Steele, Mark J; Chin, Jason M; van Golde, Celine --- "Witness Preparation and the Corruption of Memory: a Survey of Australian Trial Judges" [2024] MelbULawRw 4; (2024) 48(1) Melbourne University Law Review 152

	Home \| Databases \| WorldLII \| Search \| Feedback Melbourne University Law Review

The resolution of civil legal proceedings often depends on the evidence of witnesses of fact.^[3] There has long been an understanding by participants in the legal process that witness memory is inherently frail and potentially unreliable.^[4] The past 20 years, however, have seen an increasing appreciation that the process by which witnesses are being prepared to give evidence in civil proceedings can and does work to further undermine the reliability of that evidence by distorting witnesses’ memories of critical events.^[5] This state of affairs threatens the just resolution of disputes and public confidence in the legal system.^[6]

These concerns led, in 2021, to the publication of a practice direction for certain commercial courts in the United Kingdom (‘UK’) that more closely regulated witness preparation (Practice Direction 57AC)^[7] and, in 2022, to a report of the International Chamber of Commerce highlighting the issue and providing guidance for practitioners in arbitral proceedings.^[8]

In Australia, however, there has to date been no significant reform or systematic research addressing the potential for pre-trial interactions with witnesses in civil proceedings to unintentionally distort witness memory and the implications of that potential for fact-finding by judges.^[9] To redress this, we conducted a survey of trial judges of the Federal Court of Australia (‘FCA’) and the Supreme Courts of New South Wales (‘NSWSC’), Queensland (‘QSC’) and Victoria (‘VSC’).^[10] This article briefly puts the issue of witness memory in civil proceedings in context and then reports on and discusses the results of our survey. As will be seen, and is reflected in the epigraphs above, the judges in our sample expressed substantial concern about the impact of witness preparation practices on the reliability of the evidence given in witness statements in Australian civil legal proceedings.^[11]

Historically, witnesses in civil proceedings gave their evidence-in-chief orally,^[12] from the witness box.^[13] Since the 1990s, however, in pursuit of efficiency and transparency, the evidence-in-chief of witnesses in civil proceedings (and particularly in commercial cases) has often come to be provided in the form of a written statement or affidavit, supplied to the other parties and the court in advance of the hearing.^[14] In recent years though, it has become clear that the process by which witness statements are commonly prepared in civil proceedings can affect not just the presentation of the witness’s evidence but also the substance and integrity of the witness’s memory of the relevant events.^[15]

In criminal investigations and prosecutions, the ways in which pre-trial interactions with the witness can corrupt their evidence have been recognised for many years.^[16] This has resulted in the extensive regulation of witness interactions in criminal proceedings, including mandating procedures for eyewitness identification,^[17] recording interviews with suspects and complainants,^[18] educating police officers on the psychology of memory and training police officers in interview techniques to minimise the risk of witness memory being distorted by suggestion.^[19] It is also reflected in the use, in appropriate cases, of expert psychological evidence or judicial directions to alert the jury to aspects of the psychology and fallibility of memory which the criminal courts accept are not matters of common knowledge.^[20]

The same potential for the corruption of recollection also exists in civil proceedings. Despite this, and in marked contrast to the position in criminal proceedings, regulating the preparation of non-expert witnesses of fact in civil proceedings^[21] is essentially confined to general ethical admonitions for lawyers to observe a paramount duty to the administration of justice,^[22] to act with honesty and integrity^[23] and not to engage in conduct likely to undermine confidence in the administration of justice,^[24] while at the same time acting to ‘promote and protect fearlessly and by all proper and lawful means the client’s best interests’.^[25] More specifically, there are also prohibitions against condoning false or misleading evidence and ‘coaching’ witnesses.^[26]

These rules can be extremely difficult to apply in practice, particularly in the complex and nuanced process of engaging with a potential witness to ‘refresh’ their memory of (often long ago) events and reduce that recollection to a coherent, written narrative that serves as admissible evidence to assist a client’s case. These difficulties are further exacerbated by the imperative, in the adversarial system of justice, for lawyers to produce evidence that will advance their client’s case.^[27]

The rule against ‘coaching’ witnesses, for instance, draws a distinction between a lawyer, on the one hand, ‘advising what answers the witness should give to questions’ (which is prohibited) and, on the other, ‘questioning and testing in conference the version of the evidence to be given by a prospective witness, including drawing the witness’s attention to inconsistencies or other difficulties with the evidence’ (which is expressly permitted).^[28] In practice, this distinction can be elusive.^[29] For instance, the process of challenging a witness’s account and confronting them with anomalies, inconsistencies or implausibilities in their account will naturally and implicitly tend to encourage a witness towards a version of events that ‘adjusts’ for such matters.^[30] Similarly, while a practitioner must not mislead the court, it is generally accepted that they have a permitted role to play ‘in carefully drafting the affidavit in the best possible light, to highlight strengths and minimise weaknesses’.^[31]

In practice, as courts have recognised, there is no bright line between ‘coaching’ and ‘refreshing recollection’ or ‘questioning and testing in conference’,^[32] and the difference between permissible and impermissible conduct is ‘inevitably a matter of degree’.^[33] Further, in civil practice, there is little or no formal education or training for litigation lawyers on the nature and suggestibility of memory.^[34] Interview technique is also typically learned ‘on the job’.^[35]

Importantly, in civil proceedings, in contrast to the position in criminal prosecutions,^[36] client legal privilege also works to shield the witness preparation process from scrutiny.^[37] Litigation privilege thus protects confidential communications and documents from disclosure if they were prepared for the dominant purpose of a client being provided with legal services relating to anticipated or pending legal proceedings.^[38] This means that records or notes of interviews and drafts of witness statements are typically privileged against disclosure.^[39] The rationale for this privilege, which is distinct from the privilege that protects the confidentiality of communications relating to obtaining legal advice, is said to be a need for secrecy ‘to facilitate investigation and preparation of a case for trial by the adversarial advocate’.^[40]

To obtain data on the experiences and opinions of Australian judges about witness preparation and its pitfalls in civil proceedings, we surveyed and interviewed a sample of senior Australian trial judges. Our aim was to begin to build an evidence base for an Australian response to the problem of the capacity of the witness preparation process to corrupt witness evidence.

Judges are not themselves directly involved in the production of witness statements,^[41] but we decided to survey the views of judges on witness preparation practices for the following reasons.^[42] First, in the nature of their work, trial judges closely observe numerous civil proceedings.^[43] Moreover, judges are not merely passive observers — their role requires them to assess the veracity and reliability of the evidence of the witnesses who appear before them and to reconcile differences in the evidence to make findings of fact.^[44] This role puts judges in a position to make informed observations about the apparent impact of the way in which the evidence of witnesses is prepared and presented on witness evidence. Secondly, as our sample characteristics confirm, judges are overwhelmingly drawn from the ranks of the legal profession and so will have had direct experience of witness preparation before their appointment.^[45] Finally, judges’ views on proper legal practice are influential in setting the parameters for ethical behaviour, particularly where the rules themselves are ill-defined and inherently difficult to apply.^[46] That said, judges’ experiences and views provide only indirect access to information about witness preparation practices and their effects, and so this needs to be borne in mind when considering our results. We further acknowledge and discuss, in Part VI, the limitations inherent in asking judges about their views on these matters.

Part II describes our methodology and the characteristics of the judges in our sample. Parts III–V then present and analyse the quantitative and qualitative results from the survey, addressing, in turn, the judges’ general experience with witness statements,^[47] the judges’ views on commonly used witness preparation practices and the judges’ views on a number of potential reforms. In Part VI, we conclude with reflections on the implications and limits of our findings.

II METHODOLOGY AND SAMPLE

Our study consisted of a survey (N = 73) and interviews (N = 26) with trial judges of the FCA, NSWSC, QSC and VSC. All surveys and interviews were conducted between August and December 2022. The University of Sydney provided ethics approval.^[48] The survey was prospectively registered.^[49]

In each case, the Chief Justice of the Court gave their explicit support for the research. We invited 165 judges to participate in our survey on the practice and regulation of witness preparation in civil proceedings.^[50] Ninety-five judges responded. Twenty-two of the judges who responded disqualified themselves from participating, mostly on the basis that they sat exclusively or mainly in criminal trials or in jurisdictions that did not typically involve them hearing contentious witness evidence.^[51] Seventy-three judges completed all of, or part of, the survey.^[52] The participation rate among the judges who considered themselves qualified and able to respond meaningfully was 51%.^[53] This response rate, which is high for studies of this kind,^[54] suggests that our results are reasonably representative of the views of judges in the jurisdictions that we studied and that the topic is considered by judges to be important.^[55]

Our survey comprised 25, mostly multiple-choice, questions (some with several elements) and included ‘comments’ sections that gave the judges the opportunity to add observations in relation to questions.^[56] The judges were given the option of answering the survey either online or in hard copy.^[57] The survey was conducted on the explicit basis that the judges’ participation and responses would be anonymous.^[58]

The questions in the survey sought some background information about the judge and their judicial practice and otherwise were organised under the following headings:

1 ‘General Experience of Witness Statements’ — four questions about the judge’s recent experience with particular aspects of witness statements in cases before them;^[59]

2 ‘Witness Preparation Practices’ — a multi-part question that sought the judge’s view as to whether the use of each of 13 specific practices in witness preparation was ‘generally acceptable’ or ‘generally unacceptable’;^[60]

3 ‘Potential Reforms’ — a series of questions that solicited the judge’s view on various ways in which the preparation of witness evidence in civil proceedings might be changed, specifically:

4 ‘The Psychology of Memory’ — a multi-part question that sought the judge’s view as to whether, based on their knowledge and experience, six specific propositions relating to the psychology of memory were ‘essentially correct’ or ‘essentially incorrect’.^[62]

In addition to the quantitative data from answers to the survey questions, we also obtained qualitative commentary and observations from the judges about the matters raised in the survey in two ways.

First, the survey questions included sections in which the judge could add comments.^[63] Most judges took up this opportunity for at least some questions, with more than half of the judges providing comments on five or more questions.

Secondly, the judges who completed the survey were asked if they were prepared to be approached by the researchers to confidentially clarify or discuss their answers.^[64] Thirty-four judges agreed to be approached for such an interview; 26 were able to be interviewed by the lead author.^[65] These interviews were conducted between September and December 2022.

The interviews were relatively informal and worked from the judge’s specific survey responses. Typically, the interviews briefly covered aspects of each of the topics raised by the survey questions. They started at a general level by asking the judges whether the potential for witness preparation to distort witness memory was a topic which was ‘on their radar’ before receiving the survey. The judges were also asked whether it was their experience or expectation that many of the witness preparation practices identified by them in the survey as ‘generally unacceptable’ were nonetheless used in practice. Most interviews were recorded (with the permission of the judge involved) and transcribed, with four judges preferring the researcher to work only from notes. The interviews were generally 30–45 minutes in duration.

All of the comments and observations made by the judges in interview were collated to the specific survey question or topic to which they related. The first and second authors then organised the material under those headings into themes, resolving disagreements through discussion (eg whether a judge favoured or opposed a proposal for reform or expressed a negative, positive or neutral view about witness statements).^[66] Excerpts from that material are included below, to provide further context, depth and insight into the quantitative survey results. In selecting those excerpts, we have sought to represent the range of views expressed in relation to the subject of each survey question in a balanced way and to quote from a number of different judges.^[67]

After a summary of our sample, the following parts correspond to the three main sections of the survey, which are the subject of this article. Those parts report a summary of the results of the survey together with excerpts from the judges’ commentary and interviews.

Data were analysed using open-source software (R, version 4.2.3). We calculated 95% confidence intervals^[68] using the Sison-Glaz method for multinomial proportions.^[69] Percentages are rounded to the nearest whole number.^[70] Readers may verify our results, as our Open Science Framework page contains the survey (including the Qualtrics version), analytic code, supplementary charts and prospective registration.^[71]

The respondents were generally experienced judges, with over half reporting that they had more than five years of judicial experience. The characteristics of the sample of judges which responded to our survey are outlined in greater detail in Table 1. Columns that do not sum to 73 reflect non-responses.

The FCA was most heavily represented in our sample (n = 25), with the balance drawn from the VSC (n = 19), the NSWSC (n = 17) and the QSC (n = 10). Approximately ⅓ of respondents (n = 25) reported having over 10 years of service as a judge, ¼ reported having 5 to 10 years of experience (n = 18) and approximately ⅓ reported having fewer than 5 years of experience (n = 29).

The majority of the judges in our sample reported that their judicial practice was predominantly in commercial law or equity (n = 41), with the remainder mostly in common law (n = 12), administrative law (n = 6) or criminal law (n = 3).^[73] We also asked the judges about their work prior to serving as a judge. The majority had been barristers (n = 61), with the remainder comprising solicitors (n = 9) and one academic lawyer. Finally, we asked the judges to identify their predominant area of practice prior to becoming a judge. The majority identified this as litigation (n = 42), with the remainder being commercial law or equity (n = 14), criminal law (n = 5) or administrative law (n = 4).

The characteristics of the judges who were interviewed substantially mirrored the characteristics of the judges who responded to the survey and are summarised in Table 2.

III JUDGES’ GENERAL EXPERIENCE OF WITNESS STATEMENTS

The judges responded to four questions directed to their general experience of witness statements^[74] over the last two to five years.^[75]

First, the judges were asked how often in the past two years they had encountered a witness statement that was materially contradicted or qualified in cross-examination. Forty-two per cent of our sample said that this occurred either ‘often’ (35%) or ‘very often’ (7%).^[76] In the same period, 30% of the judges had, at least ‘often’, encountered witness statements that, after cross-examination, appeared to them to materially fail to reflect the witness’s own recollection.^[77]

Perhaps most strikingly, 65% of the judges reported that they ‘often’ (40%), ‘very often’ (21%) or ‘always or almost always’ (4%) encountered statements that, after cross-examination, appeared to them to be in words other than those of the witness.

Reflecting these concerns about the quality of written evidence-in-chief, 74% of the judges reported that, at least occasionally, they directed that all or part of the evidence-in-chief of a witness be given orally rather than in writing, with 33% reporting that they did this ‘often’ or ‘very often’ and 6% that they did so ‘always or almost always’.

Statement recounted recollection in words not those of the witness?	Never or almost never	4%	[0%, 17%]
	Occasionally	31%	[19%, 44%]
	Often	40%	[28%, 52%]
	Very often	21%	[9%, 33%]
	Always or almost always	4%	[0%, 17%]
Statement materially contradicted or qualified in cross-examination?	Never or almost never	7%	[0%, 20%]
	Occasionally	50%	[38%, 62%]
	Often	35%	[24%, 48%]
	Very often	7%	[0%, 20%]
	Always or almost always	0%	[0%, 12%]
Direct all or part of evidence-in-chief to be given orally?	Never or almost never	27%	[15%, 39%]
	Occasionally	35%	[24%, 48%]
	Often	20%	[8%, 32%]
	Very often	13%	[1%, 25%]
	Always or almost always	6%	[0%, 18%]

The overall picture that emerges from these survey responses, supplemented by the judges’ written comments and our interviews, is that a substantial proportion of the judges surveyed reported encountering serious deficiencies in the written evidence that currently emerges from witness preparation processes. This experience was echoed in observations made by a number of the judges. For instance:

Additionally: ‘Many affidavits have been over-lawyered with genuine recollection obscured by what the lawyer thinks needs to be in the affidavit to advance a party’s case.’^[80]

As reflected in these statements, the judges typically identified the source of these defects as the way in which lawyers set about interviewing witnesses and producing their written evidence. Judges thus expressed the view, for instance, that written evidence in civil proceedings is ‘overworked, over-lawyered and polished’,^[81] ‘depart[s] from the true recollection of the witness and reflect[s] what the legal adviser wants the witness to say’^[82] and is produced in a way which is ‘just nuts’.^[83] Alarmingly, a handful of the judges even doubted the extent to which witnesses were involved in the preparation of their statements, with one judge saying: ‘I suspect that the process involves the solicitor preparing a statement which is shown to the witness who is asked to agree to it.’^[84]

Unsurprisingly, a number of the judges expressed concern that the ‘over-lawyering’ of statements had serious implications for the trial process, complaining that it not only ‘leads to painful cross-examination’^[85] but could also work substantively against a party’s interests.^[86] For instance, ‘witnesses often used technical expressions such as “to be held on trust” or “the payment was a dividend” without any real understanding of the legal implications of that nomenclature’.^[87] A particularly egregious version of this defect, which was reported by several judges, is the production of witness statements in which two or more witnesses relate a contentious conversation in substantially identical terms.^[88]

A witness might, of course, of their own volition overstate or misstate their actual recollection in their witness statement and then be obliged, in cross-examination, to correct or qualify that evidence in a material particular.^[89] Accordingly, that this defect might occur ‘occasionally’ (as reported by approximately half of the judges) is perhaps unsurprising. However, for this to be encountered ‘often’ or ‘very often’ (as reported by 42% of the judges) is cause for concern.

It is also particularly concerning that 65% of the judges reported that their experience in the last two years was that witness statements ‘often’, ‘very often’ or ‘always or almost always’ appeared to recount the witness’s recollection in words that were not their own.^[90] The judges’ reported experience in this regard thus echoed the observation of Nettle and Gordon JJ in Queensland v Estate of the Late Masson:

Of course, a lawyer recasting the witness’s own words may simply be trying to improve the clarity or succinctness of the evidence.^[92] However, such recasting can also have the pernicious effect of obscuring, and even distorting, the witness’s actual recollection. As one judge observed in their written comments: ‘[t]he lawyer’s job is not to translate, but to record’ the witness’s evidence.^[93] Moreover, as Nettle and Gordon JJ point out in the above-quoted passage, the facility of lawyers with language is often not matched by that of the witness.^[94] So it is that the written evidence of a witness can come to be, not just more clearly expressed, but subtly distorted. As one judge observed of this tendency:

The existence of such concerns explains the significant incidence of judges (74%) who, at least ‘occasionally’, direct that all or part of the evidence-in-chief of a witness be given orally, rather than in writing. Asked in our survey when and why they made such directions,^[96] these judges generally said that they did so when the credit of the witness was likely to be in issue and particularly in cases of disputed conversations, where carefully crafted written evidence was felt, all too often, to obscure the truth.^[97]

In short, the view of many judges was that oral evidence was often ‘better quality evidence’ than the written evidence that emerged from current witness preparation practices.^[98] Two principal reasons were given for this. First, a number of the judges felt that oral evidence gave them ‘unfiltered’^[99] access to the authentic and ‘less tutored’^[100] voice of the witness and that this helped those judges in assessing the reliability of the evidence.^[101] Secondly, several judges also thought that oral evidence-in-chief was fairer to the witness than confining their oral evidence only to cross-examination, because it allowed the witness to settle in and express their account in their own words before being cross-examined.^[102]

However, although the judges expressed the view that requiring oral evidence-in-chief attenuated the influence of lawyers on the witness’s evidence, they also recognised that requiring oral evidence-in-chief did not eliminate that influence. As one said, ‘it doesn’t cure the problem — it changes the nature of the preparation but doesn’t prevent it happening’.^[103]

Finally, although efficiency and cost concerns are commonly identified as rationales for written evidence-in-chief, some judges expressed the view that written evidence-in-chief was typically less efficient and more costly than oral evidence.^[104] This was because oral evidence tended to be more focused^[105] and because it reduced the onerous and costly process of having to deal with, often extensive, objections to written evidence.^[106]

This general dissatisfaction with witness statements among Australian judges mirrored the sentiment expressed in a 2018 survey of UK judges and legal practitioners.^[107] In that survey, only 6% of respondents thought that witness statements ‘fully’ achieved the aim of producing the best possible evidence.^[108] Echoing our own results, 56% of respondents to the UK survey also thought that witness statements ‘failed to reflect the witness’s own evidence’.^[109] Again, a consistent theme was that witness statements were typically ‘lawyer-led’, ‘heavily crafted by solicitors’ and ‘a vehicle for the lawyer’s view of the case’.^[110]

In the UK, this dissatisfaction recently led to the imposition of mandatory ‘best practice’ guidelines on practitioners in the Business and Property Courts^[111] governing how witness statements are to be prepared and what they are to contain.^[112] In Australia, little has yet been done in this regard,^[113] but our survey results point to the conclusion that similar reform here would be strongly supported by trial judges.^[114]

IV WITNESS PREPARATION PRACTICES

In the next section of the survey, the judges were presented with a variety of practices used to prepare witnesses to give evidence in civil proceedings and asked whether they considered each practice to be ‘generally acceptable’ or ‘generally unacceptable’.

We chose the practices presented based on two criteria. First, we selected practices that are currently widely (or at least sometimes) used in witness preparation in civil proceedings.^[115] While this was based on our own research and experience, it was also confirmed by the observations of the judges that we interviewed.^[116] Secondly, we chose practices that psychological research indicates are prone to distort witness memory.^[117]

As described below, our results reveal differences among the judges as to whether particular practices are ‘generally unacceptable’.^[118] The strength of our conclusions is limited by the reality that there may be some interpretative differences arising from how one understands the practice that is being referred to in the survey^[119] or from the use of the qualifier ‘generally’.^[120] Nonetheless, our findings suggest broad consensus among the judges about the acceptability of some practices and considerable disagreement about others. Such disagreements present obvious challenges to lawyers in understanding what they should and should not be doing and also raise real questions about the feasibility of formulating more precise rules or additional guidance for the profession.

Table 4: Witness Preparation Practices Considered to Be Generally Unacceptable

Practice Posed to Judges	Percentage Indicating ‘Generally Unacceptable’	95% Confidence Interval
Preparing the witness for cross-examination by providing them with copies of the statements of supporting witnesses	87%	[80%, 96%]
Seeking to refresh recollection by providing the witness with copies of other witness statements	87%	[80%, 96%]
Seeking to refresh recollection by suggesting what may have happened based on other evidence	84%	[75%, 92%]
Seeking to refresh recollection by prompting the witness to explain what probably or could have happened	82%	[74%, 92%]
Asking leading questions to try and prompt recollection	67%	[57%, 80%]
Preparing a witness for cross-examination by providing them with the statements of opposing witnesses	67%	[57%, 80%]
Without rendering the evidence misleading, omitting from the statement discrete aspects of recollection unhelpful to the client’s case	64%	[52%, 76%]
Seeking to refresh recollection by showing the witness contemporaneous documents not created or seen by them at the time	57%	[45%, 70%]
Suggesting words for use in answering a particular question in cross-examination, to make the answer clearer, more coherent or succinct	55%	[43%, 68%]
Without rendering the evidence misleading, drafting the statement to put witness’s evidence in the best possible light for the client’s case	51%	[39%, 64%]
Putting forward language to be used in the statement other than the witness’s words to improve clarity and coherence	44%	[33%, 58%]
Practicing cross-examination on the witness’s statement	38%	[27%, 51%]
Identifying for the witness the factual issues most important to the client’s case	23%	[13%, 33%]

Starting with the areas of most agreement, more than 80% of the judges considered that it was generally unacceptable to seek to refresh the witness’s recollection by (a) providing them with copies of the statements of other witnesses, (b) suggesting to them what ‘may have happened’ based on the lawyer’s knowledge of the documentary record and the evidence of other witnesses or (c) prompting them to explain what ‘probably happened’ or ‘could have happened’. Similarly, 87% of the judges considered it generally unacceptable to prepare a witness for cross-examination by providing them with copies of the statements of supporting witnesses.

The judges were more divided, however, on other witness preparation practices; for instance, whether it was generally unacceptable to:

We suggest that this diversity of opinion is, at least in part, a reflection of the complexity and nuanced nature of the witness preparation process,^[121] which makes it inherently difficult to be definitive about what is or is not permissible in all circumstances.^[122] As discussed further below, this feature of witness preparation poses an obstacle to developing clear guidance for legal practitioners on acceptable, or even ‘best’, practice, and it poses an even greater obstacle to formulating clear ‘rules’.^[123]

The judges’ observations in interview shed some light on why many judges found these practices unacceptable. The principal theme was the perceived potential for these practices to ‘contaminate’ or ‘distort’ the witness’s evidence by suggestion. For instance:

Indeed, one judge frankly described the way in which, in their own experience as a legal practitioner, a witness’s recollection could be moulded by preparation. For instance:

In line with these observations, when asked if they agreed or disagreed with the overall proposition that ‘[p]ractices currently used in preparing witnesses of fact to give evidence in civil proceedings in Australia can distort a witness’s recollection’,^[126] a strong majority of the judges ‘agreed’ or ‘strongly agreed’ (57% of the judges, with only 4% offering any disagreement).^[127]

V POTENTIAL REFORMS

Anticipating that there may be some concerns about witness preparation practices among Australian judges, our survey also sought the judges’ views on the adequacy of existing regulation of witness preparation and potential reforms. Specifically, the judges were asked to indicate agreement or disagreement with six proposals for reform. The results are summarised in Table 6; the reforms are ordered in accordance with their popularity among the judges, beginningwith the reform with which the greatest number of the judges agreed (or strongly agreed).

Reform Put to Judges	Response Option	Percentage of Judges	95% Confidence Interval
Professional conduct rules should include additional guidance	Strongly disagree	1%	[0%, 14%]
	Disagree	3%	[0%, 15%]
	Neither agree nor disagree	22%	[11%, 35%]
	Agree	50%	[39%, 62%]
	Strongly agree	24%	[12%, 36%]
Witnesses required to state how and when recollection refreshed, identifying documents used	Strongly disagree	4%	[0%, 16%]
	Disagree	15%	[4%, 27%]
	Neither agree nor disagree	17%	[6%, 29%]
	Agree	43%	[32%, 55%]
	Strongly agree	21%	[10%, 33%]

Record witness interviews and make available to other parties if court so orders	Strongly disagree	8%	[0%, 21%]
	Disagree	38%	[26%, 50%]
	Neither agree nor disagree	22%	[11%, 34%]
	Agree	22%	[11%, 34%]
	Strongly agree	10%	[0%, 22%]
Remove litigation privilege in some cases	Strongly disagree	15%	[4%, 28%]
	Disagree	36%	[25%, 48%]
	Neither agree nor disagree	21%	[10%, 33%]
	Agree	22%	[11%, 34%]
	Strongly agree	6%	[0%, 18%]
Court to receive expert evidence in some cases	Strongly disagree	17%	[6%, 29%]
	Disagree	35%	[24%, 47%]
	Neither agree nor disagree	28%	[17%, 40%]
	Agree	18%	[7%, 31%]
	Strongly agree	3%	[0%, 15%]

Figure 1 also shows the reforms; they are ordered in accordance with the number of the judges that strongly disagreed with the reforms, beginning with the reform with which the greatest number of the judges strongly disagreed.

In summary, the judges overwhelmingly supported more guidance and education for the profession, with 74% of the judges agreeing or strongly agreeing that this was a desirable reform and only 4% offering any disagreement. Comments and observations in interview indicated that this reflected a widespread view among judges that the existing rules are not sufficiently clear and detailed about what practices should be avoided.

Proposals for more specific changes, however, generally did not attract majority support. Comments made by the judges through the survey or in interview indicated that this was either because the judges thought that the proposed reform was unlikely to be useful (eg requiring witnesses to state their level of confidence in their recollection) or because of their concern that the impact of the proposed reform on the trial process would outweigh its benefit in addressing the issue (eg receiving expert evidence on the ways in which witness preparation might distort memory).

We now review the results in more detail, starting with the least supported reforms and then moving on to those that received more support. We end with a discussion of the reforms that most divided the sample.^[128]

Starting with expert evidence, the judges were asked whether, in a case where there was a reasonable basis for believing that the recollection of a witness of fact may have been materially influenced by suggestion on the part of the lawyers, it would potentially be useful for the court to receive expert evidence as to the nature, fallibility or suggestibility of memory.^[129] Fifty-two per cent of the judges either ‘disagreed’ (35%) or ‘strongly disagreed’ (17%).

The judges who disagreed or strongly disagreed generally expressed the view that the time and cost associated with such expert evidence would outweigh any benefit, although some thought that such evidence might be more useful if it focused on the circumstances of particular witnesses rather than generalities.^[130] Other judges, in contrast, thought that an application of psychological expertise to the position of particular witnesses would be wrong in principle ‘as an abdication of the role of the tribunal or court’^[131] or ‘[de facto] fact-finding by the expert’.^[132]

On the other hand, while a handful of the judges thought that the psychology of memory involved matters of ‘common sense’,^[133] most expressed the view that knowledge of the psychological research was useful and important for judges and should be the subject of judicial education (although not expert evidence).^[134]

Two other potential reforms, which did not attract favour among the judges, were directed to introducing greater transparency into the witness preparation process. The judges were thus asked to consider the idea of removing litigation privilege over communications with non-party witnesses of fact either generally or where there is a reasonable basis for believing that the recollection of the witness may have been materially influenced by suggestion on the part of the lawyers.^[135] Fifty-one per cent of the judges either ‘disagreed’ (36%) or ‘strongly disagreed’ (15%) with that reform. The judges were also asked to consider whether parties should be required to make and retain a record of interviews with non-party witnesses and make that record available to all parties if the court so ordered. Forty-six per cent of the judges ‘disagreed’ (38%) or ‘strongly disagreed’ (8%) with this.^[136]

The judges gave two principal reasons for disagreeing with these potential reforms. First, many judges expressed concern that they would lead to ancillary disputes that would add unduly to the duration and cost of litigation: for instance, that the reforms would lead to a ‘morass’ of satellite litigation,^[137] ‘give rise to tangential, time-consuming disputes which would be a forensic nightmare’^[138] and be ‘likely to prolong trials’.^[139] Secondly, some judges were concerned that an attenuation of litigation privilege might interfere with the witness preparation process in counterproductive ways: for instance, ‘I suspect you are probably creating an impediment to a candid discussion taking place between the solicitor and the witness’.^[140]

Despite the relative lack of support for attenuating litigation privilege, however, a number of the judges accepted that there would be some salutary consequences in doing so. Most judges interviewed, irrespective of whether they opposed or supported the proposal, thus accepted that access to records of a witness’s interviews would be a useful tool in assessing the reliability of the witness’s evidence. For instance, as one judge said of their experience in having access to records of interview in criminal proceedings: ‘you can follow the questions and you can see how and why people have given answers, also what they have misunderstood when they have given an answer, the issues of language ... I find having the questions really valuable’.^[141] Additionally, the view was expressed that attenuating litigation privilege would have a salutary effect on witness preparation practice. As one judge said: ‘if more were exposed about how a witness was prepared, or could be exposed, people would be a lot more careful’.^[142] Some judges also expressed the view that it was wrong in principle for a party to be able to put evidence before a court and, at the same time, resist inquiry into how that evidence was obtained.^[143] Further, it should be noted that opinion was divided on this reform, with a substantial minority of the judges either ‘agreeing’ (22%) or ‘strongly agreeing’ (6%) with limiting litigation privilege in its application to non-party witnesses.^[144]

We now turn to a reform that received strong majority support — enhancing professional conduct rules to include additional guidance on how to prepare a witness statement so as to minimise the risk of distorting the witness’s recollection.

The section of the survey addressing that reform began by reminding judges of the existing professional conduct rules.^[145] These rules provide that practitioners (a) must not ‘coach’ a witness ‘by advising what answers the witness should give to questions which might be asked’ but (b) may question or test in conference ‘the version of evidence to be given by a ... witness, including drawing the witness’s attention to inconsistencies or other difficulties with the evidence’.^[146]

The judges were then asked whether, in their view, these rules ‘provide adequate guidance for counsel as to what is and is not permissible in preparing a witness to give evidence’.^[147] They were also asked whether the rules ‘should include additional guidance for lawyers on how to prepare a witness statement so as to minimise the risk of distorting the witness’s recollection’.^[148] Curiously, while 74% of the judges supported ‘additional guidance’, 64% also agreed or strongly agreed that the existing rules provided ‘adequate guidance’. The judges’ comments and observations in interview, however, suggest that this apparent anomaly may be more terminological than substantive. That is, many judges drew a distinction between ‘rules’ and ‘guidance’ in their responses to these two questions; generally, they favoured more ‘guidance’ but also expressed the view that this guidance neither needs to, nor should, take the form of ‘rules’.^[149]

Related to this proposed reform, the judges in interview also expressed general concern that the education and training of practitioners in witness preparation and interviewing techniques were insufficient (recognising that the training that did occur was typically ad hoc and ‘on the job’).^[150] One judge said, for example: ‘In too many cases, even in larger law firms, witness statements are prepared by lawyers who are inexperienced and insufficiently trained in the proper way to approach that task.’^[151] Although most judges were in favour of additional guidance for the profession, there were a variety of views expressed as to the form that such guidance should take.^[152] Some judges favoured stricter professional rules. For example: ‘Rules 68 and 69 are inadequate. The practices listed above should be specifically mentioned and prohibited.’^[153]

A number of the judges, however, favoured the development instead of a set of ‘guidelines’ or a ‘code of practice’. This was because, in their view, the nature of the witness preparation process did not lend itself to prescriptive ‘rules’. As one judge said: ‘I would definitely be in favour of additional guidance and education but the nature of the process may defeat “rules”.’^[154] Others doubted that even clear ‘guidelines’ were feasible and favoured more general education and training for practitioners. As one judge observed: ‘there are so many different scenarios that trying to codify what should be done or not done in all circumstances would be a fraught process’.^[155]

These doubts as to the practicality of developing clear rules, or even authoritative guidelines, are supported by the divergence of views expressed by the judges on what is and is not ‘generally acceptable’ in witness preparation.^[156] We discuss this further in our conclusion.

The other reform that gained wide acceptance was the imposition of a requirement that witnesses of fact state, if practicable, whether, and if so how and when, their recollection in relation to important disputed matters of fact had been refreshed by reference to documents and to identify those documents.

Sixty-four per cent of the judges either ‘agreed’ or ‘strongly agreed’ with this suggestion. In fact, this is a proposal that has been adopted in the UK where, since April 2021, Practice Direction 57AC has required:

The adoption of this requirement in the UK was not mentioned in the survey and none of the judges who responded to the survey or were interviewed indicated that they were aware of Practice Direction 57AC or the requirement.

Most of the comments on this proposal were provided by the minority of the judges who did not agree with it, who expressed concern that this ‘would be very complicated’,^[158] ‘would be too complex and confusing for many witnesses’^[159] and ‘seems very onerous and unnecessary’.^[160] Among the judges who agreed with the proposal, a number pointed out that it was analogous to what would occur if a witness gave their evidence-in-chief orally.^[161]

The final potential reform was the imposition of a requirement that witnesses of fact state in their witness statement how well they recall matters that are important disputed matters of fact.^[162] This proposal divided the judges, with 40% either ‘disagreeing’ or ‘strongly disagreeing’ and 32% ‘agreeing’ or ‘strongly agreeing’. Again, most of the commentary was provided by the judges who did not agree, who observed, for example, that this was ‘a matter for cross-examination’,^[163] ‘you’ll just end up with the same stock phrases and they’ll be meaningless’^[164] and that such statements would be of little use (eg ‘self-serving and carries little weight’^[165] or ‘not ... appropriate or even important’^[166]) or meaningless (eg ‘[w]hat on earth does “how well they recall” mean?’).^[167]

The judges were also asked whether there were any other potential reforms to current witness preparation practices that they thought should be considered.^[168] This yielded a range of suggestions.^[169] Two proposals were mentioned by more than one judge: (a) that more evidence-in-chief should be given orally^[170] and (b) that judges should direct the use in appropriate cases of pre-trial depositions.^[171]

VI CONCLUSION

Trial judges hearing civil cases in Australian superior courts report that they often find serious apparent defects in the witness statements and affidavits that come before them.^[172] In particular, judges often report encountering witness statements that appear to them to be expressed in words other than those of the witness, materially fail to reflect the witness’s own actual recollection, or are materially contradicted or qualified in cross-examination.^[173]

This is not a trifling matter. Every time such defective evidence is produced it has the potential to lead to an unjust result. While, as our survey results attest, these defects sometimes appear to the judge during the hearing, one cannot assume that this is always the case. Instances where it becomes apparent during the trial that a witness’s recollection has become distorted or corrupted may be just the tip of an iceberg. This danger is illustrated by those occasions when, more or less by chance, a hearing exposes the extent to which the process can go awry. For example, one judge recounted a case about whether a property had been purchased by the plaintiff for herself or for her mother (who had provided the money for the purchase).^[174] A witness swore an affidavit attesting that the plaintiff had said to the witness explicitly that the house belonged to her mother.^[175] Cross-examined, however, the witness’s oral evidence was that she simply assumed that the house belonged to the mother.^[176] Similar examples can be found in reported authorities and illustrate the potential for the witness preparation process to produce written evidence that is quite wrong.^[177]

In assessing the potential for distorted evidence to lead to an unjust result, one must also take into account the great bulk of civil cases that are resolved before trial.^[178] In such cases, the evidence is never tested but nonetheless forms the basis for advice on the prospects and risks of the litigation and drives decisions as to settlement.^[179]

In terms of what underlies these apparent defects, many judges in our sample consider that they all too often arise from the way in which current witness preparation practices operate to distort witness evidence and that many of those practices are ‘generally unacceptable’.^[180] At the same time, however, many judges considered that these ‘generally unacceptable’ practices were both in common use and permitted by existing professional rules.^[181]

This has the potential to undermine the just resolution of civil disputes, either by leading to an unmeritorious settlement or by inducing the court to place undue weight on flawed evidence. Over time, it has the capacity to undermine public trust and confidence in the judicial resolution of civil proceedings and to damage the standing of the legal system.^[182] Many judges in our sample evidently considered that this state of affairs was unsatisfactory and in need of change.^[183]

We conclude with a discussion of the limitations of this study and its implications for the path forward.

A Limitations

First, the primary limitation we would like to highlight relates to what can be drawn from the judges’ responses to our questions and prompts about the reliability of written witness evidence and the impact of witness preparation practices on that evidence. In particular, the judges’ views about such matters are necessarily impressionistic and limited by the judge’s experience and knowledge.^[184] Judges will often not know, for instance, whether a witness statement materially failed to reflect a witness’s actual recollection when it was prepared, even where that statement has been contradicted or qualified in cross-examination. Nor will they typically know with certainty whether the witness’s written and oral evidence misaligning was caused by the way in which the witness’s evidence was prepared by the lawyers involved or by some other reason. Nor can judges generally know whether their fact-finding is accurate.^[185]

While this limitation must be kept in mind, we believe that our results still provide valuable information. Trial judges are charged with the task of assessing the reliability of the evidence of the witnesses who appear before them and using that evidence to make findings of fact.^[186] This puts them in a unique position to make observations about the apparent reliability of witness evidence and the apparent impact on that evidence of the way in which it is prepared and presented. Commensurately, the fact that many judges express substantial reservations about the reliability of written witness evidence and the impact of witness preparation should provide a serious basis for concern about witness statements and the witness preparation process in civil proceedings.^[187]

Secondly, our quantitative survey results are limited in their precision. As seen throughout, the confidence intervals associated with this study’s estimates are about ±10%.^[188] Generally, however, even the low end of our estimates should be a cause for concern.

Finally, our analysis of survey comments and interview transcripts was inevitably coloured by our own backgrounds and experiences in civil litigation, evidence law and legal psychology. Conscious of this, we have endeavoured to present a balanced view of our findings and to maximise transparency in our data collection and analysis.

B The Path Forward

So what can be done to address this apparent problem? The solution most favoured by the judges who responded to our survey is additional guidance for legal practitioners on how to prepare a witness statement to minimise the risk of distorting the witness’s recollection.^[189] Opinion was divided on whether this should take the form of professional rules or less prescriptive guidelines and education. However, whether either is capable of achieving real reform is doubtful for the following reasons.

First, the involvement of lawyers in the preparation of witness evidence is widely accepted and encouraged by the courts as a means of ensuring that evidence is coherent, focused on the issues and in admissible form (ie complying with the rules of evidence).^[190] Notably, none of the judges in our survey suggested that it might be appropriate to exclude lawyers from witness preparation.^[191]

Secondly, seeking to achieve change through increased guidance and education assumes that the problem is essentially one of ignorance or misunderstanding — that it is a failing of the lawyer rather than a problem with the system. It is far from clear, however, that this is the case.

In the adversarial system, litigation practitioners have a duty (and competitive imperative), acting within ethical rules, to prepare and put forward the best possible evidence to serve their client’s interest in either winning the case or settling it on favourable terms.^[192] However, a practitioner cannot deliberately ‘coach’ the witness by telling them what evidence they should give or encourage or condone false testimony.^[193] On the other hand, however, their duty to protect and defend their client’s interests requires them to seek from the witness the most complete recollection that it is possible to obtain, challenge that recollection if it seems improbable or inconsistent with other evidence and reduce it to writing in as coherent and persuasive a form as possible.^[194] The practical reality is that, while bound to act within the existing ethical rules, as one experienced advocate and judge has observed: ‘Parties can hardly be expected to interview potential witnesses in a detached way that would minimise the damage that interrogation can do to memory.’^[195] Further, even in the modern digital era, not all important events are recorded in documents. Where they are not, practitioners have no alternative but to work with the memories of participants and to do their ethical best to try and produce favourable and persuasive evidence for their client’s case. Unprompted, however, witnesses often have little or no recollection of the relevant events.^[196] This is especially so if, as is typical in commercial cases, the events occurred some time ago and were not particularly remarkable at the time (eg routine business meetings or negotiations).^[197]

Typically, therefore, it is only by ‘refreshing’ the recollection of the witness, through extensive recourse to contemporaneous documents and other prompts, that any sort of coherent account can be obtained.^[198] By the end of that process, however, how much of the witness’s account is based on an untainted, actual memory of the original events will be difficult, if not impossible, to ascertain.^[199]

However, tainted or not, and whether or not founded in the witness’s actual experience of the events, the evidence may convincingly ‘tell the story’ of the client’s case. And, regardless, it will stand unless it is impeached in cross-examination.

The use of such prompts to ‘refresh’ the witness’s recollection is a practice that is well-entrenched and perceived by lawyers as serving their clients’ best interests.^[200] This practice is how practitioners perform the alchemy of turning the ‘dross’ of vague and fragmented memory into the ‘gold’ of coherent and persuasive evidence. Unless these prompts are prohibited by clear enforceable rules or other mechanisms are adopted to force change, such practices are likely to continue.^[201]

Thirdly, formulating clear rules to prescribe proper practice in witness preparation is a difficult, perhaps impossible, task. As many judges in our survey observed, attempting to ‘refresh’, record and test a witness’s recollection in interview is a complex and nuanced process. This process does not readily lend itself to firm guidance let alone definitive rules.^[202] This difficulty is supported by the observations of many judges to that effect.^[203] It is also supported by the division of views among the judges as to which witness preparation practices are ‘generally acceptable’ or ‘generally unacceptable’.^[204]

Fourthly, as a practical matter, changing the witness preparation process so as to avoid, or even minimise, its capacity to corrupt witness memory is not a straightforward task. The psychological research makes it clear that witness memory is inherently more fragile and malleable than most people expect and prone to distortion by even unconscious and subtle suggestion.^[205] All of the practices listed in Table 4 have that potential, and that list is far from exhaustive. Further practices that have been identified by the psychological research as having that capacity include such common and reflexive features of the questioning of a witness as (a) using qualifying descriptors in questions (eg ‘how emphatic was he?’), (b) interrupting a witness’s answers, (c) applying interrogative pressure, (d) giving implicit feedback to a witness on their answers (eg looking surprised, concerned or pleased about an answer), (e) summarising the witness’s answer and (f) using documents to ‘fill in’ the gaps in a witness’s chronology of events.^[206]

Finally, as long as litigation privilege protects the witness preparation process from view, it is practically impossible for that process to be effectively policed.^[207]

These considerations mean that, in our view, the problem of the corruption of witness memory during witness preparation is unlikely to be effectively addressed through additional rules, education or guidance alone.^[208]

An alternative to this approach that was canvassed in our survey was to make the witness preparation process more transparent. This would allow the other parties to see how an opposing witness’s evidence on a particular issue was obtained.^[209] Where that process may have distorted the witness’s recollection on an important disputed matter, a party could then bring this to the attention of the judge. This would allow the judge to better assess the reliability of that witness’s evidence and, where appropriate, voice criticism of the witness preparation process. Over time, this would be likely to lead to real changes in practice. After all, ‘sunlight is said to be the best of disinfectants’.^[210]

Consistent with these arguments in favour of increased transparency, 64% of the judges surveyed ‘agreed’ or ‘strongly agreed’ that witnesses of fact should be required to state, if practicable, whether, and if so how and when, their recollection in relation to important disputed matters of fact had been refreshed by reference to documents and to identify those documents.^[211] The difficulty with this limited approach to transparency, however, is that all that would be provided is a general description of the process and a bare list of the documents referred to. In the absence of a wider ‘lifting of the veil’ with respect to the witness preparation process, this limited approach to transparency is unlikely to provide meaningful insight into how the witness’s evidence was obtained (and potentially distorted). Worse, such an incomplete picture might actually mislead the court.

If there is a real possibility that a witness’s recollection has been corrupted in a material respect, one would need to have a much more complete picture of the information that was shared with the witness than merely knowing what documents they were shown. For example, what was the witness told about the significance to the case of the events in question? What was shared with them about the recollection of other witnesses? How did their evidence ‘evolve’ in successive interviews and drafts of their witness statement?

Armed with such information, the court may be able to see for itself, for instance, whether a witness to a critical representation recalled it spontaneously or did so only after persistent questioning and being referred to contemporaneous documents in a manner suggesting what ‘might have happened’. In such a case, the judge might even be meaningfully assisted by expert psychological evidence directed not merely to general propositions about the malleability of witness memory but, more specifically, to what happened in relation to the particular witness.^[212]

There is surely a case to be made that interviews with non-party witnesses should be recorded and excluded from the secrecy of litigation privilege (either generally or in specific circumstances). Importantly, the privileged nature of communications with non-party witnesses is not justified by the policy considerations in favour of preserving the confidentiality of lawyer–client communications but only by the far less compelling rationale that parties in adversarial proceedings should be free to prepare their cases ‘in secret’.^[213] Further, it is surely wrong in principle that a party should be able to engage in practices that can corrupt the evidence of a witness and, at the same time, shield that process from view. Such a situation unfairly leaves the parties and the court to assess the reliability of the evidence without any real insight into how it was obtained.^[214]

These considerations support an argument for the greater transparency of the witness preparation process that could be achieved by lifting litigation privilege over communications with non-party witnesses (at least where there is a reasonable basis for believing that the recollection of a witness may have been materially influenced by suggestion). This reform was canvassed by our survey but was not embraced by the majority of the judges. That majority expressed concern about whether the benefits of such a change would justify the risk of prolonging trials and adding to their cost and whether such a change might stultify frank communication with such witnesses.^[215] On the other hand, however, many judges accepted that such additional transparency would be useful in more accurately assessing the reliability of a witness’s evidence and in improving witness preparation practice. And no judge took issue with such an approach at the level of principle: that is, that it was wrong, in principle, for a party to be able to act in a manner likely to distort a witness’s recollection and, at the same time, conceal that conduct from view.^[216] Further, 28% of the judges ‘agreed’ or ‘strongly agreed’ that such a reform should be considered.^[217] A number of other judges also suggested that they might be amenable to such a reform if more education and training for the profession failed to improve the situation.^[218]

Overall, our survey adds to the basis for concern that there are significant problems with witness preparation practice in civil proceedings in Australia. It establishes that these problems are widely recognised by trial judges, who are strongly of the view that change is needed to address this situation. Our survey also points to impediments to change, such as substantial disagreement about what practices are ‘acceptable’ in witness preparation and hesitation among judges in exposing the process to greater transparency through reforming litigation privilege. The integrity of witness evidence is critical to the just resolution of civil proceedings and to public confidence in the legal system.^[219] We hope that this work^[220] will contribute to encouraging debate and building impetus for change so that witness evidence can be more effectively safeguarded from corruption through corrosive witness preparation practices.^[221]

^[*] BCom, LLB (Hons) (Melb); BCL (Oxf). Barrister and Senior Counsel, New South Wales Bar Association. This research was carried out as part of a PhD project at the University of Sydney on the topic ‘The Practice and Regulation of Witness Preparation in Civil Proceedings in

Australia’. In summary, that PhD project is examining (a) what the psychological research on the nature and suggestibility of memory tells us about how witness memory can be distorted by witness preparation practices commonly used in civil proceedings in Australia, (b) the

adequacy of existing regulation of witness preparation in that regard and (c) potential reforms to the regulation of witness preparation to address this issue. Dr Chin and Dr van Golde are Mr Steele’s supervisors for that PhD project. The lead author can be contacted by email at <mark.steele@sydney.edu.au>.

^[†] BA (Hons) (Virginia); MA, PhD (UBC); JD (Toronto). Senior Lecturer, College of Law, Australian National University.

^[‡]BSc, MSc (Maastricht); PhD (Syd). Senior Lecturer, School of Psychology, Faculty of Science, University of Sydney. The authors gratefully acknowledge the support that this research

received from the Chief Justices of each of the surveyed Courts and the judges of those Courts, who gave up valuable time to respond to the survey and, in many cases, to be interviewed. The authors are also grateful to the anonymous referees for their perceptive and helpful comments on earlier versions of this paper.

^[1]Judge 7. Quotes from judges who participated in our survey are cited anonymously in the form of either ‘Judge [number]’ (where the quote is taken from the written comments provided by a judge in their survey response) or ‘Judge [letter][number]’ (where the quote is taken from an observation made by a judge in interview).

^[3]The expression ‘witnesses of fact’ refers to witnesses giving evidence about what they have seen or heard about an event and how it happened or a matter internal to their mind (eg why they took some past decision or action) as distinct, for instance, from expert opinion evidence based on specialised knowledge or expertise: Jeremy Gans, Andrew Palmer and Andrew Roberts, Uniform Evidence (Oxford University Press, 3^rd ed, 2019) 183, 194–5. See Michelle Mattison and Penny Cooper, ‘Witness Statements for Employment Tribunals in England and Wales: What Are the “Issues”?’ (2021) 25(4) International Journal of Evidence and Proof 286, 287.

^[4] See, eg, Gestmin SGPS SA v Credit Suisse (UK) Ltd [2013] EWHC 3560 (Comm), [20] (Leggatt J) (‘Gestmin’). See generally JJ Spigelman, ‘Truth and the Law’ [2011] (Winter) Bar News 99, 108–11.

^[5]See generally Mark J Steele, ‘Witness Preparation and the Corruption of Evidence in Civil Proceedings’ (2023) 42(2) Civil Justice Quarterly 158. The focus of our survey was on the practices used in the preparation of witness statements. The process of preparing witness statements typically involves lawyers for a party working with the witness to ‘refresh’ their recollection (eg by recourse to contemporaneous documents), evaluate and test that recollection and reduce it to writing. References in this article to ‘witness preparation’ are generally to that process. It should be noted, however, that the process of preparing a witness of fact to give evidence in civil proceedings also typically involves other elements, including familiarising the witness with courtroom processes and preparing them for cross-examination: Thomas A Mauet and Les A McCrimmon, Fundamentals of Trial Technique (Lawbook, 4^th ed, 2018) 19–21. The psychological research indicates that those other elements of witness preparation also have the potential to distort witness memory, but they are not the focus of our survey or this article: see British Psychological Society, Guidelines on Memory and the Law: Recommendations from the Scientific Study of Human Memory (Report, June 2008) 2, 29 <http://www.antoniocasella.eu/dnlaw/Memory_law_2008.pdf> , archived at <https://perma.cc/33DZ-2A8D>.

^[6]‘There is a strong public interest ... that in civil cases the claimant should succeed if he is entitled to do so and should fail if he is not’: Three Rivers District Council v Governor of the Bank of England [No 6] [2004] UKHL 48; [2005] 1 AC 610, 647 [28] (Lord Scott).

^[9]Indeed, there has been very little study, in the context of civil proceedings, of the potential for pre-trial interactions with witnesses to distort witness memory. The extant body of experimental psychological research on the nature and suggestibility of memory thus focuses almost exclusively on memory in the context of witnessing and recalling criminal behaviour: see Kimberley A Wade and Ula Cartwright-Finch, ‘The Science of Witness Memory: Implications for Practice and Procedure in International Arbitration’ (2022) 39(1) Journal of International Arbitration 1, 19.

^[11]See above nn 1–2 and accompanying text. The expression ‘witness statement’ was expressly used in our survey, and is used in this article, to include both witness statements and affidavits: Mark Steele, Jason Chin and Celine van Golde, Judicial Survey (Word Survey) 2 <https://osf.io/xs8uc>, archived at <https://perma.cc/HGM9-7LGY>.

^[12]The evidence of witnesses in civil proceedings is conventionally broken up into their evidence as to the relevant facts given on behalf of the party who has called them (evidence-in-chief), then questioning by opposing parties (cross-examination) and then further questioning by the party who called them to clarify any ambiguity arising from the cross-examination (re-examination): Gans, Palmer and Roberts (n 3) 44; Evidence Act 1995 (Cth) s 28; Evidence Act 1995 (NSW) s 28; Evidence Act 2008 (Vic) s 28.

^[15]See, eg, Accuracy of Fact Witness Memory (n 8) 10–14; British Psychological Society (n 5) 29; Steele (n 5) 159.

^[20]See, eg, Dupas v The Queen [2012] VSCA 328; (2012) 40 VR 182, 195 [56]–[58] (Warren CJ, Maxwell P, Nettle, Redlich and Bongiorno JJA) (on identification evidence); Audsley v The Queen [2014] VSCA 321; (2014) 44 VR 506, 520 [56], 523 [69] (Maxwell P, Weinberg and Priest JJA) (on the impact of substance abuse, alcohol and sleep deprivation on memory); R v Bartlett [1996] VICSC 47; [1996] 2 VR 687, 690, 696 (Winneke P, Charles JA agreeing at 699, Southwell AJA agreeing at 700) (on the phenomenon of ‘recovered memory’); Hickey v The Queen (1995) 89 A Crim R 554, 560–3, 565 (Tadgell and Charles JJA, Vincent AJA) (on the impact of drug abuse on capacity for accurate recall). In relation to the dangers of identification evidence, see, eg, Evidence Act 1995 (Cth) ss 116(1), 165(1)(b), (2); Evidence Act 1995 (NSW) ss 116(1), 165(1)(b), (2); Evidence Act 2008 (Vic) ss 165(1)(b), (2); Winmar v Western Australia [2007] WASCA 244; (2007) 35 WAR 159, 163–4 [10]–[14] (Wheeler, McLure, Pullin, Buss and Miller JJA). In relation to delayed complaints, see Judicial Commission of New South Wales, Criminal Trial Courts Bench Book (77^th rev ed, 2024) 720–1 [5-080].

^[21]The preparation of expert witnesses, in contrast, has been the subject of much greater regulation than the preparation of witnesses of fact, in an attempt by the courts to reduce the influence of partisanship on such evidence. Many courts have thus promulgated ‘codes of conduct’ for expert witnesses: see, eg, Federal Court of Australia, Practice Note GPN-EXPT: Expert Evidence, 25 October 2016 <https://www.fedcourt.gov.au/law-and-practice/practice-documents/practice-notes/gpn-expt>, archived at <https://perma.cc/H7AD-W7KF>. The Federal Court Practice Note, for instance, contains provisions regulating interactions between parties and their lawyers and expert witnesses: at paras 3.1–3.4. This includes a provision that parties ‘should never attempt to pressure or influence an expert into conforming [their] views with the party’s interests’: at para 3.1. See also Uniform Civil Procedure Rules 2005 (NSW) r 31.23, sch 7; Uniform Civil Procedure Rules 1999 (Qld) sch 1C; SCGCP Rules (n 13) form 44A.

^[23]See, eg, LPUC Barristers Rules (n 22) r 8; LPUL Solicitors Rules (n 22) rr 4.1.2–4.1.4; Civil Procedure Act (n 22) s 17.

^[24]See, eg, LPUC Barristers Rules (n 22) rr 8(b)–(c); LPUL Solicitors Rules (n 22) r 5.1.2.

^[25]LPUC Barristers Rules (n 22) r 35. See also LPUL Solicitors Rules (n 22) r 4.1.1.

^[26]See, eg, LPUC Barristers Rules (n 22) rr 69–70; LPUL Solicitors Rules (n 22) r 24.

^[28] LPUC Barristers Rules (n 22) rr 69(b), 70. See also LPUL Solicitors Rules (n 22) r 24.

^[32]See above n 28 and accompanying text. As one of the judges in our sample observed in interview, ‘[t]he notion of “refreshing recollection” can be a pretty slippery proposition in practice’: Judge C14.

^[33]Majinski v Western Australia [2013] WASCA 10; (2013) 226 A Crim R 552, 559 [30] (Martin CJ, Buss JA agreeing at 563 [50]). See also New Aim Pty Ltd v Leung [2023] FCAFC 67; (2023) 410 ALR 190, 219 [119] (Kenny, Moshinsky, Banks-Smith, Thawley and Cheeseman JJ) (‘New Aim’). As has been observed, ‘the line between preparing and prompting (or “coaching,” the usual term of opprobrium) is rarely clear even for the most scrupulous’: Applegate (n 29) 279.

^[35]See Mattison and Cooper (n 3) 301. See also Witness Evidence Working Group (n 7) 11 [42]–[43].

^[36]In criminal proceedings, the duty of prosecutorial fairness requires the prosecution to inform the accused of all relevant evidence, including material relevant to the credibility or reliability of prosecution witnesses: Grey v The Queen (2001) 184 ALR 593, 599–600 [23] (Gleeson CJ, Gummow and Callinan JJ), 611 [63] (Kirby J); Mallard v The Queen [2005] HCA 68; (2005) 224 CLR 125, 133 [17] (Gummow, Hayne, Callinan and Heydon JJ), 155 [81] (Kirby J).

^[39]Public Transport Authority (WA) v Leighton Contractors Pty Ltd [2007] WASCA 151; (2007) 34 WAR 279, 282 [9], 284 [19], 287 [31], 288 [34]–[35] (McLure JA, Steytler P agreeing at 281 [1], Miller JA agreeing at 288 [37]) (‘Leighton Contractors’); Ritz Hotel Ltd v Charles of the Ritz Ltd [No 22] (1988) 14 NSWLR 132, 133–4 (McLelland J); Trade Practices Commission v Ampol Petroleum (Victoria) Pty Ltd (1994) 53 FCR 578, 579–85 (Heerey J).

^[40]Leighton Contractors (n 39) 284 [16] (McLure JA), quoting Robert J Sharpe, ‘Claiming Privilege in the Discovery Process’ in Law Society of Upper Canada (ed), Law in Transition: Evidence (Richard De Boo Publishers, 1984) 163, 164–5.

^[42]We also considered surveying practitioners in relation to witness preparation practices and may yet do so as part of our wider research project. However, obtaining direct and frank information from legal practitioners as to the practices used by them in preparing witnesses to give evidence is potentially difficult, including because that information may expose them to criticism. In the United States (‘US’), for instance, the admittedly much more ‘robust’ practices used in witness preparation (which often include the extensive rehearsal of witness testimony) have been described as ‘one of the dark secrets of the legal profession’: Applegate (n 29) 279. Witness preparation has also been called the profession’s ‘[d]irty [l]ittle [s]ecret’: Roberta K Flowers, ‘Witness Preparation: Regulating the Profession’s “Dirty Little Secret”’ (2011) 38(4) Hastings Constitutional Law Quarterly 1007, 1007; Roberta K Flowers, ‘What You See Is What You Get: Applying the Appearance of Impropriety Standard to Prosecutors’ (1998) 63(3) Missouri Law Review 699, 740. Further, the direct observation of witness interviews is problematic because such interviews are confidential and privileged and the presence of a third party might ground an argument for waiver of privilege: see Australian Law Reform Commission, Privilege in Perspective: Client Legal Privilege in Federal Investigations (Report No 107, 2008) 94 [3.68] <https://www.alrc.gov.au/wp-content/uploads/2019/08/ALRC107.pdf>, archived at <https://perma.cc/EM95-TZNC>; Leighton Contractors (n 39) 284 [19] (McLure JA). Also, consent to direct observation would be required from the client, the lawyers involved and the witness — none of whom would have any particular motivation to provide it.

^[45]See below Table 1. We recognise, however, that it is possible that some practices have changed over time.

^[46]In the UK, for instance, the recent promulgation of a practice direction for certain commercial courts more closely regulating the witness preparation process was initiated, sponsored and influenced by the judges of those courts: Steele (n 5) 172, 175, discussing Practice Direction 57AC (n 7).

^[47]The introductory section of the survey asked the judges to read references in the survey to ‘witness statements’ to include affidavits: Steele, Chin and van Golde, Judicial Survey (n 11) 2.

^[49]A prospective registration is a time-stamped record of a study protocol that can be compared to the final protocol for that study to determine if the researchers made any changes after the data were known. For the importance of prospective registration in empirical legal research, see Jason M Chin and Kathryn Zeiler, ‘Replicability in Empirical Legal Research’ (2021) 17 Annual Review of Law and Social Science 239, 243.

^[50]All trial judges in the surveyed jurisdictions were invited to participate. Judges were contacted via email to their chambers’ email addresses, which were obtained either from the court website or with the assistance of the Chief Justice. Judges who sat primarily in appeals were not invited to participate because that work does not typically involve hearing evidence from witnesses: see Richard Douglas, ‘Raising Fresh Argument or Evidence on Appeal’ [2008] (85) Precedent 8, 9.

^[51]The 22 judges who responded indicating that they would not be participating in the survey said that they would not be doing so for reasons that included sitting mostly in criminal trials (n = 9) and sitting mainly in appeals (n = 8). A full list of reasons given can be found in our supplemental material: Supplemental Materials (Supplement) <https://osf.io/cpxh5>, archived at <https://perma.cc/FZ28-BF5L>.

^[52]Not all judges responded to all questions — sample sizes for individual questions below 73 indicate that one or more judges did not respond to that question.

^[53]Our receipt of 73 survey responses out of 143 judges (165 invited judges less 22 self-disqualified responses) yields this 51% participation rate.

^[55]This is a view confirmed by observations made by a number of the judges in interview. See Judge A16: ‘I don’t think this is just an academic issue. I think it’s quite important to have some research and work done on this.’ See also Judge B16: ‘This is a topic in which judges are definitely interested.’

^[57]Qualtrics software was used for the online survey. Sixty-one judges responded online and 12 judges responded by completing a hard copy of the survey and forwarding it to the researchers via email or mail.

^[60] Ibid 7–10. The judges were also asked whether there were any other witness preparation practices that they found unsatisfactory or troubling in addition to those 13 specific practices: at 9–10. This produced a range of critical comments: see, eg, Judge 12 (‘[l]ining up witness statements ... to ensure consistency’); Judge 30 (‘[t]he practice of substantially cutting and pasting the evidence of one witness into the affidavit of another is particularly appalling, but it is not an irregular occurrence’); Judge 49 (‘[t]here are, I fear, many worse practices than those mentioned, but they include blatantly telling witnesses what or what not to say to suit the case at hand. Hopefully, that does not occur, but I fear it does’).

^[62]Ibid 20. In the interest of keeping this article focused, the results of this discrete aspect of our survey are reported in only summary form. Six propositions about the nature of witness memory were stated, with the judges asked whether each proposition seemed to them, based on their knowledge and experience, to be ‘essentially correct’ or ‘essentially incorrect’. While these propositions were generalisations, which could be subject to exceptions and differences of interpretation, our results suggested that the judges’ understanding of key aspects of the nature, fallibility and suggestibility of witness memory broadly aligned with the psychological research: see, eg, Accuracy of Fact Witness Memory (n 8) 31–40; British Psychological Society (n 5) 29. In other words, our results indicated that judges, at least at a high level, were generally well-informed as to the ways in which memory could be corrupted by suggestion. For a more detailed summary of this aspect of our survey, see ‘Witness Preparation and the Corruption of Memory: A Survey of Australian Trial Judges’ (Paper) <https://osf.io/4gc8j>, archived at <https://perma.cc/3CGK-JTX7>.

^[64]As was clear from the question asking the judges if they were prepared to be interviewed, the survey responses of the judges who agreed to be interviewed were known to the lead author. However, the interviews were conducted on the explicit basis that neither those survey responses, nor the additional observations made in interview, would be attributed to any identified judge: ibid 2–3. This was consistent with the terms of our ethics approval: see Mitchell (n 48).

^[65]There was no process of ‘selection’ by which the interviewed judges were chosen. Interviews with the other eight judges who indicated a willingness to be interviewed were not able to be conducted purely due to timing and availability.

^[66]We followed an informal version of a reflexive thematic analysis: see Virginia Braun and Victoria Clarke, ‘Reflecting on Reflexive Thematic Analysis’ (2019) 11(4) Qualitative Research in Sport, Exercise and Health 589, 594. Our starting point in analysing the interview transcripts, notes and survey comments was informed by our research concerning the quality of witness statements in civil litigation and by our experiences in practice, the study of evidence law and the study of legal psychology.

^[67]The summary descriptions in this article of the gravamen of the comments and interviews were formulated in the first instance by the lead author on the basis of a review of that material and then separately reviewed by the second author.

^[68]A confidence interval measures the precision of an estimate: Daniël Lakens, ‘Confidence Intervals’, Improving Your Statistical Inferences (Web Page, 2022) [7] <https://lakens.github.io/statistical_inferences/07-CI.html>, archived at <https://perma.cc/E32Z-SVC2>. Here, our estimates pertain to the attitudes, beliefs and knowledge of all Australian superior court judges with some experience in hearing civil proceedings. A 95% confidence interval, which is what we calculate in this article, is wide enough such that 95% of the time it includes the percentage of Australian judges that have the relevant attitude, belief or knowledge. Accordingly, confidence intervals can help indicate the ‘resolution with which an effect is estimated’: at [7.3]. Very narrow ranges provide high resolution, with larger ranges suggesting lower resolution. In this study, our confidence intervals are approximately ±10% around our estimates. We consider this to be sufficient to draw the conclusions we make because our conclusions would be warranted in most cases even if our estimates were off by ±10%. However, we invite readers to make their own judgements.

^[69]Cristina P Sison and Joseph Glaz, ‘Simultaneous Confidence Intervals and Sample Size Determination for Multinomial Proportions’ (1995) 90(429) Journal of the American Statistical Association 366, 366–7.

^[71]Jason Chin, Celine van Golde and Mark Steele, ‘Witness Preparation and the Corruption of Memory: A Survey of Australian Trial Judges’, OSF Home (Web Page, 3 September 2024) <https://osf.io/kdcft/?>. See also Steele, Chin and van Golde, Judicial Survey (n 11); Judicial Survey (Qualtrics Survey) <https://osf.io/asrt9>, archived at <https://perma.cc/H5GA-HQCV>; Australian Judges on Witness Preparation and Memory Contamination in Civil Cases (R Output, 8 June 2023) <https://osf.io/f9zp6>; Supplemental Materials (n 51); Jason Chin, Celine van Golde and Mark Steele, ‘Witness Preparation: Survey for Judges’, OSF Registries (Web Page, 15 August 2022) <https://osf.io/ej2rx>, archived at <https://perma.cc/3VBS-X6PZ>. We did not seek ethics approval to share our raw data due to the sensitivity of our sample. However, the R code shows how that raw data were analysed to produce our results: Australian Judges on Witness Preparation and Memory Contamination in Civil Cases (R Code, 8 June 2023) <https://osf.io/gm6nh>, archived at <https://perma.cc/F8YP-34KQ>. We did not perform any analyses on subsections of our sample (eg by reference to the judges’ work prior to appointment) because we considered that the samples for such subgroups would have been overly small.

^[72] Eight judges chose ‘Other’ and filled in a response. For details of all ‘Other’ responses, see Judge Practice Area: Other (Sheet) <https://osf.io/um473>, archived at <https://perma.cc/L3MU-54RE>.

^[73]The relatively small proportion of criminal judges in the sample is consistent with our recruitment approach that explained that the study was directed to their experience of witness preparation in civil proceedings: Steele, Chin and van Golde, Judicial Survey (n 11) 2.

^[74]Again, we note that the introductory section of the survey asked the judges to read references in the survey to ‘witness statements’ to include affidavits: ibid.

^[75]The first three questions were directed to the judge’s experience in the past two years and the fourth question was directed to the judge’s experience in the past five years: ibid 3–6.

^[77]It should be noted that a portion of the ‘never or almost never’ answers to these questions may reflect the fact that some judges reported that they seldom received witness evidence in written form. A point which was made by some judges in response to these questions was thus that (a) the judges who sat mainly in the common law division of some courts (eg the VSC and the QSC) seldom encountered written evidence-in-chief because the practice in that division was for witnesses to give their evidence-in-chief orally and (b) it followed that those judges would encounter the situations posited by the questions either ‘never’ or only ‘occasionally’ simply because they seldom saw witness statements: Judge 35; Judge 71. See also Uniform Civil Procedure Rules 1999 (Qld) r 390(a); SCGCP Rules (n 13) r 40.02.

^[78] In the past two years for the first three questions and in the past five years for the fourth question.

^[86]In one case, a judge reported that a critical element of the party’s claim was that ‘individuals were targeted precisely because they were vulnerable and had limited education, but their affidavits were plainly written by highly educated lawyers’: Judge 64.

^[88]See, eg, Judge 61: ‘In at least two cases this year it has been clear that passages in direct speech of conversations set out in two witness statements were so similar’ that ‘it was likely that they were prepared by the lawyer acting for the party and had been suggested to the witness rather than being based on the independent recollection of each witness’.

^[89]See Judge 63: ‘Witnesses themselves, especially parties, will in some cases distort their evidence irrespective of the conduct of legal advisers.’

^[90]See, eg, Judge 24: ‘Witness statements and affidavits are often drafted by lawyers in their own language, not the language of the witness.’

^[92]As the courts have recognised, the involvement of lawyers to assist in ensuring that witness evidence is expressed in clear language may be appropriate even in the case of expert witnesses: New Aim (n 33) 217 [112], 219 [120] (Kenny, Moshinsky, Banks-Smith, Thawley and Cheeseman JJ).

^[93]Judge 72. But see other judges’ comments, which included the statement that ‘sometimes improved expression helps, particularly with those whose use of language may not be great’: Judge 42. See also Judge B16: ‘if it’s still the truth, expressed more formally and perhaps more succinctly and more clearly than the witness could do, I’m not troubled by that’.

^[95]Judge 57. As another judge observed, witness statements ‘are usually the evidence of the solicitor’: Judge 7.

^[97]These judges thus explained that they ordered oral evidence-in-chief ‘to ensure that the true observations and recollection of the witness’ are ‘given in evidence’: Judge 34. Another reason was to have the witness’s evidence ‘unadulterated by his/her legal advisers’: Judge 50. ‘[I]t reduces the risk of the evidence being filtered through the lawyers’: Judge A42. As one judge pithily observed, the reason why they ordered oral evidence-in-chief was ‘to get to the truth’: Judge 7.

^[98]Judge A42; Judge B21; Judge D25; Judge D38. See Judge D39: ‘I do think from the judge’s point of view that there’s much greater insight into the witness’s true recollection and the witness’s honesty and integrity through that process.’

^[99]Judge A42. Similarly, ‘[o]ral evidence is reliably the evidence of the witness not the lawyers’: Judge 39.

^[101]See, eg, Judge A16: ‘Generally, I think it’s more likely to produce a genuine recollection ... to get the evidence in their words, how they speak, how they react, how they converse and then you get an idea of what is likely to have been said.’ Such comments echo the observation of Callinan J that ‘[v]iva voce evidence retains a spontaneity and genuineness often lacking in pre-prepared written material’: Concrete Pty Ltd v Parramatta Design & Developments Pty Ltd [2006] HCA 55; (2006) 229 CLR 577, 635 [175].

^[102]On the other hand, a number of the judges accepted that oral evidence-in-chief could tend to favour the confident, articulate witness over the nervous or inarticulate. The same judges, however, generally considered that they could adjust for this in their assessment of the witness ‘and make allowance for it’: see, eg, Judge B21.

^[103]Judge B38. ‘It’s not a perfect solution, but it makes it harder’: Judge D38.

^[105]As one judge said, oral evidence was ‘quicker and cheaper’ because it worked against the preparation of written statements which were too often ‘opuses ... of marginal utility’: Judge A10.

^[106]As a judge complained, ‘[witness statements] create an entire industry in taking petty evidentiary objections’: Judge 35.

^[109]Ibid 9 [30]. The majority of respondents (55.6%) either agreed (36.35%) or strongly agreed (19.25%) that a reason that witness statements did not fulfil their purpose was that they ‘fail to reflect the witnesses’ own evidence’: at app 1. See above Table 3.

^[111]The Business and Property Courts are a collective of specialist civil courts; those courts include the courts of the Chancery Division as well as the Admiralty Court, the Commercial Court and the Technology and Construction Court from the King’s Bench Division of the High Court: ‘The Business and Property Courts’, GOV.UK (Web Page) <https://www.gov.uk/courts-tribunals/the-business-and-property-courts>, archived at <https://perma.cc/BZE2-3TEQ>.

^[112]Although the guidelines are described as a ‘Statement of Best Practice’, the practice direction containing those guidelines requires the legal representative of a party with responsibility for the trial witness statement to certify that the witness statement has been prepared in accordance with the guidelines, effectively rendering them mandatory: Practice Direction 57AC (n 7) para 4.3. See generally Steele (n 5) 175–8.

^[114]This was also a view expressed by a number of the judges in interview. See, eg, Judge A8: ‘I think it’s a massive hole that people prefer not to look at.’

^[115]These practices were drawn from a variety of sources, including discussions with judges and practitioners and observations in judgments: see, eg, Gestmin (n 4) [20] (Leggatt J); Re Equiticorp Finance Ltd; Ex parte Brock [No 2] (1992) 27 NSWLR 391, 395–6 (Young J); Craig v Troy (Supreme Court of Western Australia, Ipp J, 24 February 1995) 9–10 (‘Craig’). Practices were also drawn from texts and articles by practitioners providing guidance on how to prepare witnesses as well as from existing best practice guidelines: see, eg, Mauet and McCrimmon (n 5) 19–23; Jeffrey Pinsler, Evidence and the Litigation Process (LexisNexis, 6^th ed, 2017) 786 [17.022]–[17.023]; Richard Mahoney, ‘Witness Conferences’ (2000) 24(5) Criminal Law Journal 297, 299–300; Justice Sheppard, ‘Communications with Witnesses before and during Their Evidence’ (1987) 3(1) Australian Bar Review 28, 29–32; Western Australian Bar Association (n 104) [2.1]–[2.10], [10.1]–[24.6]; Practice Direction 57AC (n 7) app paras 3.1–3.16. Finally, practices were drawn from the lead author’s own experience practicing at the commercial Bar in Sydney since 1995.

^[116]Of the 23 judges who were asked about this in interview, 20 said that it was their experience or expectation that at least some of the practices identified by them as ‘generally unacceptable’ nevertheless occurred in practice. For instance, ‘I think without a doubt they do happen’: Judge D19. ‘These practices are commonplace, both from my time as a barrister and a judge’: Judge D39. The other three judges asked for a view on this felt unable to express an opinion. See, eg, Judge C3: ‘We don’t get much insight from the Bench into what is actually happening in practice nowadays and I’ve been on the Bench for many years now.’

^[117]See, eg, Accuracy of Fact Witness Memory (n 8) 31–40. See generally British Psychological Society (n 5) 29–30; Wade and Cartwright-Finch (n 9) 14–19; Elizabeth F Loftus et al, Eyewitness Testimony: Civil and Criminal (LexisNexis, 6^th ed, 2023) ch 11.

^[119]For instance, in relation to practicing cross-examination on the witness’s statement, judges’ comments included Judge 3 responding that this would be okay as long as ‘they are not cross-examined on documents or statements’ and Judge 69 responding that it ‘[d]epends [on] what is involved’.

^[120]A number of the judges made comments to the effect that their views relied significantly on the qualification ‘generally’ and that there might be particular circumstances in which practices identified by them as ‘generally unacceptable’ might be acceptable or vice versa. See Judge 35: ‘These questions do not admit of nuance.’ Some judges emphasised that even practices identified as ‘generally acceptable’ were not necessarily considered by them as something to be

encouraged. One judge commented that ‘[e]ven though I have referred to some practices as “acceptable”, I am not suggesting it is best practice nor am I suggesting that it is always helpful for the judge’: Judge 18.

^[121]In practice, the witness interview thus functions not merely as a surrogate for oral evidence-in-chief, but as an integral part of the process by which a party and their lawyers investigate the facts and make decisions about what evidence to adduce in their case. In that process, the lawyer does not simply reduce the witness’s evidence-in-chief to writing, but challenges, probes and evaluates that evidence in order to make judgments about whether or not to deploy it: see generally Mauet and McCrimmon (n 5) 17–23.

^[122]Several judges also made observations to this effect. See Judge 57: ‘A number of these answers depend on context.’ See also Judge 44: ‘so much depends on a judgment to be made according to the nature of the case, the significance of the evidence, and the calibre of the witness.’

^[123]This difficulty is also discussed by the lead author in the context of the promulgation in the UK, by Practice Direction 57AC (n 7), of a mandatory ‘Statement of Best Practice’: see Steele (n 5) 178–80.

^[127]The fact that 38% of the judges neither agreed nor disagreed with this proposition may reflect a reservation among some judges as to their ability to be sure, in any particular case, whether a witness’s memory has in fact been distorted. When asked in interview why they had ‘neither agreed nor disagreed’ with this proposition, some judges explained that they believed that witness preparation practices could distort witness memory but could not positively verify that those practices did so or that they did so in all cases. See Judge D41: ‘I just don’t know. You might say, when you see someone talk and you realise that they didn’t speak the words recorded in the affidavit, then there’s obviously a fair bit of lawyering involved in there, but whether it actually changed the thrust of the evidence, I don’t know.’

^[128]In reporting on the attitudes expressed by the judges to the proposed reforms we have simply reported the explanations given by the judges for their agreement or disagreement in their comments or observations in interview without attempting to critically analyse or challenge those explanations (eg as to whether they might reflect an underlying resistance to change or insufficient time to fully consider the implications of those changes).

^[129]It is to be noted that our survey also asked the judges if there had been any occasion in the last five years when a psychologist had given expert evidence before them, in civil proceedings, as to the nature, fallibility or suggestibility of memory: Steele, Chin and van Golde, Judicial Survey (n 11) 17. Only six of the 72 judges who responded to this question said that this had occurred. Only two of those judges also reported on the circumstances in which this had occurred — one in a case relating to childhood recollection and the other in a case relating to recovered memory of sexual abuse.

^[130]See Judge 45: ‘[I would] expect the evidence would be too high level to be of any specific assistance in a specific case. If the evidence was directed to the particular witness, evidence and circumstances, that may be useful.’

^[133]See, eg, Judge 35 (‘a topic that approaches the obvious’); Judge 37 (‘a matter of common human experience’). Some judges also expressed stronger negative views about the utility of psychological evidence generally: see, eg, Judge 28 (‘[i]n my experience it is among the most unreliable evidence I have ever read or heard’); Judge 39 (‘[l]awyers should know more about this topic than psychologists. Psychobabble can be distracting’).

^[134]Judge 62. See also Judge A14: ‘I feel that is a general issue that every single judge and every lawyer has to understand as part of their work and so we have to have education around that.’

^[135]That is, a witness giving factual evidence (as distinct, for example, from expert opinion evidence) who is neither a claimant nor defendant in the proceedings: Gans, Palmer and Roberts (n 3) 183, 194–5.

^[136]It is to be noted that this reform would also implicitly attenuate litigation privilege, because requiring the record to be made available to the other parties would necessarily involve removing privilege from it: see Passmore (n 37) 2–3.

^[143]See Judge B16: ‘I certainly agree that it is wrong for a party to be able to engage in conduct behind closed doors which has the effect of distorting the evidence, present the evidence to the Court as if it’s the witness’s actual recollection and then resist an inquiry into what had happened.’

^[144]It is also noteworthy that 21% of the judges ‘neither agreed nor disagreed’ with the proposal, with some indicating in comments that they would need to reflect on the reform before expressing a view. See, eg, Judge 57 (‘I don’t consider I’ve been able to give this question sufficient thought to give a sensible answer’); Judge 69 (‘[w]orth looking at but have not thought through how it would work’). Further, some judges who did not embrace the reform in their survey responses acknowledged in interview that, on reflection, their initial reaction may have been too hasty. For instance, Judge A10, who ‘disagreed’ in their survey response, said: ‘I know my instinctive reaction was negative, but perhaps I was thinking about it too superficially ... Maybe it’s not a bad idea’.

^[147]Steele, Chin and van Golde, Judicial Survey (n 11) 11. Presented with this proposition, judges’ responses were ‘strongly disagree’ (1%), ‘disagree’ (24%), ‘neither agree nor disagree’ (11%), ‘agree’ (53%) and ‘strongly agree’ (11%).

^[149]See, eg, Judge 9 (‘I agree that guidance is desirable, but it could take other forms (ie need not be in the professional conduct rules)’); Judge 44 (‘[t]here are limits to how prescriptive a rule in this field can be’).

^[150]As Judge B21 recalled of their own training: ‘it was just learning on the job — you watched a solicitor sit down and carefully reconstruct for someone what they thought it was they remembered’.

^[152]See, eg, Judge A8: ‘I strongly agree that additional guidance should be provided. I think you see enough evidence that the profession doesn’t know what the proper boundaries are’.

^[153]Judge 34; Australian Bar Association (n 146) 11–12 rr 68–9. See also LPUC Barristers Rules (n 22) rr 69–70. As Judge 57 stated: ‘Without the professional conduct rules giving prominence to this issue practitioners will develop ... poor and inappropriate practices ... Rules cannot by themselves change conduct but are norm-setting and importantly so.’

^[154]Judge D25. See also Judge A15: ‘I don’t know how you reduce some of these things to rules’.

^[155]Judge A42. See also Judge C18: ‘I can’t think of anything but teaching by analogy and examples.’ Some judges suggested that seeking to develop more detailed rules might even be counter-productive: see, eg, Judge 52 (‘[b]y seeking to legislate in more detail, the result may be that the general ethical rules which govern this subject are in fact diluted’).

^[156]See above Table 4. This is an observation that has also been taken up in further detail in the context of the lead author’s recent review of Practice Direction 57AC (n 7): Steele (n 5) 178–80.

^[161]See, eg, Judge 39: ‘Just as would happen in oral evidence.’ This analogy also seems to be the underlying rationale for the requirements in Practice Direction 57AC (n 7): see at app paras 2.4–2.6.

^[162]This is also a reform adopted in the UK where, since April 2021, Practice Direction 57AC (n 7) has required that ‘[o]n important disputed matters of fact, a trial witness statement should, if practicable ... state in the witness’s own words how well they recall the matters addressed’: at app para 3.7.

^[167]Judge 28. There is also psychological research supporting the conclusion that people may, in any event, lack the metacognitive ability to assess the reliability of their own memories: see generally Timothy D Wilson and Elizabeth W Dunn, ‘Self Knowledge: Its Limits, Value, and Potential for Improvement’ (2004) 55 Annual Review of Psychology 493, 494–504, 513.

^[169]For example, one judge suggested a more inquisitorial role for the court in examining witnesses: Judge 19. Another suggestion was requiring witnesses to write out their own account of events: Judge 28. Interestingly, in a variation on the idea of requiring witness interviews to be recorded, one judge suggested a requirement for more than one practitioner to be present at witness interviews on the basis that the possibility of knowledge of what occurred at an interview becoming known to others might act as a brake on poor practice: Judge 24.

^[171]That is, allowing the parties to cross-examine witnesses under oath prior to the hearing. See, eg, Judge C3: ‘I favour the idea of “controlled” use of pre-trial depositions (ie, by leave or limited in number) to allow parties to test witness evidence before trial’. Depositions are a procedure commonly used in civil litigation in the US: see generally Michael J Legg, ‘The United States Deposition: Time for Adoption in Australian Civil Procedure?’ [2007] MelbULawRw 6; (2007) 31(1) Melbourne University Law Review 146, 146–7; Emmett (n 13) 459.

^[172]That is, 30–65% of the time, depending on the defect: see above Table 3.

^[176] Ibid. The cross-examiner prudently left it at that, so how the affidavit came to be so wrong was not explored.

^[178] Yun-chien Chang and Daniel Klerman, ‘Settlement around the World: Settlement Rates in the Largest Economies’ (2022) 14(1) Journal of Legal Analysis 80, 82, 93.

^[179]Our findings also have potential implications for the increasing use by courts of paper-based or online hearings in the wake of the COVID-19 pandemic. In particular, a concern that the taking of controversial witness evidence online blunts the ability of the cross-examiner to most effectively test the credibility of that evidence has been expressed: see, eg, Campaign Master (UK) Ltd v Forty Two International Pty Ltd [No 3] [2009] FCA 1306; (2009) 181 FCR 152, 171 [78] (Buchanan J).

^[180]See above Part IV. This is a view that is supported by a substantial body of psychological research highlighting the ways in which practices commonly used in witness preparation have the capacity to corrupt witness memory: see, eg, Steele (n 5) 161–5. See above n 117 and accompanying text.

^[181] See above Part IV. For instance, although 57% of the judges considered that it is generally unacceptable to seek to refresh a witness’s recollection by showing them contemporaneous documents not created or seen by them at the time, this has been held to be ‘normal practice’, with no ‘rule of law or professional ethics to the contrary’: Lindsay-Owen v Lake [2000] NSWSC 1046, [3] (Hodgson CJ in Eq).

^[183] See above n 60, Parts III–V. See, eg, Judge 7 (‘[there] needs to be a more complete statement of what is right and wrong’); Judge A8 (‘I think it’s a massive hole that people prefer not to look at’); Judge A53 (‘I am strongly of the view that the process of taking witness statements leads to massive distortions of memory’); Judge B21 (‘[t]his is a long, long way from where we need it to be’).

^[184]By this we mean that a judge will seldom know with certainty the extent to which a witness’s evidence was reliable or what impact the witness preparation process may have had on the reliability of the evidence. It follows that judges’ views on such matters are necessarily based on impression.

^[185]Such considerations highlight the practical impossibility of definitively establishing the real world impact of witness preparation on witness memory and, commensurately, the need to approach that question as a matter of probability (ie by comparing typical witness preparation practices with what psychological research indicates about their likely impact on witness memory) or by recourse to the judgment and experience of participants in the process (as we have tried to do through our survey): see Gary Edmond et al, ‘A Warning about Judicial Directions and Warnings’ (2023) 44(1) Adelaide Law Review 194, 224–36.

^[187]In that regard, our results also confirm and build on the evidence of the same concerns expressed in a report of the International Chamber of Commerce and in the report of the Witness Evidence Working Group in the UK that preceded the promulgation of Practice Direction 57AC (n 7): Accuracy of Fact Witness Memory (n 8) 5–7; Witness Evidence Working Group (n 7) 1.

^[189]Seventy-four per cent of the judges either ‘agreed’ (50%) or ‘strongly agreed’ (24%) with, and only 4% of the judges expressed any disagreement with, this suggestion.

^[190]‘This enables lawyers to present witnesses who are thoroughly familiar with the subject matter of their testimony and who are prepared to say what they know in a clear, coherent manner’: Craig (n 115) 9 (Ipp J). See also Thomas v SMP (International) Pty Ltd [2010] NSWSC 822, [19], [22] (Pembroke J). ‘[I]t is also clear that in Australia and New Zealand a substantial degree of witness preparation is accepted as a part of good litigation practice’: Mahoney (n 115) 299. Indeed, this is accepted as proper practice even in the case of expert evidence: New Aim (n 33) 217 [112], 219 [119]–[120] (Kenny, Moshinsky, Banks-Smith, Thawley and Cheeseman JJ). See generally Evidence Act 1995 (Cth); Evidence Act 1995 (NSW); Evidence Act 2008 (Vic).

^[191]In contrast, in civil law systems, which prefer the ‘inquisitorial’ model of justice, the parties and their lawyers are not permitted the opportunity to engage with or prepare a witness; the witness is simply called by the court and gives their evidence ‘cold’: Sergey V Vasiliev, ‘From Liberal Extremity to Safe Mainstream? The Comparative Controversies of Witness Preparation in the United States’ (2011) 9(2) International Commentary on Evidence 5:1–67, 7, 26–8.

^[192]‘The duty of counsel to his client ... is to make every honest endeavour to succeed’: Tombling v Universal Bulb Co Ltd [1951] 2 TLR 289, 297 (Denning LJ). Lawyers are thus required to ‘promote and protect fearlessly and by all proper and lawful means the client’s best interests’: LPUC Barristers Rules (n 22) r 35. See also LPUL Solicitors Rules (n 22) r 4.1.1.

^[194]This is embodied in the rule that a legal practitioner will not be acting unethically in ‘drawing the witness’s attention to inconsistencies or other difficulties with the evidence’: LPUC Barristers Rules (n 22) rr 69(b), 70; LPUL Solicitors Rules (n 22) r 24.

^[197] See Wade and Cartwright-Finch (n 9) 3; Asjeet S Lamba and Ian Ramsay, ‘Commercial Litigation in Australia: An Empirical Study’ (2015) 4(1) Journal of Civil Litigation and Practice 22, 27–8.

^[200]A lecture provided to new members of the New South Wales Bar Association during the period of their reading was said to have included a statement that ‘[t]o obtain comprehensive accounts of what took place and what was said it is often necessary by drawing upon one’s imagination to prod and stimulate a client so that [they] ultimately [are] able to give a full and convincing account’: Sheppard (n 115) 33.

^[202]See, eg, Judge C14: ‘[i]t could be quite difficult given the complex and nuanced nature of the process and its different functions — investigating, gathering evidence, testing etc’. These problems are also reflected in the frequent use in the ‘Statement of Best Practice’ appended to Practice Direction 57AC (n 7) of numerous qualifications to the prescribed rules: at app. For instance, those rules enjoin practitioners to (a) avoid ‘so far as possible’ any practice that might alter or influence the recollection of the witness: at app para 3.10; (b) use ‘[p]articular caution’ in showing a witness a document they did not create or see at the time: at app para 3.4; (c) state ‘if practicable’ whether, and if so how, the witness’s recollection in relation to important disputed matters of fact has been refreshed by reference to documents: at app para 3.7; (d) use as few drafts ‘as practicable’: at app para 3.8; and (e) generally avoid leading questions ‘where possible’: at app para 3.16. See also Steele (n 5) 178–80.

^[205] See, eg, Accuracy of Fact Witness Memory (n 8) 31–40; British Psychological Society (n 5) 29.

^[206]Accuracy of Fact Witness Memory (n 8) 22. Even experienced and trained interviewers can find it difficult to avoid such practices. A 1991–92 British study of interviews of eyewitnesses by trained police officers, for instance, found that approximately one in every six questions posed was suggestive in some way: Ronald P Fisher, ‘Interviewing Victims and Witnesses of Crime’ (1995) 1(4) Psychology, Public Policy, and Law 732, 740, citing Richard George and Brian Clifford, ‘Making the Most of Witnesses’ (1992) 8(3) Policing 185.

^[208] There is also the obvious danger that educating less scrupulous practitioners as to the extent of the susceptibility of witness memory to suggestion may exacerbate, rather than remedy, the problem.

^[210]Louis D Brandeis wrote this when arguing for the full disclosure of promoters’ fees in security offerings: Louis D Brandeis, ‘What Publicity Can Do’ (20 December 1913) Harper’s Weekly 10.

^[213]The rationale for litigation privilege, which is distinct from the privilege that protects legal advice, is thus said to be ‘the need for a protected area to facilitate investigation and preparation of a case for trial by the adversarial advocate’: Leighton Contractors (n 39) 284 [16] (McLure JA), quoting Sharpe (n 40) 165. See also J-Corp Ltd v Australian Builders Labourers Federated Union of Workers (Western Australian Branch) [No 1] (1992) 38 FCR 452, 457 (French J); Secretary of State for Trade and Industry v Baker [1998] Ch 356, 365–6 (Scott V-C).

^[214]It could also potentially be argued that putting before a court witness evidence on the implicit basis that it is the bona fide, uncorrupted recollection of the witness is conduct ‘inconsistent with the maintenance of the confidentiality of the communication’ and thereby grounds an argument for a waiver of privilege: see Mann v Carnell [1999] HCA 66; (1999) 201 CLR 1, 13 [29] (Gleeson CJ, Gaudron, Gummow and Callinan JJ). Additionally, ‘[t]he holder of the privilege should not be able to abuse it by using it to create an inaccurate perception of the protected communication’: A-G (NT) v Maurice [1986] HCA 80; (1986) 161 CLR 475, 488 (Mason and Brennan JJ). See also at 487 (Mason and Brennan JJ). As Judge A53 pondered in interview: ‘I wonder if there wouldn’t be an argument of waiver once a witness statement is relied upon, but the point is never taken because no one wants to open Pandora’s box.’

^[218]See Judge 8: ‘I think it is preferable to make the reforms to professional conduct rules’. See also Judge 34 (‘[t]he undesirable practices listed above should be first addressed in professional conduct rules’); Judge D39 (‘[m]y view is that greater guidance, clear guidance, to practitioners would do a great deal and take the issue much further forward ... I would certainly do that first’).

^[220] Author contributions are as follows: (a) conceptualisation: Mark J Steele, (b) data curation: Mark J Steele and Jason M Chin, (c) formal analysis: Mark J Steele and Jason M Chin, (d) investigation: Mark J Steele, (e) methodology: Mark J Steele, Jason M Chin and Celine van Golde, (f) project administration: Mark J Steele, (g) supervision: Jason M Chin and Celine van Golde, (h) visualisation: Jason M Chin, (i) writing (original draft): Mark J Steele and (j) writing (review & editing): Mark J Steele, Jason M Chin and Celine van Golde. Author contributions were created using the Tenzing app: see generally Alex O Holcombe et al, ‘Documenting Contributions to Scholarly Articles Using CRediT and Tenzing’ (2020) 15(12) PLOS ONE e0244611:1–11.

^[221]In the view of the authors there is considerable scope for further research in this area and we would be interested in collaborating with others on this. For instance, it would be timely and useful for there to be some empirical exploration (potentially via surveys of practitioners and judges) of the ways in which Practice Direction 57AC (n 7) has changed witness preparation practice in the UK and the perceived impact that this has had on the reliability of witness evidence there. Another opportunity for useful empirical research would be a study (potentially a survey of practitioners or a ‘case study’) to establish the extent to which civil lawyers in Australia are using practices in their day-to-day preparation of witnesses to give evidence that have clear potential to distort the memories of those witnesses.

Melbourne University Law Review