This case arises out of a "detainer," which is a request filed by a criminal justice agency with the institution in which a prisoner is incarcerated, asking that the prisoner be held for the agency, or that the agency be advised when the prisoner's release is imminent. Indiana and Michigan, along with 46 other States, the District of Columbia, and the United States, are parties to the Interstate Agreement on Detainers (IAD). See Ind.Code § 35-33-10-4 (1988); Mich.Comp.Laws § 780.601 (1979); Pub.L. 91-538, 84 Stat. 1397-1403, 18 U.S.C.App. § 2; 11 U.L.A. 213-214 (Supp.1992) (listing jurisdictions). Two provisions of that interstate agreement give rise to the present suit: Article III and Article V(c), which are set forth in the margin.1

On February 29, 1988, petitioner was charged in Jackson County, Michigan, with armed robbery, possession of a firearm during a felony, and assault with intent to murder. At the time, he was held in connection with unrelated offenses at the Westville Correctional Center in Fort Wayne, Indiana. The Jackson County Prosecuting Attorney therefore lodged a detainer against him. On September 7, 1988, the Indiana correctional authorities informed petitioner of the detainer, and he gave them his request for final disposition of the Michigan charges. On September 22, the prison authorities mailed petitioner's request; and on September 26, 1988, the Jackson County Prosecuting Attorney and the Jackson County Circuit Court received it. Petitioner's trial on the Michigan charges began on March 22, 1989, 177 days after his request was delivered to the Michigan officials and 196 days after petitioner gave his request to the Indiana prison authorities. 439 Mich. 117, 118, 479 N.W.2d 625 (1992) (per curiam ).

Prior to trial, petitioner moved for dismissal with prejudice pursuant to Article V(c) of the IAD, on the ground that his trial would not begin until after the 180-day time limit set forth in Article III(a). The trial court denied the motion, reasoning that the 180-day time period did not commence until the Michigan prosecutor's office received petitioner's request. App. 36. Petitioner was convicted on all charges except assault with intent to murder, but his conviction was set aside by the Michigan Court of Appeals, which held that "the commencement of the 180-day statutory period was triggered by [petitioner's] request for final disposition to the [Indiana] prison officials." Id., at 39. The Supreme Court of Michigan summarily reversed. 439 Mich. 117, 479 N.W.2d 625 (1992) (per curiam ). We granted certiorari. 504 U.S. ----, 112 S.Ct. 1935, 118 L.Ed.2d 542 (1992).

The outcome of the present case turns upon the meaning of the phrase, in Article III(a), "within one hundred and eighty days after he shall have caused to be delivered." The issue, specifically, is whether, within the factual context before us, that phrase refers to (1) the time at which petitioner transmitted his notice and request (hereinafter simply "request") to the Indiana correctional authorities; or rather (2) the time at which the Michigan prosecutor and court (hereinafter simply "prosecutor") received that request.

Respondent argues that no one can have "caused something to be delivered" unless delivery in fact occurs. That is self-evidently true,2 and so we must reject petitioner's contention that a prisoner's transmittal of an IAD request to the prison authorities commences the 180-day period even if the request gets lost in the mail and is never delivered to the "receiving" State (i.e., the State lodging the detainer, see Article II(c)). That still leaves open the textual possibility, however, that, once delivery has been made, the 180 days must be computed, not from the date of delivery but from the date of transmittal to the prison authorities. That is the only possibility the balance of our discussion will consider; and for convenience we shall refer to it as petitioner's interpretation.

Respondent places great reliance upon the provision's use of the future perfect tense ("shall have caused to be delivered"). It seems to us, however, that the future perfect would be an appropriate tense for both interpretations: The prisoner's transmittal of his request to the warden (if that is the triggering event), or the prosecutor's receipt of the request (if that is the triggering event) is to be completed ("perfected") at some date in the future (viewed from the time of the IAD's adoption) before some other date in the future that is under discussion (expiration of the 180 days). We think it must be acknowledged that the language will literally bear either interpretation—i.e., that the crucial point is the prisoner's transmittal of his request, or that it is the prosecutor's receipt of the request. One can almost be induced to accept one interpretation or the other on the basis of which words are emphasized: "shall have caused to be delivered" versus "shall have caused to be delivered."3

Though the text alone is indeterminate, we think resolution of the ambiguity is readily to be found in what might be called the sense of the matter, and in the import of related provisions. As to the former: Petitioner would have us believe that the choice of "triggers" for the 180-day time period lies between, on the one hand, the date the request is received by the prosecutor and, on the other hand, the date the request is delivered to the warden of the prison. In fact, however, while the former option is clearly identified by the textual term "delivered," there is no textual identification of a clear alternative at the other end. If one seeks to determine the moment at which a prisoner "caused" the later delivery of a properly completed request, nothing in law or logic suggests that it must be when he placed the request in the hands of the warden. Perhaps it was when he gave the request to a fellow inmate to deliver to the warden—or even when he mailed it to the warden (Article III(b) provides that the request "shall be given or sent by the prisoner to the warden" (emphasis added)). It seems unlikely that a legislature would select, for the starting point of a statute of limitations, a concept so indeterminate as "caused." It makes more sense to think that, as respondent contends, delivery is the key concept, and that paragraph (a) includes the notion of causality (rather than referring simply to "delivery" by the prisoner) merely to be more precise, anticipating the requirement of paragraph (b) that delivery be made by the warden upon the prisoner's initiation.

Another common-sense indication pointing to the same conclusion is to be found in what might be termed (in current political jargon) the "worst-case scenarios" under the two interpretations of the IAD. Under respondent's interpretation, it is possible that a warden, through negligence or even malice, can delay forwarding of the request and thus postpone the starting of the 180-day clock. At worst, the prisoner (if he has not checked about the matter for half a year) will not learn about the delay until several hundred days have elapsed with no trial. The result is that he will spend several hundred additional days under detainer (which entails certain disabilities, such as disqualification from certain rehabilitative programs, see United States v. Mauro, 436 U.S. 340, 359, 98 S.Ct. 1834, 1846, 56 L.Ed.2d 329 (1978)), and will have his trial delayed several hundred days.4 That result is bad, given the intent of the IAD. It is, however, no worse than what regularly occurred before the IAD was adopted, and in any event cannot be entirely avoided by embracing petitioner's view that transmittal to the warden is the measuring event. As we have said, the IAD unquestionably requires delivery, and only after that has occurred can one entertain the possibility of counting the 180 days from the transmittal to the warden. Thus, the careless or malicious warden, under petitioner's interpretation, may be unable to delay commencement of the 180-day period, but can prevent it entirely, by simply failing to forward the request. More importantly, however, the worst-case scenario under petitioner's interpretation produces results that are significantly worse: If, through negligence of the warden, a prisoner's IAD request is delivered to the prosecutor more than 180 days after it was transmitted to the warden, the prosecution will be precluded before the prosecutor even knows it has been requested. It is possible, though by no means certain, that this consequence could be avoided by the receiving State court's invocation of the "good-cause continuance" clause of Article III(a)5—but it seems to us implausible that such a plainly undesirable result was meant to be avoided only by resort to the (largely discretionary) application of that provision. It is more reasonable to think that the receiving State's prosecutors are in no risk of losing their case until they have been informed of the request for trial.

Indications in the text of Article III confirm, in our view, that the receiving State's receipt of the request starts the clock. The most significant is the provision of Article III(b) requiring the warden to forward the prisoner's request and accompanying documents "by registered or certified mail, return receipt requested." The IAD thus provides for documentary evidence of the date on which the request is delivered to the officials of the receiving State, but requires no record of the date on which it is transmitted to the warden (assuming that is to be considered the act of "causing"). That would be peculiar if the latter rather than the former were the critical date. Another textual clue, we think, is the IAD's apparent indifference as to the manner of transmittal to the warden: Article III(b) says only that the request "shall be given or sent by the prisoner to the warden" (emphasis added). A strange nonchalance, if the giving or sending (either one) is to start the 180 days. Petitioner avoids this difficulty by simply positing that it is the warden's receipt, no matter what the manner of giving or sending, that starts the clock—but there is simply no textual basis for that; surely the "causing" which petitioner considers central occurs upon the giving or sending.

Petitioner makes the policy argument that "[f]airness requires the burden of compliance with the requirements of the IAD to be placed entirely on the law enforcement officials involved, since the prisoner has little ability to enforce compliance," Brief for Petitioner 8, and that any other approach would "frustrate the higher purpose" of the IAD, leaving "neither a legal nor a practical limit on the length of time prison authorities could delay forwarding a [request]," id., at 20. These arguments, however, assume the availability of a reading that would give effect to a request that is never delivered at all. (Otherwise, it remains within the power of the warden to frustrate the IAD by simply not forwarding.) As we have observed, the textual requirement "shall have caused to be delivered" is simply not susceptible of such a reading. Petitioner's "fairness" and "higher purpose" arguments are, in other words, more appropriately addressed to the legislatures of the contracting States, which adopted the IAD's text.

Our discussion has addressed only the second question presented in the petition for writ of certiorari; we have concluded that our grant as to the first question was improvident, and do not reach the issue it presents. We hold that the 180-day time period in Article III(a) of the IAD does not commence until the prisoner's request for final disposition of the charges against him has actually been delivered to the court and prosecuting officer of the jurisdiction that lodged the detainer against him. The judgment of the Supreme Court of Michigan is affirmed.

It is so ordered.

Justice BLACKMUN, with whom Justice STEVENS joins, dissenting.

I am not persuaded that the language of Article III is ambiguous. The majority suggests that a search for the literal meaning of the contested phrase comes down to an unresolvable contest between a reading that emphasizes the word "caused" and one that emphasizes the word "delivered." But Article III contains another word that is at least as significant. That word favors petitioner's interpretation. The word is "he." The 180-day clock begins after he—the prisoner —"shall have caused" the request to be delivered. The focus is on the prisoner's act, and that act is complete when he transmits his request to the warden. That is the last time at which the inmate can be said to have done anything to "have caused to be delivered" the request. Any other reading renders the words "he shall have caused" superfluous.

Even if the provision's focus on the prisoner's act were not so clear, the statute could not be read as Michigan suggests. The provision's use of the future perfect tense is highly significant. Contrary to the majority's contention that "the future perfect would be an appropriate tense for both interpretations," ante, at ----, the logical way to express the idea that receipt must be perfected before the provision applies would be to start the clock 180 days "after he has caused the request to have been delivered." But the IAD does not say that, nor does it use the vastly more simple, "after delivery."

That this construction was intentional is supported by the drafting history of the IAD. When the Council of State Governments proposed the agreement governing interstate detainers, it also proposed model legislation governing intrastate detainers. See suggested State Legislation, Program for 1957, pp. 77-78 (1956). Both proposals contained language virtually identical to the language in Article III(a). See id., at 77. The Council stated that the intrastate proposal was "based substantially on statutes operative in California and Oregon." Id., at 76. Critically, however, neither State's provision referred to a delivery "caused" by the prisoner. The Oregon statute required trial "within 90 days of receipt" by the district attorney of the prisoner's notice, Act of Apr. 29, 1955, ch. 387, § 2(1), 1955 Ore.Laws 435, and the California law required trial "within ninety days after [he] shall have delivered" his request to the prosecutor, Act of May 28, 1931, ch. 486, § 1, 1931 Cal.Stats. 1060. If, as Michigan insists here, see Tr. of Oral Arg. 23, 26, 37, the Council's use of "caused to be delivered" was somehow meant to convey "actual receipt," then the drafters' failure to follow the clear and uncomplicated model offered by the Oregon provision is puzzling in the extreme. When asked at oral argument about this failure, counsel for amicus the United States replied that "the problem with using the verb receive rather than the verb deliver in Article III is that. . . . [t]hat would shift the focus away from the prisoner, and the prisoner has a vital role under Article III . . . because he initiates the process." Id., at 41. I submit that the focus on the prisoner is precisely the point, and that the reason the drafters used the language they did is because the 180-day provision is triggered by the action of the inmate.

Nevertheless, the majority finds the disputed language to be ambiguous, ante, at ----, and it exhibits no interest in the history of the IAD. Instead, the majority asserts that the answer to

the problem is to be found in "the sense of the matter." Ibid. But petitioner's reading prevails in the arena of "sense," as well.

I turn first to the majority's assumption that the 180-day provision is not triggered if the request is never delivered. Because "the IAD unquestionably requires delivery, and only after that has occurred can one entertain the possibility of counting the 180 days from the transmittal to the warden," ante, at ----, the majority attacks as illogical a reading under which the negligent or malicious warden—who can prevent entirely the operation of the 180-day rule simply by failing to forward the prisoner's request—could not delay the starting of the clock. Ibid. That premise is flawed. Obviously, the rule anticipates actual delivery. Art. III(b) requires prison officials to forward a prisoner's request promptly, as well. The fact that the rule for marking the start of the 180-day period is written in a fashion that contemplates actual delivery, however, does not mean that it cannot apply if the request is never delivered. Although the IAD assumes that its signatories will abide by its terms, I find nothing strange in the notion that the 180-day provision might be construed to apply as well to an unanticipated act of bad faith.1

Even on its own terms, the majority's construction is not faithful to the purposes of the IAD. The IAD's primary purpose is not to protect prosecutors' calendars, or even to protect prosecutions, but to provide a swift and certain means for resolving the uncertainties and alleviating the disabilities created by outstanding detainers. See Art. I; Carchman v. Nash, 473 U.S. 716, 720, 105 S.Ct. 3401, 3403, 87 L.Ed.2d 516 (1985); Note, The Effect of Violations of the Interstate Agreement on Detainers on Subject Matter Jurisdiction, 54 Ford.L.Rev. 1209, 1210, n. 12 (1986). If the 180 days from the prisoner's invocation of the IAD is allowed to stretch into 200 or 250 or 350 days, that purpose is defeated.

In each of this Court's decisions construing the IAD, it properly has relied upon and emphasized the purpose of the IAD. See Carchman v. Nash, 473 U.S., at 720, 729-734, 105 S.Ct., at 3403, 3408-3410; Cuyler v. Adams, 449 U.S. 433, 448-450, 101 S.Ct. 703, 711-712, 66 L.Ed.2d 641 (1981); United States v. Mauro, 436 U.S. 340, 361-362, 98 S.Ct. 1834, 1848, 56 L.Ed.2d 329 (1978). The majority, however, gives that purpose short shrift, focusing instead on "worst-case scenarios," ante, at ----, and on an assessment of the balance of harms under each interpretation. Two assumptions appear to underlie that inquiry. The first evident in the cursory and conditional nature of the concession that to spend several hundred additional days under detainer "is bad, given the intent of the IAD," ante, at ----—is that the burden of spending extra time under detainer is relatively minor. The failure to take seriously the harm suffered by a prisoner under detainer is further apparent in the majority's offhand and insensitive description of the practical impact of such status. To say that the prisoner under detainer faces "certain disabilities, such as disqualification from certain rehabilitative programs," ante, at ----, is to understate the matter profoundly. This Court pointed out in Carchman v. Nash, supra, that the prisoner under detainer bears a very heavy burden:

" '[T]he inmate is (1) deprived of an opportunity to obtain a sentence to run concurrently with the sentence being served at the time the detainer is filed; (2) classified as a maximum or close custody risk; (3) ineligible for initial assignments to less than maximum security prisons (i.e., honor farms or forestry camp work); (4) ineligible for trustee [sic ] status; (5) not allowed to live in preferred living quarters such as dormitories; (6) ineligible for study-release programs or work-release programs; (7) ineligible to be transferred to preferred medium or minimum custody institutions within the correctional system, which includes the removal of any possibility of transfer to an institution more appropriate for youthful offenders; (8) not entitled to preferred prison jobs which carry higher wages and entitle [him] to additional good time credits against [his] sentence; (9) inhibited by the denial of possibility of parole or any commutation of his sentence; (10) caused anxiety and thus hindered in the overall rehabilitation process since he cannot take maximum advantage of his institutional opportunities.' " 473 U.S., at 730, n. 8, 105 S.Ct., at 3408, n. 8, quoting Cooper v. Lockhart, 489 F.2d 308, 314, n. 10 (CA8 1973).

These harms are substantial and well-recognized. See, e.g., Smith v. Hooey, 393 U.S. 374, 379, 89 S.Ct. 575, 577, 21 L.Ed.2d 607 (1969); United States v. Ford, 550 F.2d 732, 737-740 (CA2 1977) (citing cases), aff'd sub nom. United States v. Mauro, 436 U.S. 340, 98 S.Ct. 1834, 56 L.Ed.2d 329 (1978); L. Abramson, Criminal Detainers 29-34 (1979); Note, 54 Ford.L.Rev., at 1210, n. 12. More important for our purposes, they were the reason for the IAD's creation in the first place. The majority's sanguine reassurance that delays of several hundred days, while "bad," are "no worse than what regularly occurred before the IAD was adopted," ante, at ----, is thus perplexing. The fact that the majority's reading leaves prisoners no worse off than if the IAD had never been adopted proves nothing at all, except perhaps that the majority's approach nullifies the ends that the IAD was meant to achieve. Our task, however, is not to negate the IAD but to interpret it. That task is impossible without a proper understanding of the seriousness with which the IAD regards the damage done by unnecessarily long periods spent under detainer.

The majority's misunderstanding of the stakes on the inmate's side of the scale is matched by its miscalculation of the interest of the State. It is widely acknowledged that only a fraction of all detainers ultimately result in conviction or further imprisonment. See J. Gobert & N. Cohen, Rights of Prisoners 284 (1981); Dauber, Reforming the Detainer System: A Case Study, 7 Crim.L.Bull. 669, 689-690 (1971); Note, 54 Ford.L.Rev., at 1210, n. 12. It is not uncommon for a detainer to be withdrawn just prior to the completion of the prisoner's sentence. See Carchman v. Nash, 473 U.S., at 729-730, 105 S.Ct., at 3408; Note, 54 Ford.L.Rev., at 1210, n. 12; Comment, Interstate Agreement on Detainers and the Rights It Created, 18 Akron L.Rev. 691, 692 (1985). All too often, detainers are filed groundlessly or even in bad faith, see United States v. Mauro, 436 U.S., at 358, and n. 25, 98 S.Ct., at 1846, and n. 25, solely for the purpose of harassment, seeCarchman v. Nash, 473 U.S., at 729, n. 6, 105 S.Ct., at 3408, n. 6. For this reason, Article III is intended to provide the prisoner " 'with a procedure for bringing about a prompt test of the substantiality of detainers placed against him by other jurisdictions.' " Id., at 730, n. 6, 105 S.Ct., at 3408, n. 6 (quoting House and Senate Reports).

These two observations—that detainers burden prisoners with onerous disabilities and that the paradigmatic detainer does not result in a new conviction—suggest that the majority has not properly assessed the balance of interests that underlies the IAD's design. Particularly in light of Article IX's command that the IAD "shall be liberally construed so as to effectuate its purposes," I find the majority's interpretation, which countenances lengthy and indeterminate delays in the resolution of outstanding detainers, impossible to sustain.

Finally, I must emphasize the somewhat obvious fact that a prisoner has no power of supervision over prison officials. Once he has handed over his request to the prison authorities, he has done all that he can do to set the process in motion. For that reason, this Court held in Houston v. Lack, 487 U.S. 266, 108 S.Ct. 2379, 101 L.Ed.2d 245 (1988), that a pro se prisoner's notice of appeal is "filed" at the moment it is conveyed to prison authorities for forwarding to the district court. Because of the prisoner's powerlessness, the IAD's inmate-initiated 180-day period serves as a useful incentive to prison officials to forward IAD requests speedily. The Solicitor General asserts that the prisoner somehow is in a better position than are officials in the receiving State to ensure that his request is forwarded promptly, because, for example, "the prisoner can insist that he be provided with proof that his request has been mailed to the appropriate officials." Brief for United States as Amicus Curiae 16-17. This seems to me to be severely out of touch with reality. A prisoner's demands cannot be expected to generate the same degree of concern as do the inquiries and interests of a sister State. Because of the IAD's reciprocal nature, the signatories, who can press for a speedy turnaround from a position of strength, are far better able to bear the risk of a failure to meet the 180-day deadline.2

The IAD's 180-day clock is intended to give the prisoner a lever with which to move forward a process that will enable him to know his fate and perhaps eliminate burdensome conditions. It makes no sense to interpret the IAD so as to remove from its intended beneficiary the power to start that clock. Accordingly, I dissent.