UNITED STATES of America, Appellant, v. George LABONTE, Defendant, Appellee; UNITED STATES of America, Appellee, v. David E. PIPER, Defendant, Appellant; UNITED STATES of America, Appellee, v. Alfred Lawrence HUNNEWELL, Defendant, Appellant; Stephen DYER, Petitioner, Appellant, v. UNITED STATES of America, Respondent, Appellee, 70 F.3d 1396

SELYA, Circuit Judge. After many years of study and debate, Congress passed the Sentencing Reform Act of 1984, Pub.L. 98-473, tit. II, § 212(a), 98 Stat. 1837 (1984) (codified as amended at scattered sections of 18 & 28 U.S.C.). The legislation took effect on November 1, 1987, and caused dramatic changes both in the methodology of criminal sentencing and in the outcomes produced. These changes did not go unremarked: sentencing appeals, once rare in federal criminal cases, became commonplace.

Predictably, the tidal wave of appeals loosed a flood of judicial opinions distilling the meaning, scope, and application of a seemingly boundless sea of guidelines, policy statements, notes, and commentary. And whenever it appeared that the flood waters might recede, the Sentencing Commission launched a fresh deluge of revisions that required the courts to paddle even faster in a Sisyphean effort to stay afloat. These four consolidated appeals are emblematic of the difficulties that courts face in dealing with the new sentencing regime.

All four appeals implicate Application Note 2 to the Career Offender Guideline, as modified by Amendment 506, United States Sentencing Commission, Guidelines Manual § 4B1.1, comment, (n. 2) (Nov. 1994). No appellate court has addressed the validity of Amendment 506, and, in the quartet of criminal cases underlying these appeals, two able district judges reached diametrically opposite conclusions. Although the call is close, we hold that Amendment 506 is a reasonable implementation of the statutory mandate, 28 U.S.C. § 994(h) (1988 & Supp. V1993), and is therefore valid.

Thus, after answering other case-specific questions raised by the various parties, we affirm the judgments in the La-Bonte and Piper cases; vacate the judgment in the Hunnewell case and remand for reconsideration of the appropriateness of resen-tencing; affirm the judgment in the Dyer case in respect to all non-sentenee-related matters and vacate the sentence-related aspect of that judgment, remanding for reconsideration. I. THE AMENDMENT Congress created the Sentencing Commission in 1984 to design and implement federal sentencing guidelines.

Three principal forces propelled the legislation: Congress sought to establish truth in sentencing by ehminating parole, to guarantee uniformity in sentencing for similarly situated defendants, and to ensure that the punishment fit the crime. See U.S.S.G. ch. 1, pt. A(3), & 2; see also United States v. Unger, 915 F.2d 759, 762-63 (1st Cir.1990) (explaining that the primary purposes of the Sentencing Reform Act are to provide certainty, uniformity, and fairness in sentencing), cert. denied, 498 U.S. 1104, 111 S.Ct. 1005, 112 L.Ed.2d 1088 (1991).

In addition to general guidance, see, e.g., 28 U.S.C. § 991(b), Congress also gave the Commission some specific marching orders. One such set of marching orders is conveyed by 28 U.S.C. § 994(h), which provides in part: The Commission shall assure that the guidelines specify a sentence to a term of imprisonment at or near the maximum term authorized for categories of defendants in which the defendant is eighteen years old or older and [has been convicted of a violent crime or felony drug offense and has at least two such prior convictions]. The Commission implemented

section 994(

h) through the Career Offender Guideline. See U.S.S.G. § 4B1.1, comment, (backg’d). This guideline sets forth a table of enhanced total offense levels (TOLs) — said to be a function of the “Offense Statutory Maximum” — to be employed in calculating the sentences of so-called “career offenders.” See U.S.S.G. § 4B1.1. A defendant is regarded as a career offender if he was at least eighteen years old at the time of the offense of conviction, that offense is a crime of violence or a drug-related felony, and he has two prior convictions for drug felonies or crimes of violence. See id.; see also United States v.

Piper, 35 F.3d 611, 613 n. 1 (1st Cir.1994), cert. denied, — U.S. -, 115 S.Ct. 1118, 130 L.Ed.2d 1082 (1995). When the Commission issued the Career Offender Guideline, it coined the phrase “Offense Statutory Maximum,” but did not define the phrase beyond saying that “the term ‘Offense Statutory Maximum’ refers to the maximum term of imprisonment authorized for the offense of conviction.” U.S.S.G. § 4B1.1, comment, (n. 2) (Nov. 1987). Since this definition was tautological, it proved un-illuminating.

Faced with a need to improvise, several courts of appeals concluded that the phrase encompassed not merely the statutory maximum applicable to the offense of conviction simpliciter, but also the upgraded statutory maximum that results after available enhancements for prior criminal activity are taken into account. See United States v. Smith, 984 F.2d 1084, 1085 (10th Cir.), cert. denied, — U.S. -, 114 S.Ct. 204, 126 L.Ed.2d 161 (1993); United States v. Garrett, 959 F.2d 1005, 1009-11 (D.C.Cir.1992); United States v. Amis, 926 F.2d 328, 329-30 (3d Cir.1991); United States v.

Sanchez-Lopez, 879 F.2d 541, 558-60 (9th Cir.1989). This lexicographical choice carried with it important consequences; under the courts’ construction, a defendant whose maximum possible term of imprisonment for a crime of violence or drug offense was enhanced from, say, twenty to thirty years on account of prior criminal activity, netted two additional offense levels (increasing his TOL from thirty-two to thirty-four) and found himself in a steeper sentencing range.

In Amendment 506, the Commission first meaningfully defined the phrase “Offense Statutory Maximum.” The amendment provides that the phrase, for the purpose of the Career Offender Guideline, “refers to the maximum term of imprisonment authorized for the offense of conviction that is a crime of violence or controlled substance offense, not including any increase in that maximum term under a sentencing enhancement provision that applies because of the defendant’s prior criminal record.” U.S.S.G. § 4B1.1, comment. (n. 2) (Nov. 1994).

The amended note offers the example of a defendant who is subject to a sentencing enhancement under 21 U.S.C. § 841(b)(1)(C), in which case “the ‘Offense Statutory Maximum’ for the purposes of this guideline is twenty years and not thirty years.” Finally, the Commission opted to give Amendment 506 retroactive effect. See U.S.S.G. § 1B1.10(c) (Nov. 1994).

Initially, the Commission attempted to justify the amendment as “avoiding] unwarranted double-counting as well as unwarranted disparity associated with variations in the exercise of prosecutorial discretion in seeking enhanced penalties based on prior convictions.” U.S.S.G., App. C, Amend. 506, at 409 (Nov. 1994). In addition, the Commission observed that Congress enacted the array of sentence-enhancing laws after the statutory predicate for the Career Offender Guideline had become law. See id. Subsequently, the Commission attempted to explain its newly emergent

interpretation of the Career Offender Guideline in terms of a desire to avoid unwarranted disparity and to achieve consistency. See Amendment Notice, 60 Fed.Reg. 14,054, 14,055 (1995); see also United States v. LaBonte, 885 F.Supp. 19, 23 n. 4 (D.Me. 1995). Whatever may be its provenance, it is nose-on-the-face plain that, in many instances, Amendment 506 produces lower TOLs (and, ultimately, shorter sentences) than the unembellished Career Offender Guideline (as interpreted by the courts). Due to this palliative effect, critics view it as inimical to congressional intent. II.

THE DEFENDANTS These four defendants all were sentenced in the District of Maine as career offenders prior to the birth of Amendment 506. In each instance, the prosecution filed a notice under 21 U.S.C. § 851(a)(1) signalling its intention to seek enhanced penalties for prior convictions, and the sentencing court arrived at the defendant’s “Offense Statutory Maximum” by factoring the statutory enhancement into the mix. The court then set each defendant’s TOL and guideline sentencing range (GSR) accordingly. Following the promulgation of the amendment, all four defendants tried to avail themselves of it.

We limn their individual circumstances. A.George LaBonte. A grand jury indicted LaBonte for possession of cocaine with intent to distribute in violation of 21 U.S.C. § 841(a)(1) & (b)(1)(C). After he pleaded guilty, the district court (Hornby, U.S.D.J.) sentenced him under the Career Offender Guideline. Using an enhanced statutory maximum derived from La-Bonte’s record of prior drug convictions, Judge Hornby set LaBonte’s TOL at thirty-four, granted a three-level downward adjustment for acceptance of responsibility, see U.S.S.G. § 3E1.1, arrived at a GSR of 188-235 months, and sentenced him to serve 188 months.

We affirmed. See United States v. Labonte, 19 F.3d 1427, 1994 WL 107868 (1st Cir.1994) (table). Subsequent to the promulgation of Amendment 506, LaBonte moved for resentencing. Judge Hornby determined that Amendment 506 was valid and decided to apply it. See LaBonte, 885 F.Supp. at 24. He granted LaBonte’s motion, focused on the unen-hanced statutory maximum to calculate a new TOL (thirty-two), and again deducted three levels for acceptance of responsibility. This recomputation yielded a GSR of 151-188 months, and Judge Hornby lowered La-Bonte’s sentence to the nadir of the new range. See id.

The government appeals from this disposition. B.David E. Piper. Piper pleaded guilty to a two-count information charging conspiracy to possess marijuana with intent to distribute and use of a firearm in connection with a drug offense. See 21 U.S.C. §§ 841(a)(1) & (b)(1)(B), 846; 18 U.S.C. § 924(c)(1). Utilizing an enhanced statutory maximum, Judge Hornby set Piper’s TOL at thirty-seven, subtracted three levels for acceptance of responsibility, arrived at a GSR of 262-327 months, and imposed an incarcerative sentence of 300 months. We affirmed. See Piper, 35 F.3d at 613.

Hot on the heels of Amendment 506, Piper moved unsuccessfully for resentencing. Although Judge Hornby assumed the amendment’s validity, he exercised his discretion and declined to permit Piper to benefit from it. Piper appeals from this disposition. C.Alfred Lawrence Hunnewell. A grand jury indicted Hunnewell on six narcotics counts. See 21 U.S.C. § 841(a)(1). He thereafter pleaded guilty to two counts of possessing controlled substances with intent to distribute, and the court (Carter, U.S.D.J.) dismissed the remaining counts.

Using an enhanced statutory maximum, Judge Carter set Hunnewell’s TOL at thirty-four, deducted three levels for acceptance of responsibility, arrived at a GSR of 188-235 months, and sentenced the defendant to serve 188 months. We affirmed. See United States v. Hunnewell, 10 F.3d 805, 1993 WL 483252 (1st Cir.1993) (table), cert. denied, — U.S. -, 114 S.Ct. 1616, 128 L.Ed.2d 343 (1994). After the promulgation of Amendment 506, Hunnewell beseeched the district court to trim his sentence. Judge Carter denied this motion, concluding that the Sentencing Commission lacked the authority to adopt Amendment 506.

Hunnewell appeals. D. Stephen Dyer. Dyer pleaded guilty to a charge of conspiring to possess controlled substances with intent to distribute in contravention of 21 U.S.C. §§ 841(a)(1), 846. Consulting the enhanced statutory maximum, Judge Carter set Dyer’s TOL at thirty-four, refused an acceptance-of-responsibility discount, arrived at a GSR of 262-327 months, and levied a 262-month term of imprisonment. We affirmed. See United States v. Dyer, 9 F.3d 1 (1st Cir.1993) (per curiam).

Dyer eventually filed a petition for habeas relief, see 28 U.S.C. § 2255, in which he sought to set aside his conviction or, in the alternative, to reduce his sentence by virtue of Amendment 506. Judge Carter denied and dismissed the habeas petition. Among other things, the judge, declaring Amendment 506 to be unlawful, refused to resen-tenee Dyer. Dyer protests all aspects of the district court’s order. III. THE VALIDITY OF AMENDMENT 506 We begin our analysis by discussing, generally, the methodology we will employ in examining Amendment 506.

We then proceed to tackle the two conundrums that are inextricably intertwined with the question of the amendment’s validity. A. The Methodology. Commentary authored by the Sentencing Commission that “interprets or explains a guideline is authoritative unless it violates the Constitution or a federal statute, or is inconsistent with, or a plainly erroneous reading of, that guideline.” Stinson v. United States, 508 U.S. 36, -, 113 S.Ct. 1913, 1915, 123 L.Ed.2d 598 (1993). Like the Commission’s policy statements, its commentary is binding on the federal courts. See id. at---, 113 S.Ct. at 1917-18.

In general, these interpretive materials are entitled to the same substantial degree of deference that courts routinely accord an administrative agency’s

interpretation of its own legislative rules. See id. at -, 113 S.Ct. at 1919. Thus, under Stinson, judicial scrutiny of the Commission’s commentary is limited to ensuring consistency with federal statutes (including, but not restricted to, the Commission’s enabling statute), and with the guidelines themselves. These two lines of inquiry proceed along different analytic paths. When a court ventures to determine whether the Commission’s commentary tracks the guidelines, the degree of deference is at its zenith.

In this context, commentary is not merely the end product of delegated authority for rulemak-ing, but, rather, “explains the guidelines and provides concrete guidance as to how even unambiguous guidelines are to be applied in practice.” Id. at -, 113 S.Ct. at 1918. Unless the commentary is a palpably erroneous rendition of a guideline, it merits respect. See id. at -, 113 S.Ct. at 1919; Piper, 35 F.3d at 617. The determination of whether the guidelines are consistent with positive statutory law touches a more vulnerable spot.

That inquiry implicates the traditional process of reviewing agency rules typified by the Supreme Court’s watershed opinion in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984).

Thus, while the Court has warned that Chevron does not provide an apt analogy for the process of reviewing the relationship between commentary, on the one hand, and guidelines, on the other hand, see Stinson, 508 U.S. at-, 113 S.Ct. at 1918, we believe that Chevron deference is the proper criterion for determining whether a guideline (or, for that matter, commentary that suggests how a guideline should be read) contravenes a statute. The Chevron two-step approach fits that type of inquiry like a glove. See Chevron, 467 U.S. at 842-43, 104 S.Ct. at 2781-82 (describing two-step test).

Applying this methodology here is not without complications. We limit our inquiry to the fit (or lack of fit) between the Career Offender Guideline as explicated in Amendment 506 and the applicable statute, 28 U.S.C. § 994(h). In that statute, Congress directed the Commission to ensure that certain recidivists receive sentences “at or near the maximum.” The Career Offender Guideline represents the Commission’s response to this directive. See U.S.S.G. § 4B1.1, comment. (backg’d).

Because the Commission’s understanding of its statutory mandate must be measured against the Chevron benchmark, the inquiry follows a familiar format: When a court reviews an agency’s construction of the statute which it administers, it is confronted with two questions. First, always, is the question whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter ....

If, however, the court determines Congress has not directly addressed the precise question at issue, the ... question for the court is whether the agency’s answer is based on a permissible construction of the statute. Chevron, 467 U.S. at 842-43, 104 S.Ct. at 2781-82; accord Strickland v. Commissioner, Me. Dep’t of Human Servs., 48 F.3d 12, 16 (1st Cir.), cert. denied, — U.S. -, 116 S.Ct. 145, 133 L.Ed.2d 91 (1995).

These appeals focus on a single sentence that appears in 28 U.S.C. § 994(h), a sentence that requires the Commission to adopt guidelines “that specify a sentence to a term of imprisonment at or near the maximum term authorized for [certain] categories of defendants.” This problematic sentence presents three issues of statutory

interpretation, necessitating two distinct iterations of the Chevron standard. The first application combines two issues; it concerns the explication of the word “maximum” as that word is used in

section 994(

h) and, concomitantly, the meaning of the word “categories” as used therein. The second occasion for Chevron analysis involves an exegesis of the phrase “at or near” as used in the same sentence. The two problems are interrelated, but they are somewhat different in nature. B. The First Conundrum. In the context of

section 994(h), the term “maximum” is susceptible of divergent meanings, depending, in part, on precisely what constitutes a “categor[y] of defendants.” One possible reading is that “categories” are composed of those defendants charged with violations of similar statutes against whom prosecutors have filed notices of intention to seek sentence enhancements (e.g., all repeat offender drug traffickers against whom the government has filed sentence-enhancing in-formations under 21 U.S.C. § 851(a)(1)).

On this view, the relevant statutory maximum for any such defendant would be the enhanced statutory maximum (ESM) applicable to repeat offenders. See 21 U.S.C. §§ 841(b)(1), 851(a)(1). But this reading is not linguistically compelled.

The word “categories” plausibly can be defined more broadly to include all offenders (or all repeat offenders) charged with transgressing the same criminal statute, regardless of whether the prosecution chooses to invoke the sentence-enhancing mechanism against a particular defendant (e.g., all drug traffickers, or all repeat offender drug traffickers, who are charged with violating 21 U.S.C. § 841(a)(1)). On this view, the word “maximum” refers to the unenhanced statutory maximum (USM), see 21 U.S.C. § 841(b)(1), since this represents the highest possible sentence applicable to all defendants in the category.

Since the sentencing guidelines must comport with such specific statutory directives as Congress has ordained, see United States v. Saccoccia, 58 F.3d 754, 786 (1st Cir.1995) (“It is apodictie that the sentencing guidelines cannot sweep more broadly than Congress’ grant of power to the Sentencing Commission permits.”), the question becomes whether Congress clearly intended to prefer one of these

interpretations over the other. The issue is not free from doubt. Several courts of appeals have heretofore read the word “maximum” in the former fashion (as referring to the ESM), see supra pp. 1400-01, whereas the Sentencing Commission now reads the word in the latter sense (as referring to the USM). We proceed to test this conflict in the Chevron crucible. 1. Step One: Congressional Intent. At the outset, we must determine whether Congress has spoken with sufficient clarity to foreclose alternative

interpretations. Statutory construction always starts — and sometimes ends — with the statute’s text. Here, we find Congress’s handiwork opaque. The problem is not ambiguity in definition. Rather, it is simply unclear from the bare language of the law which maxima and what categories Congress had in mind when it contrived

section 994(h). The earlier cases relating the word “maximum” to the ESM do not dictate a contrary conclusion. Those courts envisaged their primary task as interpreting the meaning of the guidelines, see, e.g., Garrett, 959 F.2d at 1010 (concluding that “the Guidelines require us to define the [term] Offense Statutory Maximum” in a particular way); Amis, 926 F.2d at 329 (stating the court’s task as “merely [to] determine the ‘Offense Statutory Maximum’ as used in guidelines § 4B1.1”), and they did so without the aid of Amendment 506.

Although two courts suggested that reading “Offense Statutory Maximum” as referring to the ESM would better effectuate congressional intent, see Garrett, 959 F.2d at 1010; Sanchez-Lopez, 879 F.2d at 559, neither of these courts held — or even hinted — that

section 994(

h) thwarted a different reading. We have found no indication that any of the courts which scrutinized the unexplicated version of U.S.S.G. § 4B1.1 detected the kind of clear, overarching congressional directive that would suffice to abort a Chevron inquiry. Even were we to believe otherwise, two abecedarian principles of statutory construction nonetheless would counsel continuation of the Chevron journey. First, courts that read a statute without the aid of an authoritative

interpretation by the agency charged with administering the statute must reexamine their reading if the agency later speaks to the point. See International Ass’n of Bridge, Structural, and Ornamental Iron-workers, Etc. v. NLRB, 946 F.2d 1264, 1271 (7th Cir.1991). Second, an agency that is charged with administering a statute remains free to supplant prior judicial

interpretations of that statute as long as the agency

interpretation is a reasonable rendition of the statutory text. See id. at 1270; see also Rust v. Sullivan, 500 U.S. 173, 186-87, 111 S.Ct. 1759, 1768-69, 114 L.Ed.2d 233 (1991) (holding that an agency is free to reverse its own previous

interpretation of a statute, sub-jeet to the same condition); Strickland, 48 F.3d at 18 (same). Hence, we trek onward. When the plain meaning of a law is not readily apparent on its face, the next resort is to the traditional tools of statutory construction — reviewing legislative history and scrutinizing statutory structure and design — in an effort to shed light on Congress’s intent. As originally envisioned,

section 994(

h) would have placed the onus of imposing sentences “at or near the maximum” directly on sentencing judges. See S.Rep. No. 98-225, 98th Cong., 2d Sess. 175 (1984), reprinted in 1984 U.S.C.C.A.N. 3182, 3358. The provision’s author, Senator Kennedy, devised it as a means of putting “[cjareer criminals ... on notice that their chronic violence will be punished by maximum prison sentences.” 128 Cong.Rec. 26,518 (1982). But that proposal did not take wing; the Senate Judiciary Committee instead approved

section 994(

h) in its current incarnation. This version, unlike the rejected proposal, addresses its command to the Commission, not the courts. The Committee obviously believed that this change would better “assure consistent and rational implementation of the Committee’s view that substantial prison terms should be imposed on repeat violent offenders and repeat drug offenders.” S.Rep. No. 98-225, supra, 1984 U.S.C.C.A.N. at 3358. We think that this history confirms that (1) in creating the Commission, Congress had an overall goal of curtailing judicial discretion in sentencing matters; and (2) in enacting

section 994(h), Congress had a specific intent to let the Commission (as opposed to individual judges) determine the best method for assuring that career offenders would receive stiff prison sentences. Past this point, the legislative archives offer no clue as to whether Congress ever recognized either the potential ambiguity of the term “maximum” or the uncertainty that might attach to the question of what constitutes a category of offenders.

Finding the relevant legislative history to be no clearer than the statute’s text, we look to the enabling legislation and the overall structure of the Sentencing Reform Act for what insights they may afford. Superficially, these considerations seem to support the government’s position that the “maximum” is the ESM. Reading “categories” narrowly enough to distinguish between offenders on the basis of whether the United States Attorney has filed sentence-enhancing informa-tions yields potentially harsher sentences in those cases, thereby promising more stringent punishment for selected repeat offenders.

That narrow reading also preserves the distinction between offenders who are subject to sentence enhancements based on prior criminal activity and those who are not — a distinction that Congress arguably delivered into the hands of prosecutors. See, e.g., 21 U.S.C. §§ 841(b)(1), 851(a)(1). Although these asseverations put the government’s best foot forward, they are at most debating points in relation to the problem at hand. They neither indicate that Congress has spoken directly to the precise issue nor reflect a sufficiently clear congressional intent to circumscribe the Commission’s interpretive powers.

Indeed, the arguments are circular; the touted advantages of the government’s reading appear to be advantageous only if one assumes the conclusion that the government is struggling to prove. We will not add hues to a rainbow. Because we find no clear congressional directive regarding the meaning of the term “maximum” as that term is used in

section 994(h), our inquiry proceeds to the second half of the Chevron two-step. 2. Step Two: Plausibility of the Commission’s

Interpretation. Where, as here, a statute is not clear, an

interpretation by the agency that administers it will prevail as long as the

interpretation is reasonable under the statute. See Strickland, 48 F.3d at 21. We believe that the Commission’s act in defining “maximum” to refer to the unen-hanced maximum term of imprisonment — the USM — furnishes a reasonable

interpretation of

section 994(h). The statute explicitly refers to “categories of defendants,” namely, repeat violent criminals and repeat drug offenders, and does not suggest that each individual offender must receive the highest sentence available against him. The Career Offender Guideline, read through the prism of Amendment 506, adopts an entirely plausible version of the categorical approach that the statute suggests. Unless one is prepared to write off Congress’s choice of the word “categories” as some sort of linguistic accident or awkward locution — and we are not so inclined — this approach is eminently supportable.

Our dissenting colleague decries the Commission’s categorical approach. He states that, indeed, “the phrase ‘categories of defendants’ is perhaps better understood ... as a ‘linguistic accident or an awkward locution.’ ” Post at 1416. To the contrary, this conclusion is foreclosed by, inter alia, the following explicit language in 18 U.S.C. § 3553: (a) ... The court, in determining the particular sentence to be imposed, shall consider— (4) the kinds of sentence and the sentencing range established for— (

A) the applicable category of offense committed by the applicable category of defendant as set forth in the guidelines issued by the Sentencing Commission pursuant to

section 99f(a)(

l) of title 28.... (Emphasis supplied). Further inescapable evidence that the term “categories of defendants” is neither an accidental nor a recent congressional usage, see post at 1416-17, appears in 28 U.S.C. § 994(b)(1): The Commission, in the guidelines promulgated pursuant to subsection (a)(1), shall, for each category of offense involving each category of defendant, establish a sentencing range that is consistent with all pertinent provisions of title 18, United States Code. (Emphasis supplied). Thus, rather than a recent slip of the legislative pen, the term “categories of defendants,” as used in

section 994(h), originated in the carefully incubated legislation mandating a guideline sentencing system that was to be promulgated and monitored by the Sentencing Commission, see 28 U.S.C. § 994, and implemented by the courts, see 18 U.S.C. § 3553.

Among the more important innovations attending the establishment of the new guideline sentencing system were certain restrictions on judicial consideration and weighting of individualized sentencing factors, see, e.g., 18 U.S.C. § 3553(a)(4), (b), (c); hence, the possibly “awkward,” but nonetheless plainly intended, usage “categories of defendants.” Given the identical statutory phrasing consistently employed by Congress in titles 18 and 28, as well as their coordinate design, we are unable to endorse the unsupported statutory

interpretation advanced in dissent. Rather, we must follow the canons of statutory

interpretation which demand that a court give meaning to each word and phrase when explicating a statute, and read the component parts of a legislative enactment as a unified whole. See United Technologies Corp. v. Browning-Ferris Indus., Inc., 33 F.3d 96, 101 (1st Cir.1994), cert. denied, — U.S. -, 115 S.Ct. 1176, 130 L.Ed.2d 1128 (1995); United States v. Ven-Fuel, Inc., 758 F.2d 741, 751-52 (1st Cir.1985); see also Greenwood Trust Co. v.

Massachusetts, 971 F.2d 818, 827 (1st Cir.1992) (“It is ... a general rule that when Congress borrows language from one statute and incorporates it into a second statute, the language of the two acts should be interpreted the same way.”), cert. denied, 506 U.S. 1052, 113 S.Ct. 974, 122 L.Ed.2d 129 (1993). Moreover, the Sentencing Reform Act places many restraints on the Commission apart from those embodied in

section 994(h). The most salient of these restraints is the requirement of sentencing consistency. See 28 U.S.C. § 994(f). The Commission adverted to this concern in promulgating Amendment 506, see U.S.S.G., App. C, Amend. 506, at 409 (Nov. 1994), and responded to it by taking a categorical approach. Similarly, Congress’s efforts to eliminate sentencing disparities can be reconciled with

section 994(h)’s exhortation for maximal sentencing only if one hears that exhortation as being addressed to categories of defendants. In the final analysis, the Commission remains fully faithful to the welter of congressional commands by choosing to treat repeat offenders as broad categories of defendants and thereby harmonizing the call for stringent punishment of recidivists with the call for consistent, non-disparate sentences. The government lodges two further objections to the plausibility of the Commission’s rationale.

First, it contends that Congress, by means of such statutes as 21 U.S.C. § 851(a)(1), intended to give prosecutors commodious discretion over the potential sentences of repeat offenders, and that Amendment 506 frustrates this intent. Though the government may well be correct in asserting that Congress did not create the Sentencing Commission with an eye toward eradicating prosecutorial abuses, it does not follow that Congress strove affirmatively to give prosecutors the keys to the kingdom.

What is more, it makes very little sense to impute to Congress a yearning for unbridled prosecutorial discretion when two major goals of sentencing reform were to “assure that sentences are fair both to the offender and to society,” S.Rep. No. 98-225, supra, 1984 U.S.C.C.A.N. at 3222, and to “avoid[ ] unwarranted sentencing disparities among defendants with similar records who have been found guilty of similar criminal conduct.” 28 U.S.C. § 991(b)(1)(B).

The government’s remaining objection to the Commission’s reading of the word “maximum” is that this reading prescribes an identical sentencing range for repeat offenders whether or not the prosecution has sought to obtain sentence enhancements. This reading, the government says, effectively eliminates prosecutorial enhancements and arrogates unto the Commission the authority that Congress explicitly vested in the United States Attorney. We find this polemic unpersuasive. We take 21 U.S.C. § 841(b)(1) as our point of departure. This

section establishes unen-hanced maximum terms applicable to all violators, enhanced maximum terms applicable to certain repeat offenders, and, in some cases, mandatory minimum terms of incarceration (enhanced or unenhanced). It is elementary that any guideline which prescribes a sentence that falls within these parameters does not conflict with the statute. What remains is a policy choice, and the Commission, by opting to emphasize the USM, has done no more than exercise its prerogative to make precisely this kind of policy choice. See Chevron, 467 U.S. at 864, 104 S.Ct. at 2792.

Furthermore, the choice is not unreasonable. The root purpose of the Career Offender Guideline, U.S.S.G. § 4B1.1, is to enhance repeat offenders’ sentences. The revamped guideline not only accomplishes that purpose but also coheres with Congress’s discernible aims in making enhanced penalties available under

section 841. While that statute establishes a possible enhanced penalty for repeat offenders if prosecutors choose, the Career Offender Guideline, as filtered through Amendment 506, ensures an actual enhancement of the TOL for all repeat offenders. This critical distinction belies the government’s lament that the amendment sounds a death knell for enhancements required by statute. The guideline,

section 4B1.1, as explicated by Amendment 506, departs from the statute,

section 841, only in the sense that the former seeks to enhance the sentences of a wider class of recidivists. This departure lacks significance. For purposes of testing the fidelity of the sentencing guidelines’ career offender provisions to the statutory scheme, it is irrelevant that some sentences beyond those mandated by Congress are also enhanced. When all is said and done, the Commission’s decision to treat the word “maximum” as meaning the unenhanced statutory maximum applicable to a category of offenders, broadly defined, is a plausible rendition of

section 994(h). We must honor the Commission’s definition. C. The Second Conundrum. As we have previously explained,

section 994(

h) contains a specific directive that, in the case of career offenders, sentences ought to be “at or near the maximum term authorized.” The government contends that, regardless of how the word “maximum” is construed, Amendment 506 is invalid because it fails to produce sentences that are “at or near” any conceivable maximum. As before, we measure this contention by wielding the Chevron yardstick. 1. Step One: Congressional Intent. At the risk of belaboring the obvious, we start from the premise that “at or near” is neither an exact nor a self-defining term.

Section 994(

h) is silent as to how “near” sentences must be to the maximum, and the legislative history is singularly unhelpful on this point. Especially since we must concentrate on the USM in calculating how “near” the Commission’s sentencing ranges are to the statutory goal, see supra

Part III(B), we are unable to divine a sufficiently clear expression of congressional intent. Thus, we quickly move to the second — and decisive — portion of the Chevron query. 2. Step Two: Plausibility of the Commission’s

Interpretation. The question of plausibility reduces to whether the Career Offender Guideline, as now interpreted by the Commission, sufficiently ensures sentences that satisfy a reasonable construction of “at or near the maximum.” In this setting, deference to the Commission is especially appropriate. “At or near” is an inherently variable phrase. In speaking with a Texan, one might say that Providence is “near” Boston, but it is doubtful if that description would (or could) be employed in speaking with a resident of, say, Cambridge or Cranston.

In all events, the phrase “at or near,” as employed in this statute, suggests a continuum of various sentences, each relatively further from, or closer to, the statutory maximum. It is also important to recognize that the career offender enhancement is not the end point of the sentencing road and, by itself, does not dictate individual defendants’ sentences.

Once the “Offense Statutory Maximum” derived from the Career Offender Guideline functions to yield a defendant’s TOL, the sentencing court must then make a myriad of individualized adjustments to the offense level, up or down, for factors such as acceptance of responsibility see U.S.S.G. § 3E1.1, role in the offense, see U.S.S.G. §§ 3B1.1, 3B1.2, and the like. It is only when all the component parts of the sentencing equation are pulled together that the court can ascertain the range of permissible sentences and, hence, settle upon the actual sentence.

Even then, the court retains authority, at least in certain circumstances, to depart downward if a particular defendant furnishes substantial assistance in the investigation or prosecution of another person who has committed an offense, see 18 U.S.C. § 3553(e); U.S.S.G. § 5K1.1, or to depart in either direction if aggravating or mitigating circumstances warrant, see 18 U.S.C. § 3553(b); U.S.S.G. § 5K2.0. Many of these prospective adjustments derive from explicit statutory commands.

See, e.g., 28 U.S.C. § 994(n) (directing the Commission to create a mechanism through which defendants will be rewarded for rendering substantial assistance). We believe that this reality has significant implications for the question at bar. First and foremost, given the labyrinthine way in which repeat offenders’ actual sentences are constructed, heightened deference to the Commission’s slant on the “at or near” language is very desirable. After all, respect for agency

interpretations is “particularly appropriate in complex and highly specialized areas where the regulatory net has been intricately woven,” Com. of Massachusetts, Dep’t of Educ. v. United States Dep’t of Educ., 837 F.2d 536, 541 (1st Cir.1988) (citation and quotation marks omitted), and the sentencing guidelines constitute a classic example of such a web.

In other words, due to the interstitial nature of the career offender calculation, a reviewing court should be generous in assessing the reasonableness of the Commission’s approximation of how “near” is “near.” The fact that the career offender adjustment does not itself directly determine any particular defendant’s actual sentence has other implications as well.

Unless one is ready to place any and all downward adjustments beyond a repeat offender’s reach — and even the government does not espouse so extreme a position — it is surpassingly difficult (if not impossible) to expect the Commission to write a rule which ensures that career offenders will invariably receive sentences “at or near” each individual’s ESM. Once a sentencing court has made such downward adjustments, it would be surprising if many defendants’ sentences came very near to the statutorily prescribed “maximum” penalties that are theoretically available (however the word “maximum” may be defined).

By like token, the very real possibility that upward adjustments to the TOL may make career offenders’ sentences more severe suggests that room should be left for play in the joints as the Commission implements the “at or near” language. Mindful, as we are, of these complexities, we think that Amendment 506 passes muster. The sentences available under the newly explicated Career Offender Guideline constitute a substantial proportion of the possible sentences permitted by statute. We can conveniently illustrate the point by reference to the four defendants who are involved in these appeals.

By operation of Amendment 506, defendants like LaBonte, Hunnewell, and Dyer now face maximum sentences of 262 months (the top of the recalculated GSR) before taking into account any individualized adjustments. A 262-month sentence represents 109.2% of the USM for these defendants’ offense of conviction. On the same basis, a defendant like Piper now faces a maximum sentence of 365 months (76% of the applicable USM).

Examining the gamut of possible sentences available against each defendant under Amendment 506, the median sentence in the range applicable to La-Bonte, Hunnewell, and Dyer (236 months) constitutes 98.3% of the USM, while the median sentence in the range pertinent to Piper (294.5 months) constitutes 61.4% of the USM.

Under any suitable definition of the word “near,” we believe that the Commission could reasonably conclude that these percentages ensure sentences sufficiently close to the USM — and sufficiently harsh — to provide a fair approximation of Congress’s desire to see that career offenders, as a group, receive maximal terms of imprisonment. IV. THE APPLICATION OF AMENDMENT 506 Having determined that Amendment 506 is a lawful exercise of the Sentencing Commission’s powers, we now address the motions for resentencing.

The principles governing motions to resentence based on newly emergent guideline amendments can be compactly cat-alogued. When the Commission amends the guidelines (or its

interpretation of the guidelines) in a manner that favors defendants, it may invite retrospective application of the new

interpretation. In such an event, a defendant who believes that the amendment, if in force earlier, would have reduced his GSR may move for resentencing. The district court, “after considering the factors set forth in

section 3553(

a) to the extent that they are applicable,” may reduce the sen-tenee “if such a reduction is consistent with the applicable policy statements issued by the Sentencing Commission.” 18 U.S.C. § 3582(e)(2). The law permits, but does not require, the district court to resentence such a defendant. See United States v. Connell, 960 F.2d 191, 197 (1st Cir.1992). Because this decision is committed to the trial court’s discretion, the court of appeals will interfere only if the record reveals a palpable abuse of that discretion. See United States v.

Par-due, 36 F.3d 429, 430 (5th Cir.1994), cert. denied, — U.S. -, 115 S.Ct. 1969, 131 L.Ed.2d 858 (1995); United States v. Tel-man, 28 F.3d 94, 96-97 (10th Cir.1994); see also United States v. Twomey, 845 F.2d 1132, 1134 (1st Cir.1988). It is plain that, under this paradigm, most resentencing battles will be won or lost in the district court, not in an appellate venue. With this brief preface, we reach the individual defendants’ cases. A. George LaBonte. In LaBonte’s case, the district court upheld Amendment 506 and applied it to reduce the defendant’s sentence. See La-Bonte, 885 F.Supp. at 24.

Although the government appeals from the reconfigured sentence, it challenges only the lower court’s validation of the reinterpreted Career Offender Guideline. Because the government has neither asserted nor argued a claim that the court abused its considerable discretion in reducing LaBonte’s sentence, we must affirm the judgment. See United States v. Zannino, 895 F.2d 1, 17 (1st Cir.), cert. denied, 494 U.S. 1082, 110 S.Ct. 1814, 108 L.Ed.2d 944 (1990). B. David E. Piper. In Piper’s case, the district court upheld Amendment 506 but refused to mitigate the original sentence.

Piper proffers a potpourri of protests to the court’s ruling. Only two of them warrant discussion. First, Piper suggests that under 18 U.S.C. § 3582(c)(2) a district court may only decide whether the policies underlying an amendment would be served by a lessened sentence. Piper misreads the statute: it authorizes the district judge to resentence when resentencing is consistent with the policies underlying the amendment, but it neither compels the judge to do so nor limits his inquiry to the consistency question.

Since the language is precatory rather than mandatory, the district court need not even consider the policy statements supporting an amendment if, “after considering the factors set forth in § 3553(

a) to the extent they are applicable,” 18 U.S.C. § 3582(c)(2), the court prefers to stand by the existing sentence. Piper’s next remonstrance suggests that the district court failed to reweigh the factors delineated in

section 3553(a), see supra note 14, and that, therefore, the court’s decision cannot constitute a proper exercise of judicial discretion. The problem with this remonstrance lies in its premise. The district judge presided over Piper’s ease from the outset. He possessed great familiarity with the odious nature of the offense of conviction (leading a “commando-style” raid on a family’s home while heavily armed, and searching for a stash of illegal drugs supposedly secreted there). Having sentenced Piper originally, he knew the intimate details of Piper’s criminal history.

At the hearing on the motion to resentenee, the judge listened to arguments that zeroed in on the very factors that Piper now claims were overlooked. In the end, Piper’s argument invites us to elevate form over substance. We decline the invitation. Where, as here, it is clear that the sentencing judge has considered the

section 3553(

a) factors, we will not interpose a further requirement that he make explicit findings as to each and all of those factors. See United States v. Savoie, 985 F.2d 612, 618 (1st Cir.1993) (holding that a district court need not make explicit findings regarding the statutory factors relevant to restitution orders “so long as the record on appeal reveals that the judge made implicit findings or otherwise adequately evinced his consideration of those factors”); United States v. Wilfred Am. Educ. Corp., 953 F.2d 717, 720 (1st Cir.1992) (similar, in respect to fines); see generally United States v.

Tavano, 12 F.3d 301, 307 (1st Cir.1993) (“As a general rule, a trial court lawfully may make implicit findings with regard to sentencing matters____”). On this record, it strains credulity to suggest that the district court neglected to take account of statutorily required items in its decisionmaking process. C. Alfred Lawrence Hunnewell. In Hunnewell’s case, the district court held that Amendment 506 was invalid, and refused to apply it for that reason. Having concluded that the lower court erred, see supra

Part III, we ordinarily would remand for further proceedings. But the government has other ideas; it asserts that the district court’s order should be construed as an exercise of discretion, and it asks us to affirm the denial of Hunnewell’s resentencing request on this basis. After a painstaking examination of the record, we reject the government’s asseveration. Calling a horse a cow does not yield milk.

Indeed, the government tacitly concedes the weakness of its position by forgoing developed argumentation on this point and instead regaling us with the reasons why the district could (or should) have declined to extend an olive branch to Hunnewell. The fact remains, however, that the discretion conferred by 18 U.S.C. § 3582(c)(2) is for the district court — not this court — to exercise in the first instance. Consequently, the denial of Hunnewell’s motion for resentencing must be set aside and the cause remanded for further consideration of that motion.

Before leaving Hunnewell’s situation, we pause to comment on the government’s suggestion that, because Hunnewell’s original sentence was still within the post-amendment GSR (albeit barely), we need not afford the district court an opportunity to decide whether to resentenee him. In its haste to validate this argument, the government distorts our holding in United States v. Ortiz, 966 F.2d 707 (1st Cir.1992), cert. denied, — U.S. -, 113 S.Ct. 1005, 122 L.Ed.2d 154 (1993).

In Ortiz, we explained that, where it appears reasonably likely that the district judge selected a sentence because it was at or near a polar extreme (whether top or bottom) of the guideline range that the judge thought applicable, the court of appeals should vacate the sentence and remand for resentencing if it is determined that the court erred in its computation of the range, notwithstanding that there may be an overlap between the “right” and “wrong” sentencing ranges sufficient to encompass the sentence actually imposed. Id. at 717-18. So it is here.

In Hunnewell’s initial sentencing hearing, both the government and the defense asked the court to impose a sentence at the bottom of the GSR. The court obliged. Giving vitality to the foundational principle on which Ortiz rests, we cannot be confident that, faced with a different range of options, the district court’s choice will remain the same. D. Stephen Dyer. Since Dyer’s and Hunnewell’s cases are virtually on all fours vis-a-vis the posture of the resentencing issue, we need not linger. For the reasons already expressed, see supra

Part IV(C), Dyer is entitled to have the district court address the merits of his request for resentencing. Y. THE

SECTION 2255 PETITION Dyer also appeals from the district court’s

summary dismissal of his

section 2255 petition. A district court may dismiss a

section 2255 petition without holding an eviden-tiary hearing if it plainly appears on the face of the pleadings that the petitioner is not entitled to the requested relief, or if the allegations, although adequate on their face, consist of no more than conclusory prognostications and perfervid rhetoric, or if the key factual averments on which the petition depends are either inherently improbable or contradicted by established facts of record. See United States v.

McGill, 11 F.3d 223, 225 (1st Cir.1993); see also 28 U.S.C. § 2255 (explaining that a hearing is unnecessary when the record “conclusively shows that the prisoner is entitled to no relief’). We believe that Dyer’s petition is both generally and specifically defective. Taking first things first, the district court noted that Dyer had not presented his factual allegations under oath, and that, therefore, he was not entitled to the relief that he sought. We agree. Dyer’s sworn petition contained nothing more than the bare statement that he received ineffective assistance of counsel.

While some additional allegations were set forth in Dyer’s memorandum of law, those allegations did not fill the void. A habeas application must rest on a foundation of factual allegations presented under oath, either in a verified petition or supporting affidavits. See, e.g., Rule 2, Rules Governing

Section 2255 Proceedings, 28 U.S.C. § 2255. Facts alluded to in an unsworn memorandum will not suffice. See Barrett v. United States, 965 F.2d 1184, 1195 (1st Cir.1992); Dalli v. United States, 491 F.2d 758, 760 (2d Cir.1974). Even were we prepared to overlook this fatal shortcoming, the petitioner would not find surcease. We review claims of constitutionally deficient performance on counsel’s part under the familiar test of Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).

According to this regime, a criminal defendant who alleges ineffective assistance must demonstrate that his attorney’s performance was unreasonably deficient, and that he was prejudiced as a result of it. See Scarpa v. Du-bois, 38 F.3d 1, 8 (1st Cir.1994), cert. denied, — U.S. -, 115 S.Ct. 940, 130 L.Ed.2d 885 (1995). When, as in this ease, a defendant has pleaded guilty to a charge, the prejudice prong of the test requires him to show that, but .for his counsel’s unprofessional errors, he probably would have insisted on his right to trial. See Hill v.

Lockhart, 474 U.S. 52, 59, 106 S.Ct. 366, 370-71, 88 L.Ed.2d 203 (1985). In light of these authorities, we think that the district court appropriately dismissed Dyer’s habeas petition. In his brief, Dyer contends, inter alia, that his trial attorney assured him that his sentence would be no more than eighteen months, and that there was simply “no way” that he would be sentenced as a career offender pursuant to U.S.S.G. § 4B1.1. Even a generous reading of this claim leaves no doubt that Dyer failed adequately to allege any cognizable prejudice.

An attorney’s inaccurate prediction of his client’s probable sentence, standing alone, will not satisfy the prejudice prong of the ineffective assistance test. See Knight v. United States, 37 F.3d 769, 774 (1st Cir.1994). Similarly, Dyer’s self-serving statement that, but for his counsel’s inadequate advice he would have pleaded not guilty, unaccompanied by either a claim of innocence or the articulation of any plausible defense that he could have raised had he opted for a trial, is insufficient to demonstrate the required prejudice. See United States v.

Horne, 987 F.2d 833, 835 (D.C.Cir.), cert. denied, — U.S. -, 114 S.Ct. 153, 126 L.Ed.2d 115 (1993); United States v. Arvanitis, 902 F.2d 489, 494 (7th Cir.1990). To add the finishing touch, the plea agreement that Dyer signed stated in so many words that he faced a maximum possible sentence of thirty years’ imprisonment. The district court reinforced this warning during the plea colloquy, and explained to Dyer that his sentence could not be calculated with certitude until the probation office prepared the presentence investigation report.

In response to questioning from the bench, Dyer acknowledged his understanding that even if he received a harsher-than-expected sentence, he would remain bound by his plea. And Dyer also assured the court that no one had made any promises to him anent the prospective length of his sentence. Thus, regardless of his counsel’s performance, Dyer was well aware of the full extent of his possible sentence when he decided to forgo a trial and enter a guilty plea. Under the applicable constitutional standard, a failure of proof on either prong of the Strickland test defeats an ineffeetive-assis-tance-of-counsel claim.

See Scarpa, 38 F.3d at 8-9. Since we find no cognizable prejudice, we need not determine what Dyer’s trial attorney did or did not tell him, or whether the attorney lacked familiarity with the sentencing guidelines to such an extent as to render his performance constitutionally infirm. We have also considered Dyer’s other assignments of error. His plaint that the district court acted precipitously in dismissing the petition without first pausing to convene an evidentiary hearing is meritless. See, e.g., McGill, 11 F.3d at 226; United States v. Butt, 731 F.2d 75, 80 n. 5 (1st Cir.1984).

His remaining claims are unworthy of detailed discussion. The lower court did not blunder in summarily dismissing Dyer’s application for federal habeas relief. VI. CONCLUSION We need go no further. For the reasons discussed herein, we affirm the judgments in the LaBonte and Piper cases (Nos. 95-1538 and 95-1226, respectively); remand for possible resentencing in the Hunnewell case (No. 95-1101); and affirm the judgment in the Dyer case (No. 95-1264) in part, but vacate it in part and remand for possible resentencing.

We intimate no view as to how the district court should resolve the remaining resen-tencing questions. So Ordered. . As we have said before, "irony is no stranger to the law.” Amanullah v. Nelson, 811 F.2d 1, 18 (1st Cir.1987). Throughout its history, the Sentencing Commission has been berated for the severity of the sentencing outcomes dictated by the guidelines. See, e.g., United States v. Jackson, 30 F.3d 199, 204-06 (1st Cir.1994) (Pettine, J., concurring) (criticizing the guidelines for fostering excessively harsh sentences); Daniel J.

Freed, Federal Sentencing in the Wake of Guidelines and Unacceptable Limits on the Discretion of Sentencers, 101 Yale L.J. 1681, 1690 (1992) ("The new sentencing guidelines are more complex, inflexible, and severe than those devised by any other jurisdiction.”); Charles J. Ogletree, Jr., Commentary: The Death of Discretion? Reflections on the Federal Sentencing Guidelines, 101 Harv.L.Rev. 1938, 1939 (1988) (criticizing the "unreasonably long sentences” produced by the guidelines). . Piper received an additional five-year sentence on the firearms count. That impost is not in issue here. .

The amendment, if applied, would have lowered Piper’s adjusted offense level from thirty-four to thirty-two, and decreased the GSR to 210-262 months. . The amendment, if applied, would have lowered Hunnewell’s adjusted offense level from thirty-one to twenty-nine, and decreased his GSR to 151-188 months. . Amendment 506, if applied, would have lowered Dyer’s adjusted offense level from thirty-four to thirty-two, and decreased his GSR to 210-262 months. .

We note in passing the suggestion by some scholars that Stinson implies an extraordinarily deferential standard of review for the entire process of evaluating guideline commentary. On this view, commentary should be honored unless it constitutes a plainly erroneous

interpretation either of a guideline or of a statute. See 1 Kenneth Culp Davis and Richard J. Pierce, Jr., Administrative Law Treatise § 6.10, at 284 (3d ed. 1994). We need not probe this possibility today. Because Amendment 506 passes muster under the Chevron test, it would clearly pass muster if we were to employ the more deferential test suggested by Professors Davis and Pierce. . Because the government does not contend that Amendment 506 is inconsistent with the guideline itself, we eschew any discussion of that point. See United States v.

Zannino, 895 F.2d 1, 17 (1st Cir.) (explaining that issues not briefed and argued are deemed abandoned), cert. denied, 494 U.S. 1082, 110 S.Ct. 1814, 108 L.Ed.2d 944 (1990). .Although we are mindful that plausible if strained

interpretations of a series of individual statutory terms might at times lead to an impermissible overall

interpretation of a statute, that is not the case here. Whether one conducts the ensuing analysis in one segment or two, the result is unaffected; the simple fact of the matter is that the Commission has developed a reasonable

interpretation of the vague and ambiguous language of

section 994(h). That said, we employ a piecemeal approach here, as we believe it better illustrates that U.S.S.G. § 4B1.1, as interpreted by the amended commentary, is a permissible construction of Congress's directive that career offenders be sentenced "at or near the maximum term authorized.” . The relevance of this somewhat arid discussion will become more apparent in

Part III(C), infra, when the need arises to determine the extent to which sentences are “at or near the maximum.” . We acknowledge the ongoing debate over the propriety, under Chevron, of going beyond plain meaning analysis and resorting to the traditional tools of statutory construction in search of a clear congressional directive. Compare INS v.

Cardoza-Fonseca, 480 U.S. 421, 446-48, 107 S.Ct. 1207, 1221 22, 94 L.Ed.2d 434 (1987) (suggesting that, under the first prong of Chevron, courts should employ "traditional tools of statutory construction”) with id. at 454, 107 S.Ct. at 1225 (Scalia, J., concurring) (rejecting this statement). This court has followed Chevron, 467 U.S. at 843, 104 S.Ct. at 2781-82 n. 9, and employed the full tool chest of statutory construction implements in attempting to detect clear congressional meaning. See, e.g., Strickland, 48 F.3d at 19. We continue that practice in this case. .

The government makes much of the fact that the Senate Judiciary Committee, in creating the Commission, disclaimed any fear that the guidelines would increase prosecutors’ discretion to reduce sentences through plea bargains. See S.Rep. No. 98-225, supra, 1984 U.S.C.C.A.N. at 3246. But Congress’s explanation (which stressed that the Commission could guard against this phenomenon because it was empowered to issue policy statements concerning the review of plea bargains, see id.), is indicative of the latitude it intended to give to the Commission. .

We think that this calculation graphically illustrates the fallacy underlying our dissenting brother's lament that Amendment 506, “effectively nullifies the criminal history enhancements carefully enacted in statutes like 21 U.S.C. § 841.” See post at 1415. . For this purpose, an "amendment” differs from a "clarification.” Clarifications explain earlier editions of the sentencing guidelines; they do not change those provisions. Because they are retrospective by nature, they do not require any special retroactivity designation. See U.S.S.G. § 1B1.11(b)(2); see also United States v.

LaCroix, 28 F.3d 223, 227 n. 4 (1st Cir.1994). In contrast, amendments do change prior guidelines and, if they are to be given retroactive effect, the Commission must so specify. See 28 U.S.C. § 994(u); U.S.S.G. § 1B1.10. This opinion deals exclusively with amendments as opposed to clarifications. . The factors set forth in

section 3553(a), insofar as they are arguably applicable to any of the instant defendants, include the nature and circumstances of the offense, the defendant’s criminal past, the GSRs, the Commission's policy statements, and the necessity of avoiding unwarranted sentencing disparities among similarly situated defendants. See 18 U.S.C. § 3553(a). . The district court initially computed a GSR of 188-235 months, and sentenced Hunnewell to serve 188 months in prison. Applying Amendment 506 to Hunnewell’s case yields a revised GSR of 151-188 months. See supra note 4.