Phraseology and Culture in English

Varieties of English around the world:
Collocational and cultural profiles
1
Christian Mair 
1. Phraseology: The “blind spot” in variety identification? 
How do we recognise varieties of English? For spoken texts, the answer to 
this question is easy: they are almost always identified, reliably and within 
a very short time, on the basis of accent features. For written texts, the an-
swer is more difficult. The number of clear orthographic or lexico-gram-
matical identifying features is small, and many of them tend to be so rare 
that we cannot count on their presence in a short passage of written text. 
What is even more problematical, however, is that most relevant lexico-
grammatical features are variety-specific only in a statistical sense. What is 
important is not the presence or absence of a particular feature as such, but 
rather its cumulative frequency in large masses of discourse, or its relative 
importance with regard to alternative variants. It goes without saying that 
such lexico-grammatical profiles in written texts are usually much less sali-
ent than accents in the spoken language. 
A specific number of pronunciation features may be said to bundle 
into an accent, and in a synchronic description of phonetic variability in 
English it is a permissible simplification to say that the language comes 
with a large number of regional and social accents which can be described 
in relative isolation from each other. This is not a fruitful approach to the 
study of variability in written English, however. A variety of written Eng-
lish cannot be defined by cataloguing a finite list of lexico-grammatical 
features which, as it were, provide the necessary and sufficient conditions 
for identifying it, and no variety of written English can be described in 
isolation from a common Standard English core. The identification of the 
regional origins of a text will by its very nature remain probabilistic and 
therefore provisional. Typically, a text of some length will present a fair-
ly characteristic statistical profile displaying variety-specific preferences 
in those areas in which present-day Standard English allows choices. 
These profiles may not always be salient to the reader but may enable
the linguist to assign a text to a certain variety with some degree of like-

---

438
Christian Mair 
lihood. Some texts, especially at the formal end of the stylistic spectrum, 
may be so neutral as to make it impossible to detect their origin from 
linguistic criteria alone, and only very few will contain lexico-grammati-
cal features which positively identify certain origins (or at least exclude 
others).
What is the implication of this state of affairs for the widespread prac-
tice of classifying world Standard English into various regional sub-varieties 
(British English, Irish English, North American English, Australian Eng-
lish, etc. – not to speak of more problematical fine-grained subdivisions such 
as Canadian English)? It works tolerably well for spoken English, but it is 
clearly inadequate for dealing with variation in written English. Varieties of 
written Standard English should not be seen as decontextualised systems 
which can be described by cataloguing their lexico-grammatical features. 
On the structural level, there is one written standard with minimal gram-
matical and very modest lexical variability. The best way to get a grip on 
what it means for a written text to be “American” or “British” (or represen-
tative of any other kind of more recent or emerging standard) is to shift the 
analysis from the level of abstract decontextualised structure to that of style 
and discourse and to describe the lexico-grammatical choices writers make 
from the range of options provided in the system in particular contexts and 
for particular purposes. 
As M.A.K. Halliday has put it in a recent essay on English as a “written 
language, standard language, global language,” the most appropriate ap-
proach to studying lexical variability in world English is to describe: 
1. not just new words, but new word-making principles; 
2. not just new words, but new word clusters (lexical sets); 
3. not just new words, but new meanings; 
4. not just new words, but new registers (functional varieties). (2003: 408) 
This shift of emphasis from isolated words as static “products” in a decon-
textualised lexicon to the process of deploying (and combining) words in 
context-embedded discourses will lead to a richer linguistic description 
because it focusses on: 
[…] ways of opening up, of expanding the semiotic potential that inheres in 
every language: opening up the creation of new terms; opening up the di-
mensions along which these terms are organized, opening up the meaning-
making resources of the lexicogrammar; opening up the modes of creating 
and transmitting knowledge, maintaining and strengthening authority. (2003: 
409)

---

Varieties of English around the world
439
Explaining variability in written English in a contextually sensitive us-
age- and performance-based framework has several advantages. First, it 
helps us get around the obvious problem that writers do not consistently 
use the features supposedly “typical” or “characteristic” of their variety 
but also use alternative forms. This is true even for British and American 
texts, that is the two varieties with the longest history of standardisation 
behind them,
2
and it is glaringly obvious in the case of the more recent 
emerging standards of, say, India and the Caribbean, where the lexico-
grammatical regionalisms almost always co-exist with British or Ameri-
can expressions (and are commonly used in a minority of instances). Sec-
ond, the discourse-based model of variability allows us to treat phrase-
ological phenomena not as exceptions and unsystematic aberrations in a 
clean and regular grammar, which they frequently appear to be on the 
level of decontextualised linguistic structure, but in a way which recog-
nises their importance in differentiating varieties of written English. 
Third, the “Britishness”, “Americanness”, etc. of a text can be appropri-
ately modelled as a matter of degree. This is necessary because the vari-
ety and frequency of lexico-grammatical regionalisms increases in infor-
mal genres, which means that informal British or American texts will 
almost always read more British or more American than their formal 
counterparts.
Last but not least, a discourse- and performance-based model of variabil-
ity in written English will also be a culturally sensitive one – an extension 
which has prompted the formulation “collocational and cultural profiles” in 
the title of the present contribution and, it is hoped, justifies its inclusion in 
the present volume. After all, idiomatic and collocational preferences are 
the most direct reflection of a community’s attitudes and pre-occupations in 
linguistic structure. 
The present contribution is organised as follows. After a review of per-
tinent scholarship (Section 2), I will present first findings from an experi-
ment in which the World Wide Web was used to establish collocational 
profiles of Britishness (Section 3). Section 4 will address a question which 
is important particularly for the study of the New Englishes, namely 
whether it is possible to identify collocational correlates of naturalness and 
idiomaticity, which would enable us to at least in part assess the degree of 
institutionalisation and/or autonomy of a particular New English on the 
basis of internal linguistic criteria rather than external sociological ones
(as has been done traditionally). Section 5 offers a brief summary and con-
clusion.

---

440
Christian Mair