UNITED STATES of America, Appellee, v. Shaun K. O'NEIL, Defendant, Appellant, 11 F.3d 292

SELYA, Circuit Judge. Concluding, as we do, that several courts of appeals have read the supervised release revocation provision (SRR provision), 18 U.S.C. § 3583(e)(3) (1988 & Supp. Ill 1991), in too crabbed a manner, we hold today that this statute permits a district court, in resentencing a person who has violated the conditions of his or her original term of supervised release, to impose a new term of supervised release in conjunction with an additional prison term, subject to certain restrictions limned in the statute itself.

Because we are staking out a position at variance with the majority view, we write at some length to explain our rationale. I. BACKGROUND OF THE CASE After having broken into a post office and stolen mail in violation of 18 U.S.C. §§ 1708, 2115 (1988), defendant-appellant Shaun K. O’Neil pleaded guilty to a class D felony. On November 9, 1990, the district court sentenced him to serve twenty-one months in prison (the top of the applicable guideline sentencing range), followed by three years of supervised release (the maximum allowed by statute). We affirmed the sentence. See United States v.

O’Neil, 936 F.2d 599 (1st Cir.1991). Soon after his release from the penitentiary, appellant committed several significant violations of the supervised release conditions, e.g., stealing a firearm while intoxicated. Dubbing appellant a “walking juvenile crime wave” who posed “a serious danger to the public,” the district judge revoked the original term of supervised release and sentenced appellant to an additional twenty-four months in prison, to be followed by a new three-year supervised release term. O’Neil appeals, asking that we vacate his sentence and remand for resentencing.

His principal allegation is that the reimposition of supervised release exceeds the district court’s statutory authority. II. THE STATUTE Passed as part of the Sentencing Reform Act of 1984, 18 U.S.C. §§ 3551-3559, 3561-3566, 3571-3574, 3581-3586, & 28 U.S.C. §§ 991-98 (1988 & Supps.), the supervised release alteration statute, 18 U.S.C. § 3583(e), of which the SRR provision is a part, authorizes a court to alter a term of supervised release in a number of ways.

A court may: (1) terminate a term of supervised release and discharge the person released at any time after the expiration of one year of supervised release ...; (2) extend a term of supervised release if less than the maximum authorized term was previously imposed, and may modify, reduce, or enlarge the conditions of supervised release, at any time prior to the expiration or termination of the term of supervised release ...; (3) revoke a term of supervised release, and require the person to serve in prison all or part of the term of supervised release without credit for time previously served on postrelease supervision, if it finds by a preponderance of the evidence that the person violated a condition of supervised release, pursuant to the provisions of the Federal Rules of Criminal Procedure that are applicable to probation revocation and to the provisions of applicable policy statements issued by the Sentencing Commission, except that a person whose term is revoked under this paragraph may not be required to serve more than 3 years in prison if the offense for which the person was convicted was a Class B felony, or more than 2 years in prison if the offense was a Class C or D felony; or (4) order the person to remain at his place of residence during nonworking hours.... 18 U.S.C. § 3583(e) (emphasis supplied).

The present controversy centers on the third of these four options. ■ The alteration statute empowers a resen-tencing court, in certain circumstances, to elongate a previously imposed term of supervised release, 18 U.S.C. § 3583(e)(2), or, in other circumstances, to revoke supervision and impose imprisonment in lieu of supervision, id. at § 3583(e)(3).

What is unclear, and what has confounded the courts, is whether an intermediate resentencing option exists: Does the statute allow a court to revoke supervision and, in effect, restructure the defendant’s sentence by imposing a combination of imprisonment plus further supervision? Although this court has never addressed the question, a minimum of six circuits have read the statute to foreclose the reimposition of a term of supervised release following revocation and imprisonment. See United States v. Truss, 4 F.3d 437, 438 (6th Cir.1993); United States v.

McGee, 981 F.2d 271, 274-76 (7th Cir.1992); United States v. Koehler, 973 Fid 132, 134-36 (2d Cir.1992); United States v. Cooper, 962 F.2d 339, 340-42 (4th Cir.1992); United States v. Holmes, 954 F.2d 270, 271-73 (5th Cir.1992); United States v. Behnezhad, 907 F.2d 896, 898-99 (9th Cir.1990); see also United States v. Gozlon-Peretz, 894 F.2d 1402, 1405 n. 5 (dictum), amended, 910 F.2d 1152 (3d Cir.1990), aff'd on other grounds, 498 U.S. 395, 111 S.Ct. 840, 112 L.Ed.2d 919 (1991). The Tenth Circuit came to the same conclusion belatedly, after reversing its field. See United States v.

Rockwell, 984 F.2d 1112, 1117 (10th Cir.) (overruling United States v. Boling, 947 F.2d 1461 (10th Cir.1991)), cert. denied, — U.S. -, 113 S.Ct. 2945, 124 L.Ed.2d 693 (1993). The Eleventh Circuit has sent mixed signals. In United States v. Tatum, 998 F.2d 893, 894-95 (11th Cir.1993) (per curiam), the court embraced the majority view. A second panel, two weeks later, bowed to Tatum on stare decisis grounds; but, in a sharp departure from customary practice, all three judges expressed their profound disagreement with Tatum’s holding. See United States v. Williams, 2 F.3d 363, 365 (11th Cir.1993).

Thus, nine circuits in all read the SRR provision narrowly. On the other side of the ledger, the Eighth Circuit stands as a waif in the wilderness. See United States v. Schrader, 973 F.2d 623, 624-25 (8th Cir.1992) (holding that

section 3583(e)(3) permits the reimposition of a term of supervised release following revocation and imprisonment); see also United States v. Levi, 2 F.3d 842, 846 (8th Cir.1993) (reaffirming Schrader). We are called upon today to add our voice to the chorus. We approach this task mindful that, while the decision to revoke a term of supervised release is ordinarily reviewable for abuse of discretion, the quintessentially legal question of whether a post-revocation sentence exceeds statutory limits necessitates plenary review. See Rockwell, 984 F.2d at 1114; see also United States v. St.

Cyr, 977 F.2d 698, 701 (1st Cir.1992) (holding that interpretive questions under the sentencing guidelines should be reviewed de novo). III. THE COMPETING

INTERPRETATIONS We start our quest by elucidating the two ways in which the SRR provision may be read as a coherent command. A To achieve the result reached by the majority of courts, the assiduous reader must proceed along the following lines. First, read the word “revoke” restrictively, i.e., in the sense of “cancel” or “annul,” so that it does not allow either the recommencing of the previously imposed term of supervision or the commencement of a new term of supervision.

Next, suppose that the word “term”, when used for the second time in the SRR provision, does not imply that there is a term of supervision in existence, but merely serves to set a temporal limit on the prison sentence that may be imposed following revocation; or, put another way, that the second use of the word “term” is to be read as if it were shorthand for a more verbose phrase like “the time period equivalent to what would have been the term.” Only if these interpretive steps are taken does it become clear, under the SRR provision, that a court may absolutely extinguish a term of supervised release and impose a new prison term, subject to certain statutory limitations, but, withal, may not impose any other or further supervision term.

B The other possible parsing of the SRR provision proceeds in three phases. At the outset, consider the possibility that the word “revoke” means simply to “recall.” See, e.g., Black’s Law Dictionary 1322 (6th ed. 1990) (defining “revoke” as “[t]o annul or make void by recalling or taking back_”). If “revoke” is read in this way, the SRR provision is not inconsistent with the recommencement of supervised release.

Next, from the fact that the SRR provision mentions a “term of supervised release” in that portion of the text following the conferral of the power to revoke, the reader plausibly can infer that the supervision term recommenced upon revocation — else there would be no term then in existence.

Finally, having posited that the supervision term is alive and well, notwithstanding the court’s order of revocation, the reader can conclude that, in authorizing the court to send a person to prison after revocation for “all or part of the term,” the SRR provision contemplates that any remaining part of the original, recalled term will be devoted to supervision. On this reading, the SRR provision allows a court to call back a term of supervised release, recommence the term, convert all or part of it into jail time (up to the statutory limit), and retain any remainder as a period of non-detentive monitoring.

Before leaving these competing versions, we wish to make two preliminary points. First, we do not regard the initial step in these analyses to be indispensable. See infra

Part IV(A). Second, each of the competing versions requires the reader to make a leap of faith beyond the four corners of the SRR provision itself. In this sense, then, the playing field is level. IV. CHOOSING AN

INTERPRETATION We turn to the difficult choice between these meanings, using the full panoply of available aids to the construction of legislative enactments. A In approaching statutory

interpretation, “it is axiomatic that the plain words and structure of the statute must be paramount.” United States v. Aversa, 984 F.2d 493, 498 (1st Cir.1993) (en banc). Most of the courts that have read

section 3583(

e) to foreclose the imposition of a post-revocation term of supervised release have done so under the banner of plain meaning. Those courts read the word “revoke” as signifying an extin-guishment so uncompromising as to preclude a post-revocation term of supervision. See, e.g., McGee, 981 F.2d at 274; Koehler, 973 F.2d at 134-35; Holmes, 954 F.2d at 272. This inflexible insistence upon a particular version of lexicographic orthodoxy seemingly overlooks that “the plain-meaning doctrine is not a pedagogical absolute.” Greenwood Trust Co. v. Massachusetts, 971 F.2d 818, 825 (1st Cir.1992), cert. denied, — U.S. -, 113 S.Ct. 974, 122 L.Ed.2d 129 (1993). In particular, “[t]erms in

an act whose meaning may appear plain outside the scheme of the statute can take on a different meaning when read in their proper context.” Id. (citing various Supreme Court precedents). The Williams court found “revoke” plain enough, but read it differently. It suggested that “revoke” could be read in the alternative sense of “call back.” Williams, 2 F.3d at 365. This sense is best illustrated by the poet William Cowper, who wrote: How readily we wish time spent revok’d, That we might try the ground again.... The Task, Book VI, 1. 25 (1784); see also supra p. 295 (quoting Black’s Law Dictionary). While we regard this approach as plausible, we do not see why even the most inelastic

interpretation of “revoke” would frustrate a reading of the SRR provision that permits imposition of a post-revocation term of supervision. If a term has been called back, it may be reimposed. If a term has been absolutely terminated, a new term still may be imposed — in the same way that, once a license is revoked, a new one may be issued. In the end, the semantic debate over the word “revoke” turns out to be no more than the swapping of heuristics.

No matter how the word is defined, the language of the SRR provision is consistent with the possibility that a post-revocation term of supervision lawfully may be imposed. We believe this linguistic intuition is verified by historical precedents. Previous Congresses used the word “revoke” in crafting the statutory forerunners of

section 3583(e)(3). See, e.g., 18 U.S.C. § 4214 (1988) (repealed 1984 anent offenses committed after November 1, 1987) (revocation of parole); 21 U.S.C.A. § 841(c) (1981 & Supp.1993) (repealed 1984) (revocation of special parole); 18 U.S.C. § 3653 (1988) (repealed 1984 anent offenses committed after November 1, 1987) (revocation of probation). Notwithstanding Congress’s use of the word “revoke,” it was widely thought that reimposition of a period of non-detentive monitoring, to commence following post-revocation imprisonment, was permitted under all three of these antecedent statutory provisions. See infra

Part IV(D). B Our structural analysis of the alteration statute and, particularly, of the SRR provision starts with the recognition that the first appellate court to interpret

section 3583(

e) rested its holding on the notion that the alteration statute is structured as a set of discrete options separated by the word “or.” Given the shape of the statute, the court reasoned, a judge may either “extend” the term under subsection (e)(2) or “revoke” it under subsection (e)(3), but not both. See Behnezhad, 907 F.2d at 898-99. Subsequent courts quickly moved beyond this restrictive rationale, realizing that it collapses into the debate over the meaning of the SRR provision and, therefore, proves nothing. See, e.g., McGee, 981 F.2d at 274; Holmes, 954 F.2d at 272. To the extent that the repeated use of the disjunctive in

section 3583(

e) sheds any light on Congress’s intent, we believe that it favors a broad reading of the SRR provision. The first principal option that the alteration statute presents to a district judge is to “terminate” the supervised release term previously imposed under subsection (e)(1). If Congress meant to “revoke” supervised release in the hard sense of the word, it could simply have used the same language twice. Most likely, then, to “revoke” as used in the SRR provision means something other than to “terminate”. C Two general principles of statutory

interpretation inform our conclusion that the SRR provision cannot be read grudgingly: the principle that the grant of a greater power necessarily includes the grant of a lesser power, unless the authority to exercise a lesser power is expressly reserved; and the principle that statutes should not be read to produce illogical results. 1. The Greater Includes the Lesser. The principle that the grant of a greater power includes the grant of a lesser power is a bit of common sense that has been recognized in virtually every legal code from time immemorial.

It has found modern expression primarily in the realm of constitutional law. See, e.g., City of Lakewood v. Plain Dealer Publishing Co., 486 U.S. 750, 763, 108 S.Ct. 2138, 2147, 100 L.Ed.2d 771 (1988) (commenting that the power to prohibit speech entirely includes the lesser power to license it at the government’s discretion); Posadas de Puerto Rico Assocs. v. Tourism Co., 478 U.S. 328, 345, 106 S.Ct. 2968, 2979, 92 L.Ed.2d 266 (1986) (holding that the power to ban casino gambling includes the lesser power to prohibit advertising of casino gambling).

While this principle has nested less frequently in the criminal law context, it is fully applicable in that milieu. To illustrate, we use an example that bears a strong family resemblance to the problem at hand. The federal sentencing guidelines originally stated that “an extraordinary physical impairment may be a reason to impose a sentence other than imprisonment.” U.S.S.G. § 5H1.4, p.s. (Nov. 1990).

Three courts of appeals, including this one, refused to understand this provision to require an all-or-nothing choice between imposing an incarcerative sentence within the guideline range or imposing no prison sentence. The courts reasoned that, despite the unvarnished language of the provision, the greater departure (no incarceration) necessarily included the lesser departure (a prison sentence below the bottom of the guideline sentencing range). See United States v. Slater, 971 F.2d 626, 635 (10th Cir.1992); United States v. Hilton, 946 F.2d 955, 958 (1st Cir.1991); United States v.

Ghannam, 899 F.2d 327, 329 (4th Cir.1990). Similarly, in this case, we are reluctant to posit an all-or-nothing choice between continuing a defendant on supervised release (with no further incarceration) and imprisoning the defendant (with no further supervision). We agree with the Eighth Circuit that if the SRR provision gives a district court the power to sentence an offender to a full term of imprisonment upon revocation, it must necessarily confer upon the court “the power under that subsection to impose a less drastic sanction.” Schrader, 973 F.2d at 625. 2. Avoiding Illogical Results.

It is also an established canon of statutory construction that a legislature’s words should never be given a meaning that produces a stunningly counterintuitive result — at least if those words, read without undue straining, will bear another, less jarring meaning. See Kelly v. United States, 924 F.2d 355, 361 (1st Cir.1991); United States v. Meyer, 808 F.2d 912, 919 (1st Cir.1987); Sutherland Stat. Const. § 45.12 (5th ed.). This principle goes back to the early days of the Republic. See M’Culloch v. Maryland, 17 U.S. (4 Wheat.) 316, 355, 4 L.Ed. 579 (1819).

In this case, the sentencing rule that emerges from a narrow reading of

section 3583(e)(3) is surpassingly difficult to defend from a policy perspective. It is hard to conceive any logical reason why Congress might authorize sentencing an offender to a non-mandatory term of imprisonment, variable in the judge’s discretion, upon revocation of a term of supervised release, but would, at the same time, withhold authority to impose a sentence of equivalent duration upon more lenient conditions. See Williams, 2 F.3d at 365; Schrader, 973 F.2d at 625. Although we could jury-rig a legislative justification for so cramped an

interpretation of the law, we think it is self-evident that barring judges from reimposing supervision following revocation needlessly inhibits the court’s sentencing options while at the same time failing to advance any of the fundamental goals of criminal sentencing. As a matter of policy, then, the implications for sentencing inherent in a stingy reading of the SRR provision go a long way toward convincing us that Congress could not have favored (or intended to compel) such a reading. D As a rule, courts should resort to legislative history and other guides to congressional intent when the words of a statute give rise to ambiguity or when they lead to an unreasonable

interpretation. See, e.g., United States v. Charles George Trucking Co., 823 F.2d 685, 688 (1st Cir.1987); Barry v. St. Paul Fire & Marine Ins. Co., 555 F.2d 3, 7 (1st Cir.1977), aff'd, 438 U.S. 531, 98 S.Ct. 2923, 57 L.Ed.2d 932 (1978). Though we believe that a generous reading of

section 3583(e)(3) best comports with plain language, statutory structure, logic, and sound policy, we are aware that ambiguity is commonly thought to' exist when statutory language is susceptible to differing, but nonetheless plausible, constructions. See United States v. R.L.C., — U.S. -, -, 112 S.Ct. 1329, 1334, 117 L.Ed.2d 559 (1992); cf. Allen v. Adage, Inc., 967 F.2d 695, 700 (1st Cir.1992) (explaining when ambiguity exists in the text of a contract). Here, as the weight of authority unquestionably attests, there is room for disagreement over the meaning of the SRR provision.

Therefore, we continue our inquiry. Where ambiguity lurks, the burial ground in which superseded statutes rest sometimes proves a fertile field for assistance in determining the meaning of .existing statutes. See Dwight v. Merritt, 140 U.S. 213, 217, 11 S.Ct. 768, 769, 35 L.Ed. 450 (1891); see also Sutherland Stat. Const. § 51.04. We think that superseded statutes are of particular value in construing provisions within the Sentencing Reform Act.

We have recognized — and we believe the Sentencing Commission has recognized — the desirability of emulating pre-guidelines practice to the extent that plain meaning does not compel change. Thus, we have repeatedly referred to pre-guidelines precedent as an aid to interpreting the sentencing guidelines. See, e.g., United States v. Emery, 991 F.2d 907, 911 (1st Cir.1993); United States v. Blanco, 888 F.2d 907, 910 (1st Cir.1989); see also U.S.S.G. Ch. 1, Pt. A, intro, comment. 3 (Nov. 1992) (stating policy that “the guidelines represent an approach that begins with, and builds upon,” pre-guidelines practice).

We believe the same principle applies in construing the Sentencing Reform Act itself. To place the genealogy of supervised release in historical context, one must first recognize that non-detentive monitoring developed along two separate lines: probation and parole. The Sentencing Reform Act, and the guidelines implementing it, swept aside both of these modalities, replacing probation with an entirely new creature bearing the same name and replacing parole (as well as its interim variant, special parole) with supervised release. See Gozlon-Peretz v.

United States, 498 U.S. 395, 400, 111 S.Ct. 840, 844, 112 L.Ed.2d 919 (1991) (noting that Congress intended to replace most forms of parole, including special parole, with supervised release). We think it is of critical importance that, prior to the sea change instigated by the Sentencing Reform Act, it was widely understood that any of the existing forms of non-detentive monitoring could follow a post-revocation sentence of imprisonment. We survey the field. 1. Probation. The debate in which we are embroiled today closely tracks an earlier debate over post-revocation probation.

The relevant pre-guidelines statute empowered a court to “revoke probation, and impose any sentence which might originally have been imposed.” 18 U.S.C. § 3653 (repealed). Under this law, five circuits viewed probation as a kind of “sentence” that could be imposed after revocation of probation. See Banks v. United States, 614 F.2d 95, 99 n. 10 (6th Cir.1980); United States v. Rodgers, 588 F.2d 651, 654 (8th Cir.1978); Nicholas v. United States, 527 F.2d 1160, 1162 (9th Cir.1976); United States v.

Lancer, 508 F.2d 719, 730-32 (3d Cir.) (en banc), cert. denied, 421 U.S. 989, 95 S.Ct. 1992, 44 L.Ed.2d 478 (1975); Smith v. United States, 505 F.2d 893, 895 (5th Cir.1974). The Tenth Circuit and a district court in the Fourth Circuit took the opposite view. See United States v. Martin, 786 F.2d 974, 976 (10th Cir.1986) (declining to overrule Fox v. United States, 354 F.2d 752 (10th Cir.1965)); United States v. Buchanan, 340 F.Supp. 1285, 1288-89 (E.D.N.C.1972).

When the smoke cleared, “the weight of authority heavily favorfed] the conclusion that reimposition of probation is permissible upon revocation of probation.” United States v. Urdaneta, 771 F.Supp. 28, 32 (E.D.N.Y.1991) (canvassing pre-guidelines case law). Under the new sentencing regime, the statute treating with post-revocation probation deals much more directly with the vexed question of reimposition. It empowers a court to “revoke the sentence of probation and impose any other sentence that was available at the time of the initial sentencing.” 18 U.S.C. § 3565(a) (1988) (emphasis supplied).

Although the question is not before us, and we, accordingly, do not rule definitively on it, it seems probable that Congress intended to depart from prevailing pre-guidelines practice and forbid reimposition of probation following the revocation of a term of probation. We draw this inference from the insertion of the word “other,” on the theory that a change in statutory language should be “read, if possible, to have some effect.” American Nat'l Red Cross v. S.G., - U.S. -, -, 112 S.Ct. 2465, 2475, 120 L.Ed.2d 201 (1992). It thus appears quite likely that the drafters of

section 3565 were aware of the pre-guidelines case law and knew how to design a statute in such a way as to address its impact head-on. 2. Parole. There was never any question that non-detentive monitoring could follow a prison sentence imposed in consequence of the revocation of a term of parole or special parole.

See, e.g., 28 C.F.R. § 2.62 app. (1993) (setting out United States Parole Commission’s policy statement to the effect that “an adequate period of renewed supervision following release from reimprisonment or reinstatement to supervision, must be available”); id. at § 2.57 (making the policy statement applicable to special parole); see also Bentsen v. Ralston, 658 F.2d 639, 640 (8th Cir.1981) (citing eases for the proposition that an erstwhile parolee serving post-revocation prison time may earn good-time credit applicable to a second parole period).

In this context, the Senate report that accompanied the Sentencing Reform Act demonstrates Congress’s awareness of the pre-guidelines practice: Under [pre-guidelines] law, if a parolee violates a condition of parole that results in a determination to revoke parole, the revocation has the effect of requiring the parolee to serve the remainder of his original term of imprisonment, subject to 'periodic consideration for re-release as required for any prisoner who is eligible for parole. S.Rep. No. 225, 98th Cong. 2d Sess., reprinted in 1984 U.S.C.C.A.N. 3182, 3306 (emphasis supplied).

We find this historical phenomenon to be especially significant in light of the wording of the provision pertaining to the revocation of special parole. The governing statute decreed that “[a] person whose special parole term has been revoked may be required to serve all or part of the remainder of the new term of imprisonment.” 21 U.S.C.A. § 841(c) (repealed). Notwithstanding that in

section 841(c), as in

section 3583(e)(3), there was no explicit authorization to commence a second non-detentive term, the Parole Commission, whose

interpretation of a provision it is charged to execute is entitled to considerable weight, see Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 844-45, 104 S.Ct. 2778, 2783, 81 L.Ed.2d 694 (1984), explicitly endorsed the reimposition of special parole. Given the obvious similarities in language, structure, and substance between

section 841(

c) and

section 3583(e)(3), we are fortified in our conclusion that

section 3583(e)(3) plausibly may bear a broader

interpretation than it heretofore has received. Moreover, it seems highly likely that Congress, in replacing a repealed provision with a new provision of hauntingly similar wording, intended that the pre-guidelines

interpretation would continue to apply. Otherwise, Congress would almost certainly have altered the language to clarify its intent — as it did in connection with probation, see supra

Part IV(D)(1). For these reasons, the historical development of non-detentive monitoring, in all its permutations, reinforces our intuition that Congress meant to leave undisturbed the widely accepted pre-guidelines practice of allowing district courts discretion to order a period of non-detentive monitoring as a part of the sentence imposed for violation of supervised release conditions. E Studying what has transpired in Congress subsequent to the passage of the alteration statute produces another possible aid to statutory construction.

The focus here is on a bipartisan quartet — comprising four senior members of the Senate Judiciary Committee — thought to have been supremely influential in the passage of the Sentencing Reform Act: Senators Thurmond, Kennedy, Biden, and Hatch. These senators uniformly favor a clarifying amendment that would remove any doubt that

section 3583(e)(3) allows reimposition of supervised release. See, e.g., 137 Cong.Rec. S10021 (daily ed. July 15, 1991) (text of S.188, sponsored by Sens. Kennedy, Thurmond, and Biden); 139 Cong.Rec. S2090 (daily ed. February 25, 1993) (S.468, sponsored by Sen. Thurmond, referred to Judiciary Committee); 139 Cong.Rec. S3054 (daily ed. March 17, 1993) (Sen. Hatch added as cosponsor to S.468). We understand that such thirteenth-hour pronouncements are of uncertain value.

Though courts may accord some weight to a subsequent enactment that reflects directly on a statute under scrutiny, see, e.g., Red Lion Broadcasting Co. v. FCC, 395 U.S. 367, 380-81, 89 S.Ct. 1794, 1801-02, 23 L.Ed.2d 371 (1969), pronouncements made in the legislative history of that subsequent statute frequently are viewed as unreliable, see Consumer Prod. Safety Comm’n v. GTE Sylvania, Inc., 447 U.S. 102, 118 n. 13, 100 S.Ct. 2051, 2061 n. 13, 64 L.Ed.2d 766 (1979), and pronouncements regarding an unpassed bill may be even more problematic, see Chapman v.

United States, 500 U.S. 453, - n. 4, 111 S.Ct. 1919, 1927 n. 4, 114 L.Ed.2d 524 (1991). Accordingly, we reach our decision today without placing significant weight on post-enactment materials. Nonetheless, courts, including the Supreme Court and this court, have occasionally thought post-enactment declarations of congressional intent possessed some probative value. See, e.g., Seatrain Shipbuilding Corp. v. Shell Oil Co., 444 U.S. 572, 596, 100 S.Ct. 800, 813, 63 L.Ed.2d 36 (1980), (relying in part on committee report relative to subsequently enacted amendment); United States v.

Ven-Fuel, Inc., 758 F.2d 741, 758-59 (1st Cir.1985) (same). We believe that if post-enactment history, short of the actual passage of a new bill, is ever to be given weight, this case is a nearly ideal candidate. The sponsors of the proposed amendments include the same senators who sponsored the enacted statute; the amendatory legislation has been characterized by a sponsor as “clar-iflying]” in nature, rather than as revisory or augmentative, see 139 Cong.Rec. S2151 (daily ed. Feb. 25, 1993) (statement of Sen. Thurmond on S.468); 137 Cong.Rec. S8892 (daily ed. June 27, 1991) (statement of Sen.

Thurmond on S.188); and, in various incarnations, the clarification has been adopted twice by the House and four times by the Senate (including twice by the Senate in the form of a freestanding bill). See 139 Cong.Rec. at S2150 (citing bills). This history strongly suggests that the amendment remains un-passed only because the vagaries of the parliamentary process are what they are.

When, as now, the two houses of Congress, in the wake of a series of judicial decisions going mainly in one direction, have repeatedly signified that an amendment is needed to clarify recently enacted legislation, it seems reasonable to infer that the courts have failed to grasp the enacting Congress’s intent. In such circumstances, the case for giving some modest weight to post-enactment history peaks. F At this point, we have marshalled the available data.

We have found neither of the contending readings to be obviously correct on the statute’s face, and we have deter-rated no direct evidence of congressional intent sufficient to capture the flag. In the end, however, three considerations persuade us that a broader

interpretation of the SRR provision is more likely what Congress intended. First, a narrow rendering is inharmonious with the statute as a whole. Second, in choosing between two plausible readings, we hesitate to select the alternative that in effect imputes to Congress a policy for which no compelling rationale can be postulated (and that, in the bargain, blindly treats a greater power as if it did not include a lesser power). Third, given a statute of protracted indeterminacy, we are inclined to favor the

interpretation that promotes continuity with traditional sentencing practice — all the more so since the preexisting practice was based in significant part upon a similarly worded statute. For these reasons, and despite our abiding respect for the courts that have gone the other way, we hold that the district judge did not err in concluding that he possessed the power to impose both a prison term and a term of supervised release following revocation of appellant’s original supervision term. V.

APPLYING THE SRR PROVISION Having determined that the court below correctly grasped the essential meaning of the SRR provision, we find, nonetheless, that it erred in fashioning appellant’s sentence. In this ease, upon revocation of the original term of supervised release, the SRR provision yields a maximum sentence length of three years.' See 18 U.S.C. § 3583(e)(3). No more than two years of that period can be devoted to incarceration. See id.

The key to these computations is that the combined limit of three years matches the length of the original term of supervision and the secondary limitation — two years in prison— matches the statutory maximum allowable for revocation of supervised release when the underlying offense is a Class D felony. See id. In light of these benchmarks, it is apparent that the sentence imposed here exceeded the maximum sentence authorized by law. Specifically, upon revocation of supervised release, the imposition of a two-year prison term followed by a fresh three-year supervision term is unlawful.

Although O’Neil’s sentence must be vacated, at least in part, the contours of the appropriate remedy remain tenebrous. On one hand, the government tells us that we should in effect lop off the last two years of the supervision term, thus bringing the sentence into statutory alignment. On the other hand, appellant urges us to vacate the whole sentence and remand for resentencing, thus permitting the district court, armed with our insights into the workings of the SRR provision, to rethink its options. While there is precedent for each of these alternatives, compare, e.g., United States v.

Vasquez, 504 F.2d 555, 556 (5th Cir.1974) (per curiam) (holding that the excessive portion of a sentence may be trimmed and the remainder left intact) with, e.g., United States v. Berkowitz, 429 F.2d 921, 928 (1st Cir.1970) (vacating entire sentence and remanding for resentencing), we believe that the latter option is preferable in this case. We explain briefly. Although subject to constitutional constraints, statutory limitations, and, now, the guidelines, sentencing is, by and large, within the province of the district court.

Sentences usually contain a variety of components, e.g., an incarcerative component, a monetary component (say, a fine or cost-of-confinement order), and a non-detentive, non-monetary component (say, supervised release). These components often interrelate. Where an appellate court unties the bundle and decides that one component must be reconfigured, it may often be better practice to enlist the district court to retrofit the package. So it is here. We think that the district court, not this court, is best equipped to gauge what the overall sentence should be. See generally United States v.

Pimienta-Redondo, 874 F.2d 9, 14 (1st Cir.) (en banc) (discussing resentencing in multiple-count case after determination that the Double Jeopardy Clause barred imposition of separate sentence on one of two counts of conviction), cert. denied, 493 U.S. 890, 110 S.Ct. 283, 107 L.Ed.2d 185 (1989). VI. CONCLUSION We need go no further.

We hold that the SRR provision, 18 U.S.C. § 3583(e)(3), permits a district court, upon revocation of a term of supervised release, to impose a prison sentence or a sentence combining incarceration with a further term of supervised release, so long as (1) the incarcerative portion of the sentence does not exceed the time limit specified in the SRR provision itself, and (2) the combined length of the new prison sentence cum supervision term does not exceed the duration of the original term of supervised release.

Since the district court overstepped these boundaries, we vacate appellant’s sentence and remand for resentenc-ing. It is so ordered. . On the majority's reading, the statutoiy limit in a given case is the lesser of (

i) the length of the original term of supervision, or (ii) the numerical limit designated by the final clause of the SRR provision vis-a-vis each specified class of offense. . On this reading of the SRR provision, there are two operative limits in any given case. First, the combined length of all post-revocation impositions (incarcerative and supervisory) may not exceed the length of the original term of supervision. Second, the incarcerative portion of the post-revocation sentence may not exceed the numerical limit designated by the SRR provision's final clause for the class of offense in question.

It will be noted that, on this reading, the concluding clause of the SRR provision places an absolute ceiling on the time a person may serve in prison following revocation of a term of supervised release and thereby ensures that the criminal justice system cannot trap an offender in its web forever. This point adequately answers those who assert that construing the SRR provision broadly sets the stage for a never-ending cycle of revocation, resentencing to prison plus supervision, and revocation again, see McGee, 981 F.2d at 275. .

This intuition was vindicated by Congress and the Sentencing Commission when, effective November 1, 1991, the phrase "other than imprisonment” was changed to read “below the applicable guideline range.” See U.S.S.G.App. C, Amend. 386 (Nov. 1991). . The fundamental goals of the Sentencing Reform Act are commonly thought to include uniformity, honesty, and proportionality. See United States v. Williams, 891 F.2d 962, 963-64 (1st Cir.1989); see also U.S.S.G. Ch. 1, Pt. A, intro, comment., at 2 (Nov. 1992). . The transition from special parole to supervised release was grotesquely complicated.

Most existing provisions for non-detentive monitoring were repealed in 1984 as part of the Sentencing Reform Act, but the repeal did not take effect until November 1, 1987. However, the special parole provision, 21 U.S.C. § 841(b)(1)(A), was repealed outright. Thus, from October 12, 1984 through October 27, 1986, neither special parole nor any substitute for it was in force. Apparently desiring to eliminate this hiatus, Congress amended the law to insert supervised release in lieu of special parole for the interval from October 27, 1986 to November 1, 1987.

Congress accomplished this feat by amending 21 U.S.C. § 841(b) (under which no provision is made for revocation). Subsequent to November 1, 1987, supervised release has been controlled by the provisions of the Sentencing Reform Act. See generally Gozlon-Peretz, 498 U.S. at 395-403, 111 S.Ct. at 844-46 (explicating historical development). . We consider it significant that no court, on either side of this debate, suggested that the statute's use of the word "revoke” might require a ban on the reimposition of a non-detentive term in sentencing defendants who had violated probation.

Instead, the debate hinged on the word "sentence” — specifically, on whether probation could be conceived as a kind of "sentence." . Even if Congress intended to preclude reimpo-sition of probation following revocation of a term of probation, that intention has no implications for supervised release. Under the Sentencing Reform Act, a term of probation may not be imposed when a defendant is sentenced to imprisonment. See 18 U.S.C. § 3553(a)(3). Since a "combined" sentence is prohibited ah initio, it would make little sense to allow a combined form of sentencing upon revocation of probation. .

For what, if any, relevance it may have, the Sentencing Commission also favors a clarifying amendment. See U.S.S.G. § 7B 1.3(g)(2) (Nov. 1992) (policy statement reading statute to allow reimposition of supervision); id. at § 7B1.3, comment, (n. 3) (advocating passage of clarifying amendment). . Senators Thurmond and Biden introduced the omnibus crime bill containing the provisions that became the Sentencing Reform Act. Senator Kennedy submitted a freestanding sentencing bill, containing nearly identical provisions, at approximately the same time. See Kate Stith & Steve Y.

Koh, The Politics of Sentencing Reform: The Legislative History of the Federal Sentencing Guidelines, 28 Wake Forest L.Rev. 223, 261 (1993). . In the process, we have considered — and rejected — the notion that the rule of lenity, a background principle that properly comes into play when, at the end of a thorough inquiry, the meaning of a criminal statute remains obscure, see Chapman, 500 U.S. at -, 111 S.Ct. at 1926, might be of help here in discerning congressional intent. See, e.g., Koehler, 973 F.2d at 135 (arguing that the rule of lenity cuts in favor of a narrow construction of the SRR provision).

The problem lies in determining whose ox may be gored. Depending on the facts of any particular defendant’s situation, a generous reading of the SRR provision can produce either a harsher or a more lenient result than a cramped reading will produce. Thus, we regard the interpretive struggle over the SRR provision as lenity-neutral. .

We are aware that the Sentencing Commission’s policy statement contemplates that the new term of imprisonment will be "less than" the maximum term of imprisonment imposable upon revocation for each class of offense, U.S.S.G. § 7B1.3(g)(2) p.s., but we use round numbers for simplicity’s sake. Moreover, although a policy statement ordinarily "is an authoritative guide to the meaning of the applicable guideline,” Williams v. United States, - U.S. -, -, 112 S.Ct. 1112, 1119, 117 L.Ed.2d 341 (1992), the policy statements of

Chapter 7 are unaccompanied by guidelines, and are prefaced by a special discussion making manifest their tentative nature, see U.S.S.G. Ch. 7, Pt. A, intro, comment. Hence, we today join six other circuits in recognizing

Chapter 7 policy statements as advisory rather than mandatory. See United States v. Thompson, 976 F.2d 1380, 1381 (11th Cir.1992); United States v. Bermudez, 974 F.2d 12, 14 (2d Cir.1992); United States v. Cohen, 965 F.2d 58, 59-61 (6th Cir.1992); United States v. Lee, 957 F.2d 770, 773 (10th Cir.1992); United States v. Blackston, 940 F.2d 877, 893 (3d Cir.), cert. denied, - U.S. -, 112 S.Ct. 611, 116 L.Ed.2d 634 (1991); United States v. Oliver, 931 F.2d 463, 465 (8th Cir.1991). On remand, the lower court must consider, but need not necessarily follow, the Sentencing Commission's recommendations regarding post-revocation sentencing.