concurring:
I agree with the result reached by the court in this case. However, I am unable to agree with all of the reasoning given by the court and therefore state my reasons for concurrence briefly and separately. A fuller exposition of my understanding of Double Jeopardy law is set forth in United States v. Smith, 929 F.2d 1453 (10th Cir.1991) (McKay, J., dissenting). It is therefore unnecessary to repeat it in this context. Furthermore, I believe the result reached by the majority is clearly correct under any Double Jeopardy analysis.
I.
As the court has indicated, Double Jeopardy analysis in the sentencing context requires us to determine whether the appellant in this case had a legitimate expectation of finality in his original sentence. See United States v. DiFrancesco, 449 U.S. 117, 137, 139, 101 S.Ct. 426, 438, 66 L.Ed.2d 328 (1980). It is now well settled that if a defendant successfully attacks a sentence or conviction, he has no expectation of finality in that particular sentence or conviction. The appellant in this case successfully challenged his sentence in a prior appeal.
In the first appeal, this court held that the original sentencing subjected the appellant to Double Jeopardy because the prosecution and the trial judge had segmented the appellant’s conduct into four purported crimes when, for Double Jeopardy purposes, that conduct can produce a sentence only as a single crime. The total sentence originally imposed for the four crimes was seven years, while the statutory maximum sentence for a single crime encompassing the same conduct is five years.
Our prior order vacated all four of the original sentences and directed the trial court to resentence the appellant on a single count which encompassed all the conduct for which he had been convicted. Thus, the requirement of the Double Jeopardy clause — that the appellant have no *919legitimate expectation of finality in the sentence — was satisfied by the appellant’s successful challenge to, and our subsequent vacation of, all his prior sentences.
On remand, the trial court resentenced the appellant to five years, the statutory maximum for a single count. This was permissible because nothing in our prior order invalidated the court’s power to sentence for all of the conduct for which it had previously sentenced the defendant. We held only that the trial court was limited to treating that conduct as a single crime with its attendant maximum.
This is a case in which the majority’s “package deal” analysis can be legitimately applied, because the trial court was specifically allowed to resentence the appellant for all the conduct of which he was convicted. The court merely “repackaged” that conduct as one crime instead of four. It is not necessary to consider in this case what limitations the Double Jeopardy clause may properly impose upon the application of the “package deal” metaphor to resentencing cases.
II.
Under the circumstances of this case, however, I believe the appellant might have attacked his resentencing as a violation of Due Process under the principles set forth by the Supreme Court in North Carolina v. Pearce, 395 U.S. 711, 89 S.Ct. 2072, 23 L.Ed.2d 656 (1979). While the appellant did not raise this issue, it bears discussion. Pearce holds that, under the Due Process clause, “vindictiveness” against a defendant may play no part in resentencing. 395 U.S. at 725, 89 S.Ct. at 2080. In order to show the absence of vindictiveness, Pearce requires that when a judge resentences a defendant to a harsher punishment, she must affirmatively state her reasons for doing do. Id. at 726, 89 S.Ct. at 2081.
When the trial court resentenced the appellant in this case to a longer sentence on one count than he had previously received on any single count, the court stated:
The original sentence was for a total of seven years and the reason for the seven years was because at the time the court was under the impression that there were four counts and that the Court needed to sentence on all four counts. I believe that the totality of the circumstances at the time deserved a seven years sentence and that is why the court entered a seven year sentence back in 1982. The Court is restricted and the Court respects the order of the Circuit Court and is limited to a five year sentence, which is the maximum that the Court can impose under count 1 and therefore imposes it.
Record, vol. 3, at 39-40.
While “vindictiveness” may encompass a retaliatory motive against a defendant, see Pearce, 395 U.S. at 725, 89 S.Ct. at 2080, I believe it may also encompass a desire on the part of the trial judge to vindicate his original sentence. Thus, the trial court’s comments in this case, at the very least, raise some Due Process questions.
I conclude that Pearce is not violated in this case, but that is only because this case differs from most cases over which there has been a dispute about the application of Pearce. In this case, we did not hold any of the grounds on which the original sentence was based to be invalid considerations in sentencing. Had we done so, I would not allow the use of the “package deal” approach to allow the trial court to impose longer sentences on the remaining grounds in order to vindicate his original sentence.
Additionally, because the appellant achieved almost 100 percent of his objectives in his original appeal (five years instead of seven, where he sought four years instead of seven), I do not believe that the appearance of vindictiveness, discussed in Pearce, arises. Id. In any event, despite my reservations, I clearly would not find the imposition of the longer sentence to be plain error. United States v. Forester, 874 F.2d 983 (5th Cir.), cert. denied, — U.S. —, 110 S.Ct. 284, 107 L.Ed.2d 264 (1989).*
*920III.
While I would approach several of the authorities cited by the majority opinion differently, I only note one difference specifically, although I view it to be dicta in the majority opinion. I cite it because I believe that the provision has been widely and uniformly used in a context for which I do not believe it was intended. I think that it is important to record my disagreement with that general usage.
In its analysis of Double Jeopardy finality, the majority cites 18 U.S.C. § 3568. My own view is that this provision is irrelevant to Double Jeopardy analysis. I have set forth in Smith a fuller analysis of why that is true. Suffice it to say that the text makes clear the provision was enacted solely to define when a sentence begins to run for purposes of calculating when it is completed. I do not believe this provision addresses finality at all.
An overarching principle of this court’s decisions in United States v. Earley, 816 F.2d 1428 (10th Cir.1987), and United States v. Villano, 816 F.2d 1448 (1987), is that an unambiguous oral sentence is sufficiently final that the court cannot alter its terms by a subsequent written order. See Villano, 816 F.2d at 1452-53 (“It is incumbent upon a sentencing judge to choose his words carefully so that the defendant is aware of his sentence when he leaves the courtroom.”). In my view, a judgment is final from the time it is pronounced unless one of the established impediments — such as the existence of a legitimate ground on which the government can and does base an appeal, or the defendant’s own successful appeal of a particular sentence or conviction — upsets the defendant’s expectation of finality. This rule is simpler to apply than the majority view and is consistent with our en banc decisions in Earley and Villano.
In order to abbreviate this separate opinion, I only note further that my views on the matter of *920the applicability of Pearce to resentencing after a prior sentence has been successfully challenged on appeal most closely parallel those of Judge Rubin, dissenting in United States v. Vontsteen, 910 F.2d 187 (5th Cir.1990).