I concur in the reversal of eight of the nine multiple-murder special-circumstance findings. As to the affirmance of defendant’s convictions, the remaining special circumstances and perforce the penalty, I dissent.
Defendant requested that his appointed attorney be replaced by retained counsel whom the trial court at one point found to be burdened with a conflict of interest by reason of counsel’s prior contact with a key prosecution witness. (See maj. opn., ante, at p. 829.) In addition, there were clear indications that defendant’s retainer agreement included a literary-rights clause in favor of counsel, placing squarely before the court the possibility of yet a second conflict of interest. Absent further inquiry by the court, and an express and effective waiver by defendant of his constitutional right to unconflicted counsel, the court had a duty to deny defendant’s motion for substitution of counsel burdened with such conflicts of interest. (See Wood v. Georgia (1981) 450 U.S. 261, 271-274 [67 L.Ed.2d 220, 230-232, 101 S.Ct. 1097]; see also Wheat v. United States (1988) 486 U.S. 153, _ [100 L.Ed.2d 140, 149-152, 108 S.Ct. 1692].) No such waiver was obtained, and the court failed to fully explore the potential for conflict on either score; yet inexplicably it granted defendant’s motion to substitute counsel and permitted the trial to proceed.
In pages of admirable scholarship, the majority set out the law applicable to a claim on appeal that defendant’s counsel was burdened by an actual or potential conflict of interest. I have no quarrel with this learned exegesis. The majority go on to roundly defeat the People’s multilayered claim that defendant waived the conflict. Again, I have no quarrel. Yet when we come to the application of the law to the facts, the majority stumble and fall. Under established principles of law (Wood v. Georgia, supra, 450 U.S. at pp. 272-274 [67 L.Ed.2d at pp. 230-232]; United States v. Winkle (10th Cir. 1983) 722 F.2d 605, 608-612; Brien v. United States (1st Cir. 1982) 695 F.2d 10, 14-15), defendant’s convictions must be vacated and remanded to the trial court for a determination (1) whether a literary-rights agreement actually existed and (2) whether there was an actual conflict of interest arising from such agreement and from counsel’s prior relationship with witness Munro adversely affecting counsel’s performance such as to require reversal.
*859I am well aware of the pain and horror that this defendant has inflicted on so many. I too feel that his crimes are so shocking and heinous that it seems an outrage to question the lawfulness of the verdict in any way. Yet I am convinced that an impartial application of the law to the facts requires us to remand the matter to the trial court for an examination whether counsel operated under an actual conflict of interest which adversely affected his performance.
The majority find as to the possible existence of a literary-rights agreement that “[t]he court cannot be deemed to have known, or to have had reason to know, of the possibility of a conflict in this regard.” (Maj. opn., ante, at p. 838.) They hold that “a court can be held to have knowledge or notice of the possibility of a conflict only when ... it is provided with evidence of the existence of a conflict situation . . .” (id. at p. 838, italics in original, citing Wood, supra, 450 U.S. 261), and that no such “evidence” was present here.
It is difficult to understand what “evidence” the majority would require. The general rule has been that the duty of inquiry arises when the court “knows or reasonably should know” that a potential conflict exists. (See Cuyler v. Sullivan (1980) 446 U.S. 335, 347 [64 L.Ed.2d 333, 346, 100 S.Ct. 700]; see also United States v. Burney (10th Cir. 1985) 756 F.2d 787, 791; United States v. Winkle, supra, 722 F.2d at p. 611; Brien v. United States, supra, 695 F.2d at p. 15, fn. 10.) Wood itself established a duty to inquire into the potential for conflict when there is simply an “appearance” or “suggestion” of conflict. (Wood, supra, 450 U.S. at pp. 272, 273 [67 L.Ed.2d at pp. 230-231].) In Wood, the potential conflict was latent in the proceedings; the high court held that the trial court should have suspected a potential for conflict from defense counsel’s strategy in the case.
The defendants in Wood were employees at a pair of retail pornography establishments. They were convicted of distributing lewd materials and given heavy fines, the payment of which formed a condition of their probation. When defendants failed to pay the fines, their probation was revoked. Defense counsel leveled a constitutional attack against the statute under which defendants were convicted, rather than pleading for leniency. The court knew that the defendants’ lawyer, who had also represented them at trial, had been hired by their employer, and that the defendants had believed that their employer would also be paying their fines. The state’s representative requested that the court inquire into the possibility of a conflict of interest arising from defense counsel’s dual loyalties, but no such *860inquiry was made. The United States Supreme Court held that “the possibility of a conflict of interest was sufficiently apparent at the time of the revocation hearing to impose upon the court a duty to inquire further” (450 U.S. at p.272 [67 L.Ed.2d at pp. 230-231], italics & fn. omitted), and remanded the case “to determine whether the conflict of interest that this record strongly suggests actually existed. . . .” (Id. at p. 273 [67 L.Ed.2d at p. 231].)
The trial court in this case was on notice that an indigent defendant in a notorious case had retained counsel, who refused to deny the charge that the retainer included a literary-rights agreement. When the court tentatively questioned defendant’s retained counsel about the existence of such an agreement, he gave answers which, while nonresponsive in the strict sense, were as clearly suggestive of the existence of a possible conflict as the facts before the court in Wood.1 Yet the court inquired no further. To say that these facts merely “lay a basis for speculation” as to the existence of a literary-rights agreement (maj. opn., ante, at p. 838) is to dodge the clear implications of the Wood holding.2
A potential conflict of interest implicates more than just the defendant’s interest in effective assistance of counsel. The Supreme Court has recently recognized that a trial court justifiably may refuse to accept a defendant’s waiver of his attorney’s potential conflict of interest, in order to vindicate the state’s interest in just verdicts, in preserving the appearance of fairness, and in preventing unethical practices by attorneys. (Wheat v. United States, supra, 486 U.S. 153,_ [100 L.Ed.2d at pp. 149, 151].) The court recognized that potential conflicts may be difficult to evaluate in the murky light of pretrial hearings, but maintained that “a court confronted with and alerted to possible conflicts of interest must take adequate steps to ascertain whether the conflict warrants [different] counsel.” (Id. at p._ [100 L.Ed.2d at p. 149].) Under the circumstances of this case, the People’s *861charge that defense counsel’s retainer was in part a literary-rights agreement was utterly plausible. The state’s interest in the substance and appearance of justice, as well as its interest is preventing unethical practices by attorneys, required a full inquiry into the existence of such an agreement.
With respect to counsel’s prior relationship with witness Munro, the majority concede that the trial court committed error under Wood (supra, 450 U.S. 261) by ordering the substitution of retained counsel in place of defendant’s appointed counsel without any attempt to obtain a waiver from defendant of his constitutional right to conflict-free counsel.3 The majority misinterpret Wood, however, to require that even when a trial court has failed to inquire fully into a potential conflict, unless defendant can show from the record that counsel’s conflict of interest adversely affected counsel’s performance, his convictions must stand. (Maj. opn. at p. 843.)4 “[A]ssum[ing] for argument’s sake” that counsel had an actual conflict of interest (maj. opn. at p. 843), the majority declare themselves unable to “find or even conjecture any failing on Charvet’s part that could be attributed to any information he or his partner Stewart could conceivably have received from Munro when they discussed the possibility of representation.” (Id. at p. 843.) The reason that the majority cannot fully determine whether there was in fact an actual conflict arising from counsel’s representation of Munro is that although the trial court determined that counsel had a brief contact with Munro, it failed to inquire into and resolve the factual disputes raised by the district attorney and counsel regarding the extent to which this representation presented a potential for a conflict of interest affecting counsel’s representation of defendant. In fact, the court seemed more concerned that counsel’s contact with Munro presented some unfairness to the People, who intended to rely upon him as a central prosecution witness.
When the trial court fails to inquire into an apparent conflict of interest and fails to resolve the question whether there is an actual conflict, it leaves behind a record which is inadequate for a determination whether there was an actual conflict which adversely affected counsel’s representation of the defendant. The lesson of Wood v. Georgia, supra, 450 U.S. 261, is that when such a defective record prevents meaningful appellate review, the reviewing court should vacate the judgment below and remand the matter to the trial *862court for a hearing to determine whether there was an actual conflict which adversely affected counsel’s representation of the defendant. I would therefore vacate all counts and remand to the trial court to determine (1) whether there was a literary-rights agreement between defendant and his retained counsel, and (2) whether the conflict of interest resulting from such agreement, if it existed, and from counsel’s prior relationship with witness Munro adversely affected counsel’s performance.
Appellant’s petition for a rehearing was denied March 2, 1989.
When questioned by the court about the existence of a literary-rights agreement between defendant and himself, counsel would neither confirm nor deny it. He argued in defense of such agreements, however, and stated that the prosecution “has absolutely no right to go into my fee arrangement. . . .” (See maj. opn., ante, at pp. 828-829.)
It is true that unlike the trial court in Wood, supra, 450 U.S. 261, the court below did not know what arguments counsel planned to make on defendant’s behalf and thus could not know whether counsel’s manner of representation would in some way indicate the existence of an actual conflict. This is due simply to the timing of defendant’s request to substitute counsel, however, and does not diminish the strength of the indications that such a conflict was possible. As the high court pointed out in Wheat, the court normally must examine a potential conflict of interest “in the murk[y] pre-trial context when relationships between parties are seen through a glass, darkly.” (Wheat, supra, 486 U.S. at p._ [100 L.Ed.2d at p. 151].)
Indeed, they defend the point vigorously and at length. (See maj. opn. at pp. 838-841.)
The majority’s discussion on this point is flawed by its failure to consider vacation and remand as an alternative to outright reversal. Such an omission is puzzling since this was the disposition in Wood itself, and was also central to the discussion of that case in United States v. Winkle, supra, 722 F.2d at pages 611-612 and Brien v. United States, supra, 695 F.2d 10, 15, footnote 10, cited by the majority.