with whom BARRETT, Circuit Judge, joins in dissenting:
Overriding the district judge’s characterization of his own sentencing proceeding, the majority opinion applies a general, rebuttable legal presumption to relieve a confessed, convicted, career criminal of any separate punishment on his drug dealing conviction. I find that result, on the facts of this case, to be offensive both to justice and the administration of justice. It trivializes the role of the sentencing district judge and frustrates the imposition of criminal penalties which Congress prescribed and intended to be imposed.
When a result in a particular case is as unpalatable as the one here the legal process from which the result flows becomes suspect. Analysis of that process shows the suspicion to be justified. Further justification for a critical analysis of the process employed in the majority opinion is found in the fact that Congress has passed a law outlawing use of the presumption in question in cases like this. The new law is *1435effective in a few months; thus defendant Earley escapes its application. But when Congressional sentiment against use of the presumption is so strong, the judiciary ought to take the hint and more critically examine both the presumption itself and whether it must be used in a particular case. That is especially true when the presumption itself is acknowledged to be rebuttable.
In short, both the result in this case and a clear message from Congress make it our duty to explore every legitimate avenue for not using the presumption, and place a heavy burden of persuasion on those who insist there is no other choice. The majority opinion assumes no such burden.
I contend first that consideration of the presumption does not even arise in this case because there is no reasonable doubt on this record that the sentencing judge intended to punish Earley separately on his drug dealing conviction. That is, the record as a whole leaves no reasonable doubt that the drug dealing sentence was consecutive to the sentence imposed by another judge fifteen years earlier. Second, even if the presumption enters into consideration, it does not compel us, as a matter of law over which we have no discretion, to relieve Earley of his separate punishment on his drug dealing conviction. Finally, double jeopardy does not immunize Earley from the administration of justice.
The basic facts, again, are as follows. The sentencing judge in this case failed to indicate in exact words in his pronouncement of sentence whether two consecutive five-year sentences for drug offenses were to be served concurrently with or consecutively to a federal sentence imposed some fifteen years earlier. In the earlier proceeding, another judge in another state had sentenced Earley to twenty-five years in prison for interstate racketeering and bomb possession. Earley had been paroled after serving ten years of the prior sentence, but his parole had been revoked in 1982 after he had committed the drug offenses for which he subsequently was sentenced, in this case, on January 4, 1985. Under the terms of the parole revocation, he was not eligible for re-parole on the underlying offense until early 1985.
The exact words used at the point of conclusion of the sentencing proceeding January 4, 1985, are: “on each of Counts 1 and 3, ... imprisonment for a term of five years to run consecutively for a total of ten years and [the defendant] to serve a special parole term of life.” R. Yol. II. at 13. The written judgment and commitment order reflects the oral statement. R. Vol. I at 29. The majority essentially bases its opinion on the omission at that point in the sentencing proceeding of an express statement that the sentence was to be served consecutively to the underlying sentence.
I.
As indicated, I believe the sentencing proceeding as a whole discloses beyond a reasonable doubt that the district judge intended Earley’s sentence for drug dealing to run consecutively to the underlying sentence. Thus, I maintain that the presumption of concurrency bears no relevancy to this case.
In exploring that point it must be remembered that no rule exists which limits the imposition of sentence to certain rote words at a particular narrow point in the sentencing proceeding. Also, there is no doubt of our authority to examine the sentencing proceeding as a whole, and related facts, to determine what sentence was given, or to seek the judge’s intent in that regard. Certainly, we are as duty-bound and as much at liberty to search for manifest intent in sentencing as we are to seek legislative intent, or, for that matter, the intent of the framers of the Constitution. The search for intent is an exercise in which we continually engage. “[T]he Constitution does not require that sentencing should be a game in which a wrong move by the judge means immunity for the prisoner.” Bozza v. United States, 330 U.S. 160, 166-67, 67 S.Ct. 645, 649, 91 L.Ed. 818 (1947).
The most urgent reason for our review of the sentencing proceeding as a whole is the sentencing judge’s formal, solemn, sincere characterization of his own proceed*1436ing. He tells us that he manifestly intended consecutive sentences; his recorded statements and actions in sentencing Earley fairly disclose that intention. His October 25,1985 order denying the petition to vacate his May 3,1985 clarification order does not admit to any change in or enhancement of sentence. It does not concede that the sentencing proceeding as a whole was in any way ambiguous.1 The majority passes over the matter of deference to the sentencing judge with slight comment. But the role of the district judge in sentencing is not to be so lightly dismissed. At the point of sentencing the judge speaks for society, acting to do the bidding of citizens’ representatives in Congress in imposing punishment. The interests of society demand our deference to that function. When the district judge tells us what he or she expressly intended in sentencing and tells us to look at his or her words and actions, our duty of deference makes clear how we should incline our review.
With that framework for review, we turn to the record. A key portion of the sentencing transcript is cited in the majority opinion in a footnote as if it were not at the very heart of the dispute. I am bringing it front and center to compel a detailed analysis of the words so that their clear import can be fully understood.
THE COURT: Mr. Earley’s pre-sentence report is one of the more remarkable ones that I have reviewed in nine years plus of attempting to find fair and appropriate sentences in criminal cases. It is one of the most serious and extensive criminal records that I have seen. I know nothing about Mr. Earley other than what I read in the report and what I have heard from what’s been said in this proceeding.
But, it does strike me that from an early age and throughout his lifetime, there has been just a succession of criminal activity that gives me absolutely no basis for expecting or hoping for any kind of rehabilitation in the case.
R. Vol. II at 12 (emphasis added). It is unmistakable from those comments that the judge regarded Earley as a hardened, career criminal with “absolutely ” no basis for hope of rehabilitation of “any kind.” Those observations follow directly after the court’s reference to what sentence will be appropriate, considering that Earley has one of the most extensive and serious criminal records the judge has seen in nine years on the Bench. Thus, our attention is directed to what sentence the judge deems appropriate for a lifetime criminal who is beyond rehabilitation. Does that suggest the court is or would be disposed to no separate sentence (i.e., a concurrent sentence) for drug dealing? Parole eligibility in less than five months? Certainly not. In that same unmistakable vein, the judge continues:
Ordinarily, and in any criminal case, a plea of guilty itself gives the sentencing judge some basis for positive expectations in regard to the Defendant. There is hardly any basis for that in this case.
Mr. Earley has certainly made a very favorable plea bargain, however, and severely limited his total exposure to punishment as a result of the favorable plea agreement. And, there’s nothing wrong with that, it is encouraged and it’s perfectly legitimate. In doing so, he has limited his exposure very substantially and has the benefit of having done so. But, beyond that, there is simply no reason that I can find to credit Mr. Earley with any reduction in time. His probation violation argument is not persuasive in light of the fact that he has actually just been required to serve out the punishment for another offense. All he had to do to avoid that was not break the law. And, he saw fit to engage in narcotics trafficking, which is a very serious offense.
Id. at 12-13 (emphasis added). This portion of the judge’s remarks relates back to a prior dialogue about whether the offend*1437er should be given any credit in his current sentencing for the thirty months he served for his parole violation. The judge is observing that parole violation is a separate offense from drug dealing and was punished separately. The point being made is that neither the petitioner’s plea bargain nor his time served for parole violation gave the judge any reason to be lenient in imposing sentences for drug dealing. The judge then makes clear his view that the drug trafficking conviction is “very serious.” Does that, coupled with the judge’s statement that “there is simply no reason that I can find to credit Mr. Earley with any reduction in time,” suggest an intent not to impose any separate sentence? Again, certainly not.
The judge then concludes:
And, under all these circumstances, I simply find that he gets the benefit of his plea agreement. But there is no basis for any other consideration.
Id. at 13 (emphasis added). The last sentence bears repeating: “But there is no basis for any other consideration” (emphasis added).
Consistent with the clear intent to impose a maximum punishment the judge then gave Earley just that — the maximum punishment allowed, five years on each count, to be served consecutively, plus a special parole term for life. At that point the judge failed to address the underlying sentence imposed fifteen years earlier by another judge in a different state. But his intention in that regard is absolutely unmistakable.
Here, then, is the summary of the judge’s manifest perception of Earley’s drug dealing conviction vis-a-vis sentencing: (1) one of the most serious and extensive criminal records the judge had ever seen; (2) lifetime of criminal activity; (3) absolutely no basis for hope of rehabilitation; (4) receipt already of a favorable plea bargain; (5) recognition that when Earley was last on parole he engaged in criminal activity; (6) guilt for the very serious offense of drug trafficking; and, (7) no basis for any other consideration. Add to that the type of criminal conduct in the underlying conviction: interstate racketeering and bomb possession! It is shocking to me that anyone could read the statements by the sentencing judge, in this factual context, and conclude that the judge’s intent that Earley be sentenced separately on his drug trafficking conviction is not manifest beyond a reasonable doubt. Conversely, it is untenable, even absurd, to assert, by presumption or otherwise, see Maj. Op. at 1431 n. 4, that the judge may have intended lesser or no separate punishment and may have been inclined to put Earley in line for parole as early as possible — perhaps in three to five months!
We can deduce without reservation that had the judge explicitly elaborated on his intent in the actual pronouncement of the sentence he would have ordered the two drug sentences to be served consecutively with the prior sentence. This is the only reasonable deduction. The alternative, that the judge’s intent was unknowable or unclear, as the majority suggests in footnote 4, requires the majority necessarily to assume that the judge would want directly to contradict and undermine his intent to impose the maximum sentences for the drug offenses. To intend two contrary results simultaneously is irrational, and there is no evidence that the judge wished to produce an irrational result. To elaborate the point again, Earley became eligible for parole on his prior sentence less than three months after beginning to serve his drug sentences. Had the two later sentences run concurrently with the prior sentence, he could have received (and initially did receive) a release date resulting in imprisonment of less than five months on his consecutive, maximum drug sentences of ten years. The result is to erase the judge’s sentence for the drug offenses. That is where we are: the majority opinion “presumes” this judge intended no separate sentence for Earley on drug dealing! That result will not be understood nor appreciated by our fellow citizens. It does indeed make sentencing a game “in which a wrong move by the judge means immunity for the prisoner.”
*1438II.
We are not compelled to apply the presumption of concurrence in any event. The majority opinion cites no controlling authority, and there is none. Even if the presumption applied it would be rebutted on the facts of this case.
A.
With no analysis, the majority states flatly that the presumption of concurrence “has in effect become a rule of law,” citing Subas v. Hudspeth, 122 F.2d 85 (10th Cir.1941), and a comment from the Senate Report on the Crime Control Act which also cites to Subas.2 But the cited comment from Subas was only dicta.3 The holding in Subas was that the presumption did not apply. And, Senate committee reports do not make law for us, especially when they miscite supporting authority, i.e., Subas.
The fact is that no case in this circuit has held that the presumption of concurrence must apply to a case like this. No Supreme Court case has so held. Only two circuits, the Fifth and the D.C. Circuit, have applicable holdings (leaving aside the question of rebuttability). See Schultz v. United States, 384 F.2d 374 (5th Cir.1967); Borum v. United States, 409 F.2d 433 (D.C.Cir.1967), cert. denied, 395 U.S. 916, 89 S.Ct. 1765, 23 L.Ed.2d 230 (1969). Furthermore, the result under the Borum facts can not be duplicated in the D.C. Circuit because Congress promptly thereafter enacted a D.C.Code provision outlawing the presumption.4 The Fifth Circuit is left as the sole circuit with analogous holdings.5
To be sure, there is dicta, spread through dozens of cases, referring generally to a “rule of law” regarding a presumption of concurrence. But it is just that: both general and dicta. When federal courts actually face the necessity of deciding the direct issue before them, they frequently avoid the presumption or apply it in discernible categories of fact situations, the overwhelming majority of which are not analogous to this case. A review of textbooks on sentencing discloses a vast silence on the subject of the supposed “rule of law” or at most a passing comment so general as to be worthless for application to specific cases with differing fact situations.6 That reflects the trouble with grasping “rules of law” from the ether. They are too vaporous to apply to specific cases.
Therefore, I challenge the majority opinion premise that we are required in this case to apply the presumption of concurrence because the presumption is a “rule of law” which controls us.
Detailed analysis begins with a fundamental fact distinction which exists in presumption of concurrence cases. First, cases involving multiple sentences imposed by the same judge at the same time can be differentiated from sentences imposed by the same judge at different times. Second, an even greater distinction exists where sentences are imposed at different times by different judges — which is the case before us. These distinctions are so pivotal that they have formed the basis for recent Congressional legislation on the subject.7 The new statute requires that sentences im*1439posed for a different crime at a different time must be served consecutively to sentences imposed for other crimes at an earlier time, unless the sentencing judge specifies that the later sentence is to be served concurrently with the earlier sentence. Multiple sentences imposed by the same judge at the same time run concurrently if the sentencing judge does not specify that they are to be served consecutively (or if a statute does not provide otherwise). In analyzing the case authorities that follow, it must be emphasized that the case before us involves the most remote fact situation: a sentence imposed on Earley for his drug dealing conviction, wholly separate from a sentence imposed fifteen years earlier by a different judge in a different state for different offenses. I turn now to the cases.
The dicta from Subas which the majority opinion cites as if it is controlling authority was a summary by the court of one of the defendant’s arguments. Subas referred to “the salutary rule” regarding the presumption of concurrency on its way to holding that the presumption should not apply. Furthermore, the reference to the presumption was followed by language from United States v. Daugherty, 269 U.S. 360, 363, 46 S.Ct. 156, 157, 70 L.Ed. 309 (1925), that the “elimination of every possible doubt [in the sentencing language] cannot be demanded,” which was the point of the Subas opinion. That, plus the fact that Subas involved sentences handed down by the same judge at the same time,8 sufficiently places Subas beyond any possibility of being controlling authority in this case.
Deeper analysis is even more damaging to the majority’s attempt to use Subas as authority in a case like that before us, involving sentences by different judges at different times. Subas dicta, cited here by the majority, reads, “[ajbsent clear language to the contrary, it is presumed that sentences imposed on more than one offense at the same time, or at different times, will run concurrently.” Subas, 122 F.2d at 87. At this point in the Subas text, a footnote appeared, containing a string of citations to cases supposedly in support of the dicta. Examination of the fifteen cases in the supporting footnote reveals that ten of them involve situations where multiple sentences were imposed by the same judge at the same time and not by different judges at different times. Furthermore, in a majority of the ten cases the presumption is explicitly stated in terms restricting it to the factual situation of multiple sentences by the same judge at the same time. See, e.g., Hode v. Sanford, 101 F.2d 290 (5th Cir.1939); United States v. Remus, 12 F.2d 239 (6th Cir.) (referring to United States v. Patterson, 29 F. 775 (C.C.1887)), cert. denied, 271 U.S. 689, 46 S.Ct. 640, 70 L.Ed. 1153 (1926). Additionally, approximately half of the ten cases declined to apply the presumption of concurrence to the actual facts of the case. See, e.g., United States v. Daugherty, 269 U.S. 360, 46 S.Ct. 156, 70 L.Ed. 309 (1926); Remus, 12 F.2d 239.
In partial contrast, two other cases cited in the footnote do concern sentences imposed by different judges at different times, but neither applies the presumption. McNealy v. Johnston, 100 F.2d 280 (9th Cir.1938); Zerbst v. Walker, 67 F.2d 667 (10th Cir.1933). McNealy states the presumption broadly in terms that do not explicitly limit the presumption to the same judge at the same trial. Its statement is dicta, however, because McNealy construed an ambiguous sentence to uphold the judge’s intent that it run consecutively with an earlier sentence imposed by an earlier judge. Similarly, we declined to apply an extended presumption in the context in Zerbst, where a prisoner had escaped and was not actually serving his previous sentence, and where the judge was not aware of the prior sentence at the time of the subsequent sentencing. In Zerbst this *1440court acknowledged that the Fifth Circuit had applied an extended presumption to situations where sentences were imposed by different judges at different times but we did not specify under what circumstances we would have applied such a presumption. Even if we had, it would have been dicta. Thus, the question of how we would rule on the kind of issue facing us in this case was left open.
The Subas footnote contains only three cases in which the presumption is actually applied where there are multiple sentences imposed at different times. One of those cases was before our court, and we justified application of the presumption on the grounds that knowledge of the first sentence could be presumed because multiple sentences had been imposed by the same court even though at different times. White v. Kwiatkowski, 60 F.2d 264 (10th Cir.1932). The presumption was not being applied where sentences were imposed by different courts. The other two cases are both Fifth Circuit cases, and they are the only cases cited which apply the presumption of concurrent sentences where one judge is silent with respect to a prior sentence by a different judge. Aderhold v. McCarthy, 65 F.2d 452 (5th Cir.1933); Zerbst v. Lyman, 255 F. 609 (5th Cir.1919). Neither cites any federal court cases outside the Fifth Circuit in support of the extended presumption. Neither gives any explanation for presuming that judges are as conscious of sentences imposed by other judges at earlier times as they are of their own multiple sentences; nor does either decision explain why the prisoner’s interest in liberty should be protected in cases where completely separate judgments issue from completely separate trials and convictions.9 Instead, Aderhold cites to Zerbst, and Zerbst cites to three state court decisions. In turn, those decisions cite either no cases at all or inapposite cases as support for their determination to allow sentences imposed at different times to run concurrently.10
What emerges from careful analysis of the Subas footnote is only the most meagre legal support for the expansiveness of the presumption as stated somewhat cavalierly in the body of the Subas opinion. Subsequent to Subas, our court considered application of the presumption of concurrence in Owensby v. United States, 385 F.2d 58 (10th Cir.1967). The factual setting in Owensby involved an attempt by a sentencing judge to correct a faulty sentence — one of four separate sentences imposed at the same time by him. The correction of the faulty sentence, while valid in itself, did not specify the relationship of the earlier sentences to the newly corrected sentence. In the original sentencing, each of the sentences was to be served consecutively. We stated, without citation to any source, that “the valid sentence imposed October 28 in the Oklahoma case [correcting the faulty sentence], absent a specific provision to the contrary, is deemed to run concurrently with other sentences.” Id. at 60. The *1441effect was to undo the original consecutiveness of the faulty sentence. However, the presumption was invoked in Owensby in a situation where the same judge imposed the sentences in question. Therefore, Owensby is not analogous to the factual setting before us here.
No other Tenth Circuit case on point discusses the presumption of concurrence.11 There are no holdings on point. The same is true for Supreme Court cases. In short, no law exists which compels us to apply the presumption of concurrence in this case.
As previously indicated, the Fifth Circuit has authority arguably analogous to the instant case,12 as did the D.C. Circuit for a brief period until effectively overruled by Congress.13 That is the sum total of persuasive authority actually on point, available to the majority for support. Furthermore, those cases have facts sufficiently different from those in this case that they lose any persuasiveness they may otherwise have, especially in light of the sentencing transcript available here. The closest additional authority is found in dicta from United States v. Wenger, 457 F.2d 1082 (2d Cir.), cert. denied, 409 U.S. 843, 93 S.Ct. 44, 34 L.Ed.2d 83 (1972). In Wenger the Second Circuit, without citation of any supporting authority, said it was merely willing to “assume” that the presumption of concurrence would apply to situations involving sentences by different judges at different times. However, the court in Wenger refused to apply the presumption on the facts of that case. Thus, except for the Fifth Circuit, and some dicta from the Second Circuit, the circuits appear not to have addressed the presumption of concurrence in factual circumstances similar to the case before us.
In summary, we find ourselves in the indefensible position of expanding a common law presumption in our circuit at a time when Congress has ordered the courts to retreat from the presumption of concurrence not only in cases where sentences are imposed by different judges at different times, but even in cases where sentences are imposed by the same judge at different times. Given the lack of any actual authority supporting the majority opinion, I view its holding as a distressing disregard of the express desire of Congress, creating an expanded rule which can only last for the few months between the issuance of this opinion and the effective date of the legislation later this year, the ultimate effect being a gratuitous and unmerited indulgence to Earley. In short, the majority not only does not search for reasons why Earley should not be relieved of separate punishment for his drug dealing conviction, but it affirmatively searches for reasons why he should be relieved of that conviction. In doing so, it not only does not join the ranks of other federal circuits, as it seems to believe, but it extends the presumption of concurrence to a situation which causes our holding to now stand alone with those in the Fifth Circuit.
B.
The final argument for rejection of the presumption is that it is rebuttable where the ambiguity can be resolved or where the facts cry out for rejection of the presumption. Rebuttability is, of course, a reprise of the argument that the presumption should not apply at all (section I). But it more accurately addresses the majority opinion on its own ground, i.e., silence delivers concurrent sentences.
Where justice would be distorted by rigid application of the presumption, appellate courts have shown broad willingness to acknowledge some version of the “rule” in dicta, then find it inapplicable on the facts *1442by examining judicial intent. See, e.g., United States v. Daugherty, 269 U.S. 360, 46 S.Ct. 156, 70 L.Ed. 309 (1926) (court acknowledged only that one circuit had invoked presumption but then overturned circuit where silence was with respect to the order in which consecutive sentences were to be served); United States v. Wenger, 457 F.2d 1082 (2nd Cir.) (assuming existence of expanded presumption but refusing to apply it where explicit statements by defendant’s counsel provided evidence that defendant understood the judge’s silent intent that terms would run consecutively), cert. denied, 409 U.S. 843, 93 S.Ct. 44, 34 L.Ed.2d 83 (1972); Buie v. King, 137 F.2d 495 (8th Cir.1943) (affirming correction of judgment after a hearing to determine the judge’s intent and to resolve ambiguous silence in the original sentence). Cf. Puccinelli v. United States, 5 F.2d 6 (9th Cir.1925) (presumption invoked where no record evidence of intent was available to justify amending a sentence so that prison terms ran consecutively; record evidence may include any relevant item including an entry, note, or memorandum from the records or quasi records, showing the initial judgment was intended to be consecutive).
The majority opinion suggests the presumption would not apply, despite silence, if Earley clearly was aware the sentences were consecutive.14 Thus, rebuttability of the presumption is recognized. The majority opinion, however, demonstrates the same defect in legal analysis on rebuttability that it shows in its approach to the presumption itself. First, it essentially assumes this circuit is controlled by judicial expressions elsewhere, and it makes no attempt to acknowledge that it is announcing a rule of law on rebuttability (sentence is consecutive if defendant is clearly aware) for the first time in this circuit. Second, it assumes that the subject matter of rebuttability in general and the outer limits of rebuttability on the point in question (defendant’s knowledge) have been established in the three cited cases.
The parameters of the presumption are hardly so clearly fixed or immutable. If immutable, why, for instance, would the Seventh Circuit be observing a serious erosion in the presumption by the Second Circuit. See United States v. Solomon, 468 F.2d 848, 850 n. 4 (7th Cir.1972), cert. denied, 410 U.S. 986, 93 S.Ct. 1513, 36 L.Ed.2d 182 (1973). Immutable things are not erodable. Nor do they reflect differences from circuit to circuit both in expressions of the “rule” and in its application to fact situations. The majority opinion cites, and research discloses, no case which is factually on all fours with this case. There does not even appear to be a case on all threes. But the majority opinion makes no attempt seriously to explore the rebuttability of the presumption of concurrent sentences on the facts of this case. As shown in section I of this dissent, such an exploration is made imperative by a common sense reading of the sentencing proceedings and deference to the sentencing judge’s characterization of those proceedings.
Finally, as already has been observed, the majority opinion, without the slightest deference, rejects the sentencing judge’s own fact finding that defendant Earley knew that the sentence imposed was consecutive to the underlying sentence. The majority insists that there be objective evidence that Earley was subjectively “aware” of what his sentence was. Maj. Op. at 1431 n. 4. Three cases are cited in support of that conclusion.15 Nevertheless, the language in those cases is not binding on this circuit, and the conclusion deserves to be challenged. If the sentencing judge’s subjective intent is irrelevant, the same should be true with regard to the defendant. If it does not matter what the judge thought or understood, why does it matter what Earley actually thought or understood? Applying to Earley the majority’s objective standard for the judge, the ques*1443tion becomes: what was Earley entitled to understand as to the judge’s intention from an objective review of the record? What would an objective fact finder determine that reasonable people should have understood from the sentencing proceedings as a whole? Was Earley entitled to believe that the judge did not intend to give him a separate sentence for his conviction for drug dealing, after a major portion of the counts had been plea bargained away, and after the judge’s comments with respect to Earley’s record?
It is significant that in requiring evidence that the defendant clearly understood the sentence to be consecutive, the majority applies a stricter standard for rebutting a common law presumption than the Supreme Court uses for constitutional purposes. In considering defendants’ expectations of finality in a sentence subject to appeal, the Supreme Court has specifically disregarded what defendants personally know or do not know. In United States v. DiFrancesco, 449 U.S. 117, 136, 101 S.Ct. 426, 437, 66 L.Ed.2d 328 (1980), the Supreme Court stated: “The defendant, of course, is charged with knowledge of the statute and its appeal provisions, and has no expectation of finality in his sentence until the appeal is concluded or the time to appeal has expired.” I contend that Earley’s personal “awareness” is irrelevant, and that he is charged with an apprehension of what the proceedings show on their face — that concurrent sentences were not intended.16
The constant underlying theme in sentencing is not a search for the defendant’s intent. Rather, it is a search for judicial intent, a point which I have purposely stressed over and over in this dissent. The theme is aptly picked up by the Eleventh Circuit in United States v. Purcell, 715 F.2d 561 (11th Cir.1983), where the court stated:
After a careful review of the record, we find that the trial judge’s original intent was to sentence Purcell to an enhanced prison term of ten years for Count II, not an unenhanced excessive term under the normal sentencing statute. In determining the terms of a sentence, it is the intent of the sentencing judge which controls, and that intent is to be determined by reference to the entire record.
Id. at 563 (emphasis added). Our court at least gave lip service to that approach in the companion case of United States v. Villano, 816 F.2d 1448 (1987), where we stated: “[i]f there is an ambiguity in the sentence, then such extrinsic evidence as ... the judge’s intentions ... may be consulted.” Id. at 1453 (footnote omitted). The opinion in Villano then qualified this remark by limiting the situations in which ambiguity arises. One of those situations was said to be where the plain meaning of the judge’s words led to an irrational or absurd result. Id. at 1453 n. 6. Silence was not among the situations listed as creating an ambiguity, but clearly, unmistakably, the judge’s silence here leads to an irrational or absurd result if the sentencing transcript is not allowed to clarify the judge’s intent. Under such circumstances I see no difference between ambiguous silence and ambiguous words.
In sum, the majority opinion once again demonstrates not only unwillingness to acknowledge that it is plowing new ground in this circuit, and extending the presumption in question, but unwillingness to explore all possible avenues for avoiding a blind and rigid application of this common law presumption which will die, by statute, in a few months.
III.
After briefly finding that Earley received a presumed concurrent sentence, the majority devotes the remainder of its opinion to whether the Double Jeopardy Clause barred the judge from “changing” to a consecutive sentence five months later. In its double jeopardy analysis the majority *1444makes sweeping determinations as to the position of this circuit subsequent to the Supreme Court opinion in United States v. DiFrancesco, 449 U.S. 117, 101 S.Ct. 426, 66 L.Ed.2d 328 (1980).
In DiFrancesco the Supreme Court held that the Double Jeopardy Clause did not prevent a statutorily authorized government appeal of concurrent sentences under 18 U.S.C. § 3576. It also held that “the increase of sentence on review under § 3576” did not constitute “multiple punishment in violation of the Double Jeopardy Clause.” 449 U.S. at 138, 101 S.Ct. at 438. In the process of so holding the Supreme Court made crystal clear “that a sentence does not have the qualities of constitutional finality that attend an acquittal.” Id. at 134, 101 S.Ct. at 436. The Court also expressly left open the question of judicial power to increase a sentence after the defendant has begun to serve that sentence. Id.17
The majority sees nothing new in DiFrancesco, describing its teaching only as an “appropriate framework” for applying existing double jeopardy analysis. According to the majority the prevailing constitutional concept has always been that double jeopardy attaches at the point where judicial power to sentence ends, inclusive of appeals and retrials. As a generality directed to the “finality of judgments,” there is probably not much wrong with such a view. But it begs the question, i.e., when and under what circumstances does the Double Jeopardy Clause dictate an end of judicial power to sentence, including clarifications and corrections of manifest omissions or errors? The majority opinion answers the question by perceiving support in DiFrancesco for the “bright-line” approach to double jeopardy. Thus, DiFrancesco is viewed merely as emphasizing “that the location of bright-line constitutional finality in the sentencing context may vary with the varying statutory provisions granting jurisdiction to the courts.” Maj. Op. at 1433 (emphasis added). The majority then commits this circuit to the following rigid bright-line rule for determining when constitutional finality is achieved and double jeopardy attaches:
Unless a statute or rule extends the sentencing process further, the limit imposed by double jeopardy analysis is coterminous with the limit of the court’s original sentencing authority — stopping the sentencing court’s power “at the jailhouse door.”
Maj. Op. at 1434.18
No such rule can or ought to be gleaned from DiFrancesco. While the case dealt with a statute, it did not limit its teachings on constitutional finality to statutes and rules.19 Furthermore, it opened, and expressly refused to close, the door relating to the constitutional effect, if any, of transfer to executive custody. Refusing the invitation to re-examine the subject, the majority now closes that door with scant analysis.
Respectfully, the double jeopardy rule announced by the majority is antique and inadequate, deriving its rigidity from the *1445discredited notion that a sentence is the constitutional equivalent of an acquittal. It is advanced by the majority at a time when other circuits are re-examining and overturning historic double jeopardy positions due to DiFrancesco and pursuing more flexible approaches.20 The majority acknowledges, that since DiFrancesco, most courts “have focused on whether the defendant had a legitimate expectation that his sentence was final.” Maj. Op. at 1438.21 But it refuses to pursue that inquiry here unless a statute or rule of procedure permits. The result of the majority’s announced rule will be to isolate this circuit both in concept and rigidity of double jeopardy analysis and condemn us in future cases to rationalizing our analysis to the “framework” imposed here.
I believe that the reasoning of DiFrancesco and its careful attention to the originating cases of Ex parte Lange, 85 U.S. (Wall) 163, 21 L.Ed. 872 (1874), and United States v. Benz, 282 U.S. 304, 51 S.Ct. 113, 75 L.Ed. 354 (1931), make clear that “bright-line” finality need not occur at the moment of executive custody, even in the absence of a rule or statute.
As the DiFrancesco Court notes, Ex parte Lange is cited as the primary authority for the principle that double jeopardy attaches where the federal government attempts to punish an offender twice for the same offense. As with Subas, both the facts and the actual holding of Ex parte Lange matter greatly.22 In Ex parte Lange, the offender had been erroneously sentenced to both serve a jail term and pay a fine, while the statute authorized only one or the other. The offender had begun to serve his jail term and had completely finished paying his fine, when the judge tried to undo the prior illegal sentence by imposing the jail sentence alone, in effect attempting to withdraw the fine. The ' Lange Court said: “We are of the opinion that when the prisoner ... had fully suffered one of the alternative punishments to which alone the law subjected him, the power of the court to punish him further was gone.” 85 U.S. (Wall) at 176 (emphasis added). The Lange Court also observed that resentencing the prisoner to serve a year’s sentence when he had already served five days of that sentence was to impose a sentence of one year and five days, which also had the effect of punishing the criminal twice for the same offense.23 Stirring language about the necessity of preventing a criminal from being punished twice for the same offense must be read in this context. The finality under the double jeopardy clause, according to Ex parte Lange, was that which attached with respect to the sentence or portion thereof that had actually been served, not finality that attached as soon as executive custody began. See United States v. Busic, 639 F.2d 940, 951 (3d Cir.1981) (“the Court has applied the Double Jeopardy Clause only *1446when the defendant has satisfied the punishment imposed” (emphasis original)), cert. denied, 452 U.S. 918, 101 S.Ct. 3055, 69 L.Ed.2d 422 (1981).24
In United States v. Benz, 282 U.S. 304, 51 S.Ct. 113, 75 L.Ed. 354 (1931), the Supreme Court found no finality in an excessive sentence. Despite the fact that the prisoner was in executive custody, the Court found no constitutional impediment to reduction of his sentence. That is the salient fact situation in Benz, and its language suggesting that double jeopardy might attach were the sentence to be increased rather than decreased while an offender was in executive custody is dicta. The dicta was based on an expanded reading of Ex parte Lange, which as DiFrancesco noted “states no such principle.” 449 U.S. at 138, 101 S.Ct. at 438.25
The DiFrancesco Court duly and properly limited Benz and Lange to their narrow holdings. In turn and as would be expected, its holding also was narrow and limited to the facts before it. In framing the issue, however, the Court stated broadly:
[O]ur task is to determine whether a criminal sentence, once pronounced, is to be accorded constitutional finality and conclusiveness similar to that which attaches to a jury’s verdict of acquittal. We conclude that neither the history of sentencing practices, nor the pertinent rulings of the Court, nor even considerations of double jeopardy policy support such an equation.
Id. at 132, 101 S.Ct. at 435. DiFrancesco then went on to hold that no double jeopardy problem attached to a statutorily authorized appeal by the government of a sentence already being served by a prisoner, even though the appeal might result in an increased sentence.
In the current situation, there is no reason to apply a bright-line test of executive custody as the measure of finality for two reasons: the Supreme Court, in DiFrancesco, has done a convincing job of showing that Ex parte Lange and United States v. Benz compel no such result and of showing that no such bright line exists in a whole host of situations involving appeals, retrials, and revocation of probation. Such a range of exceptions seriously undercuts any argument that a criminal perceives the length of his sentence as determined once and for all when he begins to serve it. That range of exceptions can be extended easily to cover a situation where silence in a sentence creates no uncertainty and doubt in the prisoner’s mind as to the judge’s intent and where, therefore, no presumption regarding concurrent sentences need be invoked. A clarification of the judge’s sentence that is supported by the sentencing transcript constitutes but one part of a unitary and continuing sentencing proceeding. It does not unfairly subject the prisoner to the embarrassment, expense, and ordeal of a second trial — one of the primary objections to a second prosecution after acquittal. Cf. Swisher v. Brady, 438 U.S. 204, 98 S.Ct. 2699, 57 L.Ed.2d 705 (1978). Neither does it subject the prisoner to the anxiety of imposition of an improperly motivated or vindictive sentence that is stiffer than the first — an objection to increasing a sentence upon retrial. See North Carolina v. Pearce, 395 U.S. 711, 89 S.Ct. 2072, 23 L.Ed.2d 656 (1969).
The majority opinion cites to United States v. Jones, 722 F.2d 632 (11th Cir.1983), in support of the continuing validity of a bright-line test. United States v. Jones, however, involved a change in an explicit sentence, not a case of silence by one judge with respect to the sentence imposed by a previous judge. In Jones, the *1447sentencing judge erred in his understanding of the facts of the case and upon discovering his error decided that the prison term initially imposed had been too short. He attempted to increase the prison term, and the Eleventh Circuit found that the offender’s rights under the double jeopardy clause had been violated. The Jones court found that the resentencing did not impose multiple punishments for the same offense, since the entire time served under the first sentence (14 days) plus the time imposed under the resentencing (4 years) did not exceed the maximum permitted by Congress. See Jones, 722 F.2d at 638 n. 5. However, Jones held that the defendant’s legitimate expectations with respect to the length of his sentence had been frustrated by the resentencing and such expectations could not be prejudiced once the defendant had begun serving his sentence. In our case, there is no basis for asserting that Earley’s expectations legitimately extended to the firm belief that his drug sentences were concurrent with his earlier sentences.26 We also note that Jones appears to stand virtually alone alongside recent cases that “have generally found that a defendant has no reasonable expectation of finality at the time he begins to serve his sentence." United States v. Arrellano-Rios, 799 F.2d 520, 523 (9th Cir.1986).
The court in DiFrancesco commenced its double jeopardy analysis with the following observation about that clause: “That its application has not proved to be facile or routine is demonstrated by acknowledged changes in direction or emphasis.” Id. 449 U.S. at 127, 101 S.Ct. at 432. At the conclusion of its opinion the court made another revealing comment: “It has been observed elsewhere that sentencing is one of the areas of the criminal justice system most in need of reform.” Id. at 142, 101 S.Ct. at 440. Yet, in the face of that theme, the majority insists that DiFrancesco does not require “a change in the approach we have heretofore taken,” Maj. Op. at 1433, and installs, through bright-lines, a “facile or routine” approach.
The matter before us spanned a period of less than five months, beginning with a sentencing proceeding which the majority perceives as ambiguous (and I declare to be conclusive), and ending with the judge’s communication to the parole board. Constitutional considerations do not and should not attach to what at most could be described as an incomplete, therefore unclear, pronouncement of sentence. If necessary, the sentence is subject to the judge’s clarification. Throughout such a clarification process Earley would be protected by the right to constitutional due process. The majority instead discerns constitutional double jeopardy reliance on Earley’s part and a safe harbor of executive custody, with the result being constitutional immunity for Earley on double jeopardy grounds. I view that result as inimical to, not in furtherance of, constitutional principles. “Bright-lines” for constitutional themes more nearly reflect a judicial desire for simplicity and convenience than constitutional doctrine.
IV.
Finally, the majority argues that under a due process argument, the defendant Earley was entitled to a belief in the finality of his sentence at the time he appeared before the parole board on March 28, 1984 and was given a tentative release date. See *1448Maj. Op. at 1434 n. 8. Under parole hearings, however, a criminal always knows that a tentative release date is subject to change if additional evidence is obtained by the board that throws into question the basis for its judgment.27 Moreover, no “crystallized expectation” of early release can arise (1) where the sentencing transcript provides evidence that the intent of the sentencing judge was to sentence the prisoner to the maximum sentences allowed under the sentencing statute and (2) there was no need to invoke a presumption of concurrent sentences given the factual circumstances of this case. Cf. United States v. Lundien, 769 F.2d 981, 986-87 (4th Cir.1985) (no violation of due process where defendant had not “served so much of his sentence that his expectations as to its finality have crystallized and it would be fundamentally unfair to defeat them”; no double jeopardy violation either where clarification order enhanced sentence from ten to twenty years, even though defendant had been incarcerated for five days before enhancement), cert. denied, — U.S. -, 106 S.Ct. 815, 88 L.Ed.2d 789 (1986).
CONCLUSION
The distressing result in this case stems from linking a defective and shaky presumption to an overly rigid rule of double jeopardy analysis in sentencing situations. Clearly, Earley was not entitled to any expectation of constitutional finality in this proceeding.
What we have in the majority opinion is “the exaltation of form over substance,” something to be avoided, as noted in DiFrancesco. 449 U.S. at 142, 101 S.Ct. at 440. While the majority may have thought that such exaltation was based on federal court practice and precedent in other circuits, that practice and those precedents have not been critically examined in the few sentencing cases that have come before us. They have been substantially altered by the DiFrancesco ruling and undermined by examination of the legal genealogy of the presumption of concurrence. With this opinion, which is stubbornly rooted in the past, our circuit crosses paths with the Supreme Court and other circuits. Unfortunately, we are travelling in the wrong direction.
. The judge’s May 3, 1985 clarification order was to clarify "any ambiguity as to the court’s intention’’ and was made in response to a request. It is clear that the judge himself believed the sentencing transcript cleared up any ambiguity in the actual sentencing language.
. SJRep. No. 225, 98th Cong., 2d Sess. 126, reprinted in 1984 U.S.Code Cong. & Admin.News 3182, 3309.
. Ironically, just prior to quoting dicta from Subas as controlling law, the majority rejects, on grounds that it is dicta, language from our opinion in Scarponi v. United States, 313 F.2d 950 (10th Cir.1963), which undercuts the majority position that an ambiguous sentence is not an illegal sentence under Fed.R.Cr.P. 35(a). Maj.Op. at 1431.
. In 1970 Congress enacted D.C. Code provision title 23, section 112, which provides that a sentence imposed shall run consecutively to other sentences unless the sentencing court expressly provides otherwise.
. See Schultz v. United States, 384 F.2d 374 (5th Cir.1967); Aderhold v. McCarthy, 65 F.2d 452 (5th Cir.1933); Zerbst v. Lyman, 255 F. 609 (5th Cir.1919).
. See, e.g., H. Bross & A. Van Hirsch, Sentencing (1981); A. Campbell, Law of Sentencing, 249-50 (1978); M. Frankel, Criminal Sentences: Law Without Order (1972); G. Mueller, Sentencing: Process & Purpose (1979); J. Williams, Law of Sentencing & Corrections (1974).
. 18 U.S.C. § 3584, effective November 1, 1987.
. In Subas the judge sentenced the offender to seven years imprisonment on each of three counts, ‘‘[s]entence not to run concurrently.” The petitioner argued unsuccessfully that the reference in the singular to “sentence” meant that while the first and second prison terms were to run consecutively, the second and third terms were to run concurrently with one another. The petitioner also argued, equally unsuccessfully, that since the words "not to run concurrently” did not indicate the sequence in which the sentences were to be served they, therefore, must run concurrently.
. Lack of familiarity with the underlying sentence or multiple sentences is obviously one reason for distinguishing between sentences handed down at different times, and by different judges. The majority would doubtless point out that the sentencing judge here was made aware of the underlying sentence; therefore the reason for the distinction disappears. I would merely respond that the judge’s lack of familiarity is what obviously led to the problem before us, and rest my case. There is a further reason for the distinction, however: separate punishment for separate crimes committed at different times. Congress affixed penalties to the commission of federal crimes. A separate criminal conviction, at a different time, merits separate (consecutive) imposition of the punishment mandated by Congress, unless the court clearly declares that punishment should not be separately imposed (concurrent). That is the driving force behind the new legislation, and ought to be clear to us in this case even without the new statute on the subject.
. See Ex parte Green, 86 Cal. 427, 25 Pac. 21 (1890) (citing no cases in support); Ex parte Gafford, 25 Nev. 101, 57 Pac. 484 (1899) (same court imposing sentence at different times; citing no cases in support); Ex parte Black, 162 N.C. 457, 78 S.E. 273 (1913) (citing to Gafford and other inapposite cases in which the presumption had been invoked where multiple convictions had issued from the same court at the same time or during the same term). In none of these cases was their any indication that the result of the concurrence would be to allow the prisoner to escape serving the subsequent prison term.
. There are other cases dicussing the meaning of silence in such contexts as state court-federal court sentences and in the context of silence regarding the sequence in which consecutive sentences are to be served. These cases declined to apply the presumption. See, e.g., Milter v. Willingham, 400 F.2d 873 (10th Cir.1968); Mills v. Hunter, 204 F.2d 468 (10th Cir.1953); Hill v. United States, 186 F.2d 669 (10th Cir.1951); Williams v. Hunter, 165 F.2d 924 (10th Cir.1947); and Wall v. Hudspeth, 108 F.2d 865 (10th Cir.1940).
. See supra note 5.
. See supra note 4.
. Maj.Op. at 1431 n. 4, citing United States v. Naas, 755 F.2d 1133, 1136 (5th Cir.1985); United States v. Wenger, 457 F.2d 1082, 1084-86 (2nd Cir.), cert. denied, 409 U.S. 843, 93 S.Ct. 44, 34 L.Ed.2d 83 (1972); United States v. Bussey, 543 F.Supp. 981, 984 (E.D.Va.1982).
. See supra note 14.
. Doubt, if there is any, as to the meaning of the proceedings could be resolved by a further hearing at which all available relevant evidence on the subject, including the sentencing transcript, could be evaluated by either the same or another district judge, see, e.g., Downey v. United States, 91 F.2d 223 (D.C.Cir.1937).
. Whether the Double Jeopardy Clause applies to resentencing after confinement at all is a question which need not be confronted directly in this case. There is at least some doubt on the question. See United States v. Lundien, 769 F.2d 981 (4th Cir.1985).
. The jailhouse door rule announced by the majority is the first post-DiFrancesco holding on that point in this circuit. None of the cases cited by the majority so held. See United States v. Lawson, 670 F.2d 923 (10th Cir.1982); United States v. Preston, 634 F.2d 1285 (10th Cir.1980); and United States v. Davidson, 597 F.2d 230 (10th Cir.1979). It is noteworthy that in Lawson, we expressly left the question open, citing DiFrancesco. 670 F.2d at 929.
. Prior to its reformulation of the jailhouse door rule, which it would apply except in the face of a statute or rule, the majority acknowledges that other circuits no longer maintain such a rigid view of finality attaching at the point of executive custody. See Maj.Op. at 1433 n. 6. See also the majority’s discussion of United States v. Bishop, 774 F.2d 771 (7th Cir.1985) (court’s inherent power to correct a sentence induced by fraud extends beyond the line of executive custody). Thus, the majority must concede at the outset either that its "statute or rule” bright line is unique to our circuit, or else that it immediately recognizes exceptions to its own rule while avoiding an explanation as to why the constitution absolutely prohibits an exception under the circumstances of this case.
. United States v. Bello, 767 F.2d 1065 (4th Cir.1985); United States v. Jefferson, 714 F.2d 689 (7th Cir.1983), vacated on other grounds, — U.S. -, 106 S.Ct. 41, 88 L.Ed.2d 34 (1985); McClain v. United States, 676 F.2d 915 (2d Cir.), cert. denied, 459 U.S. 879, 103 S.Ct. 174, 74 L.Ed.2d 143 (1982); United States v. Busic, 639 F.2d 940 (3d Cir.), cert. denied, 452 U.S. 918, 101 S.Ct. 3055, 69 L.Ed.2d 422 (1981).
. In United States v. Bello, 767 F.2d 1065 (4th Cir.1985), the Fourth Circuit expressly rejected the "narrowest reading” of DiFrancesco to the effect "that the Double Jeopardy Clause does not preclude an increase in a sentence where the Government directly appeals that sentence pursuant to specific statutory authority.” Id. at 1069. Such a narrow reading is essentially the one proposed by the majority here. The Bello court opted for a broader interpretation of DiFrancesco, stating: "As discussed above, DiFrancesco directs the Court’s inquiry to whether the defendant had a legitimate expectation of finality as to the severity of his sentence, in order to determine whether an increase in the sentence is essentially a multiple punishment for the same offense.” Id. at 1070.
. It is not just the Subas case that has been extended beyond its borders to serve the purposes of subsequent courts. The Supreme Court in DiFrancesco made similar observations with respect to the misapplications of Ex parte Lange and United States v. Benz, See subsequent discussion.
. Later cases have indeed clarified that credit must be given for time served on an existing sentence when that sentence is later reimposed or its length increased after retrial. See, e.g., North Carolina v. Pearce, 395 U.S. 711, 89 S.Ct. 2072, 23 L.Ed.2d 656 (1969).
. The majority’s reliance on United States v. Jones, 722 F.2d 632 (11th Cir.1983) and United States v. Bishop, 774 F.2d 771 (7th Cir.1985) is misplaced. Both are relying on an offender’s legitimate expectation of finality as if it ordinarily attached at the moment of executive custody, a notion clearly undercut by a close reading of Ex parte Lange and DiFrancesco’s narrowing of Ex parte Lange.
. Our reliance on Benz in Owensby v. United States, 385 F.2d 58 (10th Cir.1967) (disallowing a resentencing where sentence was already "being served” since prisoner had been in executive custody for three days) seems clearly misplaced in the subsequent light of DiFrancesco. Therefore, Owensby does not serve as any kind of precedent for Earley.
. Four decisions that have imposed a bright-line test of executive custody when a judge attempted to clarify his prior silence can and should be distinguished from the case at hand. United States v. Naas, 755 F.2d 1133 (5th Cir.1985); Borum v. United States, 409 F.2d 433 (D.C.Cir.1967), cert. denied 395 U.S. 916, 89 S.Ct. 1765, 23 L.Ed.2d 230 (1969); Owensby v. United States, 385 F.2d 58 (10th Cir.1967); and Schultz v. United States, 384 F.2d 374 (5th Cir.1967). The latter three cases have been previously distinguished in the text of this dissent. In addition, all three are pre-DiFrancesco cases, a fact which provides additional grounds tor distinguishing them. Naas, although a post-DiFrancesco case, concerns a situation in which the sentencing judge forgot to indicate whether five sentences were to run concurrently with or consecutively to each other. In other words, the silence in Naas is the common type in which the presumption of concurrency is typically invoked and is unlike the factual situation before us. Also, Naas never discusses DiFrancesco, as the majority notes, a fact that weakens any persuasiveness it might otherwise have.
. In a similar vein, DiFrancesco noted that "[t]he Double Jeopardy Clause does not provide the defendant with the right to know at any specific moment in time what the exact limit of his punishment will turn out to be.” 449 U.S. at 137, 101 S.Ct. at 437.