LINGUIST List 7.718

Tue May 21 1996

Sum: -y in English

Editor for this issue: T. Daniel Seely <>


  1. Dror Ben-Arie, Sum: -y in English

Message 1: Sum: -y in English

Date: Tue, 21 May 1996 08:23:53 +0700
From: Dror Ben-Arie <P85013VM.BIU.AC.IL>
Subject: Sum: -y in English
Dear Linguists,
About 10 days ago I posted a query on the pronunciation of "-y" in
different varieties of English. I have received responses from 21 people,
some sharing their personal observations and some supplying theoretical
analyses of the phenomena.

I want to thank all those who contributed, and to apologise for not
having answered personally to all of them:

 Dorine S. Houston, Lance Eccles, Toyomi Takahashi, Charles T. Scott,
 Wenchao Li, Larry Trask, Paul Woods, Glynis Baguley, Geoffrey Sampson,
 Jakob Dempsey, John Coleman, Dom Watt, Marc Picard, Dale Russell,
 Mark A. Mandel, Chris Miller, Charles Rowe,,
 Waruno Mahdi, Laurie Bauer, Robbie Petterson.

Several people asked to see a summary of the responses. I am glad to
share it with you. For obvious reasons, I could not quote here all

The attached summary consists of the following sections:
1. General
2. Theory
3. England - RP
4. England - regional variation
5. English vs. American pronunciation
6. Variation outside England

Dror Ben-Arie
PhD student in linguistics
English department, Bar-Ilan University

My original query was:
> Date: Wed, 08 May 1996 08:19:12 +0700
> From: P85013VM.BIU.AC.IL
> Subject: "-y" pronunciation in British English
> I have noticed that "-y" is transcribed as /i/ in American dictionaries
> and as /I/ in British dictionaries. Yet, the British English speakers I
> know find such pronunciation odd. I would like to ask:
> - in what dialects is "-y" indeed /I/ (e.g., "happy", "nicely") ?
> - in those dialects, can /I/ turn into /i/ in certain circumstances,
> for instance when followed by a vowel (e.g. "the angry American") ?
> I would appreciate your answers. Please reply to .

1. General
- --------
John Coleman:
> The best discussion I know of is J. Windsor Lewis (1990) HappY land
> reconoitred: the unstressed word-final -y vowel in General
> British Pronunciation. In S. Ramsaran, ed. Studies in the
> Pronunciation of English. Routledge. 159-167.
> There is a paper by Local in the same volume on the pronunciation of
> -y in Tyneside. In K. Lodge's book there are many examples from
> Stockport, in which the vowel in question is normally pronounced [E]
> (open-mid front vowel).

Chris Miller:
> Wells, J.C. 1982. Accents of English, vol. 1 [of 3], Cambridge
> University Press.

2. Theory
- -------
Charles Scott:
> Chomsky & Halle (1968) analyze unstressed, word-final -y as a glide
> phonologically, i.e. /y/, assume a rule of glide vocalization (i.e. to
> [I]) after stress has been assigned, then a rule of Word-Final Vowel
> Tensing ([I] ---> [i]). The claim, then, is that unstressed word-final
> high or mid vowels are predictably tense, which would then give them
> appropriate glides, i.e. [iy], [ey], [uw], [ow], e.g. city, tissue,
> fellow (no examples of unstressed word-final [ey]).
> So, the [I] transcription might unwittingly be a phonological
> representation, though in fact there are some speakers who seem to have
> a real phonetic word-final [I] also.

John Coleman:
> You must be careful not to confuse the question of the phonemic analysis
> of "-y" from the pronunciation of "-y". Even for speakers who pronounce
> it as [i], (e.g. in American and many varieties of British English),
> there are many good arguments that it is an instance of the phoneme /i/.
> In fact, it has been extensively studied, and is known in English
> descriptive phonetics as the "happY vowel" (stet.)

Toyomi Takahashi:
> According to the Longman Pronunciation Dictionary by John C. Wells, your
> question is related to the issue of neutralisation of high weak vowel.
> You can find a more precise account of this neutralisation on pages
> 152-153 and 476 of LPD.

Lance Eccles:
> There can be such a range of variants (from /i/ to /e/) because front
> vowels don't contrast at the end of English words.

Geoffrey Sampson:
> It strikes me in any case that transcribing [i] v. [I] in that
> circumstance might be as much a conventional decision as a reflection of
> an objective difference, given that one would expect some mutual
> influence between the vowels anyway.
> This is something I've often wondered about. Firstly, my own analysis
> of what happens in my speech (which is pretty much core RP), is that
> "-y" is realized as [i], that is, with the quality (more or less) of
> the /i:/ phoneme and the length (more or less) of the /I/ phoneme. The
> easiest (simple-minded) phonological analysis, I think, is just to say
> that "-y" is /I/, realized as [i] in word final position. The phoneme
> should be /I/, since there is a contrast with /i:/; I can't think of a
> precise minimal pair offhand, but the contrast is between the suffixes
> "-y" and "-ee" (secondary stress confuses the issue, I suppose; a true
> minimal pair might be "goaty" and "goatee", but now I'm not sure
> whether I would distinguish them!). Of course, this might mess up
> one's pet phonological theory, which probably says that the lax vowel
> phonemes can't occur in open syllables. I haven't looked to see how
> real phonologists analyse this problem!

3. England - RP
- -------------
Larry Trask:
> Now, you're aware that there has existed for some generations a
> standard accent in England called Received Pronunciation, or RP. RP
> is usually characterized as a purely class-based accent having no
> regional associations, but in fact in most respects it is much closer
> to southern speech than to northern. Traditionally, though, there was
> one glaring exception: RP-speakers used northern [I] instead of
> southern [i] in the words under discussion. Hence the traditional RP
> pronunciation of `city' is with [I] in both syllables, and this is the
> pronunciation given in older descriptions of RP.
> Now, only a tiny percent of people in England have an RP accent, and
> very likely most of your English friends have typical southern
> accents rather than RP, so they'll assuredly have [i]. BUT, and this
> is a big but, RP itself has changed in the last generation or so, and
> most RP-speakers now use [i] rather than [I] in these words.
> Accordingly, the use of [I] is now old-fashioned, and you are
> unlikely to hear it from a younger RP-speaker. You will, however,
> hear it from elderly or conservative RP-speakers: certain actors,
> generals, Oxbridge dons, and members of the royal family. The Queen
> has [I], but the Queen's version of RP is remarkably conservative: it
> has been described as being "stuck in a time warp".
> But already the [I] pronunciation is considered so outdated that John
> Wells, in his _Longman Pronunciation Dictionary_, recognizes only [i]
> and confines [I] to the briefest of footnotes in his introduction.
> The RP change from [I] to [i] has apparently been rapid, and it is
> now complete.

Geoffrey Sampson:
> I'm an Englishman (age 51) whose
> family background was lower-to-middle middle class in the London area,
> and I believe my speech would be regarded in other respects as conforming
> to the RP system, without sounding particularly "posh" in terms of the
> precise etic values of the vowels; but I definitely have [i] rather than
> [I] for "-y" in all environments. Yet, in novels of the 1940s or 1950s
> with an Indian background, if the author wanted to imitate in writing
> the speech of a Eurasian (a mixed-race white/Indian), a very standard
> way to > do it was with spellings like "happee" for "happy". I remember
> discussing this once (I think it would have been in the early 1970s)
> with John Wells (Professor of Phonetics at Univ College London),
> who felt it was clear that in standard English the pronunciation is [I],
> and, as I listened to him speak, it seemed to me that with respect to
> his own speech this was quite correct. But then John would definitely
> be categorized by fellow Englishmen as in terms of his speech several
> notches of the social scale above me! My guess is that this is a fairly
> recent change whereby a variant that used to occur only in some clearly
> non-standard varieties of the language has spread into currency within
> more or less standard English (I'm sure I'm not unusual in using [i]
> rather than [I] for "-y" though otherwise conforming to RP). Much
> more recently there have been other features which have transited from
> varieties of English that were highly stigmatized in my childhood into
> social acceptance (you may have encountered recent discussions of
> "Estuary English", containing features of lower-class London speech
> such as glottalized T but now spoken by members of the younger
> generation up to and including some of the royals, not that that's
> saying much these days alas!). And this is not just a question of a
> changing distribution of pronunciations over a stable class structure;
> it is also very obvious if one lives in England that whereas in the
> 1950s large numbers of middle class speakers would have liked their
> speech to mark them out as upper-middle-class, in the 1990s that is no
> longer a usual aspiration.

Dom Watt:
> . . . Young speakers of RP (if such a thing exists any more) use [i]
> or [Ii] at the ends of words like _happy_ or _nicely_, as far as I can
> tell. Of course, hardly anyone in Britain speaks RP anyway.

Chris Miller:
> I can vouch from personal experience that it is probably an RP
> pronunciation, perhaps typical of upper classes and those who emulate
> them, including my mother. I remember distinctly when I was about six or
> seven years old, a year or two after immigrating to Canada from England
> (I think this was in the context of my father's deciding definitively to
> stay), asking my mother whether this meant that now we should start
> saying (I think this was the word) "funn[i]" instead of "funn[I]".
> Leaving aside this early example of metalinguistic awareness that shows
> I was probably foreordained to become a linguist (and a phonologist to
> boot), I think this is good evidence for this pronunciation as a
> perceived norm among at least some segments of the population in
> southern England in the 60s.

Waruno Mahdi:
> I'm not a native (nor even a non-native) British English speaker,
> but I listen to and watch the BBC in regular intervals, and have
> gathered the impression that among the announcers (who, I do presume,
> have been schooled to speak a more or less standard version of
> British English) one can hear /i/ as well as /I/. The /I/-pronouncing
> dialect seems to be somewhat more "distinguished" than the /i/-pronoun-
> cing. The "uppitiest" dialect even pronounces /e/ (e.g. Margret
> Thatcher), perhaps a bug one gets infected with in Eton? It's the
> same dialect which almost lets the lower jaw drop out of its hinges
> in order to pronunce IPA ae-ligature low enough.

4. England - regional variation
- -----------------------------
Larry Trask:
> Generally speaking, in regional speech, the south of England uses [i],
> but the north uses something else. For most northern speakers this is
> [I], but in parts of the north, and above all in Lancashire, it is
> much more open than [I] and approaches the vowel of `bed', or even the
> vowel of `bad' (to my moderately trained ear). So, for example, in
> `city', a southerner has a higher vowel in the second syllable, a
> typical northerner will have the same vowel in both syllables, and a
> Lancashire speaker will have a lower vowel in the second syllable.
> (The Lancashire pronunciation is really startling if you've not heard
> it before. Try pronouncing `sit at' without the final [t], and you'll
> get a reasonable approximation.)

Paul Woods:
> Some northern speakers use /I/ for words like "city", although slightly
> diphthongised.

Chris Miller:
> Having heard an example of this pronunciation on the news just two days
> ago, I can attest that it is fairly common that this word-final
> pronunciation is realized as a diphthong [II>] where the second
> element is a centralized offglide.

Glynis Baguley:
> In some accents, the sound certainly is /i/; the
> speech of North-east England (Newcastle-upon-Tyne especially) springs to
> mind. In other areas it's emphatically not /i/; in West Yorkshire for
> example it even goes some way towards /e/. In RP I'm pretty sure that
> phonemically it's /I/, though phonetically it seems to be between [I]
> and [i] though nearer the former. (Caveat: I'm not a phonetician and I
> don't really know how feasible that is.)
> If you get the chance to see old British movies - WWII ones perhaps -
> note the pronunciation of `posh' characters, eg high-ranking
> officers in the armed forces. This is _miles_ away from /i/. Not many
> people sound like that these days.
> Subjectively, I feel that I make the same sound twice in words like
> `silly' or `pretty', and different sounds in words like `greedy' or
> `beefy'. I'm from the North-West of England, but have lived in the south
> for the last 18 years, and as far as I'm aware my `-y' sounds don't
> carry any characteristically northern or north Midlands flavour.

Laurie Bauer:
> Mine, for a start. I was brought up in Yorkshire, one of the few areas
> to retain this feature.
> Yes, for me it does alternate before vowels, but I've been in NZ for so
> long that the local /i/ prounciation may be influencing me.

Dom Watt:
> The phonetic values of the -y in these words varies enormously between
> different accents in the UK: in Scotland it's something like [e], in
> Newcastle and Wales a very close [i:], in Yorkshire and Lancashire []
> (schwa), in Birmingham [i], and so on.
> As to whether it is ever [I]: the realization of "-y" varies quite a lot,
> and is a useful indicator of some accents. In some accents from London
> and its environs, (particularly stigmatized ones!), it's diphthongized
> and retracted to something like [I-] (schwa, retracted I). I think
> that in some Northern accents it's [I]; but I don't know what happens
> before vowels. (Your posited situation is what I would expect to
> happen, but I'd be lying if I claimed to know that it did!)

5. English vs. American Pronunciation
- -----------------------------------
Mark Mandel:
> Dragon Systems builds speech recognition software, and we develop
> extensive pronouncing dictionaries of the languages we work with. After
> we built an early recognizer for British English based on our American
> pronunciations, our British researchers made this same claim to me and
> asked to change the final /i/s to /I/.
> On listening to them, I had to agree. To me, the British final -y (in
> several dialects*) does not sound identical to the /I/ in the word
> "British" (either of them), but some phonetic difference can be allowed
> for final position and an open syllable, and it is acoustically (well,
> auditorily: I didn't make spectrograms) closer to /I/ than to /i/. If
> you used the same /I/ finally as internally, I'm not surprised they
> found it odd.

6. Variation outside England
- --------------------------
Marc Picard:
> You can get some interesting info on *-y* in American English in

Larry Trask, Lance Eccles, Laurie Bauer:
> Final -y. as in `happy' and `nicely', is universally [i] in American
> English, and also in most or all Southern Hemisphere accents
> (Australia, New Zeland).

John Coleman, Charles Scott:
> The most famous American pronunciation dictionary, Kenyon and Knott,
> notes both variants under the entry for "-y", but prefers the /I/
> representation over /i/ throughout the text.
> See also Kenyon "American Pronunciation".

Dorine S. Houston:
> /i/ Mid-Atlantic Coast English
> /I/ the PA Dutch country people in the Pennsylvania Dutch country of
> Lancaster country
> I (an urban dweller) tend to associate this pronunciation with the PA
> Dutch whose real first language is actually PA Dutch, and related rural
> groups. However, I must add that the children and grandchildren of
> those who say [happI], when they become more urbanized, change their
> pronunciation to [happi].

Dale Russell:
> . . . We almost all came to the conclusion that (at least for American
> mid-Atlantic speakers, and I can now confirm for mid-Westerners as
> well), this word-final sound is not exactly like /i/ in other
> contexts, nor is it exactly like /I/. We're conditioned to think of
> it as /i/, and maybe it is in slow careful pronunciation (as well as
> in certain phonetic contexts, as you observed), but in many contexts
> in connected speech, /I/ is probably closer, and has the advantage of
> calling our attention to the divergence of reality from expectations.

Charles Rowe:
> There are some dialects in the American South (generally the coastal
> ones--which tend to correspond generally to the so-called "r-less"
> dialects in the South) which use a lax [I] for /i/ in this morpheme.
> As far as American English goes, the phenomenon you speak of ([I] for
> /i/ for the morpheme +y+) is probably not restricted to Southern
> Coastal-type varieties, on second thought.
> Specifically: I've noticed Carl Kassel's (?sp) lect on NPR. I do not
> know where he hails from--or indeed, whether he is American (he may be
> Canadian, for example). At any rate, Kassel's speech is rather strongly
> characterized with this feature.

Jakob Dempsey:
> such a lowered value seems to be common in some sort of rural (Southern?
> Scots-Irish?) version of American English, vis. all those country-western
> and rock singers (and British imitators such as Mick Jagger) who go
> even further and give a value like -e or even -E (lower).

Robbie Petterson:
> New Zealand English has its roots in British English, and we NZ-ers
> definitely pronounce "-y" as /i/, not /I/. The NZE /I/ is nearly a
> schwa in other positions, so it would sound like a "-er" ending if it
> were at the end of a word.
> Australian English, though similar, has a pronunciation of /I/ which is
> much closer (if not equal) to /i/, and I have heard them describe their
> "-y" ending as a /I/.
> A point of interest: in popular songs that we hear sung by Americans,
> we hear the "-y" ending very often being pronounced as /I/, especially
> the word "baby". Some of us even imitiate it (and other American
> pronunciation features) when we sing those songs.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue