This appeal requires us to address a series of pretrial delays (the causes and consequences of which are disputed) in light of the Sixth Amendment right to a speedy trial and the factors made relevant to that inquiry by the Supreme Court's landmark opinion in Barker v. Wingo, 407 U.S. 514, 92 S. Ct. 2182, 33 L. Ed. 2d 101 (1972). Inasmuch as the underlying case involves a state prisoner's attempt to secure a writ of habeas corpus, our task proceeds under the deferential standard of review mandated by the Antiterrorism and Effective Death Penalty Act (AEDPA), Pub.L. No. 104-132, 110 Stat. 1214 (1996). Upon careful perscrutation of a tangled record, we conclude that petitioner-appellee Rahim RaShad did little to seek a speedy trial, and, concomitantly, suffered no cognizable prejudice from the period of delay attributable to the Commonwealth. With two of the four Barker factors weighted against the petitioner, we hold that the state court's rejection of his speedy trial claim was neither contrary to, nor an unreasonable application of, clearly established Supreme Court precedent. Accordingly, we reverse the district court's order granting habeas relief and direct that the petitioner's conviction be reinstated.

The events giving rise to the petitioner's conviction transpired on the night of May 26, 1984. The petitioner and his ex-girlfriend, Denise Rodriguez, made plans to attend a movie that evening. The show was sold out, so the petitioner suggested that the couple repair to his Boston apartment. At this juncture the participants' stories diverge.

Rodriguez says that she resisted the invitation, but was forced to go along. The petitioner initially took her to his place of employment, where he pushed her down a flight of stairs and began to beat her. He then led her to his apartment while twisting her arm. Once there, he raped her repeatedly. When that phase of the encounter ended, he tied her to a bedpost and raped her again. Afterwards, he stuffed a sock and a towel into her mouth, securing them with tape. When he left to run an errand, Rodriguez escaped and telephoned the police.

The petitioner's version is considerably different. He maintains that Rodriguez went to his apartment eagerly, snorted cocaine with him, and voluntarily engaged in recreational sex. The next morning, he went to the grocery store. When he returned, Rodriguez was gone. A friend told him that the police were looking for him. Police officers subsequently knocked on his door, but he did not respond. Instead, he gathered some belongings (including his cache of drugs) and, after the police left the premises, fled to Dorchester. He never returned to his apartment.

Much of what transpired thereafter is uncontested. Around the end of May, the petitioner learned that a rape complaint had been issued against him. He spent the next twenty-seven months avoiding the authorities. At various times during that period he resided in Dorchester, Brockton, and New York. His peregrinations ended in August of 1986, when the Boston police arrested him on an unrelated charge. He used a pseudonym ("Charles McCrary"), and was convicted under that name.1 The court sentenced him to serve a six-month term in a Massachusetts state penitentiary.

In September of 1986 (while serving that sentence), he sent a handwritten note to the clerk's office of the Boston Municipal Court. The note disclosed his true identity ("Larry Graham") and stated that he "would like to see about being brought forward" on the pending charges. On February 5, 1987 — while the petitioner was still incarcerated — a Suffolk County grand jury indicted him on charges of kidnapping, aggravated rape, and assault and battery. The next month, he completed serving his prison term, and the Massachusetts authorities immediately extradited him to Texas to face other unrelated charges. Thus, the petitioner was unavailable for arraignment on the new indictment. Aware of the problem, the presiding magistrate issued a default warrant "in order to get the process working under the [Interstate Agreement on Detainers] to bring [the petitioner] back to Massachusetts" so that he could stand trial. The Commonwealth, however, neither lodged a detainer nor took any other steps to regain custody of the petitioner while he was incarcerated in Texas.

The petitioner spent the next forty-two months in a Texas jail. During that time, he eschewed any contact with the Massachusetts authorities, although he claims to have tried to contact his lawyer anent the status of his case. Texas released him in August of 1990. When he thereafter attempted to obtain a Texas driver's license, the authorities came across the outstanding default warrant and detained him briefly. After checking with their Massachusetts counterparts, however, they informed the petitioner that, although there were charges pending against him in Massachusetts, there was no basis for detaining him in Texas.

On the record before us, the petitioner's whereabouts for the next fifteen months are a mystery (there is some evidence that he was placed in a Texas pre-release program, but the duration and nature of that program is uncertain). Thus, our tale resumes in November of 1991, when the petitioner returned to Boston.2 For the following four months, he lived openly under his own name. During this period, he was twice stopped for traffic violations and once posted bail for a friend. Each time, the pending kidnapping/rape/assault charges escaped the authorities' attention.

The petitioner claims that he spoke to Rodriguez during this time frame, and that she told him that she did not think the charges against him were still pending. The petitioner made no effort to verify this fact (and, as matters turned out, Rodriguez was dead wrong).

In March of 1992, the chickens came home to roost. The petitioner was rousted during an Immigration and Naturalization Service (INS) sweep. He gave his name and birth date, but, when the INS uncovered the outstanding default warrant, he repeatedly denied any knowledge of the kidnapping/rape/assault charges. Although the exact wording of his statements is in dispute, the petitioner has admitted that he denied being the person named in the warrant (and the underlying indictment) in order to avoid being prosecuted for the offenses.

The INS delivered the petitioner to the Boston police, and the long-dormant indictment came to life. Trial was delayed for a span of roughly seven months due to the petitioner's serial motions for continuances. On October 7, 1992, with his trial finally set to begin, the petitioner moved to dismiss the case on the ground that the Commonwealth had abridged his Sixth Amendment right to a speedy trial. The state trial justice held an immediate evidentiary hearing at which the petitioner testified to many of the events just described. The timing of the motion and the dispatch with which the trial justice sought to address it left little time for preparation, and the Commonwealth did not present any evidence. In any event, the trial justice denied the motion without issuing written findings. Later that month, a jury convicted the petitioner of simple rape while acquitting him on the other charges (including aggravated rape).

The Massachusetts Appeals Court affirmed the conviction in an unpublished memorandum opinion. See Commonwealth v. Graham, 41 Mass.App.Ct. 1113, 671 N.E.2d 1016 (1996) (table). The court considered the petitioner's speedy trial claim under the Barker framework. It found the length of the pretrial delay sufficient to warrant further inquiry, but concluded that the other Barker factors preponderated in the Commonwealth's favor. In particular, the court attributed a sizable portion of the delay to the petitioner's actions; concluded that he had done very little to promote a speedy trial; and rejected his claim that the delay had caused actual prejudice. On December 31, 1996, the state supreme court denied further appellate review. Commonwealth v. Graham, 424 Mass. 1102, 674 N.E.2d 1084 (1996) (table). The petitioner then explored other state post-conviction relief procedures, the last of which terminated on April 23, 1998.

On June 29, 1998, the petitioner sought a writ of habeas corpus in the federal district court. After the application was narrowed to a single claim — that the Commonwealth had abridged the petitioner's Sixth Amendment right to a speedy trial — the district court found in his favor. RaShad v. Walsh, 204 F. Supp. 2d 93, 115 (D.Mass.2002).3 In reaching this result, the court first determined that the aggregate delay was sufficient to warrant a full-blown speedy trial inquiry. Id. at 102. In its view, the Commonwealth bore full responsibility for most of this delay, especially for (1) the five months between the petitioner's September 1986 letter to the Boston Municipal Court and his ensuing indictment; (2) the forty-two months that the petitioner spent in a Texas jail; and (3) the nineteen months between the expiration of the petitioner's Texas sentence and his eventual arrest. Id. at 103-08. The court then concluded that the petitioner had effectively asserted his speedy trial right, id. at 108-09, and that the length of the delay created a presumption of prejudice, id. at 110-11. The court added that, in all events, the totality of the petitioner's specific allegations tended to support a finding of presumptive prejudice. Id. at 112. Based on these assessments, the court deemed the state court's decision partly an unreasonable application of Supreme Court precedent and partly a product of insupportable factual determinations. Id. at 103. Accordingly, the court issued the writ.

This appeal ensued. We stayed the district court's order pending the completion of our review.

Before we begin our traverse of this case, we pause to survey the legal landscape. Because this is a habeas corpus proceeding brought by a state prisoner, we not only must weigh the strength of the petitioner's constitutional claim but also must consider whether the state court's determination that no constitutional infraction occurred, even if incorrect, was objectively reasonable. See Ouber v. Guarino, 293 F.3d 19, 31 (1st Cir.2002). We describe the guideposts that inform each part of this analysis.

The Sixth Amendment provides that "[i]n all criminal prosecutions, the accused shall enjoy the right to a speedy and public trial...." U.S. Const. amend. VI. Barker is the seminal Supreme Court case interpreting this directive. It delineates a quadripartite balancing test for use in evaluating potential speedy trial violations. See Barker, 407 U.S. at 530-33. The first prong involves the length of the delay between arrest or indictment, on the one hand, and the date of trial, on the other hand. See United States v. Loud Hawk, 474 U.S. 302, 310, 313-14, 106 S. Ct. 648, 88 L. Ed. 2d 640 (1986). The case law indicates that short periods of delay — say, appreciably less than one year — ordinarily are insufficient to justify further inquiry. See Doggett v. United States, 505 U.S. 647, 652 n. 1, 112 S. Ct. 2686, 120 L. Ed. 2d 520 (1992). Delays that exceed this floor generally require additional investigation. See id. As the length of the delay increases, its relative importance also increases. See id. at 652; United States v. Trueber, 238 F.3d 79, 87 (1st Cir.2001).

The reasons for the delay comprise the second factor in the calculus of decision. See Barker, 407 U.S. at 531. This element seeks to ensure that courts not concentrate on the sheer passage of time without also taking account of the etiology of the delay. The inquiry into causation involves a sliding scale: deliberately dilatory tactics must be weighed more heavily against the state than periods of delay resulting from negligence. Id. By like token, to the extent that valid reasons cause delay, the delay does not count against the state at all. Id. So too delay that is caused by the defendant. See Davis v. Puckett, 857 F.2d 1035, 1040-41 (5th Cir.1988).

The third prong of the Barker framework addresses the timeliness and frequency of the defendant's assertions of his speedy trial right. See Barker, 407 U.S. at 529 (instructing courts "to weigh the frequency and force of the objections as opposed to attaching significant weight to a purely pro forma objection"). Although a defendant does not waive his constitutional right to a speedy trial by failing to assert it at a time when the state could have nipped the violation in the bud, his failure to do so means that he must make a much stronger showing on the other factors in order to succeed in his claim. See id. at 532.

The fourth, and final, Barker factor implicates the extent to which the defendant was prejudiced by the delay attributable to the state. Id. In dealing with this factor, an inquiring court should keep in mind that the speedy trial right is intended to prevent oppressive pretrial incarceration, minimize the accused's anxiety, and limit the possibility that the passage of time will impair the accused's ability to mount a defense. Id. The latter concern is the most serious one. Id.

As a general rule, the defendant bears the burden of alleging and proving specific ways in which the delay attributable to the sovereign unfairly compromised his ability to defend himself. See United States v. Aguirre, 994 F.2d 1454, 1455 (9th Cir.1993). In aggravated cases, involving grossly excessive delay, prejudice may be presumed despite the defendant's inability to identify particular testimony or evidence that has become unavailable due to the passage of time. Doggett, 505 U.S. at 655-56.

Because this habeas petition was filed after April 24, 1996, the AEDPA controls. Lindh v. Murphy, 521 U.S. 320, 322, 336, 117 S. Ct. 2059, 138 L. Ed. 2d 481 (1997). Thus, we may grant relief only if the state court proceeding

(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or

(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the state court proceeding.

28 U.S.C. § 2254(d).

Under Williams v. Taylor, 529 U.S. 362, 120 S. Ct. 1495, 146 L. Ed. 2d 389 (2000), a state court decision is considered contrary to Supreme Court precedent only if it either applies a test that is inconsistent with one announced by the Court or reaches the opposite conclusion on materially indistinguishable facts. Id. at 405-06. That branch of the AEDPA standard is only marginally involved in this appeal — the state court correctly deduced that Barker constituted the controlling Supreme Court precedent — and so we do not dwell upon it.

That leaves the "unreasonable application" clause. A decision falls victim to that clause if the state court applies the correct legal standard in an objectively unreasonable manner, unreasonably extends a Supreme Court precedent to an inappropriate context, or fails to extend such a precedent to an appropriate context. Id. at 407-08. In this tamisage, it is the strength of the state court's ultimate conclusion, rather than its announced rationale, that must be evaluated. Ouber, 293 F.3d at 34. Importantly, the test does not demand infallibility: a state court's decision may be objectively reasonable even if the federal habeas court, exercising its independent judgment, would have reached a different conclusion. Williams, 529 U.S. at 411; Williams v. Matesanz, 230 F.3d 421, 425 (1st Cir.2000).

Whether a particular state court decision warrants that seal of approval must be decided primarily on the basis of Supreme Court holdings that were clearly established at the time of the state court proceedings. Williams, 529 U.S. at 412. Nevertheless, factually similar cases from the lower federal courts may inform such a determination, providing a valuable reference point when the relevant Supreme Court rule is broad and applies to a kaleidoscopic array of fact patterns. See Ouber, 293 F.3d at 26; O'Brien v. Dubois, 145 F.3d 16, 25 (1st Cir.1998). The doctrinal framework erected in Barker constitutes such a rule.

There is another aspect to the reasonableness inquiry in federal habeas cases. The AEDPA allows relief from a state court judgment if that judgment is based on an unreasonable determination of the facts. 28 U.S.C. § 2254(d)(2). Under this standard, "the state court's factual findings are entitled to a presumption of correctness that can be rebutted only by clear and convincing evidence to the contrary." Ouber, 293 F.3d at 27; see also Sanna v. Dipaolo, 265 F.3d 1, 7 (1st Cir. 2001) (explaining that the standard applies only to the determination of "basic, primary, or historical facts"). Unless the petitioner can carry this heavy burden, a federal habeas court must credit the state court's findings of fact — and that remains true when those findings are made by a state appellate court as well as when they are made by a state trial court. E.g., King v. Bowersox, 291 F.3d 539, 540 (8th Cir. 2002); Everett v. Beard, 290 F.3d 500, 507 (3d Cir.2002).

We proceed to assay the petitioner's speedy trial claim in terms of the four Barker factors as they apply here. Once that is accomplished, we consider the district court's views, and, finally, determine whether the state court's decision satisfies the AEDPA standard.

The speedy trial clock begins to tick only when "a defendant is indicted, arrested, or otherwise officially accused." United States v. MacDonald, 456 U.S. 1, 6, 102 S. Ct. 1497, 71 L. Ed. 2d 696 (1982). The application of this directive to the instant case is no trivial matter: the Commonwealth preferred charges against the petitioner in May of 1984, took him into custody for an unrelated crime in August of 1986, and indicted him in February of 1987. Despite the thirty-three month delay between the complaint and the indictment — most of which resulted from the petitioner's abscondment — neither the state court nor the federal district court ventured a definitive statement as to which date wound the constitutional clock.

The search for an answer to this puzzle is complicated by the MacDonald Court's interchangeable use of the terms "indictment" and "filing of charges." Id. at 6-7. Other cases, however, leave no doubt that only a "public accusation" animates the right to a speedy trial. United States v. Marion, 404 U.S. 307, 321, 92 S. Ct. 455, 30 L. Ed. 2d 468 (1971). Consistent with this approach, the case law suggests that, in the absence of either an indictment or an information, "only the actual restraints imposed by arrest and holding to answer a criminal charge" engage the speedy trial right. Loud Hawk, 474 U.S. at 310; accord United States v. Mala, 7 F.3d 1058, 1061 (1st Cir.1993). Because the preferment of charges here was unaccompanied by any public accusation or act of detention, that date is irrelevant for speedy trial purposes.4

This brings us to the petitioner's arrest in August of 1986. Although arrest may trigger the right to a speedy trial, it does not do so unless the arrest is the start of a continuous restraint on the defendant's liberty, imposed in connection with the same charge on which he is eventually put to trial. Acha v. United States, 910 F.2d 28, 30 (1st Cir.1990) (per curiam); United States v. Stead, 745 F.2d 1170, 1172 (8th Cir.1984). Thus, the fact that the petitioner was in state custody prior to his indictment is of no consequence unless that detention was related to the charges on which his speedy trial claim is based. Here, however, no portion of the petitioner's pre-indictment detention was caused by the pendency of charges arising out of the Rodriguez incident. Consequently, the date of the indictment is the starting point for the speedy trial analysis that must be conducted in this case.

The period between the petitioner's indictment in February of 1987 and his trial in October of 1992 aggregates five years and eight months. This delay is more than sufficient to justify inquiry into the remaining three Barker factors. See Doggett, 505 U.S. at 652 n. 1. It is to that inquiry that we now turn.

The next question relates to how much responsibility the Commonwealth and the petitioner, respectively, bear for the sixty-eight months of delay. During this time frame, the petitioner spent forty-two months in a Texas jail, fifteen in parts unknown, four living under his own name in Massachusetts, and seven engaged in pretrial preparation. The petitioner accepts responsibility for this last segment. Responsibility for each of the remaining three intervals is contested. We analyze each separately.5

1. The Texas Time. The petitioner was incarcerated in Texas on an unrelated charge for some three and one-half years. We are convinced that the Commonwealth must shoulder significant responsibility for this period of delay.

Our conclusion is driven primarily by the Supreme Court's decision in Smith v. Hooey, 393 U.S. 374, 89 S. Ct. 575, 21 L. Ed. 2d 607 (1969). There, a prisoner was indicted in Texas on state charges while serving a sentence at a federal penitentiary. For the next six years, the defendant made repeated requests for a speedy trial but the Texas authorities did nothing to advance the matter (even though they knew the prisoner's whereabouts). When the petitioner sought a dismissal, the state asserted that it had no obligation to gain custody of the defendant while he was imprisoned by another sovereign. The Supreme Court disagreed. We glean from its opinion that when a defendant demands a speedy trial, the state is obliged to make a diligent effort to obtain custody over his person even if another sovereign ultimately might decline to grant such custody.6 Id. at 383.

Here, the Commonwealth knew that the petitioner was incarcerated in another jurisdiction, yet it made no attempt to bring him back for trial during the forty-two months of his Texas confinement. To this extent, the case is analogous to Smith. Given that precedent, the Commonwealth must shoulder significant responsibility for this period of delay.7

Here, moreover, the Commonwealth's failure to act was aggravated by the fact that both Massachusetts and Texas were parties to the Interstate Agreement on Detainers (IAD). See Mass. Gen. Laws ch. 276, App. § 1-1 et seq.; Tex.Code Crim. Proc. art. 51.14. The IAD contains provisions that allow the transfer of an incarcerated defendant from one jurisdiction to another in order to stand trial. See 18 U.S.C., App. § 2, art. IV. The IAD contemplates that a state in which untried charges are pending may lodge a detainer against a defendant incarcerated in another state. Once that happens, the defendant must be informed of the detainer, and, if he so requests, he must be remitted for trial within the next 180 days. Id. at art. III. If the defendant does not make such a request, the detainer ensures that he will be transferred into the custody of the state that lodged the detainer upon the completion of his prison term. See Alabama v. Bozeman, 533 U.S. 146, 148, 121 S. Ct. 2079, 150 L. Ed. 2d 188 (2001).

Although the IAD contains no explicit requirement that a state lodge a detainer whenever it learns that a person under indictment is being held in another jurisdiction, the compact was crafted with the policy and purpose of "encourag[ing] the expeditious and orderly disposition of [untried] charges." 18 U.S.C., App. § 2, art. I. Thus, even though the failure to lodge a detainer is not itself a per se violation of a defendant's speedy trial right, it is a significant misstep, for which the state must bear responsibility.

In this case, the record does not suggest that the failure to lodge a detainer was deliberate. It was, however, plainly negligent; the Commonwealth knew why the petitioner did not appear for arraignment, and the presiding magistrate envisioned that the Commonwealth would proceed under the IAD. The failure therefore cuts in favor of the petitioner's speedy trial claim. See Doggett, 505 U.S. at 656-57. Holding otherwise would allow a state to circumvent the IAD with impunity — a result that would contravene both the intent behind the IAD and the Supreme Court's admonition that "the primary burden ... to assure that cases are brought to trial" rests with prosecutors, not defendants. Barker, 407 U.S. at 529.

2. The Fifteen-Month Hiatus. Responsibility for the period of delay following the petitioner's egress from the Texas jail is quite a different matter. The state court considered the petitioner a fugitive for this entire period. The petitioner assails this determination, asseverating that the Commonwealth should be held fully responsible for this interval.

The petitioner bases his asseveration upon Doggett, a case in which the defendant was absent from the United States for two and one-half years following his indictment. Doggett, 505 U.S. at 648. He subsequently returned to this country and lived openly under his own name for almost six years until he was arrested and tried. Id. at 649-50. The government defended its actions, claiming that it had lost track of Doggett shortly before he returned to the United States and had not learned his whereabouts until the Marshal's Service ordered a credit check. Id. The Supreme Court rejected the government's explanation; it determined that the government had not exercised due diligence in locating Doggett and counted this period of delay against the government. Id. at 652-63. One thing that figured importantly in this determination was the fact that Doggett knew nothing of the pending indictment prior to his arrest. Id.

The record of this case presents a far different picture as to the fifteen months following the petitioner's release from the Texas jail. One striking difference is the lack of relevant information. The petitioner apparently spent an indeterminate amount of time at a halfway house in Texas as part of a pre-release program, but otherwise we know very little. For aught that appears, the petitioner vanished into the Bermuda triangle. The record does not tell us where he was, what he did, whether he lived under his own name, or whether he endeavored to conceal his identity (as he had done on other occasions). Consequently, we cannot estimate how easy it would have been for the Massachusetts authorities to locate him. We do know that, unlike in Doggett, the petitioner had been informed that there were charges pending against him, see infra note 8, and he never asked the Massachusetts authorities if those charges were still live.

In the habeas context, a petitioner who fails to adduce any evidence regarding a segment of pretrial delay cannot rebut the presumption of correctness to which the state court's finding against him is entitled. See Wilson v. Mitchell, 250 F.3d 388, 394-95 (6th Cir.2001) (affirming, under the AEDPA, a state court's finding that the defendant was a fugitive and therefore responsible for a period of delay when he had failed to produce evidence to the contrary); cf. Mala, 7 F.3d at 1061-62 (explaining that lack of evidence relevant to the Barker analysis cuts against the defendant as to issues on which he bears the burden of persuasion). The petitioner has not satisfied that burden in regard to this fifteen-month traipse. We therefore lack any basis for disturbing the Massachusetts Appeals Court's finding that the petitioner was a fugitive during that period (and, thus, responsible for the attendant delay).

3. The Next Four Months. The four months immediately preceding the petitioner's 1992 arrest present a closer parallel to Doggett, and the petitioner, represented here by able counsel, rides that horse for all it is worth. He emphasizes that he lived under his own name in Massachusetts during that time frame, and that the authorities squandered several opportunities to detain him.

Although it was incumbent upon the Commonwealth to seek the petitioner with diligence, see Doggett, 505 U.S. at 656-57, even a diligent investigator might have taken months to uncover the fact that the petitioner had quietly returned to Massachusetts. The petitioner's failure to contact the authorities or inquire about the status of his case during the brief period of his residence in Massachusetts, along with his post-arrest attempt to evade the indictment, leave no doubt that he did nothing to facilitate the Commonwealth's task. Under these circumstances, the state court was not obliged to assign responsibility for this relatively short period of delay to the Commonwealth.8 We therefore refrain from counting these four months in the petitioner's favor.

We next consider the extent to which the petitioner affirmatively sought a speedy trial. See Barker, 407 U.S. at 531-32. The petitioner argues that he discharged his burden of asserting the speedy trial right when he wrote to the Boston Municipal Court in September of 1986, inquiring about the possibility of being "brought forward" on the charges. We emphatically disagree.

It must be recalled that, when the petitioner authored that letter, there was no indictment pending against him — a circumstance largely attributable to the fact that he had been a fugitive from justice for upwards of two years. Thus, the communique was sent before the Sixth Amendment right to a speedy trial attached. See supra Part III(A). Although we are willing to give a prisoner's handwritten note a reasonably liberal interpretation, it is surpassingly difficult to read this letter — asking that the petitioner be "brought forward" on charges — as a demand for a speedy trial.

The timing is critical. The usual rule is that a notice sent before the formal commencement of a criminal case is deemed premature (and, therefore, carries little weight) for speedy trial purposes. See United States v. Henson, 945 F.2d 430, 438 (1st Cir.1991) (finding premature a "letter request sent to the district court before any federal charges had been lodged"). This is because one characteristic of a speedy trial demand is that it "notifies the state forum that a violation of constitutional proportions may be ongoing, and it gives the state an opportunity to remedy the situation." Hakeem v. Beyer, 990 F.2d 750, 765 (3d Cir.1993). For this reason, a pre-indictment notice is not particularly influential in the speedy trial context. Cf. Barker, 407 U.S. at 528 (rejecting a rule that would give weight to a premature or pro forma speedy trial demand). We see no reason why this sensible construct should not apply in the circumstances of the instant case.

Putting the September 1986 note to one side, the petitioner's argument collapses like a ruined soufflé. During the entire sixty-eight months that elapsed between the petitioner's indictment and his trial, he never requested a prompt disposition of his case. Although he claims to have called his lawyer in regard to the matter during his sojourn in a Texas jail, he does not say that he instructed his lawyer to seek a speedy trial — nor is there any evidence that his lawyer did so. In short, from and after the date of the indictment, the petitioner took no action of any kind either to accelerate the proceedings in his case or to stimulate an expeditious adjudication. This pattern of avoidance is fairly strong evidence that the last thing that the petitioner wanted was a trial. Cf. United States v. Johnson, 579 F.2d 122, 124 (1st Cir.1978) (drawing such an inference from the fact "that [the defendant] did not reassert his alleged desire for a speedy trial when he learned he was going to be transferred to [prison in another state]").

In an effort to parry this thrust, the petitioner protests that he thought Massachusetts was no longer seeking to prosecute him once he had completed serving his Texas sentence. That protest rings hollow. If, in fact, the petitioner harbored any such impression, he easily could have checked with the proper authorities (or with his own lawyer, for that matter) to verify the accuracy of his belief. Instead, he chose to keep a low profile, apparently hoping that the indictment would die of old age.9 Such a strategy is tailored to frustrate, not further, the goal of a speedy trial. Courts should be very hesitant to reward a defendant who, like the petitioner, has gambled with his speedy trial right and lost. See Look v. Amaral, 725 F.2d 4, 6-8 (1st Cir.1984); see also Aguirre, 994 F.2d at 1457 ("The Speedy Trial Clause primarily protects those who assert their rights, not those who acquiesce in the delay....").

In sum, the record is clear that the petitioner failed to seek a speedy trial with anything remotely approaching diligence. This counts significantly in the speedy trial calculus. See United States v. Bergfeld, 280 F.3d 486, 488 (5th Cir.2002). To make a bad situation worse, the petitioner's actions actually delayed any possibility of a trial. He spent over two years evading the authorities after charges were first filed against him, and then, once he was released from jail in Texas, he spent fifteen months outside Massachusetts without contacting the authorities. Even after the Commonwealth apprehended him and got the case back on track, the petitioner caused a further delay of almost seven months in order to prepare his defense. This period, too, is on his tab. See Hakeem, 990 F.2d at 765 (finding that "unreadiness to proceed to trial" counted against the defendant even when he explicitly asserted his speedy trial right).

To say more on this point would be supererogatory. Bearing in mind that any assertion of speedy trial rights "must be viewed in the light of [the defendant's] other conduct," Loud Hawk, 474 U.S. at 314, the third Barker factor militates strongly against the petitioner.

The fourth Barker factor deals with the degree to which the passage of time has caused undue prejudice to the defendant. The most prevalent form of prejudice involves the extent to which the passage of time has hampered the preparation and presentation of a defense. See id. at 532. Although this seems self-explanatory, the inquiry is more complicated than simply matching delay with prejudice. To the extent that a defendant bears responsibility for causing periods of delay — such as when he goes to ground in an effort to evade prosecution — any prejudice resulting therefrom is his own fault and cannot redound to his benefit. See Aguirre, 994 F.2d at 1456 n. 3. Consequently, under this branch of the prejudice rubric, we focus on the question of whether the delays attributable to the state compromised the preparation or presentation of the defendant's case.

We first address the petitioner's principal contention: that prejudice must be presumed. Before we reach the merits of this argument, however, we must deal with a threshold question. In the absence of a waiver or some other extraordinary circumstance, a federal court's habeas jurisdiction over a petition filed by a state prisoner extends only to claims that have first been presented to the state courts. Rose v. Lundy, 455 U.S. 509, 515, 102 S. Ct. 1198, 71 L. Ed. 2d 379 (1982); Nadworny v. Fair, 872 F.2d 1093, 1095 (1st Cir.1989); 28 U.S.C. §§ 2254(b)-(c). In this instance, the Commonwealth asserts that the "presumptive prejudice" claim is unexhausted (and, therefore, not justiciable) because the petitioner did not present it "face-up and squarely" to the state courts. Martens v. Shannon, 836 F.2d 715, 717 (1st Cir.1988). We do not agree.

The exhaustion doctrine honors hallowed principles of federal-state comity. It serves to ensure that the state courts are sufficiently apprised of a federal claim to have a meaningful opportunity to address that claim. Picard v. Connor, 404 U.S. 270, 275, 92 S. Ct. 509, 30 L. Ed. 2d 438 (1971). In order to achieve exhaustion, the petitioner must have presented both the factual and legal underpinnings of his federal claim to the state's highest court. Scarpa v. DuBois, 38 F.3d 1, 6 (1st Cir. 1994). A claim of non-exhaustion presents a purely legal question, engendering de novo review. See Adelson v. DiPaola, 131 F.3d 259, 262 (1st Cir.1997).

Here, all the relevant facts were presented to the Supreme Judicial Court of Massachusetts in the petitioner's unsuccessful application for leave to obtain further appellate review (ALOFAR). In addition, the ALOFAR openly exposed the petitioner's speedy trial claim and specifically alleged that "the Commonwealth failed to rebut the `presumption of prejudice' due to the unusually lengthy delay." Given this presentation, we think it is likely that a reasonable jurist would have recognized the federal constitutional dimensions of the petitioner's claim. No more is exigible to satisfy the exhaustion requirement.10 See, e.g., Barresi v. Maloney, 296 F.3d 48, 52-56 (1st Cir.2002); Nadworny, 872 F.2d at 1096-99.

On the merits, Doggett is the mainstay of the petitioner's "presumptive prejudice" argument. See Doggett, 505 U.S. at 655 (explaining that "excessive delay presumptively compromises the reliability of a trial in ways that neither party can prove or, for that matter, identify"). Doggett establishes that, when the pretrial delay is grossly excessive, the fourth Barker factor can tilt in the defendant's favor even though no showing of actual prejudice has been made. See Aguirre, 994 F.2d at 1455. This does not mean, however, that the presumption is either automatic or inexorable. See Doggett, 505 U.S. at 655-56. Indeed, this case aptly illustrates why the presumption sometimes may be unwarranted.

To be sure, the overall length of the delay here is comparable to that experienced in Doggett. Nevertheless, the petitioner was responsible for much of the delay and he knew of the charges all along but did little to assert his right to a speedy trial. In these respects, this case is unlike Doggett — and these distinctions are significant. Where delay, though protracted, results in material part from a defendant's unexcused inaction, he is not entitled to a presumption of prejudice. Aguirre, 994 F.2d at 1458.

The presumption is even less appropriate when a defendant takes affirmative steps to delay his trial. See Wilson, 250 F.3d at 396; Robinson v. Whitley, 2 F.3d 562, 570 (5th Cir.1993); Aguirre, 994 F.2d at 1456 n. 3. This is such a case. The responsibility for more than half of the pretrial delay lies at the petitioner's doorstep: he was a fugitive for twenty-seven months, acted much like a fugitive for another fifteen months, laid low for the next four months, and then delayed trial for seven months in order to prepare his defense.

That disperses of the petitioner's "presumptive prejudice" argument. While the state bears some responsibility for forty-two months of delay, see supra Part III(B), there is no rational basis for presuming that the state-caused delay added significantly to the harm caused by the petitioner's dilatory tactics. Thus, the petitioner must corroborate his claim of prejudice by explaining how specific events beyond his control impaired his defense. See, e.g., Aguirre, 994 F.2d at 1455.

Each of the petitioner's allegations of actual prejudice falls short of the mark. First, he complains of a purported inability to locate a security guard who saw him enter his apartment building with Rodriguez on the night of the rape. We agree with the state court that the petitioner failed to explain how the guard would corroborate his story and also failed to make a showing of futile or frustrated efforts to locate the guard. These defects defeat the claim of prejudice. See, e.g., Henson, 945 F.2d at 438 (finding vague and unspecified claims of a missing alibi witness insufficient to justify a finding of prejudice). To them, we add that the guard's testimony (whatever it might have been) is of dubious relevance. After all, the petitioner was acquitted on the kidnapping charge, and it is unclear how the guard — who, at most, could have testified regarding the manner in which Rodriguez and the petitioner proceeded upstairs to the apartment — was in a position to enlighten the jury about the alleged rape. Cf. Look, 725 F.2d at 7 (refusing to find prejudice where the missing witness's testimony could not have been exculpatory).

The petitioner's suggestion that the delay caused the Commonwealth to lose track of relevant evidence (the rope, sock, towel, and tape that he allegedly used to restrain Rodriguez) is similarly flawed. The Commonwealth rediscovered the evidence in a storage locker on the day before trial, and the jury had a full opportunity to inspect it. Moreover, the petitioner refused the trial justice's offer of a continuance to enable him to conduct tests on the evidence. Because the petitioner has failed to show how the temporarily misplaced physical evidence would have assisted his case if produced at an earlier date, he has not demonstrated any cognizable prejudice.

The petitioner next suggests that the passage of time dimmed Rodriguez's memory. Even assuming, for argument's sake, that this suggestion is true, the clouded recollection of a key prosecution witness would seem to be helpful, rather than harmful, to the defense. (Indeed, the petitioner successfully impeached Rodriguez with her memory lapses at trial.) Not infrequently, "delay is a two-edged sword," so loss of memory on the part of a witness does not automatically count in a defendant's favor under the Barker analysis. Loud Hawk, 474 U.S. at 315. Because the petitioner fails to explain how the deficiencies in Rodriguez's testimony prejudiced his defense, we find the claim insufficient to demonstrate actual prejudice.

There are two other recognized forms of prejudice in the speedy trial context. One stems from the prolongation of pretrial incarceration; the second involves the great anxiety that, in some special circumstances, attends the pendency of an indictment. Barker, 407 U.S. at 532-33. While the petitioner makes no claim of unduly oppressive detention or excessive anxiety per se, he muses in his appellate briefs that — even though his incarceration in Texas bore no direct relationship to the charges he faced in Massachusetts — this period of pretrial delay deprived him of an opportunity to serve concurrent sentences for the Texas and Massachusetts offenses.

This argument was not raised in the state court proceedings, and we are reluctant to entertain it here. While variations in the legal theory urged by a habeas petitioner usually are permitted as long as "the ultimate question for disposition" remains the same as the one presented to the state courts, Picard, 404 U.S. at 277-78, a federal habeas court ought not to consider completely new and different theories presented in support of a constitutional claim. See Turner v. Fair, 617 F.2d 7, 11 (1st Cir.1980) (finding that the petitioner's argument amounted to a "new theory" to support an alleged Sixth Amendment violation and therefore was not exhausted); see also Trigones v. Bissonnette, 296 F.3d 1, 8-9 (1st Cir.2002) [No. 00-2504, slip op. at 12-13] (refusing to entertain a new argument in support of a Confrontation Clause claim). Because the petitioner made no reference to this particular form of prejudice in the state court proceedings, we decline to consider it here.

Our decision to treat the petitioner's argument as unexhausted is fortified by the fact that the issue he seeks to raise is one that the state courts were in the best position to adjudicate. After all, the likelihood that a concurrent sentence might have been imposed absent the delay is a matter that depends heavily on state sentencing practices. This idiosyncracy makes it all the more advisable that we not reach out for the previously unadjudicated question. See Granberry v. Greer, 481 U.S. 129, 134-35, 107 S. Ct. 1671, 95 L. Ed. 2d 119 (1987) (explaining that when a "case presents an issue on which an unresolved question of fact or of state law might have an important bearing, both comity and judicial efficiency may make it appropriate for the court to insist on complete exhaustion"); Kokoraleis v. Gilmore, 131 F.3d 692, 694 (7th Cir.1997) (similar). Under these circumstances, we need not give weight to this belated assertion of "sentencing prejudice."11

That ends this aspect of the matter. When all is said and done, we conclude that the petitioner failed to establish any actual prejudice sufficient to satisfy the fourth Barker factor.

The district court found the state court's resolution of the Barker factors so far afield as to warrant habeas relief. See RaShad, 204 F.Supp.2d at 115. We believe that this is an overly harsh assessment — and one that is, in the final analysis, insupportable.

The district court's rationale is flawed in several respects. For one thing, the court placed great emphasis on what it perceived as the state court's failure to assign responsibility to the Commonwealth for the period of the petitioner's Texas incarceration, deeming this to be an unreasonable application of the Barker standard. Id. at 103. But that ascribes to the Massachusetts Appeals Court a more detailed analysis than the court in fact undertook. The Appeals Court stated only that the petitioner was responsible for a "sizable" portion of the aggregate elapsed time between indictment and trial without assigning responsibility for each individual period of delay. In all events, the AEDPA requires that a federal habeas court evaluate the state court's ultimate conclusion, not its announced rationale. Ouber, 293 F.3d at 34; Bui, 170 F.3d at 243. That means that the state court's overall balancing of the Barker factors, not its handling of any single factor, is the proper focal point in this case.

For another thing, the district court declared that the state court's assignment of responsibility to the petitioner for the nineteen-month period of delay following his release from confinement in Texas was premised on untenable findings of fact. RaShad, 204 F.Supp.2d at 103. In particular, the district court quarreled with the notion that the petitioner was a fugitive during that period. See id. at 105-06. In staking out that position, the district court approached the question as though the petitioner had been living openly in Massachusetts during the entire nineteen-month period. See id. This approach lacks a solid footing. The record is silent as to the petitioner's whereabouts during the first fifteen months of that period, see supra Part III(B)(2), so the most that can be said is that the petitioner was in Massachusetts for the last four months. Because the record is barren of any evidence as to the petitioner's activities and whereabouts during the first fifteen months (other than some inconclusive references to a Texas pre-release program), we must presume that the state court's finding was correct as to that period. See 28 U.S.C. § 2254(e)(1); see also Ouber, 293 F.3d at 27 (explaining that, under the AEDPA, "the state court's factual findings are entitled to a presumption of correctness that can be rebutted only by clear and convincing evidence to the contrary"). To the extent that the district court failed to give the Commonwealth the benefit of such a presumption, it went against the grain of the AEDPA's well-honed standard of review.12

The district court also criticized the state court's finding that the petitioner, when eventually apprehended, "denied his identity." RaShad, 204 F.Supp.2d at 107. This is no more than a semantic quibble. It is undisputed that the petitioner gave his true name and birth date while at the same time vehemently denying that he was the person charged in the underlying indictment. Rather than resolve such a question de novo, as the district court did, see id. at 108, we give credence to the state court's plausible characterization of the event. We add that, contrary to the district court's implication, such deference is due despite the fact that the contested finding was made by a state appellate court rather than by a state trial court. See, e.g., Everett, 290 F.3d at 507.

Our most significant disagreements with the district court's rationale concern the third and fourth Barker factors. As to the third factor, the state court found that, "[e]xcept for the [petitioner's] notice to the court in September 1986," his course of conduct was characterized by "inaction and avoidance." The district court interpreted this reference to the September 1986 missive as a finding that the petitioner had "asserted his speedy trial right clearly and unequivocally in his letter." RaShad, 204 F.Supp.2d at 108. This misreads the record: the Massachusetts Appeals Court made no such finding, and the record will not support so bold a characterization of the petitioner's note. The Supreme Court has warned against assigning talismanic significance to a lone assertion of the speedy trial right, see Barker, 407 U.S. at 528-29, and the assertion relied on here (coming, as it did, before any indictment had been handed up) is especially problematic. Considering the evidence in its totality — including the evidence suggesting that the petitioner did not really want a speedy trial — the state court's finding of "inaction and avoidance" on the petitioner's part seems entirely supportable.

As for the fourth prong of the Barker framework, the district court concluded that the state court's approach was contrary to Barker and Doggett to the extent that it failed to presume prejudice. RaShad, 204 F.Supp.2d at 110 n. 21. This allowed the district court to place the burden of disproving prejudice on the Commonwealth and to enter a finding of prejudice despite its acknowledgment that the petitioner's specific prejudice allegations might not suffice to show actual prejudice. See id. at 110-12.

The district court's treatment of this factor rises or falls on the validity of its conclusion that this case demands a presumption of prejudice. We think it must fall: the Massachusetts Appeals Court did not employ such a presumption, and we have determined that a presumption of prejudice is unwarranted here. See supra Part III(D). The state court's failure to extend Doggett to a case in which the defendant (1) knew of the pending charges, and (2) himself caused much of the delay was not unreasonable. See Williams, 529 U.S. at 407-08.

We are left with the question of whether the state court's balancing of the four Barker factors was reasonable. In answering this question, the AEDPA requires that we cede substantial deference to the state court's legal and factual conclusions. See 28 U.S.C. §§ 2254(d)(1)-(2), (e)(1). This deference is heightened in a Barker-type case, because constructing a balance among the four factors "is more judicial art than science." Look, 725 F.2d at 8.

As the district court observed, the opinion of the Massachusetts Appeals Court is not very comprehensive. We find it disconcerting that the opinion did not discuss explicitly the consideration most favorable to the petitioner — the Commonwealth's negligence in not bringing him to trial during his stint in a Texas jail. It is not our function, however, to grade a state court opinion as if it were a law school examination. Rather, we review the state court's ultimate findings and conclusions to ascertain whether they constitute an unreasonable application of clearly established Supreme Court precedent. Ouber, 293 F.3d at 34; Bui, 170 F.3d at 243.

In this instance, the state court supportably determined that "a sizable portion of the delay was caused by the defendant," that his conduct after he was indicted reflected "inaction and avoidance" on his part (rather than an intent to press his speedy trial claim), and that he had failed to demonstrate any actual prejudice. Under the applicable Supreme Court precedents, these determinations suffice to ground a denial of the petitioner's speedy trial claim. See Barker, 407 U.S. at 532 (explaining that a defendant who cannot show that he "wanted a speedy trial" must carry a heavy burden on the other factors).

We need go no further. When all the relevant factors are given proper weight, the state court's ultimate conclusion that the petitioner suffered no deprivation of his constitutional right to a speedy trial may not be inevitable, but that conclusion nonetheless falls within the sphere of objective reasonableness. Consequently, we reverse the district court's contrary determination and direct that court to reinstate the petitioner's conviction.

Reversed and remanded.