(dissenting). My vote in this case is to affirm the order upholding the judgment of conviction, upon a jury verdict, for manslaughter in the first degree, attempted murder in the second degree and several other associated crimes. I disagree with the Court’s application of the per se reversal and new suppression hearing remedy for the pretrial Rosario violation (People v Rosario, 9 NY2d 286, rearg denied 9 NY2d 908, cert denied 368 US 866). This is not simply a refusal to create "another exception” to the per se remedy, but rather represents a new variety of expansion of that extreme remedy into the pretrial phase of the criminal adjudication process (contrast, People v Jackson, 78 NY2d 638; People v Bin Wahad, 79 NY2d 787).
Thus, while I would not conclude that the Sheridan and Yella instances constitute Rosario trial violations on this record meriting a new trial, I refrain from joining issue in those respects with the majority of the Court because only the pretrial Rosario aspect, with whose resolution, analysis and discussion, I thoroughly disagree, constitutes for me the crucial, new feature of this case.
A revisitation to the origins of the Rosario rule graphically portrays how far the pendulum has swung. In that seminal case, this Court held that the People must disclose all prior statements of its witnesses based on a "right sense of justice” (People v Rosario, 9 NY2d, at 289, supra). The Court then affirmed a conviction for murder in the first degree, cautiously and wisely observing that:
"This court is exceedingly] slow to disregard error as harmless under section 542 of the Code of Criminal Procedure, particularly in a capital case. However, in the one before us, we are as convinced as judges may ever be, in view of the overwhelming proof of guilt and the absence of any real inconsistency between prior statement and trial testimony, that the jury would not have decided the case differently even if defense counsel had had the use of the statements in question. In other words, it may not be said that any substantial right of the appellant was prejudiced by the trial court’s erroneous ruling” (id., at 291).
That articulation and application by the Court, which created *623the Rosario rule, thus instructively reflected a respect for the common law and legislative mandates for proportionality. Decisions subsequent to Rosario continued, for a while, to apply Rosario’s own harmless error proportionality (see, People v Hawa, 13 NY2d 718; People v Fasano, 11 NY2d 436, after remittitur 18 AD2d 1137, mod 14 NY2d 43; People v Pereira, 11 NY2d 784, cert denied 370 US 962; People v Hernandez, 10 NY2d 774, rearg denied 10 NY2d 885, cert denied sub nom. Ramirez v New York, 369 US 831; People v Turner, 10 NY2d 839, cert denied 369 US 807; People v Hurst, 10 NY2d 939). With respect to Rosario trial errors, the new per se era dawned with People v Consolazio (40 NY2d 446, cert denied 433 US 914), People v Perez (65 NY2d 154, after new trial 154 AD2d 485, lv denied 75 NY2d 774), and, in my view, reached its nadir with People v Ranghelle (69 NY2d 56), and People v Jones (70 NY2d 547).
This tracing of the history leads me to the conclusion that the Court has strayed far from the course plotted and intended by the primary precedent, Rosario itself. This Court’s premier purpose and strength, as a case-by-case common-law adjudicatory body, suffers when it increasingly surrenders its process to automatically triggered rubrics. To be sure, the Court should and does make precedential principles, but it does so incrementally, not by sweeping edicts which imitate legislative black-letter ukases. Individualized justice — Rosario’s own "right sense of justice” — is abandoned in both the jurisprudential and practical sense by such per se nostrums. Instead, what emerges is an antithesis of a "right sense of justice” — per se-ness.
I believe it is persuasive, though not controlling, that the Court only recently and painstakingly explored and decided on policy grounds not to inject the per se remedy into the CPL article 440 postjudgment stream of criminal matters (People v Jackson, 78 NY2d 638, supra). Distinct weighty reasons, including the historical perspective and empirical wisdom from repeated applications of the Rosario rule, warrant corresponding judicial forbearance from forcing this latest form of the per se remedy into pretrial phases of criminal proceedings.
While I agree that a pretrial Rosario violation occurred in this case when investigating Officer Toto’s memo book was not turned over for cross-examination at the Huntley hearing, I note also that another police officer’s book was mistakenly turned over and this miscue was not discovered — or revealed —by anyone, including defense counsel, at that time. Then, *624the correct cross-examination material was turned over to defense counsel at trial and was used to cross-examine Toto before the petit jury. Thus, the trial court was not left to speculate as to "what portions may [have been] useful to the defense” or how "single-minded” defense counsel might have used those portions at the pretrial violation stage (see, People v Perez, 65 NY2d 154, 160, supra). Since defense counsel was able to use the materials to attempt to discredit the People’s witness at trial, a record template exists for the trial court to fashion an appropriate on-the-spot remedy or for appellate courts to gouge the harm, if any, caused by the absence of the material at the pretrial proceedings (contra, majority opn, at 615). It therefore cannot be said, under these circumstances, that "there is no way, short of speculation, of determining how [the materials] might have been used or how [their] denial to counsel might have damaged defendant’s case” (People v Jones, 70 NY2d 547, 552, supra). Rather, courts in such instances have better than usual tools and methods to conduct an effective prejudice/harmless error analysis "weighing the potential impeachment value of the withheld material” (majority opn, at 615).
Further support for a more proportioned approach may also be found in a variety of this Court’s precedents dealing distinctly with the pretrial versus trial dichotomy. In People v Chipp (75 NY2d 327, cert denied 498 US 833), this Court stressed that the claimed right of the defendant to call the complainant as a witness at the Wade hearing must be evaluated in light of defendant’s ability "to examine the witness and explore issues relevant to identification at trial” (id., at 338). We further noted that "[a]ny improprieties in pretrial identification procedures can be investigated under the ’time-honored process of cross-examination’ ” at trial (id., at 338, quoting Watkins v Sowders, 449 US 341, 349). In People v Castillo (80 NY2d 578 [decided today]), the Court states that ”[t]he due process requirements for a hearing may be less elaborate and demanding than those at the trial proper” (id., at 582 [citations omitted for other such illustrations]; see also, People v Morales, 80 NY2d 450, 457).
The balance of remedies approach supports a prejudice/ harmless error analysis for whatever range of corrective ac tian is contemplated where the focus of concern in the pretrial hearing — adequate cross-examination — is effectively reproduced at trial before the critical adjudicator of guilt or inno*625cence — the jury. In this case, Toto’s credibility was finally and fully explored and subjected to challenge with all required and relevant materials at the time when it really counted. That opportunity subsumes, at least in part, the failure to allow for the earlier exploration on pretrial cross-examination. While there is no full substitute for compliance in the earlier instance, neither is there justification to add another sweeping per se corrective remedy for a demonstrably unmeaningful pretrial Rosario lapse. To do so once again catapults the Rosario remedy into a paramount hierarchical ranking above all other kinds of procedural protections and violations (see, People v Jones, 70 NY2d 547, 555, supra [concurring opn]).
The case before us deals with a pretrial hearing on the voluntariness of the defendant’s admissions to the crimes charged. The determination of that question is made by the court alone at the pretrial stage, and then is reviewed again and may be resolved differently by the jury under the dictates of CPL 60.45 and CPL 710.70 (3) (see, People v Graham, 55 NY2d 144). Thus, this additional procedural safeguard further warrants some proportionality in implementing the Rosario remedy in this circumstance because, apart from being able to assess possible harm against a record template, at least some of that harm, if any, based on the pretrial Rosario violation may be directly mitigated and addressed by defense counsel’s use before the petit jury of the Rosario material. A proportionate remedy, short of a new hearing, may, for example, include informing the jury of the pretrial violation during the trial cross-examination and in the instructions to the jury. We have said in another, apt Rosario connection: "The determination of what is appropriate is committed to the trial court’s sound discretion, and while the degree of prosecutorial fault may be considered, the court’s attention should focus primarily on the overriding need to eliminate prejudice to the defendant.” (People v Martinez, 71 NY2d 937, 940.)
The urgency for the Legislature to act and remedy this situation cannot be overemphasized. Five years ago, in People v Jones (70 NY2d 547, supra), I expressed a strongly held concern that: "The new per se error rule unavoidably plants an uncertainty into every tried criminal case. It is a law enforcer’s nightmare and a perpetrator’s delight. Insofar as the rule is not constitutionally rooted, I believe it would be useful for the Legislature to consider a specification in CPL 470.05 and article 240, especially CPL 240.20 and 240.45, overcoming the per se-ness of this exalted court-made rule” *626(id., at 557). That observation was, in part, premised on another perceived flaw:
"This relatively new absolute rule even blinks the legislative direction * * * that '[a]n appellate court must determine an appeal without regard to technical errors or defects which do not affect the substantial rights of the parties’. It must be supposed that the court overrides the statutory prescription by the fiat that all Rosario errors affect 'substantial rights’ because the court says so. Since the court is the final word, that may be so, even if it is also tautological. The finding in both courts below that the particular nondisclosure was truly harmless, merely duplicative and trivial, accents the extravagance of a per se rule made by a fallible institution which tolerates no imperfection in the law enforcement authorities whose conduct they govern. While the mission of the courts is to render justice, the 'right sense of justice’ referred to in Rosario itself implies an evenhandedness, not the imbalance created by a per se rule of this kind” (id., at 556-557).
United States Supreme Court Justice Robert Jackson sagely admonished in an analogous context that applies as well to this Court: "[The Supreme Court is] not final because [it is] infallible, but [is] infallible only because [it is] final.” (Brown v Allen, 344 US 443, 540 [Jackson, J., concurring], reh denied 345 US 946.)
Perhaps the Legislature agrees with this Court’s toe-in-the-water per se probe in Consolazio, which has since cascaded into a full body immersion (People v Ranghelle, 69 NY2d 56, supra; People v Perez, 65 NY2d 154, supra; People v Jones, 70 NY2d 547, supra; and majority opn herein). If so, the Legislature ought at least to harmonize the court-made rule that countermands the legislative direction to the courts to disregard "technical errors or defects” (CPL 470.05 [1]).
On the other hand, the Legislature might more comprehensively consider halting the wide swing of the Rosario pendulum and restoring it to a proper place of perspective and proportion. One way to accomplish this and effect a sensibly balanced remedy would be open-file pretrial discovery on the criminal side. New York prosecutors, long opposed to this reform, may at long last recognize the lesser price paid by *627society and victims from prosecutors opening their files to defense counsel in the pretrial discovery phase compared to risking the turmoil and uncertainty of the massive Rosario mine field in countless individual cases. In the end, such a modulated procedural reform would more surely bring about fairness for and less games-playing by both sides, and would definitively restore "a right sense of justice” as an antidote to the spreading threat of the per se virus.
Acting Chief Judge Simons and Judges Titone, Hancock, Jr., and Smith concur with Judge Kaye; Judge Bellacosa dissents and votes to affirm in a separate opinion.
Order reversed, etc.