Available at

4.2.2.3 Numbers 
The following linguistic item to be analysed is numbers, with 27 occurrences in the 
source speeches and 148 in the students’ interpretations. Many studies say that 
the error rates can reach 40% to 70% for trainees (Korpal, 2016; Korpal & 
Stachowiak, 2018; Timarová, 2012) so I would expect such an error rate in this 
present study. Interpreters also use strategies such as reformulation (Jones, 2002; 

---

Results and discussion 
 
page 79 
Pinochi, 2009) or approximation (Kalina, 1992: 254) so I would expect the share of 
expanded and reduced renditions to be quite significant. 
Table 21 – Renditions for numbers 
STU01 STU02 STU03 STU04 STU05 STU06 STU07 STU08 Total 
Close 20/27 
24/27 
13/16 
13/17 
9/13 
16/23 
15/21 
4/4 
78.86% 
Div. 
3/27 
3/27 
1/16 
1/17 
2/13 
3/23 
3/21 
/ 
9.63% 
Zero 2/27 
/ 
1/16 
3/17 
2/13 
2/23 
3/21 
/ 
8.71% 
Exp. 
1/27 
/ 
1/16 
/ 
/ 
1/23 
/ 
/ 
1.72% 
Red. 1/27 
/ 
/ 
/ 
/ 
1/23 
/ 
/ 
1.01% 
Table 21 does not confirm my hypotheses. Indeed, divergent renditions only 
account for 9.63% of all renditions, and expanded renditions represent 1.72% of 
them. However, it is important to shed light on the numbers analysed in this study. 
Most of them are expressed using one or two words and do not seem to be a 
problem for interpreter trainees. 
(78) twelfth century (Item #34 – IN03) 
(79) ten laws of art (Item #55 – IN03) 
(80) four of the smartest people (Item #23 – IN04) 
Students dealt with numbers very well as more than three-quarters of the 
occurrences were correctly translated (78.86%). I expected students to use two 
strategies, namely reformulation and approximation, but it was not the case at all 
since expanded and reduced renditions respectively account for 1.72% and 1.01% 
of all renditions. It is also interesting to see that STU03 and STU04 have the 
lowest percentage of divergent renditions among all students, with 6.25% and 
5.88% of all renditions. Furthermore, STU02 (88.88%) and STU03 (81.25%) have 
the highest percentage of close renditions and, apart from STU08 (excluded of the 

---

Results and discussion 
 
page 80 
analysis as STU08 only interpreted four occurrences), they are the only two 
students to have a higher percentage of close renditions than the average for this 
linguistic item (78.86%).