* * * * * * * * * * * * * * * * * * * * * * * *
LINGUIST List logo Eastern Michigan University Wayne State University *
* People & Organizations * Jobs * Calls & Conferences * Publications * Language Resources * Text & Computer Tools * Teaching & Learning * Mailing Lists * Search *
* *
LINGUIST List 17.1861

Fri Jun 23 2006

Sum: English 'Should Of' Construction

Editor for this issue: Jessica Boynton <jessicalinguistlist.org>

To post to LINGUIST, use our convenient web form at http://linguistlist.org/LL/posttolinguist.html.
        1.    Mark Jones, English 'Should Of' Construction

Message 1: English 'Should Of' Construction
Date: 22-Jun-2006
From: Mark Jones <markjjoneshotmail.com>
Subject: English 'Should Of' Construction

Regarding query: http://linguistlist.org/issues/17/17-1819.html#1

I recently posted a query to LinguistList about the 'should of' construction in
speech). Although I was familiar with the occurrence of orthographic to
represent reduced 'have' orthographic <'ve> (both phonetic [v], where [] =
schwa) in writing, I hadn't seen any references to the use of a realisation with
a LOT vowel (back rounded or unrounded, depending on dialect) in stressed
positions in speech., i.e. a full form of 'of'. Use of the full form OF occurred
in stressed positions in phrases like "He didn't score, but he SHOULD OF" in TV
interviews with two British English footballers (Alan Shearer, 35, born
Newcastle, and Wayne Rooney, 20, born Liverpool).

This summary is in two parts: the first part is a collection of responses, and
the second is a comment on the comments. Many thanks to all who responded -
names at the end.

Part 1 - Respondents' comments

Geographical extent: UK (East Yorkshire, south and NW England, Nottinghamshire,
London (1940/1950's), Liverpool, Reading, Newcastle); US (NE AmEng, US
colloquial English, Kentucky/Tennessee/Appalachian English, Georgia, New York);
Southern Hemisphere (New Zealand Eng, Ausralian Eng, South African Eng).

Respondents often suggested the evolution of the form was due to reanalysis as
unstressed of = 've phonetically. This I think is probably correct, and

Some questioned whether the full form was not occasionally a transcription
error, or in fact simply a "long schwa". This comment only works for Englishes
in which the LOT vowel is back unrounded, i.e. not British English or Southern
Hemisphere. See comments below.

One respondent suggested that stressed HAVE was impossible for him in his New
York speech.

I think we can safely conclude that stressed SHOULD OF in speech is not a
geographically restricted dialectal feature, nor is it limited to writing.

The following references turned up - there are others by Tim Stowell and Cleo
Condoravdi, but no detailed information was given. I haven't checked Coates.


- Coates, Richard. (1989). ''A solution to the 'must of' problem''. York Papers
in Linguistics 14.159-167.
- Kayne, Richard. (1997). ''The English complementizer of''. The Journal of
Germanic Linguistics 1.43-54.
- Cheshire, Jenny, Viv Edwards, and Pamela Whittle. (1993). "Non-standard
English and
dialect levelling." In Milroy, James and Lesley Milroy (eds.). Real English. The
grammar of
English dialects in the British Isles. Longman, London: 53-96.
- (2005). ''Syntactic surprises in some English letters: the underlying progress
of the language''. Language History from Below - Linguistic Variation in the
Germanic Languages from 1700 to 2000, Clifton, Bristol. URL:

Aidan Coveney reports similar phonetically-driven reanalysis in written French
with "que l-on" > "qu'on l'on".

Part 2 - Comments on the comments

Some phonetically-oriented comments are required. Unstressed 'of' and ''ve' are
phonetically identical. This undoubtedly underlies the reanalysis, but
unstressed forms with schwa ([] below) tell us nothing about which form is
used. The LOT vowel ([O] below) is needed. I can put this no better than David
Denison in his paper on 'Syntactic Surprises in some English letters' (2005: 9):

"If the spelling represents unstressed [v] then we have a purely graphic
phenomenon? It doesn't represent indubitable proof that the writers were not
identifying this form with the perfect auxiliary have, though it is suggestive?
However, we know from the present day that many speakers genuinely identify the
word with of and not with have and - crucially - are happy to give it the
stressed pronunciation [Ov]. When that happens, we know for sure that we have a
significant reorganisation of the auxiliary system."

So written forms of for unstressed <'ve> are like orthographic confusions
of or etc., and no indication of anything
other than an ad-hoc solution to writing speech. This could be a separate

Kayne (1997) comments that may be pronounced with a LOT vowel, but
otherwise deals with unstressed instances (with schwa?), and hence no
intepretation on the use of either ''ve' or 'of' is reliably possible. Kayne
specifically rejects the grammaticality of a stressed OF (example 35).

Patrick Honeybone mentioned data on the acceptability of 'should of' in a
written questionnaire on dialect grammar in Britain in a survey carried out
1986-1989 (Cheshire et al. 1993: 66). 73/80 schools across Britain reported the
form. Unfortunately, the questionnaire is written and the context is unstressed
(schwa likely). Cheshire et al. (1993: 66) say that the 'full form' is
pronounced with a LOT vowel, and suggest the form may be recent, and that
it may have emerged in writing. David Denison (2005) mentions the 19th century
use of (in an 1814 letter from Keats and poetry from Reynolds), and the
occurrence of written seems common in children acquiring literacy.

Even if the UNSTRESSED forms may occasionally sound more like they contain the
LOT vowel to a transcriber, this is no guarantee of an 'of' reanalysis. Schwa is
a very malleable vowel, and its precise quality will be affected by neighbouring
vowels and consonants. This is not necessarily a problem if corpus studies pay
attention to the degree of labial and lingual coarticulation to neighbouring
consonants or the height and rounding of adjacent vowels, but this is unlikely.
Transcribers could be misled, though the indeterminate quality of schwa could be
a legitimate factor in reanalysis by naïve language users: a usage-oriented
account might argue that if a language user hears many cases of more LOT-like
instances of schwa in ''ve' (because the majority of the possible neighbouring
vowels condition a perceptual movement in that acoustic direction), an 'of'
reanalysis is more likely.

Either way, as Denison (2005) says, the stressed forms with a LOT vowel for 'of'
(and also possibly lacking /h/) are essential to identify a case of reanalysis.

To really demonstrate the OF reanalysis in stressed forms in a maximally
objective way, transcription is no use, and we should use spectrographic data on
1) the presence/absence of [h], and 2) a qualitative comparsion of the putative
OF vowel with other definite cases of stressed OF and (H)AVE. I'm not
suggesting all analysis requires spectrograms, but where there is doubt these
methods can help. There may even be subtle differences between unstressed ''ve'
and 'of' in production (cf. Lavoie's instrumental acoustic work on the
differential reduction patterns of phonologically identical 'for' and 'four' in
AmEng, Journal of the IPA 32(2), 2002: 175-202), though reanalysis suggests any
effects are too subtle/ variable for many listeners to

Arguably, the General American realisation of 'of' (with a back unrounded vowel)
is easier to associate with schwa than the British and Southern Hemisphere
English LOT vowel. This doesn't seem to have affected the possibility of
reanalysis for naive native language users, but it might bear on how salient the
effect is to those listening out for it in an auditory impressionistic analysis,
i.e. non-North American forms may be easier to identify as definite cases of
stressed OF, cf. comment on 'long schwa' above.

I hadn't intended this summary to become a pro-phonetics rant, but the phonetic
forms underlie the rest of the analysis (both by linguists and language users),
so we must pay close attention to them and to influences upon them.


Thanks (alphabetically) to:

Don, Joan Beal, Alex Bellem-Hussein, Theresa Biberauer, Karen Chung, Aidan
Coveney, Michael Covington, David Denison, David Eddington, William Edmondson,
Marc Fryd, Rurik Greenall, Elizabeth Hogbin, Patrick Honeybone, Steven Hartman
Keiser, Paul Kerswill, Evan J. Kidd, Andrew McIntyre, Lise Menn, Bruce Morén,
Tim Nisbet, Cinnamon Nolan, Michael Putnam, Karl Rein, Larrry Rosenwald, Charley
Rowe, Nicole Russell, Ann Sawyer, Karen Stanley, Tim Stowell, Michael T. Swan,
Richard Terezopoulos, Rob Truswell, David Tugwell, Kevin Watson.

Mark J. Jones
Department of Linguistics
University of Cambridge

Linguistic Field(s): General Linguistics
Respond to list|Read more issues|LINGUIST home page|Top of issue

Please report any bad links or misclassified data

LINGUIST Homepage | Read LINGUIST | Contact us

NSF Logo

While the LINGUIST List makes every effort to ensure the linguistic relevance of sites listed
on its pages, it cannot vouch for their contents.