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
first
3
0.05
4
0.05
finally
17
0.29
6
0.08
then 
18
0.31
26
0.34
second
2
0.03
2
0.03
next
3
0.05
secondly
5
0.09
2
0.03
last
4
0.07
lastly
3
0.05
firstly
5
0.09
in the first place
3
0.05
TOTAL
63
1.09
40
0.52
Table 2-1a: Listing (enumerative) conjuncts in both corpora
No. of words 
in all texts 
57,819
Non-native 
speakers 
77,533
Native
speakers 
Types of 
conjuncts
No.
Norm. frequency
No.
Norm. frequency
furthermore
10
0.17
5
0.06
in addition
4
0.07
10
0.13
moreover
21
0.36
2
0.03
in particular
5
0.09
3
0.04
similarly
12
0.21
14
0.18
also
5
0.09
2
0.03
further
2
0.03
additionally
1
0.01
too
1
0.01
TOTAL
59
1.02
38
0.49
Table 2-1b: Listing (additive) conjuncts in both corpora

---

40
Since listing conjuncts as a whole clearly prevail in the NNSC, where they 
have the average frequency of occurrence 2.11, in comparison with 1.01 in 
the NSC (cf. also Table 2-1 above), the following example illustrates several 
tokens of both enumerative (namely 
first, second
and three tokens of 
finally
) 
and additive conjuncts (i.e
. moreover, similarly
) typically applied in non-native 
speakers’ academic writing:
(3) NNSC, Text 2A
The length tendencies become even more apparent if we trace the 
number of tokens. Here again, we decided to scrutinize the two extremes 
of a cline. 
First
, we focussed on the number of tokens of the longest 
paragraphs (exceeding 10 C). Second, we surveyed the number of single-
C paragraphs. Paragraphs exceeding ten C turned out to be extremely 
rare in both the journalism subcorpora (one paragraph in the newspaper 
subcorpus out of 345 and six paragraphs in the magazine subcorpus). 
They grew in significance in children’s fiction and natural sciences (11 
instances each). However, the longest paragraphs turned out to be rather 
common in humanities (48 tokens) and in adult fiction (60 instances). …
Finally
, we investigate the average length values. The mean paragraph 
length in the overall research corpus corresponds to 4.49 C. Examining 
individual style corpora shows some significant tendencies. The academic 
corpus typically featured rather long paragraphs (5.70 C per paragraph 
on average). It was followed by the fiction corpus with an average 4.39 C. 
Finally
, journalism paragraphs tended to be the shortest of all (3.37 C).
Moreover
, we have noticed some striking differences even among the 
subcorpora examined. The mean length value of the humanities subcorpus 
(6.36 C per paragraph) exceeded its natural sciences counterpart (5.04 
C). 
Similarly
, the magazine subcorpus displayed on average longer 
paragraphs (4.31 C) than its newspaper counterpart (2.42 C). 
Finally, 
with its 5.29 C to a paragraph, the adult fiction subcorpus prevailed in 
length over the children’s fiction counterpart (3.49 C).
Example (3) provides evidence that the semantic category of listing conjuncts 
is often used in order to help the reader(s) understand the path through a text full of 
mostly quantitative information and thus help them interpret the text as a coherent 
piece of discourse. As for other voices entering the text, it is important to mention 
here that the text in the above example has just one author although she uses the 
personal pronoun 
we
when referring to herself, i.e. to the single author of the text. 
Probably this happens under the influence of the author’s mother tongue, since 
in the Czech academic tradition it is quite common to use the authorial plural 
pronoun 
we
even when the RA is single-authored (cf. Dontcheva-Navratilova 
2012).

---

41
The most frequent listing conjunct of all in the NSC (0.34) and the second 
most typical in the NNSC (0.31) – the enumerative conjunct 
then 
– is illustrated 
in Example (4). Although 
then
tends to be rather informal, the author applies it 
relatively often, for example when describing editorial procedure. Its common 
use is probably connected with a slightly more informal style which usually 
characterizes native speakers’ academic writing.
(4) NSC, Text 5
To begin, it may be useful to summarize briefly the whole editing procedure, 
insofar as it concerns the English for Specific Purposes Journal. It is as 
follows. When a manuscript is received, it is prepared for blind review 
by having the name(s) of the author(s) removed, plus occasionally any 
references that may give strong clues about the author’s identity. The 
editor 
then
 selects two reviewers, usually members of the editorial board 
who have expressed interest in the subject area, but also occasionally 
people from outside the board with a particular expertise or interest in 
the topic of the manuscript. The ‘blinded’ manuscript is 
then
 sent out to 
the reviewers, who in turn send back their reviews within a specified time 
period-usually two to three months. The editor will 
then
 read through the 
manuscript and the reports and write the editorial letter, the main purpose 
of which is to convey the editor’s decision about the manuscript and to 
provide both a summary of the reviews and suggestions for revision of the 
manuscript.
Finally, let me point out that one token of a listing conjunct, namely the 
additive conjunct 
in addition
, is shown in Example (2) above. Since its 
normalized frequency is slightly higher in the NSC (0.13) than in the NNSC 
(0.07), it represents together with 
then
those conjuncts that are slightly more 
frequent in the NSC, all the other listing conjuncts being more typically applied 
in the NNSC (cf. Tables 2-1a and 2-1b above).