If the Appellate Division, in reversing the motion court’s grant of suppression, had performed an attenuation analysis and consequently concluded that defendant’s act of physically contacting a police officer was not in fact directly attributable and proportionately responsive to the preceding official illegality, I would agree that the appeal should be dismissed since it would then present an inquiry turning in essential part on factual findings by the Appellate Division unreviewable by this Court (see CPL 450.90 [2] [a]). The Appellate Division, however, decided the matter “on the law,” and while that is not from this Court’s perspective a binding characterization, it is accurately descriptive of the Court’s decision. The problem now presented—a purely legal one—is that while purporting to make a finding of attenuation, the Court eschewed the analysis upon which such a finding must rest.

The evidence adduced at the hearing upon defendant’s suppression motion showed that, upon observing defendant walking in the early morning in an area near a public housing project known for its high incidence of crime, Police Officer David Porras asked defendant to stop. Officer Porras inquired whether defendant lived in the project, and defendant replied that he did not. The officer then asked defendant for photo identification, which defendant provided. Officer Porras examined defendant’s identification and found it satisfactory; he acknowledged that his inquiry had at that point run its course. Nonetheless, he did not return defendant’s photo identification and allow him to go on his way. Rather, he retained the identification and stood by without interceding as one of his partners approached and repeated the inquiries he, Porras, had just completed. Defendant reportedly became agitated and, at the approach of yet a third police officer, either pushed or punched Officer Porras1 in what the officers understood as an attempt “to go.” A scuffle ensued in the aftermath of which defendant was arrested for *843assault and disorderly conduct, and then searched. Drugs were found on his person.

In granting defendant’s motion to suppress the drugs thus recovered, the motion court found that defendant had been illegally detained and that the illegality had not been attenuated by defendant’s attempt to get past the police officers “blocking [his] egress.” The court observed,

“[e]ven if there was a basis for initially requesting information from defendant . . . any such justification was exhausted after he answered Porr[a]s who was obligated to return the identification and allow him to leave. The continued detention was unlawful and the reaction of defendant proportionate to the circumstances. It does not attenuate the unlawful detention and render the contraband admissible.”

The Appellate Division, while disagreeing with the motion court’s conclusion that the primary official illegality had not been attenuated by defendant’s conduct, was assertedly disinterested with the nature and quality of the illegality or of the attenuating conduct. Indeed, the Court expressly found it unnecessary to decide whether there had been any official illegality (74 AD3d 520, 522 [1st Dept 2010]), since “[o]nce defendant punched Officer Porras, any allegedly unlawful conduct in stopping and questioning defendant was attenuated by his calculated, aggressive and wholly distinct conduct” (id. at 521 [emphasis added]). And, while the Appellate Division characterized the attenuating conduct as a “punch,” that characterization was ultimately unimportant to its analysis, since the Court found it “of no moment whether defendant punched or pushed Officer Porras, because . . . the police officers did not initiate or attempt to initiate physical contact with defendant” (id. at 522).

The dissenting Justices, by contrast, after a thorough review of the record facts to ascertain precisely what had transpired and the manner in which defendant’s alleged attenuative conduct was related to his immediately preceding detention, concluded:

“In view of the officers’ concession that defendant was trying to get away from them, the documentary evidence and Porras’s eventual admission at the hearing that defendant pushed him, the limited physical force used against Porras by defendant was *844an immediate response to his unjustified detention. It does not constitute an independent act sufficiently attenuated from the unlawful detention so as to dissipate the illegal taint associated with it (cf. Townes, 41 NY2d at 101-102), but was an immediate and direct consequence of that unlawful detention” (74 AD3d at 529).

Ordinarily, a finding of attenuation in the search and seizure context entails an analysis taking into account the specific nature of the illegal intrusion and the timing and quality of the subsequent, allegedly attenuating act or circumstance. Where, as in the present matter, it is claimed that the defendant’s act attenuated the primary illegality, “[t]he test to be applied is whether defendant’s action . . . was spontaneous and precipitated by the illegality or whether it was a calculated act not provoked by the unlawful police activity and thus attenuated from it” (People v Wilkerson, 64 NY2d 749, 750 [1984]).

Here, the Appellate Division simply concluded that any alleged illegality, no matter how extreme or provocative,2 must have been attenuated by defendant’s act of initiating physical contact with Officer Porras, even if that act consisted only of a push to get past the three officers illegally blocking his way. This was, in the end, not an attenuation analysis at all, but simply the announcement of an arbitrary rule that any physical contact with a police officer—no matter what its force or purpose—if not preceded by an attempt by the officer to contact the defendant, will be deemed distinct and unattributable to any precedent official illegality, no matter how provocative. Certainly, there is nothing in our cases that could be construed as permitting the substitution of such a rule for an attenuation analysis.

I do not say that the Appellate Division was bound to resolve the attenuation issue as the motion court had and the dissenting Justices would have, only that if it was to decide the question differently it had to perform an attenuation analysis—that is to say it had to make a conscientious effort to ascertain whether defendant’s allegedly attenuating act was or was not directly precipitated by and a proportionate response to his *845closely antecedent illegal detention.3 To do this, it would have been essential for the Court to make some finding as to the nature and quality of the official illegality and defendant’s following act. Inasmuch as the Court expressly found that such findings were either dispensable or irrelevant, and instead merely concluded that defendant’s initiation of contact was per se attenuative, whatever the nature of the contact or of any antecedent official provocation, I believe that the correct disposition of this appeal is to remit to the Appellate Division for a proper determination of the attenuation issue.

In instead pigeonholing this appeal as one involving a “mixed question,” the Court makes a choice that is not only unsound jurisdictionally, but erosive of this Court’s role in articulating the law governing police-civilian encounters. The doctrine of attenuation in the search and seizure context is of course nothing more than a closely limited exception to the general, dominant rule that police intrusions must be justified at their inception (see Terry v Ohio, 392 US 1, 19-20 [1968]; People v Cantor, 36 NY2d 106, 111 [1975]; People v Moore, 6 NY3d 496, 498 [2006]). If the exception is not to swallow the rule, care must be taken to assure that the doctrine is correctly employed. When courts with the factual jurisdiction to make attenuation findings employ facile analytic shortcuts operating to shield from judicial scrutiny illegal and possibly highly provocative police conduct, an issue of law is presented that is, I believe, this Court’s proper function to resolve. The alternative is to turn a blind eye to “tactics . . . [under which] any person might be approached, detained, intimidated, harassed, even provoked into a display of aggression and thereupon arrested, effectively eviscerating Fourth Amendment protections and ‘abandoning] the law-abiding citizen to the police officer’s whim or caprice’ (Cantor, 36 NY2d at 112)” (74 AD3d at 530 [Tom, J., dissenting]). This is not an exaggerated or purely academic concern in a jurisdiction where, as is now a matter of public record, hundreds of thousands of pedestrian stops are performed annually by the police,4 only a very small percentage of which actually result in the discovery of evidence of crime.

*846Judges Graffeo, Read, Smith, Pigott and Jones concur; Chief Judge Lippman dissents in an opinion in which Judge Ciparick concurs.

Appeal dismissed, in a memorandum.

OPINION OF THE COURT

Memorandum.

The appeal should be dismissed upon the ground that the reversal by the Appellate Division (74 AD3d 520 [1st Dept 2010]) was not on the law alone within the meaning of CPL 450.90 (2) (a).*

Here, the Appellate Division’s reversal of Supreme Court’s order granting suppression, while termed “on the law,” was actually predicated upon a differing view concerning the issue of attenuation, which is a mixed question of law and fact. A reversal *842on a mixed question typically does not meet the requisites of CPL 450.90 (2) (a) (see People v Mayorga, 64 NY2d 864, 865 [1985] [dismissing an appeal from an Appellate Division order of reversal involving mixed question of whether “there has been an attenuating break in an interrogation”]; People v Lawrence, 74 NY2d 732 [1989]; People v Howard, 74 NY2d 943 [1989]; People v Hinton, 81 NY2d 867 [1993]).