Phraseology and Culture in English

part of a larger variable unit. The kind of phrase structures outlined by 
Kennedy (1996) have been explored to some extent both in corpus-based 
research (Aijmer 1996) and in the area of pragmatics (Sadock 1974; Brown 
and Levinson 1987). They are often discussed in a speech act framework 
for analysis and related to particular illocutionary functions in discourse. 
The second type of clustering is the seemingly random accumulation of 
different modal items and instances of vague language in certain types of 
discourse. This has also received some attention (Channel 1994). An example 
of this kind of clustering can be found in the following extract, which has 

---

260
Svenja Adolphs
been taken from the five million word CANCODE corpus, a corpus of spo-
ken discourse that will be described in more detail in the next section: 
Do you have any hopes and fears for the future for yourself 
Em. As you get older perhaps maybe about old age. I do I think
About growing old or 
Yeah. 
 you know not necessarily but you may not but I’m just saying do
you 
This type of clustering is extremely common in spoken discourse and acts 
as a means of hedging opinions and different kinds of speech acts. 
The third type of clustering is explored by Möllering (2001) and also 
Hoye (1997) and deals with relatively stable combination of a limited num-
ber of consecutively occurring modal items that leads to a particular mean-
ing. It is this type of idiomatic modality clustering that will be the focus of 
the present study. In order to analyse such clusters, the study draws on a 
five-million word corpus of spoken English, outlined in the following section. 
4. The CANCODE corpus 
CANCODE stands for the Cambridge and Nottingham Corpus of Discourse 
in English and is a collaborative project between the University of Notting-
ham and Cambridge University Press, with whom sole copyright resides. 
The corpus was assembled in the 1990s and consists of five different con-
textual categories, which are based on the relationship that holds between 
the speakers. The speaker relationships constitute a cline of proximity and 
distance between the participants in the recorded interactions (for a detailed 
description, see McCarthy 1998). The categories have thus been labelled as 
follows: intimate, socio-cultural, professional, transactional and pedagogic. 
The intimate category accounts for conversations between close family or