28 In Johnson v Johnson (2000) 201 CLR 488 the High Court considered the issue of a trial judge making a statement during the course of the trial in respect of which one of the parties took objection and applied for the judge to disqualify himself on the grounds that his remarks had given rise to an appearance of bias. The trial judge refused to do so. In holding that the judge's statement did not create an appearance of bias the majority (Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ) made the following observations regarding the relevant principles to be applied:
11 It is not contended that Anderson J was affected by actual bias. It has been established by a series of decisions of this Court that the test to be applied in Australia in determining whether a judge is disqualified by reason of the appearance of bias (which, in the present case, was said to take the form of prejudgment) is whether a fair-minded lay observer might reasonably apprehend that the judge might not bring an impartial and unprejudiced mind to the resolution of the question the judge is required to decide: eg Re Lusink; Ex parte Shaw (1980) 55 ALJR 12; 32 ALR 47; Livesey v New South Wales Bar Association (1983151 CLR 288; Vakauta v Kelly (1989)167 CLR 568; Webb v The Queen (1994) 181 CLR 41.
12 That test has been adopted, in preference to a differently expressed test that has been applied in England (cf Locabail (UK) Ltd v Bayfield Properties Ltd [2000] QB 451), for the reason that it gives due recognition to the fundamental principle that justice must both be done, and be seen to be done (cf R v Sussex Justices; Ex parte McCarthy [1924] 1 KB 256 at 259 per Lord Hewart CJ). It is based upon the need for public confidence in the administration of justice. "If fair-minded people reasonably apprehend or suspect that the tribunal has prejudged the case, they cannot have confidence in the decision." ( R v Watson; Ex parte Armstrong (1976) 136 CLR 248 at 263 per Barwick CJ, Gibbs, Stephen and Mason JJ). The hypothetical reasonable observer of the judge's conduct is postulated in order to emphasise that the test is objective, is founded in the need for public confidence in the judiciary, and is not based purely upon the assessment by some judges of the capacity or performance of their colleagues. At the same time, two things need to be remembered: the observer is taken to be reasonable; and the person being observed is "a professional judge whose training, tradition and oath or affirmation require [the judge] to discard the irrelevant, the immaterial and the prejudicial" ( Vakauta v Kelly (1988) 13 NSWLR 502 at 527 per McHugh JA, adopted in (1989) 167 CLR 568 at 584-585 per Toohey J).
13 Whilst the fictional observer, by reference to whom the test is formulated, is not to be assumed to have a detailed knowledge of the law, or of the character or ability of a particular judge ( Webb v The Queen (1994) 181 CLR 41 at 73 per Deane J) the reasonableness of any suggested apprehension of bias is to be considered in the context of ordinary judicial practice. The rules and conventions governing such practice are not frozen in time. They develop to take account of the exigencies of modern litigation. At the trial level, modern judges, responding to a need for more active case management, intervene in the conduct of cases to an extent that may surprise a person who came to court expecting a judge to remain, until the moment of pronouncement of judgment, as inscrutable as the Sphinx. In Vakauta v Kelly (1989) 167 CLR 568 at 571 Brennan, Deane and Gaudron JJ, referring both to trial and appellate proceedings, spoke of "the dialogue between Bench and Bar which is so helpful in the identification of real issues and real problems in a particular case." (See also Re Lusink; Ex parte Shaw (1980) 55 ALJR 12 at 15 per Murphy J; 32 ALR 47 at 53). Judges, at trial or appellate level, who, in exchanges with counsel, express tentative views which reflect a certain tendency of mind, are not on that account alone to be taken to indicate prejudgment. Judges are not expected to wait until the end of a case before they start thinking about the issues, or to sit mute while evidence is advanced and arguments are presented. On the contrary, they will often form tentative opinions on matters in issue, and counsel are usually assisted by hearing those opinions, and being given an opportunity to deal with them.
14 There was argument in this Court, prompted by Anderson J's explanation of what he intended to communicate, about whether the effect of a statement that might indicate prejudgment can be removed by a later statement which withdraws or qualifies it. Clearly, in some cases it can. So much has been expressly acknowledged in the cases eg Re JRL; Ex parte CJL (1986) 161 CLR 342 at 372 per Dawson J; Vakauta v Kelly (1989) 167 CLR 568 at 572 per Brennan, Deane and Gaudron JJ, 577 per Dawson J. No doubt some statements, or some behaviour, may produce an ineradicable apprehension of prejudgment. On other occasions, however, a preliminary impression created by what is said or done may be altered by a later statement. It depends upon the circumstances of the particular case. The hypothetical observer is no more entitled to make snap judgments than the person under observation.