In my posts for ‘B’ and ‘D’ I talked about bilabial and dental consonants. These terms are an indication of where the sounds in each category are made in the mouth. In other words, they are an indication of the place of articulation. In contrast, the term fricative is an indication of how a consonant sound is made. That is to say, it indicates the type or manner of articulation. This two-way split is reflected in a chart that many of you will be familiar with in one form or another, especially those of you who did phonetics at university or during teacher training. The horizontal axis shows place of articulation and the vertical axis reflects manner.
All consonant sounds are made in the same way, which is by restricting the air as it travels from our lungs out of our body through our mouth or nose. There are various ways in which we can restrict the passage of the outgoing air, but they mostly involve bringing the moveable parts of our mouth (the bottom lip, the bottom teeth, and the tongue) into contact or near contact with the fixed parts of our mouths (top lip, top teeth, alveolar ridge, hard palate, soft palate. A complete restriction or obstruction of the airstream, for example, gives rise to /p/, b/, /t/, /d/, /k/ and /g/, the consonants known as the stops (or plosives). In contrast, a narrowing of the airflow, gives rise to the fricatives. The term fricative refers to the turbulence or friction that this narrowing creates as the air travels out of our mouth.
English has nine fricatives – /f, v/, /θ, ð/, /s, z/, /ʃ, ʒ/ and /h/. Eight of these work in voiceless–voiced pairs and differ from each other in place of articulation. One member of each of the four pairs is voiceless, with the shading in the table above indicating which of the two is voiced. The ninth fricative, /h/, is voiceless and has no voiced equivalent.
So far, this is all probably revision for you. But as with the dentals and bilabials of my earlier posts, in real life things are not that easy. Quite a lot of the fricatives of English are not part of the phonological systems of other languages, and learners from these other languages will need to ‘fill in the gaps’ with something that is part of their language, and which they feel is a good replacement.
In the post on dental consonants, for example, we saw that /θ/ and /ð/, the English dental fricatives, are regularly replaced either with the dental plosives /t̪/ and /d̪/, or the labio-dental fricatives /f, v/, or the alveolar fricatives /s, z/. When you first come into teaching you can easily be at a loss to explain why this might be, but if you go back up to the consonant table, you will see how close these ‘alternative’ sounds are in terms of their place or manner of articulation with respect to the place and manner of articulation of /θ/ and /ð/. It is really does just take a slight shift either in the place or manner to get from /θ/ or /ð/ to their most common substitutions.
These slight shifts lie behind a lot of what we pick up on in our learners’ pronunciation errors. With Japanese–L1 learners of English, for example, the labio-dental fricative /f/ is not phoneme of Japanese and is replaced by the closest alternative, which is the voiceless, bilabial fricative /ɸ/. This is the sound Japanese speakers make at the beginning of words like ‘foto’ or ‘Fuji’, but it doesn’t exist in English. As a result, listeners align what they hear with the only voiceless, bilabial phoneme English does have, which is /p/. The end result is that coffee can sound like copy or fur like purr. This sort of substitution can have a serious impact on intelligibility and needs dealing with. Learners need showing what a fricative is and how it differs from a plosive, together with discrimination work to check that they are actually perceiving the difference between the two. To do this minimal pairs like coffee/copy, fur/purr or feel/peel can be useful.
Through a similar process, both Japanese and Spanish speakers of English are regularly heard to pronounce vote like boat. The explanation lies in the fact that neither language has /v/ as a phoneme, and so speakers from either background reach for what they do have, which is the voiced, bilabial fricative /β/. Again this is not a sound of English, so the listener aligns it with the voiced bilabial plosive /b/. As before, treatment of this problem begins with perception work – if your learners can’t hear the difference between /b/ and /v/ then they are going to struggle to produce it. Once you are confident that they are hearing the difference (and remember to work with sounds in words and not just isolated sounds), demonstrate the precise articulation of /v/, with the lower lip raised until it is light contact with the upper teeth, plus the idea that it is a voiced sound.
Of the other fricatives of English, /z/, /ʃ/ and /ʒ/ can be problematic for learners from various language backgrounds, including Mandarin Chinese, Japanese, Malay, Spanish, and Greek. The confusion of /s/ and /ʃ/ is quite common, for example, and is problematical in the sense that both sounds are frequent in English and so represent the difference between a significant number of pairs of words such as sew and show, mess and mesh, or rust and rushed. The pronunciation of /ʒ/ is also problematic for learners from a number of L1 backgrounds, but unlike /s/ and /ʃ/, is not a common sound of English. This is a useful idea – this sound isn’t very common so it won’t get in the way of intelligibility too often, so I can leave it alone.
Technically here we are referring to functional load, an idea that is usually attributed to the work of Catford (1987) and Brown (1991). In an interesting experiment to test out the validity of the notion of functional load in pronunciation, Canadian researchers Murray Munro and Tracy Derwing (2006) found that a group of listeners with no training in phonetics found that pronunciation errors with a low functional load reduced comprehensibility much less than errors with a high functional load. ‘In fact, sentences with even three low FL errors were judged more comprehensible than sentences with a single high FL error’ (Derwing and Munro, 2015: 75). This notion of functional load is really useful to us. Some sounds are simply not worth the effort as they don’t get used often enough to contribute significantly to the intelligibility of the message.
As an alternative to simply ignoring a problem sound, however, I’m a big fan of the solution that David Deterding puts forward in Teaching the Pronunciation of English as a Lingua Franca (Walker, 2000: 106). Because even quite proficient Mandarin Chinese speakers of English replace /ʒ/ with /r/ (making usually sound like [jurəli]), Deterding proposes that these learners be deliberately encouraged to use /ʃ/ to replace /ʒ/ as an alternative to their natural tendency to use /r/. Doing this means that usually sounds like you Shirley. However, in Deterding’s opinion (an opinion based on many years of experience teaching in SE Asia, together with a deep understanding of the pronunciation of English for international communication), this /ʃ/ for /ʒ/ is a practical solution that his learners can manage.
His other solution for the problem of the /ʒ/ in usually, I like even more. Get your students to say generally, instead.
Brown, A. (1991). Functional load and the teaching of pronunciation. In A. Brown (ed.), Teaching English pronunciation: A book of readings. London: Routledge, 211–224.
Catford, J. C. (1987). Phonetics and the teaching of pronunciation: A systemic description of English phonology. In J. Morley (ed.), Current perspectives on pronunciation. Washington, DC, TESOL, 878–100.
Derwing, T. M. & Munro, M. J. (2105). Pronunciation Fundamentals. Evidence-based Perspectives for L2 Teaching and Research. Amsterdam, John Benjamins Publishing Company.
Munro, M. J. & Derwing, T. M. (2006). The functional load principle in ESL pronunciation instruction: An exploratory study. System, 34, 520–531.
Walker, R. (2010) Teaching the Pronunciation of English as a Lingua Franca. Oxford: Oxford University Press.