familiar to our interlocutor (irrespective of whether this is the ‘standard’ or not) we are more likely to be understood. This is perhaps why speakers who share the same L1 are more easily understood amongst themselves when speaking an L2, than speakers who don’t share the same L1.
Sorry about not writing my name below my response.
-Sulabha S.
hi all
there’s a new scale for EAP learners based on comprehensibility [http://www.ucl.ac.uk/ioe/news-events/news-pub/may-2017/ioe-launches-worlds-first-comprehensibility-scale/] where comprehensibility is defined as listener “effort” to understand
not sure if that makes it any easier to assess though : )
also there is a nice paper comparing the conceptions by Larry Smith and Jennifer Jenkins which is worth a read – World Englishes, English as a lingua franca, and intelligibility
Margie Berns, 2008
ta
mura
Thank you very much for the scale and paper.
no problem Justin : )
Thanks, Mura, for this link. Significant, I think that the designer of the scale says: “Many current rating scales, such as IELTS, are so general when referring to ‘comprehensibility’ that it is difficult to pinpoint the factors that make someone easier or harder to understand at different scale levels.” It’s also interesting that they retain the notion of ‘error’ (as in ‘Errors minimally interfere with the message’) without, at first glance, at least, defining that an error is, although they insist that native-like accent is not the goal. What constitutes a pronunciation error, in that case?
hi Scott
according to the pron scale descriptor, errors include – “misplaced word stress, sound substitutions, not stressing important words
in a sentence”
as to what “constitutes a pronunciation error” are you asking about to what “standard” the raters will judge performance?
if so that “standard” would be the EAP “contexts” that the raters work in i guess?
it is interesting that scale does not ask raters “to compensate for their familiarity with L1 accented” which the authors point to as a limitation (as familiarity will affect scores) but on the other hand recognizes the reality of everyday assessment of students
ta
mura
Thanks, Mura. I guess the problem that testers face – using this or any scale – is that the criteria can become somewhat circular. Descriptor: errors did not impede intelligibility. Q: What is an error? A: Anything that impedes intelligibility.
Thanks for this, Scott.
Just to throw in another perspective and set of references, ‘Intelligibility’ was also the focus of an ELT Journal Key Concept piece (issue 69/2 in 2015), freely available at https://academic.oup.com/eltj/pages/key_concepts . Hope it’s of interest (and apologies for the shameless ELTJ plug).
It’s comforting to see how my incessant questions continue to haunt you a year after the end of the summer session ;). In all seriousness, though, what helpful food for thought. Thank you for this post!
Thanks, Autumn – who knows what other questions might come back to haunt you? 🙂
I’m still processing so much of the material and questions you raised during our class–and for that I am forever grateful!
If you’re autumn, I guess summer is always going to haunt you 🙂
Truly, I am a haunted individual.