Abstract:

Courtroom discourse is widely studied (Cotterill, 2003; Eades, 2008; Heffer, 2005; Matoesian, 1993; Shuy, 2006) in the forensic linguistics and law and language fields. This thesis extends existing research on courtroom questioning in a new setting, that is, Malaysia’s adversarial criminal courts. The issue of how questioning in direct and cross-examination exercises neocolonial power and control over witnesses is raised in this thesis. The linguistic impact of colonialism (Eades, 2008; Powell, 2008) is therefore an important dimension of Malaysian courtroom talk. While Malaysia has a hybrid trial system, which is based on the Anglo-American system due to British colonialism, in 1995 it moved to a non-jury system with judges giving verdicts, providing an opportunity to examine continuing effects of a post-colonial context for lawyers’ discourse. The investigation focuses on the forms and discourse-pragmatic functions of questioning and responses in the Malaysian criminal courtroom and barristers’ rhetorical strategies when constructing narratives of events in front of judge(s). This thesis seeks to answer the following research questions:

1. What are the discursive practices used by barristers and witnesses in Malaysian criminal trials? and
2. What is the relationship between question types and responses?

A corpus-based forensic discourse analysis approach is used to investigate a pilot corpus (the Shipman trial) and then to investigate 16 Malaysian criminal cases (i.e., MAYCRIM). MAYCRIM is a specialised corpus with a size of 326,785 words and consists of a range of criminal offences such as murder, drug trafficking, human trafficking and others which were collected from the Sessions and High Courts of Malaysia. MAYCRIM is unique because it features Bahasa Malaysia, Malaysian English and mixed code.  

The corpus-based analysis reveals interesting patterns of lawyer questioning and witness resistance. Probing questions, that is wh-questions and indirect can you questions paired with material and verbal ‘process types’ (Halliday, 1985; Halliday & Matthiessen, 2004) maximise witnesses’ productivity, while challenge questions, such as SAY-questions and invariant tag questions, coerce through personalisation and quoting strategies that face-threaten witnesses in cross-examination. Firstly, probing questions that are used by barristers to maximise witnesses’ productivity produced two important patterns: (1) wh-prompt + you + material/mental/verbal processes (e.g., What was your purpose of entering Malaysia?) and (2) modal can + you + material/verbal processes (e.g. Can you still identify the report that you have lodged?). The wh-prompt makes the prosecution narrative and the witness more credible because it can be used to probe and control witnesses to recall and elaborate specific details of events. Moving to the second type of probing question, which is the can you indirect question, lawyers are able to present witnesses as either helpfully productive in direct or lacking knowledge and credibility in cross-examination, by pairing the can you part with verbal processes such as explain, tell, confirm, inform, elaborate. On the other hand, the SAY-question is another combative tool used by barristers in cross-examination to indicate that witnesses’ existing answers are problematic. The most common patterns of SAY-questions found in this analysis are are you saying and are you telling. The SAY-question which is embedded in yes/no and declarative questions is not only used to limit a witness’s response but as a linguistic device to imply witnesses’ answers are problematic.

The finding that invariant tag questions are dominant might suggest that Malaysian barristers are less able to perform power and control with witnesses in cross-examination. However, despite lacking polarity, I argue that invariant tag questions with do you agree, correct/betul, agree/setuju, particle tak/not, and do you know have the same potential to perform control and power as canonical tag questions. The declarative + agree/setuju with the highest affirmatory function was mostly used by lawyers to get affirmative answers from witnesses. The linguistic marker to achieve this function is via a factive sentence with tag agree/setuju that can be used as a request for agreement from the hearers. Secondly, declarative + correct/betul is used by defence lawyers to refute prosecution evidence by exposing inconsistencies and inaccuracies in witnesses’ answers. Lawyers perform the illocutionary force of checking, through a combination of reported speech or writing with the correct/betul tag, that enacts prior evidence testified or reported by witnesses. It is also found that both the invariant tags agree/setuju and correct/betul are very conducive because they make inferences based on existing evidence. Thirdly, we know that a reversed polarity tag question has a strong force for agreement and is highly coercive; however in the MAYCRIM corpus, the declarative + do you agree is also highly coercive. This is because it allows lawyers to impose their version of facts or presuppositions to coerce hearers to accept their assertion.

In response, witnesses demonstrate resistance via disagreement, correction, evasion and challenge, demonstrating that witnesses are able to overcome the power asymmetry that is particularly pronounced in cross-examination, though not without costs. Resistant answers can become a defensive strategy to mitigate lawyers’ dominant power, or can be used as a diversion strategy to disrupt lawyers’ sequences of questions. Interestingly, they can even be used to open up linguistic negotiation between lawyers and witnesses in interaction. A continuum of witnesses’ resistance is suggested for legal practitioners to understand how their questions affect witnesses and at the same time help to prepare their witnesses for courtroom examination. This continuum of witnesses’ resistance is developed from quantitative distributions and qualitative analysis of the data. The continuum suggests that when a least controlling question such as a wh-question is used, the probability of witnesses resisting is higher compared to controlling and coercive questions such as yes/no, declarative and invariant tag questions. The continuum provides a functional framework for lawyers to assess witnesses’ answers and can be a point of reference for lawyers to instruct witnesses and/or defendants on what they might encounter when they are called to testify in the courtroom.

This study makes three original contributions to theory, methodology and practice: 1) to enhance the field of courtroom questioning and pragmatics 2) to propose a range of corpus search terms that are useful for investigating courtroom questioning and 3) with implications for legal practitioners in general, and for Malaysian legal counsels in particular, and where the findings can be a point of reference for legal counsels and legal educators. To conclude, the present study has moved beyond the mere description of linguistic variation in the Malaysian legal system to the range of linguistic phenomena that can have social consequences within the legal system. Its key contribution is not just for academic purposes but will hopefully also have significant implications for lawyer/witness interaction and legal training.

Author: Nurshafawati Ahmad Sani