Includes very interesting history of training in Japan (a system that is agency-run with graduates tied to the agency and yet a system that is very successful – imagine suggesting that in Europe!) and a summary in English of Michaela Albl-Mikasa’s very important PhD on Note-taking as an inter-language.
To Know How to Suggest …: Approaches to Teaching Conference Interpreting
Dörte Andres & Martina Behr (eds.)
Great book for interpreter trainers, or anyone looking for a basic introduction to the history of interpreter training and research.
In the following extract from her essay, Intonation in the production of and perception of simultaneous interpretation, Miriam Shlesinger describes a small experiment which suggests very clearly that unnatural intonation from an interpreter can impair the listener’s understanding of a given passage.
The moral of the story… make an effort to speak normally and with normal intonation patterns when in the booth. Record yourself onto a dictaphone to check how you’re doing.
The extract is taken from…Intonation in the production of and perception of simultaneous interpretation. Shlesinger, Miriam. In Lambert and Moser-Mercer (Eds.). Bridging the Gap. Empirical Research in Simultaneous Interpretation. 1994 Benjamins. Schlesinger, Miriam.
2. Procedure and Apparatus
Ten excerpts were taken at random from the recorded, real-time output of professional interpreters in actual conference settings. Each excerpt was approximately 90 seconds long (15-20 typed lines). Six were in Hebrew (interpreted from English); four were in English (interpreted from Hebrew). Two of the English excerpts were the output of non-native speakers, whose Hebrew (native-language) interpreting output is also included in the corpus; i.e., the ten passages were the output of eight professional interpreters.
2.1. Production: Isolating the Salient Features
Repeated listening to the interpreted output revealed a set of features which seemed to differ markedly from those of spontaneous discourse. To validate the intuitive judgements on intonation in the interpreted passages, elicitation of the same passage by the same speakers in a non-interpreting setting was called for; i.e., ideally, it would have been useful to obtain the same texts from the same informants. Since it was not feasible to elicit these in the form of spontaneous speech, a second-best option was chosen: the passages were transcribed, and each interpreter was asked to read “his/her” passage aloud. It should be noted that a minimum of three years had elapsed between the actual interpretation and the experiment in Question; only two of the ten interpreters recognized the passage – and this only as a result of marked lexical cues. Thus, it may safely be assumed that that the reading, for all intents and purposes, was prima vista, was in no way expected to reveal those features which typify intonational interpretation.
2.2. Perception: Effect of Interpretational Intonation on Comprehension and on Recall
Once the features characteristic of simultaneous interpreting production had been pinpointed, their effect on perception could be assessed. Towards this end, recordings of three passages were played to two groups of subjects (eight in Group A and seven in Group B), matched for fluency in the languages concerned, and for familiarity with interpreting.
Both groups of judges listened to recordings of the same three passages. Group A listened to the interpreted version, whereas Group B heard the readaloud one. Since the text, the speaker (interpreter/reader) and the format (a recording) were identical for each passage, it was assumed that the sole distinguishing factor would be the intonation, and that by comparing the test results of the two groups, it would be possible to establish the differential effect of interpretational intonation.
After each recording had been played, three questions on each of the three passages were presented in written form to each of the two groups of judges. (The same nine questions were presented to both groups.) The nine questions were meant to determine the effect of the mode of delivery. The passages did not lend themselves to inferential questions, but rather to a straightforward test of recall and comprehension, given that they averaged no more than fifteen lines and were presented out of context.
The salient features of interpretational intonation, described in the analysis below, have been grouped into four broad categories: (1) tonality – the distribution of an utterance into distinctive contours, or information units; (2) tonicity – which syllable carries maximum pitch prominence in the tone group; (3) tone – the means of marking the opposition between certain and uncertain polarity. If polarity is certain, the pitch of the tonic falls; if uncertain, it rises (Halliday 1967); (4) prosody – duration and speed.
All but one of the features described below – acceleration – was found in the output of at least four different interpreters; i.e., only one of the three features was possibly idiosyncratic. Moreover, although languages (and: even dialects) do differ in various aspects of intonation (Ladd 1990; Cutler 1987; Rameh > 1985), the features which came to the fore in the present study’ seemed to cut across (at least these two) languages.
The data below refer to occurrences of the given feature in the interpreted passage only; i.e., whenever the feature appeared in the read-aloud mode as well, it was not counted. It should be noted that each of the interpreted texts in the corpus was longer in duration than its read-aloud counterpart; i.e., the interpretation was slower (took more time) than the reading.
Passage 2 and passage 6 were produced by interpreters who, although working into English, are native speakers of Hebrew. The same two interpreters also produced the Hebrew passages 1 and 7, respectively. The data below do not reveal a significant pattern of differences between the two passages in the case of either interpreter, though further study is needed to determine whether directionality does in fact play a role in interpretational intonation.
3.1.1. Tonality (Chunking, Pauses, Division into Tone Groups). The range of grammatical structures which pauses tend to separate is relatively small and constant (Crystal 1969). Functional pauses serve to divide discourse into tone groups and organize it into information units (Halliday 1985; El-Menoufy 1988). Nonfunctional pauses caused by hesitations, on the other hand, tend to lower the congruence between chunking and syntax, since the ensuing junctures are nongrammatical.
The data below would seem to indicate that pauses within grammatical structures are by far the most salient feature of tonality in interpretation; i.e., interpreters are prone to introduce a disproportionate number of pauses in “unnatural” positions, which are liable to impede understanding (Alexieva 1987).
As for pauses at the clause or sentence boundary, while the interpreted passages did generally include pauses at sentence boundaries (in conformity with the original), these tended to be tentative rather than final, and were often coextensive with a low-rise intonation (pitch movement 3). Tentative pauses are frequently used in the middle of a primary contour, usually serving a parenthetic function (Ladd 1975; Pike 1945); thus, a high incidence of such pauses in nonparenthetical position is anomalous. Moreover, since the tentative pause correlates with an attitude of uncertainty, the cumulative pragmatic effect is bound to be altered. As in other instances of introducing an element which runs counter to the listener’s expectations, understanding is likely to be affected.
A test was devised to determine the extent of comprehension and recall by each of the two groups of subjects. The same test was administered to both groups. It consisted of three questions on each of three passages (a total of nine questions). The total number of correct answers in Group A (8 subjects) was 17 out of a possible 72 (21 %) and in Group B (7 subjects) 26 out of a possible 63 (41 %). The difference between the proportions of correct answers in the sample was 20%; i.e., the proportion of correct answers in the second group was approximately double that of the first.