Cohesive substitution in English and German 225
comparison with the functionally similar occurrences of ellipsis would be neces-
sary in order to check whether it is the linguistic function of substitution or rather
the conceptual relation of type reference/co-denotation that plays a minor role for
the cohesiveness of the texts under investigation.
The second observation is that the figures for the two English corpora are
higher than those for the German corpora. Comparisons between the English
originals and their German translations suggest that this difference partially stems
from realizations of the same conceptual relation in German by other cohesive
devices. Again, a fine-grained study of the occurrences of ellipsis as well as their
specific functions would be required to obtain a comprehensive picture. It may
then turn out that German texts are semantically different in their use of rela-
tions of co-denotation, and/or in the choice of linguistic realizations. In terms of
Table 4, at least the latter seems to be the case. But additionally, translations into
both languages score higher in substitution then in their original texts.
Table 4. General distribution of cohesive substitution
EO ETRANS GO GTRANS
Total number of devices 269 161
Distribution in relation to all
tokens per subcorpus
0,08% 0,05%Excluding interviews 217 244 123 153
Distribution in relation to all
tokens per subcorpus
0,07% 0,08% 0,04% 0,05%Tables 5 (English) and 6 (German) show the total numbers for different forms of
substitution and their distributions in percentages.
Table 5. English subcorpora: distribution of devices of substitution
EO ETRANS
Total
numbersPercentages Total
numbersPercentagesone 83 30,86% 62 25,41%
ones 29 10,79% 32 13,11%
the same 21 7,81% 26 10,66%
that 1 0,37% 4 1,64%
those 34 12,64% 38 15,57%
do 69 25,65% 41 16,80%
do so 25 9,29% 35 14,34%
so 7 2,60% 6 2,46%