(with whom Wilkins, J., joins, dissenting). The record of this case discloses a dangerous vigilantism, not to be condoned even if it began out of understandable feelings of frustration. The response of the police detective in charge of the official investigation to these private activities was maladroit or worse.1 Later, the same officer and others deliberately obstructed counsel’s access to the defendant when the defendant had dire need of advice. At a trial following such events, constitutional protections should have been accorded to the accused with particular scruple. The able trial judge tried conscientiously to give the accused his constitutional due, but I think two of his rulings were faulty. The rulings were (I) that the statements made by the defendant to his kidnappers at the Sears parking lot around 6:30 p.m. on December 9, 1971, were voluntary and thus admissible, and (II) that the Commonwealth could use for impeachment purposes the statements made by the defendant to the police at the hospital early the following morning, at a time when the defendant’s counsel was being kept from him by the police. Because the court upholds these rulings, I am obliged to dissent.
I
This case must be the first in our jurisprudence in which incriminating statements, made by a kidnapped person to his kidnappers while still in their grip, have been adjudged to be acts of free will. How does the court justify such an extraordinary conclusion here?
All members of this court accept the basic facts — the historical or subsidiary facts — as found below. The dispute is as to the conclusions to be drawn from those facts, a matter on which this court, as an appellate court dealing with constitutional rights, is required to make its own independent judgment. See Commonwealth v. *706Murphy, 362 Mass. 542, 550-551 (1972)2 (concurring opinion of Hennessey, J.); Napue v. Illinois, 360 U. S. 264, 271-272 (1959).3 The majority of this court reach *707their conclusion by a train of reasoning that declines to acknowledge the natural inferences flowing from the subsidiary facts, and constructs instead a wholly speculative theory to explain the defendant’s behavior.
That the statements given up to 4:15 p.m. of December 9 were coerced, is not disputed. But we have to sum up the circumstances of that coercion because they bear on the defendant’s situation when he made the further statements two hours later.
A large number of hostile pursuers, all the more fearsome because not quite identifiable, had been harassing the defendant over a period of fifteen months, making threatening appearances at unpredictable times at his home, school, and places of work. The insistent surveillance broke out into episodes instinct with violence. Toward the end the defendant would have ground for believing that his tormentors had already convicted him of murder and sought only an opportunity to enforce their own law. Finally came the kidnapping at Mt. Ida. The physical hurt was compounded by the uncertainties of a long trip to an unknown destination. Arrival in the dead cold of winter at an isolated, snowbound place must further have shaken the defendant. The threat of the bread knife was upon him throughout the night.
Starting in the early morning and for some six to eight hours the defendant was questioned by three and then five antagonists whose determination to break him may have been intensified by an apprehension that they could not “justify” the kidnapping, if called to account for it, unless they managed to extract some tangible results. This may have underlain the severity of the questioning: in any event, it was extended, repetitious, nagging, interspersed with extremely rough language and threats to take the defendant’s life, threats that he would never leave the place alive. In confronting this inquisition, the defendant was alone, without benefit of friends or advice. At length, the defendant’s will was broken. He made incriminating statements to Ferreri and Campbell.
*708It is conceded that these statements were coerced. But the defendant still withheld the revelation of the exact location of the body. Instead he offered to lead Ferreri to the gravesite. The kidnappers debated the defendant’s offer; only after argument among themselves (Ferreri pressing one view and Fontacchio and Heard the other) did they decide to accept the offer and take the defendant with them to Boston, rather than to continue to hold him at the cabin until he revealed the location of the gravesite and the information could be verified. But the defendant was not to be released until the body was found. Thus, the kidnapping and imprisonment were not brought to an end by the defendant’s initial statements at the cabin, but would continue until he satisfied his captors’ ultimate demand. On these facts, I conclude that the defendant’s acceptance of the condition that he reveal the gravesite was as much coerced as his initial statements. His statements at the Sears parking lot were thus made within a continuing constraint and compulsion.4
In light of the natural conclusion from the subsidiary facts that the defendant remained under the heel of the kidnappers through the 6:30 p.m. statements, it may be unnecessary to apply those tests which have been used in more doubtful cases to measure how far coercion or illegality has been attenuated by later events. But if those tests are applied here, the conclusion is reinforced.
As to whether there has been an insulating “break in the stream of events” between successive statements, the cases point to certain central, objective considerations. Among these — besides the elementary question of the length of time between the statements, here quite short — are the factors: whether in the interval the defendant *709had an opportunity to see his family or friends (Reck v. Pate, 367 U. S. 433, 444 [1961]), or to consult with counsel (Darwin v. Connecticut, 391 U. S. 346, 349 [1968]; Clewis v. Texas, 386 U. S. 707, 709, 711 [1967]); whether he has been throughout the period continuously in the hands of those who obtained the first statement (Beecher v. Alabama, 389 U. S. 35, 38 [1967]); and whether the later statement was given to the same persons as the original, coerced statement. (Lyons v. Oklahoma, 322 U. S. 596, 604 [1944].) According to these objective indicators, there is no basis for discovering here a material break in the stream of events.
Next, as to “cat-out-of-the-bag,” we observe that by 4:15 p.m. the defendant had already made statements involving himself in the death of the victim; he had not divulged the exact location of the grave, but he had given up its approximate location. The main secret was out. There is nothing to suggest that the defendant knew that under the law his statements to that point were inadmissible; indeed, such mention as the defendant is supposed to have made of his chances in case of trial indicate that he thought his statements could and would be used against him. But if he believed that his first statements were beyond recall — and realistically they were, regardless of their exact legal position at trial — the defendant would see little point in withholding the rest of his story. So the conclusion is well justified that the coercion which produced the pre-4:15 p.m. statements was also the cause of the post-4:15 statements. And here, to repeat, we have the added, overriding factor that the defendant was under great continuing pressure to make the final disclosure of the gravesite as a means of getting free of the kidnappers.
My assessment of the subsidiary facts seems to me within the reasoning of the passage from Mr. Justice Jackson in United States v. Bayer, 331 U. S. 532, 540 (1947), and the remarks by Mr. Justice Harlan in Darwin v. Connecticut, 391 U. S. 346, 350-351 (1968), quoted by *710the court.5 Again, in United States v. Gorman, 355 F. 2d 151, 157 (2d Cir. 1965), cert. den. 384 U. S. 1024 (1966), the Second Circuit considered a “situation in which, after a first confession has been extracted from a man previously professing innocence by means calculated to break his will, a second confession is more politely secured.” Judge Friendly wrote, “In such a case, there is a strong basis both in logic and in policy for drawing the inference that the second confession was the product of the first, and for permitting that inference to be overcome only by such insulation as the advice of counsel or the lapse of a long period of time.” Compare Fisher v. Scafati, 439 F. 2d 307, 310-311 (1st Cir. 1971), cert. den. 403 U. S. 939 (1971), where Chief Judge Aldrich suggested that Miranda warnings after a first invalid confession may not themselves make a second confession admissible unless accompanied by advice about that prior invalidity and inadmissibility.6
There is analogy in a case decided by the Supreme Court last term, Brown v. Illinois, 422 U. S. 590 (1975). After a warrantless arrest without probable cause, the defendant was given Miranda warnings and then, about 9 p.m. , made an inculpatory statement. The defendant then went with the police to look for an alleged confederate, Claggett, and at 3 a.m. the next morning, after *711repeated Miranda warnings, gave a second statement. The court held that the illegal arrest vitiated the defendant’s first statement despite the warnings: “Brown’s first statement was separated from his illegal arrest by less than two hours, and there was no intervening event of significance whatsoever.” 422 U. S. at 604. As to the second statement, the court said of it that it was “clearly the result and the fruit of the first.” Id. at 605. “The fact that Brown had made one statement, believed by him to be admissible, and his cooperation with the arresting and interrogating officers in the search for Claggett, with his anticipation for leniency, bolstered the pressures for him to give the second, or at least vitiated any incentive on his part to avoid self-incrimination.” Id. at 605, n. 12.7
In the present case, the problem for the trial judge, and for this court in following him, was how to reconcile a conclusion that the defendant’s statements after 4:15 p.m. were voluntary, with (a) the earlier conceded coercion by the kidnappers, (b) the effect on the defendant’s mental state of his having made the initial confession, (c) the determination by the kidnappers, well understood by the defendant, to hold the defendant until he completed his confession by revealing the gravesite, and (d) the kidnappers’ possession and control of the defendant until he actually did so. The trial judge and the *712court have responded to this challenge by simply introducing a kind of deus ex machina: they assert that a sudden and complete change occurred after the initial statements; with the cessation of overt intimidation on the part of the kidnappers, the defendant abruptly became friendly and trustful toward them, so that his actions and statements thereafter were manifestations of his free will, uninfluenced by the previous coercion. In attempted support of this inference, the court seizes upon a number of incidents after 4:15 p.m. With occasion to cry out to the two hunters or at the toll stations on the way back to Boston, the defendant remained silent. Similarly, the defendant made no attempt to attract the attention of passersby at the parking lot, and did not seize any opportunities to escape that may have presented themselves while he was descending to the tracks and returning (with the advantage of the open penknife). It is said that the defendant warmed to Ferreri and spoke of those to whom he first confessed as his “friends,” he took credit with Ferreri for not appealing to the hunters, and he contributed some change to pay a toll. Just before reaching the parking lot, he volunteered to Ferreri an indication of the bus stop figuring in his confession, and later talked easily to Heard about his legal chances. Even so, we encounter the fact that during the ride to Boston the defendant secretly unscrewed the door lock plunger on his side of the car in order to provide himself with physical proof that he had been kidnapped; he complained that the GTO automobile was bugged; and he resisted the walk down the path.
All this behavior does not lead to the inference that the defendant was free of compulsion or of its effects; on the contrary, his behavior is entirely consistent with a broken will and indeed is to be expected from one in that condition. As was said in a case where a defendant had been intimidated and beaten by private parties and shortly thereafter made statements to the police: “Torture destroys not only physically but psychologically. Ele*713ments of despair, fatigue, craving for companionship, identifying one’s interrogator as a friend and source of aid,8 and suggestions of guilt were all present in a crude, haphazard form in this case.” People v. Berve, 51 Cal. 2d 286, 292 (1958). It is all too easy, reading this record in retrospect, with control of one’s faculties and with time and capacity to think clearly, to point out one or another moment when the defendant might have escaped. But such heroics are more likely in the movies than in real life. If the defendant had the ability to think of escape, he might also have thought that it would only result in recapture by the five vigorous young men or ultimately by others of the concerned group. In all probability, planning escape was out of the question for the defendant. Suffering from the bewildering and frightening events since his abduction, and already deeply committed by the pre-4:15 statements, he most likely was incapable of further resistance, though he might yet retain sufficient presence of mind for such sporadic acts as taking the door lock plunger. At the same time, the facts demonstrate that the defendant was not beyond some gestures to ingratiate himself with his tormentors; it may be inferred that he felt these a means of preventing further mistreatment or of gaining his final release. Once he had told his story, the “cat-out-of-the-bag” syndrome explains his telling it again, and his further remarks to Heard were nothing but self-comforting braggadocio. Any inference that the defendant was lighthearted after 4:15 p.m. because he thought the kidnappers accepted his story of a blow struck in anger, is dispelled when we note that, even if that statement were believed, the defendant would still be in very serious trouble: consider here his admission, as part of the story, *714that he had deliberately concealed the body, and then suppressed the truth for fifteen months.
In evaluating the historical facts to reach a conclusion, we should recall that it is not the defendant’s burden to establish that his statements were coerced; the burden is on the government to prove the contrary, that the statements were freely willed. Jackson v. Denno, 378 U. S. 368, 376-377 (1964). Lego v. Twomey, 404 U. S. 477, 489 (1972). It is submitted that the inference of abrupt and total transformation of the defendant, from hostility and resistance to an attitude of voluntary cooperation, simply is not made out on the basis of the historical facts. Rather, the most modest conclusion that emerges from the facts is that the post-4:15 p.m. statements were substantially conditioned and influenced by the coercion directed at the defendant throughout the period during which he was held.
II
The police conduct surrounding the questioning of the defendant at the M. G. H. on the morning of December 10 violated the defendant’s constitutional right to the assistance of counsel. The trial judge so held, and the court concedes the point. When the questioning began, Detective Gawlinski, in charge of the case,9 knew that *715the defendant’s counsel was trying to reach him. Yet he neither took steps to inform the defendant of that fact nor returned counsel’s calls; instead he tried in a highly suspicious (if clumsy) way to conceal or avoid his responsibility for this breach of the Constitution by absenting himself from the interrogation that he knew to be going on.10 There was, then, in the words of the trial judge, “a course of conduct calculated to circumvent . . . [the defendant’s] constitutional rights,” conscious “treading on constitutional thin ice,” “deception and circumvention” by the principal investigating officer.
Nevertheless, the trial judge ruled that the statements obtained at the hospital could be used for impeachment purposes if the defendant testified in his own defense, and the court affirms. I think the ruling is not required by the decided cases and is fundamentally wrong. I could, with some difficulty, sympathize with such a decision if the violation of constitutional right involved was accidental or of a minor or technical nature. Here it was deliberate and of serious consequence.
*716The court Tests its conclusion on two cases, Harris v. New York, 401 U. S. 222 (1971), and Oregon v. Hass, 420 U. S. 714 (1975), Neither is a sufficient prop. In Harris (a five to four decision), the Supreme Court held that, notwithstanding the failure of the police to give full Miranda warnings (Harris was told of his right to counsel, but not of his right to court-appointed counsel), a statement obtained during custodial interrogation could be used for impeachment when Harris testified in his own behalf at trial. Weighing the promotion of truth through allowing the impeaching use of the statement, against the possible added deterrence of police misconduct that would flow from denial of such use, the court thought the truth-seeking interest prevailed. It is a ground for distinguishing the Harris case from the present that Harris did not involve the direct violation of a constitutional right but only the violation of a prophylactic rule safeguarding the right.11 Beyond that, however, it is vital to observe that the questioning in Harris took place before Miranda was decided, so that the violation of the defendant’s right was unintentional.
We followed Harris v. New York in Commonwealth v. Harris, 364 Mass. 236, 239-240 (1973). In doing so, we took note of the objections raised by the dissenters in Harris v. New York,12 and quoted from Riddell v. Rhay, 404 U. S. 974, 976 (1971) (Douglas, J., dissenting from denial of certiorari): “[T]he possible use of tainted statements . . . opens the door to a calculated risk by police *717interrogators.” The risk referred to is that involved in intentionally violating constitutional rights in hopes that damaging statements will be obtained useful for impeachment. We said, “The present case does not require us to enter into this dispute. . . . [T]he record is entirely barren of any indication that police or prosecutor took any ‘calculated risk’; there seems rather to have been an inadvertent defect in the Miranda warnings given.” Thus it is evident that neither Harris v. New York nor our own Commonwealth v. Harris reaches the present case which is the paradigm of police deliberateness and calculation in infringing constitutional rights.
The court argues also from Oregon v. Hass. That case posed the question (in the words of the Supreme Court): “When a suspect . . . states that he would like to telephone a lawyer but is told that this cannot be done until the officer and the suspect reach the station, and the suspect then provides inculpatory information, is that information admissible ... for impeachment purposes . . .?” 420 U. S. at 714-715. The Supreme Court held (six to two) that it was admissible, as long as no “abuse” occurred making the statement involuntary or untrustworthy. Hass was arrested at his home for bicycle theft, and, after full Miranda warnings, agreed to show the police where he had left the bicycle. “[The police officer] and Hass then departed in a patrol car for the site. . . . On the way Hass opined that he . . . would like to telephone his attorney .... [The officer] replied that he could telephone the lawyer ‘as soon as we . . . [get] to the office.’ . . . Thereafter . . . [Hass] pointed out a place . . . where the bicycle was found.” 420 U. S. at 715-716.
The Supreme Court avoided characterizing the police violation of Hass’s rights as either accidental or deliberate. If the sketchy facts are read as implying that the police acted in good faith, or at least without design to evade the Constitution, then the case is like Harris v. *718New York and, like that case, does not reach the present situation. Two recent Supreme Court cases, decided since Harris v. New York, suggest that Hass should be so interpreted.
In Michigan v. Tucker, 417 U. S. 433 (1974), a defendant, not informed of his right to appointed counsel, made a statement which led the police to a witness. In holding that the witness was properly allowed to testify, Mr. Justice Rehnquist wrote that “[w]e consider it significant to our decision in this case that the officers’ failure to advise respondent of his right to appointed counsel occurred prior to the decision in Miranda.” He explained that “ [t]he deterrent purpose of the exclusionary rule necessarily assumes that the police have engaged in willful, or at the very least negligent, conduct which has deprived the defendant of some right. By refusing to admit evidence gained as a result of such conduct, the courts hope to instill in those particular investigating officers, or in their future counterparts, a greater degree of care toward the rights of an accused. Where the official action was pursued in complete good faith, however, the deterrence rationale loses much of its force.” 417 U. S. at 477. See the discussion of the case at 27 U. of Fla. L. Rev. 302, 309-310 (1974).
Emphasis on the importance of good-faith behavior of the police, as a factor in decision as to admitting or rejecting a suspect’s statement, appears also in Brown v. Illinois, 422 U. S. 590 (1975). This came three months after Hass; both were written by Mr. Justice Blackmun. Setting out the considerations with regard to admitting or excluding a statement made after an arrest which violated the Fourth Amendment, the Justice said that the giving of Miranda warnings after the arrest was “an important factor ... [b]ut . . . [t]he temporal proximity of the arrest and the confession, the presence of intervening circumstances . . . and, particularly, the purpose and flagrancy of the official misconduct are all relevant.” *719422 U. S. at 603-604.13 It seems unlikely that a court which singled out for emphasis “the purpose and flagrancy of the official misconduct” in one case analyzing whether an exclusionary rule should be applied, would totally ignore the presence of purposeful and flagrant misconduct in another case also dealing with exclusion. Thus the failure of the Hass majority to weigh the deliberateness and seriousness of the police conduct in the case before it suggests that they believed no deliberate police misconduct — or at least no very invidious deliberate police misconduct — was present.14
Presumably this court disagrees and sees in Hass more than a casual misprision in that the police continued the patrol car on its course to the area of the crime, rather than turning it back to the station, at the moment when the suspect stated his desire for counsel. This court then takes the Hass case a step further and reads it as covering also as aggravated a situation as we find in the case at bar.
Hass, if interpreted to cover intentional interference by the police with suspects’ access to counsel, would en*720courage the most objectionable kind of “risktaking” by the police. Mr. Justice Jackson said that any qualified lawyer will tell a suspect not to give a statement to the police.15 Hence, faced with an attempt by a suspect to see a lawyer, as in Hass, or of a lawyer to see his client, as here, the police will have a choice: if they accede and allow a meeting with counsel, they will be obeying the Constitution, but they will get no statement from the suspect; if they deliberately prevent the contact, they will be scorning the Constitution, but they will have the chance of getting a statement from the suspect that can be used to impeach (and may have other practical uses). In sum, the police will have nothing to lose, and much to gain, by intentionally flouting the Constitution.16 A rule of law presenting such a temptation to the police is radical and unwise.
It is said that permitting impeaching use of the statement furthers the truth-seeking function because only defendants bent on perjury will refrain from taking the stand through fear of being impeached. But the rule would in practice operate also against suspects who are trying to tell the truth throughout. Even one attempting to be as truthful as possible may recall certain facts incorrectly or fail to recall other important ones: the time after arrest is confused and pressure-filled; there is indeed *721a subtle coercion that is inherent in all police interrogation. That a defendant at trial tells a story not on all fours with his prior statement to the police does not mean that he is committing perjury, but the inconsistency evident to the. trier may nevertheless be devastating to the defendant’s case. The possibility of initial error by even a conscientiously truthful suspect and the later embarrassment at trial are reasons why counsel will advise his client to remain silent and not to accommodate the police. All this very much qualifies the notion that allowing the impeaching use will further the search for truth.17
This court is, of course, at liberty to adopt a higher standard than that which the Supreme Court has applied to the States under the Federal Constitution. See Cooper v. California, 386 U. S. 58, 62 (1967). If, indeed, the Hass case goes so far as to hold that a statement obtained from a suspect by deliberate and calculated police obstruction of his right to counsel may be admitted for impeachment purposes at a State court trial, then we should decline to adopt such a rule and we should hold, instead, as a matter of Massachusetts law, that the statement is inadmissible for any purpose. It is instructive that at least two States have already rejected the milder doctrine of Harris v. New York and have imposed upon themselves a rule more protective of the accused. See State v. Santiago, 53 Hawaii 254 (1971); Commonwealth v. Triplett, 462 Pa. 244 (1975). See also, e.g., State v. Brown, 262 Ore. 442 (1972) (interpreting State Constitution’s double jeopardy clause independently of Federal Constitution); People v. Brisendine, 13 Cal. 3d 528 (1975); State v. Kaluna, 55 Hawaii 361 (1974) (both *722interpreting State constitutional protection against unreasonable search and seizure to be broader than the guaranty found in the Federal Constitution by the decisions of United States v. Robinson, 414 U. S. 218 [1973], and Gustafson v. Florida, 414 U. S. 260 [1973]).
To conclude: The lawlessness of the “concerned group” is here matched by official lawlessness. Both brands of anarchic behavior deserve solemn rebuke. Out of the welter came a trial so beset by error that the conviction should be reversed and judgment entered for the defendant.
See n. 9, infra.
At part II, section 4, of its opinion, the court chides this dissent for refusing to accept the trial judge’s findings and in effect adopting contrary findings. The criticism is misdirected and ignores constitutional requirements. As will be evident, we do indeed differ from the trial judge in his “finding” (quoted by the court) that the defendant was “completely free from fear.” But to call that and similar statements by the trial judge subsidiary findings and thereby to foreclose reexamination of them here would subvert the process of review in constitutional cases. Those statements are merely reformulations in other words of the judge’s conclusion that the defendant acted voluntarily after 4:15 p.m., and are the very constitutional issue that must be reassessed by this court. Particularly pertinent is the closing remark in the following passage by Hennessey, J., concurring in the Murphy case, cited in the text: “[T]he ultimate findings and rulings of a judge may give rise to a meaningful appeal, even in a case where his subsidiary findings are beyond practical challenge. This is true because the ultimate conclusions of a judge on identification issues may be of constitutional proportions. This court must, where justice requires, substitute its judgment for that of a trial judge at the final stage. . . . The mere recital of appropriate phrases denoting constitutional acceptability may serve only to obscure error in admitting the evidence.” 362 Mass. at 551. See Frankfurter, J., in Watts v. Indiana, 338 U. S. 49, 50-51 (1949), and Culombe v. Connecticut, 367 U. S. 568, 603-606 (1961).
The court said in Napue: “The duty of this Court to make its own independent examination of the record when federal constitutional deprivations are alleged is clear, resting, as it does, on our solemn responsibility for maintaining the Constitution inviolate. Martin v. Hunter's Lessee, 1 Wheat. 304 [1816]; Cooper v. Aaron, 358 U. S. 1 [1958]. This principle was well stated in Niemotko v. Maryland, 340 U. S. 268, 271 [1951]: ‘In cases in which there is a claim of denial of rights under the Federal Constitution, this Court is not bound by the conclusions of lower courts, but will reexamine the evidentiary basis on which those conclusions are founded.’ It is now so well settled that the Court was able to speak in Kern-Limerick, Inc. v. Scurlock, 347 U. S. 110, 121 [1954], of the Tong course of judicial construction which establishes as a principle that the duty rests on this Court to decide for itself facts or constructions upon which federal constitutional issues rest.’ As previously indicated, our own evaluation of the record here compels us to hold that the false testimony used by the State in securing the conviction of petitioner may have had an effect on the outcome of the trial.” See Drope v. Missouri, 420 U. S. 162, 174 (1975).
See the distinction suggested in Commonwealth v. McGarty, 323 Mass. 435, 438 (1948), between an officer’s saying to a suspect during questioning that he will not be beaten, and the officer’s saying he will not be beaten if he confesses to the crime. See also Commonwealth v. Femino, 352 Mass. 508, 514 (1967).
See, further, Stewart, J., in Harrison v. United States, 392 U. S. 219, 224-226 (1968), and Harlan, J., dissenting in the same case and further explaining his position in the Darwin case. 392 U. S. at 227, note. See also Ruffin v. United States, 293 Atl. 2d 477, 480-481 (D. C. App. 1972).
It may serve in some measure to explain the trial judge’s error in admitting the post-4:15 statements, that in his original findings he omitted entirely to deal with the factor of “cat-out-of-the-bag” and paid insufficient attention to the factor of “break in the stream of events.” Accordingly, this court entered an order directing the trial judge to address himself to these two factors. The judge’s “supplementary findings” do not add to the subsidiary facts and asseverate his earlier conclusions without adding any fresh appreciation of the defendant’s predicament before or after 4:15 p.m.
This court took a similar approach in Commonwealth v. Spofford, 343 Mass. 703 (1962), a case involving not the admissibility of a subsequent confession, but rather the effectiveness of a consent for a search, given after a prior illegal search had turned up incriminating evidence, as the defendant knew. We emphasized that, given the prior search and its consequences, the defendant was “in no environment to make a free choice,” and held that the consent obtained was “an offshoot of the original unreasonable search and seizure” and so did not validate the subsequent search. 343 Mass, at 707, 708.
That Brown and Spofford both relate to an inquiry into the lasting effect of a Fourth Amendment violation, while the case at bar involves a Fifth Amendment violation, is of no consequence for our inquiry, as the court appears to recognize. See, supra, 688, n. 32.
This warming of the pursued toward his pursuer appears in imaginative literature in the relation of Jean Valjean to the detective Javert in Les Miserables, and of Raskolnikov to police inspector Porfiri Petrovich in Crime and Punishment.
Gawlinski’s earlier connection with the case sheds light on his actions and motivation during the hospital interrogation.
Gawlinski knew of the extensive surveillance of the defendant by the concerned group and he also became aware of the exacerbated incidents such as the one at Henry F. Bryant & Son, Inc., which ended in a physical encounter, with Ferreri or Fontacchio saying, “George, we know what you did and you’re going to pay for it,” and “You think you got away with it this time but you didn’t — we’ll get you,” or words to that effect. Yet Gawlinski took no decisive action against any of this activity. His attitude is further illustrated by an incident that occurred in July, 1971. Arthur M. Pascal, a private investigator employed by the father of the victim, learned that Erwin *715Katz, a “concerned” person, was planning to “pick up” or “kidnap” the defendant for questioning at which the father would be present. Pascal called Gawlinski to ask whether Gawlinski had given Katz the “green light” (as Pascal had been told by others); Pascal pointed to the danger of violence by the father. Gawlinsld indicated that he knew what was going to be done, yet insisted simply that there should be no “rough stuff”; if there was, he said, he would prosecute. The illegality and violence latent in the entire situation might have been more evident to an independent police officer. Gawlinski’s independence, however, had been impaired by his too close association with the father and brother; it is symptomatic that Gawlinski’s many meetings with the father took place at the father’s residence rather than in official quarters. Gawlinski’s hunger for results — for leads from any source — evidently overcame his respect for legality and orderly behavior. It comes, then, as no surprise that Gawlinski engineered to prevent the defendant from seeing his counsel on the morning of December 10.
In fact, Detective Sheehan, who did participate in the interrogation, also knew that counsel had been trying to reach Gawlinski.
See Michigan v. Tucker, 417 U. S. 433, 443-444 (1974); Michigan v. Payne, 412 U. S. 47, 53 (1973).
See Harris v. New York, 401 U. S. at 226-232 (Brennan, J., dissenting). Adverse commentary on the Supreme Court’s decision in Harris was copious and severe. See, e.g., Dershowitz & Ely, Harris v. New York: Some Anxious Observations on the Candor and Logic of the Emerging Nixon Majority, 80 Yale L. J. 1198 (1971); The Supreme Court, 1970 Term, 85 Harv. L. Rev. 40, 44 (1971); 10 Duquesne L. Rev. 128 (1971); 40 Fordham L. Rev. 394 (1971); 45 Temple L. Q. 118 (1971); 33 Pitt. L. Rev. 135 (1971).
The concurring opinion of Mr. Justice Powell, in which Mr. Justice Rehnquist joined, developed at some length the distinction between “technical” and “flagrant” violations of the Fourth Amendment and the consequences of the distinction on the admission or exclusion of statements later given. 422 U. S. at 606.
The flagrancy of official misconduct as a determinant of whether a statement should be suppressed is also adopted by the A.L.I.’s Model Code of Pre-Arraignment Procedure (1975 Approved Draft). In § 150.3 (1) and (2), the Code takes the position that “[a] motion to suppress a statement . . . [obtained in violation of the Code’s procedural protections which include right of access to counsel] shall be granted ... if the court finds that the violation upon which it is based was substantial .... A violation shall ... be deemed substantial if . . . [t]he violation was gross, wilful and prejudicial to the accused. . . .” According to § 150.3 (3), a violation, not meeting the foregoing test, may nevertheless be found substantial if it satisfies another definition of which material elements are “the extent of deviation from lawful conduct,” and “the extent to which the violation was wilful.”
Watts v. Indiana, 338 U. S. 49, 59 (1949) (Jackson, J., concurring in part and dissenting in part).
The egregiousness of Hass, if it is taken to extend to permitting the use for impeachment of statements gained by deliberate denial of the right to counsel, is shown by comparing it with a rule that would apply Harris v. New York to allow impeaching use of statements gained by deliberate denial of proper Miranda warnings. If the latter rule were in force, the police would still have significant incentive to give the warnings, since many suspects give statements, which are fully admissible, after being given warnings. But in the situation of deliberate denial of the right to counsel there is no deterrance whatever of the illegal police conduct, since, as noted, an attorney if given access to his client will advise him to make no statement.
Compare United States v. Hale, 422 U. S. 171 (1975), forbidding the prosecutor from asking a defendant why he did not tell the police at the time of his arrest the facts amounting to alibi that he testified to at trial. The Hale court reasoned that the inference the jury could draw — that the alibi was a contrivance — was so prejudicial to an honest defendant that questioning on the point must be forgone even though it might expose a perjurer.