Research Report 14 - Deaf jurors' access to court proceedings via sign language interpreting: an investigation (by Jemina Napier, David Spencer and Joseph Sabolcec)

A project funded by the NSW Law Reform Commission and
the Macquarie University External Collaborative Grant Scheme.

6.1 In this section, we discuss the analysis of the translation and the interpretation of the comprehension test results.

6.2 The purpose of analysing the translation was to assess the degree to which important concepts in the original summation were translated. In doing so, a note of caution must be raised. The usual method for such an analysis, as used here, is to conduct a back translation from the target language (Auslan) to the source language (English) in order to see the degree to which there has been any shift in meaning. Although conducted by a qualified Auslan/English interpreter, there is still the potential for meaning to shift during the back translation process. It could be argued that even the reading of the original English text will introduce the potential for something to be misread or emphasis to be altered from that originally intended by the author.

6.3 A full transcript of the analysis can be seen in Appendix H. The legal concepts identified in the original text will now be examined.

6.4 The issue of the difference between murder and manslaughter, in particular, the mens rea or intention to commit a crime, was accurately translated. However, the “reasonable person” test was not well translated. The summation clearly directed the jury to consider “the measure of responsibility being that of a reasonable person stepping off a train at Redfern”. The back translation stated, “You can however make a decision about his words, language, his behaviour”. Although, the use of the word “objectively” was used in the very next sentence, it is uncertain if the average juror would understand the term “objectively” without the use of the example of the reasonable person. This appears to be a problem with this part of the translation.

6.5 No problems were identified with the translation of the first element of manslaughter. However, in relation to the second element of manslaughter, a subtle difference was identified between the summation referring to acts that “caused the death” of the victim compared with the back translation which refers to acts that “contributed to the death” of the victim. An act or acts that contribute to someone’s death is different to an act or acts that cause someone’s death. The accused may have contributed to the death but his act may not have caused the death.

6.6 A major problem was identified in relation to the translation of this concept. The summation clearly left the interpretation of the evidence to the jury:

6.7 The back translation was found to be better, in terms of plain English, than the summation. A minor comment could be made that the summation asks the jurors to consider first if there was a threat at all:

Key legal concept No 5 – clarifying the written directions: what did the accused do?

6.8 The summation leaves the question of what the accused did to the jury. The judge goes close to commenting on the truth of the evidence, but ultimately leaves it to the jury by opining, “ I do not think it is in the least controversial…”, whereas the back translation states on this issue, “All would agree that the victim was directly targeted…” This important difference is the difference in leaving facts for the jury to determine and the judge directing the jury on accepted facts – accepted by him and, by implication, the jury.

Key legal concept No 6 – clarifying the written directions: did the accused’s actions give rise to a well-founded apprehension of physical harm?

6.9 The order of the translation was found to be problematic. The following sentence:

Key legal concept No 7 – clarifying the written directions: was it reasonable for the victim to seek to escape?

6.10 The back translation used the word “causally”, which was assessed to be incorrect. This was not thought to be hugely damaging, but confusing, as the issue of well-founded apprehension or fear is not determinative alone of the issue of causation – it merely contributes to the issue of the presence of causation. Otherwise, there appear to be no other problems with the translation.

Key legal concept No 8 – clarifying the written directions: was the method of escape reasonable and proportionate?

6.11 A reference to the victim escaping because of a “weakness in character” was deemed to be an incorrect insertion, as it was not mentioned in the original summation. This is an objective test, not an assessment of the victim’s character as to whether it was reasonable to escape the threat posed by the accused. Otherwise, no other problems were evident.

6.12 From the analysis of the back translation, it can be seen that there were some problems with translations of the broad legal concepts central to the case. Yet, the problems were largely subtle shifts between legal definitions or objectively presented facts, to interpretations of the meaning of such statements. As stated in paragraph 5.28, the overall accuracy of the interpretation was 87.5%, which is relatively high. Given that candidates who sit the NAATI Interpreter Level test are required to achieve a pass mark of 70% for successful accreditation and to be considered safe to practice, 87.5% accuracy is more than acceptable.

6.13 An overall pattern that can be seen in the responses to the comprehension test is the difference between responses to questions of fact and questions relating to legal concepts. Overall, most respondents answered questions of fact correctly. In the case of deaf respondents, this means that the facts of the case had been interpreted clearly and correctly and had been understood. When asked to comment on the comprehension test, four participants specifically mentioned the facts of the case as being one of the easier aspects of the activity.

6.14 When factual errors did arise, they sometimes arose in respondents who otherwise provided correct answers to more complex questions. An example of this can be seen in Question 9. All the deaf respondents provided correct answers as to the accused being physically restrained by the women, while only four of the hearing respondents provided a correct answer. One hearing respondent answered the question correctly but drew a conclusion as to the intentions of the accused to free himself from the grasp of the women. First, this response was not called for by the activity but offered up by the respondent and, secondly, it is inconclusive in terms of the accuracy of the conclusion drawn. One hearing respondent failed to answer the question and guessed incorrectly, yet this respondent had previously answered all of the questions about the legal concepts and the factual matrix correctly.

6.15 A similar pattern can be found with Question 11, where one deaf respondent answered with:

6.16 A possible explanation for this observation is that the facts of the case were sparsely distributed in the body of the text, often arising incidentally within a discussion about a legal concept. These facts were also rarely repeated during the excerpts selected for this study, therefore making them easier to miss. This suggests that in an actual trial, where evidence is presented over a longer period of time and in a more systematic manner, these comprehension errors may decrease as jurors would have time to absorb evidence and arguments before hearing a judge’s summation.

6.17 Discussion will now turn to responses to questions where a pattern of misunderstanding may be found. As already suggested in paragraph 5.28 - 5.57, the answers to Questions 4, 5, 10 and 12 were problematic, and focus on the legal concepts raised in the summation.

6.18 Five out of six hearing respondents answered Question 4 correctly, with one additional respondent giving the correct answer, but being confused by the reference to murder, stating that the summation had been addressing the issue of manslaughter. Only three out of the six deaf respondents answered this question correctly, however, with one of the respondents who answered correctly admitting that he or she did not understand the interpretation of the phrase “beyond reasonable doubt”. When the researcher simplified this to “without doubt”, the respondent confirmed the correct answer. Overall, this may indicate a low level understanding of a basic threshold concept in criminal trials. If respondents cannot grasp this basic threshold concept, then the rest of the evidence may well be lost or misinterpreted by the jury, whether hearing or deaf, in the jury room. However, the level of misunderstanding is comparable between the two sample groups, meaning that the concept or the form of the question was difficult for both groups. This issue will be discussed further below.

6.19 Turning to responses to Question 5, only one of the hearing respondents answered this question incorrectly, compared to three of the deaf respondents. This response is a concern, as this question is a threshold question that distinguishes between the two elements of murder, that is, the mens rea (the intention to kill), and the actus reus (the act of killing). The prosecution has the burden of proof to prove beyond a reasonable doubt that the accused intended to kill the victim and did kill the victim. If deaf jurors have difficulty understanding the difference between the act and the intention to kill, and that lack of understanding is allowed ultimately to determine the outcome of the case in the mind of that juror, then unsafe verdicts are possible.

6.20 It should also be noted that Question 5 was also the longest of the questions asked of the participants. This raises the challenge of modality in the design of comprehension materials.¹⁰⁸ In this study, this becomes an issue not only for participants responding to complex verbal information, but also signed multiple choice items which must also be held in memory as the participant assesses the question, the options and his or her recollections.

6.21 Questions 10 and 12 resulted in the highest number of errors. All hearing respondents answered Question 10 incorrectly. Most were close to the correct answer, but none of the hearing respondents stated the rule correctly. Whether this is fatal to their ultimate understanding of the concept and would lead to unsafe acquittal or conviction is hard to tell. The following elements of their responses show how close the responses were to the correct answer:

6.24 All of the deaf respondents answered the question incorrectly. The closest deaf respondent stated, “we need to think about what is fair and proportionate”. The respondent in question did not go on to explain what he or she meant by “fair” and “proportionate”. We may assume that “proportionate” meant whether the victim’s response was proportionate to the threat by the accused. However, “fair” probably does not mean invoking the objective test. It is too easy for a respondent not to understand the difference between the subjective test (“fair” as judged by the victim’s demeanour and “fair” as judged by the reasonable person). It is the lack of adequate explanation that makes even this response an incorrect answer. Only two of the deaf respondents raised the element of reasonableness. The other three deaf respondents totally missed the point and talked about the facts, as did the two above-mentioned hearing respondents.

6.25 Again, this is a complex legal concept that is difficult to grasp, but an important concept that effectively proves the causation issue so central to a successful manslaughter conviction. The high level of misunderstanding by both hearing and deaf respondents is a concern. Referring to the comments made by participants at the end of the comprehension test may provide some further illumination of the above results.

6.26 When asked what they found difficult about the activity, six respondents (four hearing, two deaf) responded with reference to the legal language used, the ‘wordy’ explanation, or the apparent repetition within the excerpts used in the study, for example, “there seemed to be a lot of repeating so you almost get lost in it”. This suggests a level of fatigue may have played a part during the viewing of the selected excerpts. One deaf participant noted:

6.28 When asked if they would be interested in being involved in a jury after the experience gained in this study, all deaf participants responded positively while noting the negatives. Three participants added that they saw jury participation as the responsibility of all citizens and that deaf people should not be treated any differently:

Table of contents