(concurring).
Plaintiff William Brady and his wife, co-plaintiff Theresa Brady, have jointly advanced the federal claim that when law enforcement officers learn to, a certainty that a person they are holding in custody pursuant to a valid arrest warrant is in fact not the perpetrator of the offensé giving rise to the arrest warrant, the law enforcement officers then have an obligation under the Constitution to release that person' from custody, an obligation assertable in a law suit brought pursuant to 42 Ú.S.C. § 1983. (The plaintiffs have also advanced certain claims under Massachusetts law, but the proper disposition of those claims is not at issue on- this appeal). Since Theresa Brady’s share of the federal claim is entirely derivative from, and hence contingent on the viability of, the share of the federal claim put forward by her husband, I will, for convenience of presentation in the discussion which follows, refer to the plaintiffs’ federal claims in the singular — “the Brady claim,” or • “Brady’s claim”. Defendant police officers Maryann Dill, Kenneth J. Hudson, Jr., Douglas Mendes and Steve Yrona have questioned whether the Brady claim has any constitutional footing; further, defendants have contended that — assuming arguendo that the Brady claim is a cognizable constitm tional claim — the constitutional principle undergirding that claim was not clearly established at the time (March 4th to March 6th, 1995) William Brady was arrested and held in custody, and hence that the defendants would be shielded from any putative liability by qualified immunity.
The District Court, in ruling on defendants’ motion for summary judgment, concluded that Brady’s claim is firmly rooted in the Fourth Amendment: “If a jury finds that defendants knew that Brady was not really the man wanted by- the warrant, then they could also conclude that the defendants violated Brady’s Fourth Amendment right to be free from unreasonable seizure.” Brady v. Dill, 24 F.Supp.2d, 129, 135 (D.Mass.1998). Fur*118ther, the District Court found that the Fourth Amendment right sought to be vindicated was clearly established; the District Court rested this aspect of its analysis on the Supreme Court’s recognition, in Maryland v. Garrison, 480 U.S. 79, 86-87, 107 S.Ct. 1013, 94 L.Ed.2d 72 (1987), of the obligation of police officers executing a search warrant to desist from searching premises lying outside the scope of the search warrant the moment the officers recognize that they have travelled beyond the warrant’s boundaries. Accordingly, the District Court — while granting summary judgment in favor of two defendant officers who clearly played no role in Brady’s allegedly extended detention — declined to grant summary judgment in favor of Officers Dill, Hudson, Mendes and Vro-na. The question whether there came a time at which “defendants did know ... that they had the wrong man” was, in the District Court’s view, “the quintessential jury question.” Brady v. Dill, supra, 24 F.Supp.2d at 135 (emphasis in original).
Disagreeing with both aspects of the District Court’s ruling, this court concludes that (1) Brady has not presented a viable constitutional claim, and (2) even if the claim were to be regarded as having persuasive doctrinal footing, the published case law which can be said to demonstrate some measure of antecedent credence for the claim is so meager as to be “too slender a reed to ward off a finding of qualified immunity.” In sum, the court is of the view that defendants are doubly entitled to summary judgment. (However, this court’s reversal of the judgment of the District Court does not necessarily close this litigation: this court’s opinion is, properly, at pains “to direct the District Court to dismiss Brady’s state-law claims without prejudice for want of federal jurisdiction,” so that any potentially cognizable state-law claims may be asserted in the Massachusetts courts).
On the question whether Brady has presented a viable constitutional claim I respectfully disagree with the position announced by this court. On the question whether defendants are entitled to qualified immunity, I respectfully disagree with the position announced by the District Court. This means that I concur in the judgment of this court but not in the court’s opinion. I write separately to explicate the line of thinking that leads me to differ with both courts.
I. Is the Brady claim a viable constitutional claim?
i
This court’s opinion states: “We believe that [the Brady] claim ... is foreclosed by Baker v. McCollan, 443 U.S. 137, 99 S.Ct. 2689, 61 L.Ed.2d 433 (1979).” In Bakerüie Supreme Court held that a person arrested in conformity with a validly issued warrant who was held in custody for three days by a sheriff who made no real effort to investigate the detainee’s fully justified and readily documentable insistence that, although named in the warrant, he was in fact not the person who committed the offense charged, had no cognizable § 1983 claim against the sheriff for the three days of inappropriate detention. While deeming Baker to be adversely dispositive of the claim advanced in the case at bar, this court recognizes that Brady — and the District Court — have posited a reading of Baker that is not preclusive of the Brady claim:
In an effort to escape Baker’s, deadly embrace, Brady urges us to read the Court’s words narrowly. In his view, Baker stands only for the rule that when a person in custody protests his innocence, police officers have no affirmative obligation to investigate. Baker, Brady maintains, does not address situations in which police officers, after an arrest, come to possess “actual knowledge” that the detained person, though named in an outstanding warrant, is a victim of mistaken identities. In such an eventuality, Brady posits, the Constitution requires the officers to release the detainee without further ado. The district court en*119dorsed this cramped reading of Baker and refused to grant the appellants’ motion for summary judgment disposition because it perceived a factual dispute as to whether they had actual knowledge of Brady’s innocence.
Turning to what the Supreme Court wrote in Baker in the paragraph which concludes its Fourteenth Amendment analysis, I find it difficult to acquiesce in this court’s characterization of the Brady/district court interpretation of Baker as a “cramped reading.” What the Supreme Court wrote was this:
The Fourteenth Amendment does not protect against all deprivations of liberty. It protects only against deprivations of liberty accomplished “without due process of law.” A reasonable division of functions between law enforcement officers, committing magistrates, and judicial officers — all of whom may be potential defendants in a § 1983 action — is entirely consistent with “due process of law.” Given the requirements that arrest may be made only on probable cause and that one detained be accorded a speedy trial, we do not think a sheriff executing an arrest warrant is required by the Constitution to investigate independently every claim of innocence, whether the claim is based on mistaken identity or a defense such as lack of requisite intent. Nor is the official charged with maintaining custody of the accused named in the warrant required by the Constitution to perform an error- • free investigation of such a claim. The ultimate determination of such claims of innocence is placed in the hands of the judge and the jury.
443 U.S. at 145-146, 99 S.Ct. 2689 (footnote omitted).
The quoted paragraph announces two linked propositions of constitutional law. These propositions govern claims brought, under § 1983, by one arrested pursuant to a valid arrest warrant who, notwithstanding that he is concededly the person named in the warrant, promptly and continuously asserts his innocence, but who, nonetheless, is not promptly released by the law enforcement officers who arrested and/or subsequently detained him. The propositions — rendered in paraphrase— are these: (1) The “liberty” clause of the Fourteenth Amendment (“[N]or shall any State deprive any person of life, liberty, or property without due process of law”)14 does not obligate a law enforcement officer carrying out an arrest under a valid warrant to investigate claims of innocence voiced by or on behalf of the person arrested. (2) The “liberty” clause does not obligate an officer who, post-arrest, has custody of a person validly arrested, to conduct an “error free” (443 U.S. at 146, 99 S.Ct. 2689) inquiry into claims of innocence advanced by or on behalf of the *120person arrested and subsequently detained. (This second proposition is not, as a matter of syntax, entirely devoid of ambiguity: the proposition may be taken to mean that a custodial officer is under no constitutional obligation to conduct an inquiry, perfect or otherwise, into claims of innocence; or, alternatively, the proposition may be taken to mean that, whether or not there is deemed to be a constitutional obligation to inquire, a custodial officer who undertakes to inquire has no constitutional obligation to insure that the findings arrived at are “error free.”)
Neither of these two propositions addresses in terms — or, I submit, by necessary implication — the question posed in the case at bar: when (1) law enforcement officers undertake to investigate claims that a person in custody pursuant to a valid arrest is innocent, and (2) the investigation produces information that provides the officers with actual knowledge that the person in custody is not the person who committed the crime, does the “liberty” clause impose on the law enforcement officers an obligation to release that person forthwith — notwithstanding that he is the person named in the arrest warrant — because they know to a certainty that he is innocent?
It may be urged that in transposing the quoted Baker paragraph into two paraphrased propositions I have given the Court’s language short shrift by ignoring the paragraph’s closing sentence: “The ultimate determination of such [i.e., the detainee’s] claims of innocence is placed in the hands of the judge and jury.” 443 U.S. at 146, 99 S.Ct. 2689. Quoting that sentence, this court properly observes that “Baker thus venerates the separation of functions among various government actors.” So understood, the distinction drawn by the Supreme Court between the conventional role of law enforcement officers and the conventional role of courts is clearly responsive to the issue posed in Baker — namely, whether a law enforcement officer has a constitutional duty to investigate, and arrive at reliable findings with respect to, claims that a person properly taken into custody is in fact (as Baker was in fact) not the person who committed the alleged offense. The answer to that question is in the negative for the reason that, in the American scheme of things, it is generally the case that determining innocence vel non is a judicial function rather than a police function. But that general description of the American scene does not answer the question posed by the case at bar: in a circumstance in which, going beyond their constitutional duty, law enforcement officers do investigate and in the process learn that the person detained is indisputably not the perpetrator of the offense charged, are not the law enforcement officers then under a constitutional duty forthwith to release the person detained? In arguing that this is a question that is not, as the court holds, “foreclosed” by Baker,15I would note that Baker’s, attri*121bution to the judiciary of institutional responsibility for “[t]he ... determination of ... claims of innocence” characterized that determination as the “ultimate” determination. Assigning to the judicial branch “ultimate” responsibility for determining innocence would not appear to preclude an earlier finding of innocence by law enforcement officers of the executive branch. It might, indeed, be thought to invite such executive intervention in the circumstance — the rare circumstance — in which agents of the executive branch acquire actual knowledge (as opposed to information establishing a high degree of probability yet in some modest measure still open to further inquiry) of the innocence of the person detained.
ii
In the foregoing subsection of this opinion I have undertaken to summarize the broad construction of Baker presented in the court’s opinion; and then, to counter that broad construction, I have undertaken to present, in summary form, a reconstruction of Baker positing what I think is a plausible, albeit less spacious, reading. Doubtless some will tend to regard this revisionist reading as less a reconstruction than a deconstruction — or, more charitably, a desiccation. I must acknowledge that the court has advanced some good arguments in favor of its broad reading. To do justice to the court’s arguments, which are developed with considerable care, I quote from those arguments at some length (footnote omitted):
Important considerations undergird the separation of functions recognized in Baker and Thompson [v. Olson, 798 F.2d 552 (1st Cir.1986) (see supra, notes 1 and 2) ], and those considerations are at their zenith when a person who is named in a facially valid warrant, supported by probable cause, is arrested pursuant to that warrant. When such a person asserts that he is a victim of mistaken identities, he in effect is pressing a claim of innocence in fact-^a claim not analytically distinct from any other factual defense (say, an alibi defense or a defense premised on a lack of specific intent) tendered by a person whom the police arrest in pursuance of a warrant issued by a judge or magistrate. Regardless of the merits of the defense, our legal system simply does not rely on police officers to determine its bona fides, even though they may have information bearing on that ultimate question and even though they may harbor strong and informed opinions one way or the other. To the contrary, once probable cause has been established, a warrant issued, and an arrest perfected, the ordinary course is for the prosecutor to decide whether to go forward, and if he elects to proceed, for the judicial branch to make the final ascertainment of guilt or innocence — not for the police to take matters into their own hands.
*122It is, moreover, impossible to subscribe to Brady’s (and the district court’s) supposition that police officers who must make conclusions from conflicting evidence gathered during a post-arrest investigation may come to “know” that an arrestee is innocent.... Absent personal knowledge of Brady’s innocence — and Brady does not aver that any of the appellants were percipient witnesses to the drunk driving or its immediate aftermath — the worst one can say about the troopers who continued to hold Brady in custody is that they came to believe, with some degree of subjective certainty, that the man they had arrested, though named in the warrant, was innocent of the underlying charge. This is not knowledge — and this kind of subjective belief, without more, is generally insufficient to justify a police officer’s unilateral release of a person who has been lawfully arrested pursuant to a valid judicial order.
These are cogent arguments. But they are not, I think, incontrovertible.
Let us consider just what it is that Brady contends. He contends that: (1) the facially valid warrant pursuant to which he was arrested named him rather than the real culprit because the real culprit tricked the police by pretending to be Brady; and (2) in response to Brady’s protestations of innocence the police inquired into the matter and soon came to realize that Brady was not the man they wanted; but, (3) rather than releasing him at once, the police kept Brady in custody for thirty-six hours until, finally, Brady was brought before a magistrate who directed that he be freed.
Now let us measure Brady’s claim of innocence against the analysis propounded by the court. According to the court, when a person arrested pursuant to an arrest warrant that names him “asserts that he is a victim of mistaken identities, he in effect is pressing a claim of innocence in fact — a claim not analytically distinct from any other factual defense (say, an alibi defense or a defense premised on a lack of specific intent) tendered by a person whom the police arrest in pursuance of a warrant issued by a judge or magistrate.” I have no quarrel with the proposition that such claims are not “analytically distinct.” But I would argue that such claims also are — or at least may be — pragmatically very different. Thus — subscribing in this respect to the court’s argument — I find it very difficult to posit a scenario in which a claim of “lack of specific intent” could, within a few or even several days of dedicated police investigation, be so painstakingly explored, intricately unraveled and subtly evaluated as to lead the investigating officers to have absolute assurance of the innocence of the person in custody. But I have little difficulty in positing such a scenario with respect to a claim of “mistaken identities.” And I do not have to resort to putting together a hypothetical case. Our doctrinal raw materials offer us a real world case — the case of Baker v. McCollan. Linnie Carl McCol-lan — so the Court narrated — was victimized by his brother Leonard. Leonard fabricated a copy of Linnie’s driving license with Leonard’s photograph superimposed on it. With the aid of this counterfeit identification, Leonard, when arrested in October of 1972 in Potter County, Texas, for a narcotics offense, orchestrated matters with such fraternal devotion that he was booked and bailed — and subsequently failed to appear — as Linnie. Wherefore, when, on December 26, 1972, Linnie was pulled over by a Dallas police officer for going through a red light, “[a] routine warrant check revealed that Linnie Carl McCollan was wanted in Potter County, and respondent was taken into custody over his protests of mistaken identification..” 443 U.S. at 141, 99 S.Ct. 2689. For reasons not elaborated on in the Supreme Court’s opinion, or in Justice Stevens’s dissent, notwithstanding the fact that the Dallas Police Department promptly notified the Potter County Sheriffs Department of Linnie’s arrest, four days were to elapse before Linnie was taken from Dallas to Potter County. What the transpired, as the Court narrates it, was this:
*123On December 30, Potter County deputies took custody of respondent and placed him in the Potter County Jail in Amarillo. He remained there until January 2,1973, when officials compared his appearance against a file photograph of the wanted man and, recognizing their error, released him.
Ibid.
Like the defendant police officers in the case at bar, those who had Linnie Carl McCollan in charge were not “percipient witnesses” of the offense charged against the person in custody: that is, they had no “personal knowledge of [that person’s] innocence.” But Linnie Carl McCollan’s custodians did not conclude that they had to wait for a judicial officer to dismiss the charges and order Linnie released. “[RJecognizing their error,” they “released him.” Of course the Supreme Court did not say that it was the constitutional duty of McCollan’s custodians to release McCollan on “recognizing their error.” That issue — which Brady’s claim presents to this court — -was not before the Baker Court. But it is, I think, of at least passing interest that the Supreme Court’s narrative of the way in which McCollan’s (belated) release from custody was accomplished seems to treat it as a matter of course.
It is my submission that the action Lin-nie Carl McCollan’s custodians took to release him when they knew to a certainty that he was not their man was their constitutional duty. The court disagrees. In terms of ultimate practical difference, the gap separating my constitutional position from that of the court is not a wide one, for the reason that, were my position to prevail, there would, I think, be no more than a handful of cases in which a plaintiff would be able to demonstrate to a fact-finder that police officers kept someone in custody beyond a point in time at which they knew to a certainty that the person in custody was not the perpetrator of the offense charged. I employ the phrase “knew to a certainty” advisedly. That would be my translation of the phrase “actual knowledge” — the phrase contended for by Brady and relied upon.by the District Court. Moreover, ;in explicating “knew to a certainty,” I would instruct the fact-finder that the operative constitutional principle is the following: An affirmative duty to release arises only if an arresting or custodial officer ascertains beyond a reasonable doubt that the suspicion (probable cause) which forms the basis for the privilege of arrest is unfounded. (I have adapted this formulation — almost verbatim — from the formulation employed by this court in Thompson v. Olson, supra, 798 F.2d at 556, to describe the duty of an arresting officer “following a legal war-rantless arrest based on probable cause.” The Thompson formulation — which was drawn from § 134, Comment f, of the Second Restatement of Torts — is stated by this court in today’s opinion to be a principle of Maine law, not of federal constitutional law. Possibly so; but, as I explain in footnote 2, supra, I am not as agile as the court is in tracing, let alone walking, the fine line between state and federal law which the court discerns in its examination of Thompson. In any event, whatever the scope — just Maine law or also federal law — of the Thompson formulation, I regard it as a useful source for the constitutional principle I would deploy in cases such as the case at bar).
If Brady’s case were to go to trial, I would regard it as highly unlikely that a jury charged in the manner I have just described would find in his favor. I agree with the court that “the worst one can say about the troopers who continued to hold Brady in custody is that they came to believe, with some degree of subjective certainty, that the man they had arrested, though named in the warrant, was innocent of the underlying charge.” But although that is my view of the facts of record at the summary judgment stage, I am persuaded that a jury might conceivably view the facts differently. Which means that I agree with the District *124Court’s assessment that this is “the quintessential jury question.”16
If the constitutional principle I have proposed were to be adopted, it would be my expectation that concrete applications of the principle would turn out to be few and far between. Does the expectation that the proposed principle would rarely have occasion to be vindicated mean that the proposed principle would really be a sort of whimsical superfluity that would trivialize, not augment, the Constitution? I am persuaded that the answer is no. And I say this because, so it seems to me, to reject the proposed principle — to conclude that there is house room in the American scene for law enforcement officers to keep a person in custody for any appreciable length of time after they know to a certainty that the person is not the culprit— is to diminish what is after all the central value of the text and of the spirit of the Constitution: liberty.17
II. Is the constitutional principle underlying Brady’s claim sufficiently clearly established so as to defeat the defendants’ assertion of qualified immunity?
At the end of its opinion, the court examines the question whether, assuming ar-guendo the viability of Brady’s constitutional claim, the constitutional principle that Brady has sought to vindicate is sufficiently clearly established so as to defeat the defense of qualified immunity deployed by the defendant law enforcement officers. The court answers that question in the negative. I agree. The court’s analysis seems to me entirely apt. I would only add that the strength of the court’s arguments against the viability of Brady’s claim, coupled with the dearth of case law supporting my counter-arguments, seem to me to demonstrate decisively that the Brady claim, even if the court were to accept its theoretical cognizability, could not properly be pursued in this case against these defendants. That is to say, when these defendants did the things of which Brady has complained, they had little ground for supposing that the Constitution imposed on them the duty that I have attempted to formulate.
Conclusion
Although I disagree with the court’s rejection of the cognizability of Brady’s constitutional claim, I acquiesce in the court’s view that the defendants are, in any event, shielded by qualified immunity. Accordingly, I concur in the judgment of the court.
. The Baker Court, two paragraphs before the paragraph I have quoted in the text, pointed out'that Baker had been arrested "pursuant to a warrant conforming, for purposes of our decision, to the requirements of the Fourth Amendment.” 443 U.S. at 144, 99 S.Ct. 2689. Therefore, since the Fourth Amendment’s strictures with respect to "seizures” — which, pursuant to the Fourteenth Amendment, are deemed to be binding on the states — were satisfied, the Court’s analysis of Baker's constitutional claim was keyed to post-arrest ingredients of Fourteenth Amendment “due process” (e.g., the Sixth' Amendment's "right to a speedy and public trial”). Building on Baker, the opinion of this court in the case at bar, views Brady's claim as one to be assessed under the broad Fourteenth Amendment flag of post-arrest “due process” rather than the narrow Fourth Amendment ensign of "seizure.” This analytic approach seems to me sound — a view which puts me at odds, conceptually, with the analytic approach taken by the District Court. Notwithstanding my perception that this court is correct in by-passing the District Court's Fourth Amendment route and opting for the more direct "due process” path, it is my ultimate conclusion, which I endeavor to explicate in this section (section I) of this concurrence that the District Court, in determining that Brady presented a cognizable claim, arrived at the right constitutional destination (albeit through a Fourth Amendment analysis to which I do not subscribe), and that this court, in rejecting the District Court's determination, winds up at a constitutional dead end.
. In amplification of why Brady’s claim is "foreclosed” by Baker, the court observes that the Supreme Court’s "respect for the separation of functions ... explains why the Baker Court declined to impose on police officers an affirmative duty of investigating claims of innocence. The same principle also 'explains why the Court deemed adherence to the archetypical post-arrest due process guarantees sufficient to protect McCollan’s rights (and, ultimately, to defeat his section 1983 claim ).”
The court then goes on to say that ”[w]e hewed to this very line in Thompson [v. Olson, 798 F.2d 552 (1st Cir.1986)], which, as a precedent of this court, further binds us in our consideration of the case at hand. ... Thompson involved a warrantless arrest in which the police officer himself had made the initial probable cause determination. The plaintiff then filed a section 1983 suit founded on a theory of substantive due process, coupled with a pendent state-law false imprisonment claim. In discussing the latter, we drew upon Fourth Amendment jurisprudence and noted the general rule, which we considered applicable to cases of warrantless arrests, that ’once the arrest has been properly effected, it is the magistrate and not the policeman who should decide whether probable cause has dissipated to such an extent following arrest that the suspect should be released.’ ... If the rule we recognized in Thompson applies to a warrantless arrest, it must apply, a fortio-*121n, to an arrest of a person named in a facially valid arrest.” The court then proceeds, by way of footnote, to acknowledge that “Thompson spoke of a limited exception to the general rule, namely, that a police officer has an-affirmative duty to release a detainee following a warrantless arrest if the officer ‘ascertains beyond a reasonable doubt that the suspicion (probable cause) which forms the basis for the privilege to arrest is unfounded.’ ” However the court undertakes to draw the sting of this “limited exception to the general rule” by stating that “the Thompson court recognized this exception only in the context of Maine’s law of false imprisonment, and not as a matter of constitutional law.” ' With all respect, I think the court draws the line between the “general rule” (presented as constitutional law) and the “limited exception” (presented as Maine law) a bit too finely. Both are considered in the fourth paragraph of a five-paragraph discussion of Maine law. As this court states, that discussion was informed by "Fourth Amendment jurisprudence." There is, however, no obvious highway sign which
advises the jurisprudential voyager, at the close of the formulation of the federal "general rule,” that the next phase — the formulation of the “limited exception” — crosses the jurisprudential boundary into the nether realm of state law.
. I have not undertaken to address the court's observation that, given that in Baker "the Supreme Court pronounced a three-day detention as failing to constitute a deprivation of liberty without due process, it would take circumstances much more egregious than Baker's for us to conclude that a weekend detention of approximately thirty-six hours, accompanied by a concerted effort on the part of the police to secure the detainee's release, resulted in a wrong of constitutional dimensions.” Suffice it to say that the Baker Court did not hold that it is constitutionally permissible to detain a person for three days after his custodians know to a certainty that the detainee should not be in custody, for that question, as I have sought to make clear, was not before the Baker Court.
. I appreciate that the court's holding deáls only with the question whether Brady has a claim under § 1983 — a question the court answers in the negative. It appears to me, however, that an almost inescapable corollary of the court’s holding is that, had Brady in his penultimate hour of custody petitioned for habeas corpus — supporting his petition by un-controverted proof that his custodians knew to a certainty that Brady was not the wrongdoer — the Great Writ would not have opened the door of Brady's cell.