Coherence and Cohesion in English Discourse

No. of words
in all texts 
57,819
Non-native 
speakers 
77,533
Native
speakers 
Types of conjuncts
No.
Norm. frequency
No.
Norm. frequency
accordingly
3
0.04
as a consequence
1
0.01
as a result
1
0.02
2
0.03
consequently
6
0.10
3
0.04
hence
4
0.07
4
0.05
now
1
0.01
of course
2
0.03
(and) so
5
0.09
19
0.25
so that
5
0.09
6
0.08
therefore
44
0.76
59
0.76
thus 
61
1.06
36
0.46
then
16
0.28
39
0.50
for this reason
1
0.02
in that case
1
0.02
in this case
5
0.09
2
0.03
in this respect
4
0.07
in this way
1
0.02
1
0.01
TOTAL
154
2.66
178
2.30
Table 2-4: Resultive/inferential conjuncts in both corpora
Drawing on the results given in Table 2-4, it can now be postulated that some 
resultive/inferential conjuncts are very frequent in both corpora, in particular 
therefore 
(with a frequency rate of 0.76 in both corpora) and 
thus
(with the 
frequency 1.06 and 0.46 in the NNSC and NSC respectively), which means 
that these conjuncts represent more than two thirds of all resultive/inferential 
conjuncts in the NNSC and more than half of all those found in the NSC. It is 
also worth noting here that 
therefore
and 
thus 
(both shown when used in the 
NNSC in Example (10) below), along with contrastive/concessive 
however
and 
appositive 
for example
, also illustrated below, have been found in Biber et al. 
(1999: 885) to represent the most typical conjuncts in all academic texts written 
by native speakers of English; these findings are in agreement with the results 
drawn from my analysis, although the appositive conjunct 
e.g.
and the resultive/
inferential 
then
have also been found in my data to be very frequent in the texts 
written by native speakers of English and, in addition, the listing conjuncts 
i.e
.,
 
e.g. 
and
 namely
occur with a noteworthy frequency in the texts written by Czech 
speakers of English. 

---

48
(10) NNSC, Text 5
As it follows from the above mentioned, the originator or producer of 
newspaper discourse cannot be viewed as an individual; therefore, we 
can hardly speak of the ‘sender’ and his or her intentions, which we 
would normally consider in spoken interaction, for example. In our 
view, it is not possible to compare the ‘communication’ that takes place 
in 
newspaper discourse between the ‘writer’ and ‘reader’, as we would 
analyze it in face-to-face conversation because with newspaper discourse 
the 
negotiation of meaning is excluded (for negotiation of meaning in 
face-to-face conversation, cf. Povolná 2009). The traditional sender/
receiver model is thus insufficient for news discourse analysis and, as 
Scollon (1998) suggests, should be abandoned. The terms ‘writer’ and 
‘reader’ need to be understood as general concepts, which do not denote