The effects of input rate on translation approaches

How the input rate affects the translation approach of interpreters is a rarely investigated issue. To our knowledge, only two empirical studies examined the effects of input rate on interpreters’ choice of translation approach and both of them used the method of product

46

analysis. Dam (2001) set out to test the hypothesis that a more difficult speech leads to a more form-based strategy. A group of student interpreters who had just finished training and passed an exit exam was asked to perform SI from Spanish into Danish on two speeches with different levels of difficulty. The difficulty levels were measured by the use of technical terms, numbers, average sentence/clause lengths, and input rates. However, there was only a small difference between the input rates of the two source speeches – 119 wpm vs. 125 wpm. Dam (2001) conducted a comparative analysis on the degree of formal equivalence at the lexical level between the source and target language segments. Form-based interpreting was operationalized as “lexical similarity” between the source and target segments and meaning-based interpreting as “lexical dissimilarity” (p. 35). Target text segments were

divided into five categories: similar, similar (dissimilar), similar/dissimilar, dissimilar (similar) and dissimilar, representing the degree of formal equivalence from high to low. Contrary to the assumption of interpreting researchers, results showed that there was a higher proportion of form-based interpreting in the less difficult speech (the slow speech) and a higher

proportion of meaning-based approach in the more difficult text (the fast speech). Dam (2001) concluded that the more difficult the source speech, “the more interpreters tend to deviate from the form-based approach and move towards the meaning-based approach” (p.50). This peculiar finding might stem from the fact that “additions” in the output – information not given in the source speech – were considered a part of lexical dissimilarity; that is, as the meaning-based approach. If interpreters made more additions in the more difficult text, resulting in more lexical dissimilarity, they could deviate not only from the form but also from the “meaning” of the source speech. In this case, it is paradoxical to call it the

“meaning-based” approach. Given this doubt, the jury is still out on whether more difficult source texts or faster speeches lead to more meaning-based interpreting.

Zhang (2010) investigated the effects of input rate and source text difficulty on the deverbalization of student interpreters when performing consecutive interpreting with notes

47

from English to Chinese. Deverbalization was examined at the syntactic and discourse levels, respectively defined as the “absence of source text syntactic interference on target language production” and “integration of contextual knowledge into the discourse representation” (p.

iii). Participants were asked to interpret eight English texts under four conditions (easy slow, easy fast, difficult slow, difficult fast). The input rates of 110 wpm and 150 wpm were used for slow and fast speeches. The difficulty level of the source text was manipulated with reference to the scores of Flesch-Kincaid Readability Index. Comparative analysis between the source text and the target text was conducted on three types of experimental sentences embedded in the source texts: sentences with adverbial phrases, sentences with adverbial clauses, and sentences with discourse-inappropriate words. For the first two types of experimental sentences, the main clause was always placed before an adverbial phrase or clause. To interpret English sentences with the structure of “main clause + adverbial phrase or clause” into Chinese, the participants had to change the order into “adverbial phrase or clause + main clause” so as to produce grammatical Chinese sentences. The proportions of ApM (adverbial phrase + main clause) and AcM (adverbial clause + main clause) in Chinese target texts were used as one of the dependent variables, representing the degree of deverbalization at the syntactic level. At the discourse level, when the participants changed

discourse-inappropriate (DI) words in English texts into discourse-appropriate (DA) ones in the Chinese texts, it was considered a manifestation of deverbalization. It was hypothesized that a higher proportion of deverbalization at both syntactic and discourse levels would be observed in the slower, less difficult texts. Results showed that this was indeed the case.

When the degree of deverbalization was analyzed in general as well as at separate levels of syntax and discourse, the main effects of source text difficulty and input rate were significant.

The opposite results observed in these two studies may be attributed to the differences in the level of analysis, mode of interpreting and measurements of source text difficulty. Dam (2001) focused on the lexical level in SI, whereas Zhang (2010) examined the syntactic and

48

discourse levels in consecutive interpreting. There was a small difference of input rates in Dam’s (2001) study, but in Zhang (2010), the difference between the fast and the slow speech was 40 wpm. Since the present study is aimed at exploring the effects of input rate on

linguistic interference at the syntactic level, it remains to be seen if the findings of Zhang (2010) can be generalized to the mode of SI.

Based on the findings of studies reviewed in the above sections, input rate can affect the accuracy and fluency of simultaneous interpreters’ output. However, for the effects of input rate on each parameter of accuracy and fluency as well as on EVS, the output rate and the translation approach of interpreters, results were divergent across studies. This divergence painted a rather complex picture of the influence of input rate on interpreting performance, which merits more investigation. Moreover, due to small sample size, inferential statistics were not employed by many studies, making it difficult to compare and generalize the findings. Therefore, the present study is intended to reexamine the influence of input rate on several linguistic and temporal parameters that have been adopted by previous studies and add one new parameter – lexical diversity. It is hoped that further understanding of the role of input rate can be gleaned by taking a more comprehensive approach in examining different aspects of the interpreting output and by comparing three different speeds.

49

Chapter Three Method

This chapter is divided into six sections. The first section presents an overview of the study. The second section outlines the research design. The third section provides background information of the participants. The fourth section describes the selection criteria and the design of input materials. The fifth section explains the procedures of the SI experiment. The last section illustrates methods of data scoring and analysis.