Coherence and Cohesion in English Discourse

Different 
speech event 
categories
Office hours
(16,176 words)
Discussion 
sections
(15,542 words)
Study groups
(15,483 words)
Total
(47,201 words)
No. Norm. 
freq.
No. Norm. 
freq.
No. Norm. 
freq.
No. Norm. 
freq.
Reformulation
6
0.37
1
0.06
9
0.58
16
0.34
Explanation 
5
0.30
17
1.09
32
2.07
54
1.14
Delay device
3
0.19
5
0.32
27
1.74
35
0.74
Total 
14
0.87
23
1.48
68
4.39
105
2.22
Table 3-1: Pragmatic functions of 
I mean
in different 
speech event categories in 
academic spoken discourse (MICASE)
Table 3-1, which provides overall results from the three speech event 
categories studied in this chapter, indicates that the marker 
I mean
tends to be used 
most frequently in study groups, where it reaches the normalized frequency rate 
4.39, which means a frequency that is considerably higher than that in the other 
two types of speech events, since the normalized frequency rate in the categories 
of discussion sections and office hours is 1.48 and 0.87 respectively. However, 
the differences in the frequency of occurrence between different categories are 
not considered most important here, since it is assumed that they can be caused 
by differences in individual speakers’ communicative habits and preferences 
rather than differences
in speech event categories. For example, the number of 
speakers in the office hours and study groups under investigation is two to five, 
whereas in the discussion sections this is as many as 18 to 22. Of course, not all 
speakers are really active, nor do they frequently use the marker 
I mean
. Based 
on my results, it can now be concluded that the number of discourse participants 
does not influence the frequency of occurrence of 
I mean
: the frequency rate in 
discussion sections (1.48), which have the highest number of participants (18 to 
22), is somewhere between the frequency rates of office hours (0.87) and study 
groups (4.39), which have two to five and four participants respectively.

---

67
With regard to what has already been stated, it seems that what is most 
interesting here are the differences concerning possible 
pragmatic functions 
I 
mean
can perform in academic spoken discourse. My results prove that most 
speakers apply 
I mean
in particular when they intend to introduce explanations 
or clarifications, in which case the normalized frequency rate in the corpus as 
a whole is 1.14 tokens per 1,000 words, i.e. approximately 51 per cent of all 
occurrences. By using 
I mean
in this function the current speaker (cf. Speaker 5 
below) attempts to help his/her partners in a given speech event understand what 
he/she wants to communicate, as evidenced by the example that follows: 
(1) (DIS495JU119.7)
S1
media. okay so tell me how th- some of those things work. 
 
what did you learn from your parents, about politics?
 
Brett?
S5
they usually tell you, well they usually brainwash you and tell you to vote 
their, <