665 F.3d 160 (2011)
UNITED STATES of America, Plaintiff-Appellee,
v.
Armarcion D. HENDERSON, Defendant-Appellant.
No. 10-30571.
United States Court of Appeals, Fifth Circuit.
December 15, 2011.
Cristina Walker, Robert Watts Gillespie, Asst. U.S. Attys., Shreveport, LA, for Plaintiff-Appellee.
Patricia Ann Gilley, Gilley & Gilley, Shreveport, LA, for Defendant-Appellant.
Before SMITH, SOUTHWICK and GRAVES, Circuit Judges.
PER CURIAM:
The court having been polled at the request of one of the members of the court, and a majority of the judges who are in regular active service and not disqualified not having voted in favor (FED. R.APP. P. 35 and 5TH CIR. R. 35), the petition for rehearing en banc is DENIED.
In the en banc poll, 7 judges voted in favor of rehearing (Stewart, Dennis, Elrod, Southwick, Haynes, Graves, and Higginson), and 10 judges voted against rehearing (Jones, King, Jolly, Davis, Smith, Garza, Benavides, Clement, Prado, and Owen).
HAYNES, Circuit Judge, joined by DENNIS, Circuit Judge, dissenting:
I respectfully dissent from the court's decision to deny rehearing en banc. Two issues raised by the panel's opinion merit the full court's attention: (1) the nature of the error that can be corrected under Federal Rule of Criminal Procedure 35(a); and (2) the timing of when the "obviousness" of plain error is judgedat the time of the error or at the time of the appellate decision.
I.
On the first issue, while the panel cites the appropriate standard"errors which would almost certainly result in a remand of the case to the trial court for further action"it applies this standard in a way that puts the opinion at odds with our own precedent, Watkins, and that of other circuits. Federal Rule of Criminal Procedure 35(a) provides that "[w]ithin 14 days after sentencing, the court may correct a sentence that resulted from arithmetical, technical, or other clear error." FED.R.CRIM.P. 35(a). The Advisory Committee's notes provide that "[t]he authority to correct a sentence under this subdivision is intended to be very narrow and to extend only to those cases in which an obvious error or mistake has occurred in the sentence, that is, errors which would almost certainly result in a remand of the case to the trial court ...." FED.R.CRIM.P. 35 advisory committee's note.[1]
The Advisory Committee's notes also explain that Rule 35(a) was intended to codify the results in United States v. Cook, 890 F.2d 672 (4th Cir.1989), and United States v. Rico, 902 F.2d 1065 (2d Cir.1990), subject to a more stringent time requirement (now 14 days) for correcting sentencing errors. FED.R.CRIM.P. 35 advisory committee's note; see also United States v. Ross, 557 F.3d 237, 239-41 (5th Cir.2009). In Cook, the appellate court upheld the district court's decision to amend a sentence that was not authorized under the sentencing guidelines as they existed at the time. 890 F.2d at 675. Similarly, in Rico, the Second Circuit upheld the district court's decision to correct a sentence that mistakenly applied a plea agreement. 902 F.2d at 1068. Thus, Rule 35(a) is intended to allow a district court to correct a sentence that was unlawful. See Cook, 890 F.2d at 675; Rico, 902 F.2d at 1068; FED.R.CRIM.P. 35 advisory committee's note. However, "[t]he subdivision is not intended to afford the court the opportunity to reconsider the application or interpretation of the sentencing guidelines or for the court simply to change its mind about the appropriateness of the sentence." FED.R.CRIM.P. 35 advisory committee's note.
Other than the panel's opinion, only one published Fifth Circuit case has addressed whether a Rule 35(a) motion can preserve error. See United States v. Watkins, 450 F.3d 184 (5th Cir.2006) (per curiam). In that case, the defendants filed a timely Rule 35(a) motion to raise their claim that application of a firearm adjustment to their sentences would violate their Sixth Amendment rights. Id. at 185. They had not raised that point of error before the district court announced their sentences. Id. Our court concluded that their Rule 35(a) motion was sufficient to preserve the error. Id.
In distinguishing Watkins, the panel looked beyond the facts set out in the Watkins opinion.[2] The underlying record in Watkins indicates that the defendants filed a Rule 35 motion because a Supreme Court case issued three days after their sentencing rendered their sentence unlawful. That Supreme Court opinion was issued during the period that the district court could have corrected its error under Rule 35. The panel distinguished Watkins because here, the Supreme Court's opinion in Tapia v. United States, ___ U.S. ___, 131 S.Ct. 2382, 180 L.Ed.2d 357 (2011), was issued after the case had already been appealed and after the fourteen-day time period during which the district court could have corrected the error had expired. Watkins itself, however, made no such distinction, and I do not think it is appropriate to "go behind" the published opinion to introduce facts not therein expressly relied upon.
Several other circuits have indicated that Rule 35 permits a district judge to correct errors of law. See Cook, 890 F.2d at 675 (noting that the district court could correct a sentencing error because the original sentence "was not a lawful one"); Rico, 902 F.2d at 1068 (upholding a sentencing modification because the original sentence was an "illegal sentence"); United States v. Himsel, 951 F.2d 144, 147 (7th Cir.1991) (noting that "the district judge had authority to vacate [a defendant's] first sentence if that sentence was illegal"); United States v. Quijada, 146 Fed.Appx. 958, 971 (10th Cir.2005) (unpublished) (concluding that a mistake or violation of the law was clear error). The panel opinion represents a divergence (if not a split) from those cases, worthy of the full court's consideration.
Moreover, it would seem odd not to interpret "clear error" to mean "legal error." If the district court could not correct a legal error, Rule 35's "other clear error" would seem to have little meaning since "arithmetical" and "technical" are already listed. If this court concludes that "clear error" means "legal error," then the district court would have had the authority to correct Henderson's sentence at the time Henderson filed his Rule 35(a) motion. Even under the law as it existed at the time of Henderson's Rule 35(a) motion, Henderson's sentence would likely have been considered unlawful. Certainly, Tapia makes clear that it is. Additionally, this is not a situation where the district court would have simply "changed its mind" or made a different discretionary call about Henderson's sentence, as it could have found that the sentence originally imposed was unlawful under 18 U.S.C. § 3582(a).
Thus, practically speaking, it makes little sense not to construe Rule 35 to permit correction of legal errors within the 14 day period. One could construe the panel opinion to mean that even though the district court realizes a legal error, the parties must still go through a time-consuming and expensive delay to fix it. Rule 35's strictures seem more directed to avoiding "flip-flopping" than to avoiding correction of legal errors. It would seem strange that a point of legal error actually raised to the district court and able to be ruled upon by that court while the court still was within the time for correcting the error[3] would be considered the same way as a point never raised at all in the district court. Therefore, I would recommend concluding that Henderson's Rule 35(a) motion preserved the point of error adequately, and this court should review Henderson's claim de novo. United States v. Oliver, 630 F.3d 397, 413 (5th Cir.2011) (noting that an issue raised and rejected in the district court is reviewed de novo), cert. denied, ___ U.S. ___, 132 S.Ct. 758, ___ L.Ed.2d ___, 2011 WL 3205411 (Nov. 28, 2011). Under a de novo standard of review, Henderson would be entitled to a new sentencing hearing.
II.
If the panel correctly determined that the Rule 35(a) motion did not preserve the error, then the question is raised whether the "obviousness" of the error made is judged at the time of the error or at the time of appeal. The panel opinion, with little discussion, concludes that error is judged at the time of the proceeding in question (here, pre-Tapia, at the sentencing hearing). In so doing, the panel opinion does not acknowledge either our intra-circuit split or the inter-circuit split on this question. To understand this issue, a bit of history is necessary. The Supreme Court established in Johnson v. United States, 520 U.S. 461, 468, 117 S.Ct. 1544, 137 L.Ed.2d 718 (1997) that "where the law at the time of trial was settled and clearly contrary to the law at the time of appeal[,] it is enough that an error be `plain' at the time of appellate consideration."
The Government argued in this case that "if the law at the time of trial is not settled, it is not enough that the error be plain at the time of appellate consideration." The Government cites no authority for its contention[4]; instead, it simply assumes that because the Supreme Court stated that "where the law at the time of trial was settled and clearly contrary to the law at the time of appeal[,] it is enough that an error be `plain' at the time of appellate consideration," Johnson, 520 U.S. at 468, 117 S.Ct. 1544, the converse of that statement must also be true. However, the issue is not nearly so clear, as the Supreme Court has left open the question of whether plain error would be established "where the error was unclear at the time of trial but becomes clear on appeal because the applicable law has been clarified," United States v. Olano, 507 U.S. 725, 734, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993), and our sister circuits have split over whether Johnson applies to the plain error analysis when the law was unclear at the time of trial and later becomes clear. The Ninth and District of Columbia Circuits hold that if the law is unclear at the time of trial and later becomes clear, the exception laid out in Johnson does not apply, and the error is evaluated based on the law as it existed at the time of trial. See, e.g., United States v. Gonzalez-Aparicio, 648 F.3d 749, 757 (9th Cir.2011) ("When the state of the law is unclear at the time of trial and is then clarified by subsequent authority, the district court's error is still not considered plain .... Therefore, plain error `normally means error plain at the time the district court made the alleged mistake.'"); United States v. Mouling, 557 F.3d 658, 664 (D.C.Cir.) ("We therefore hold that where, as here, the law was unsettled at the time of trial but becomes settled by the time of appeal, the general rule applies, and we assess error as of the time of trial."), cert. denied, ___ U.S. ___, 130 S.Ct. 795, 175 L.Ed.2d 559 (2009).
In contrast, the First, Second, Sixth, and Eleventh Circuits hold that Johnson applies whether the law was clear or unclear at the time of trial; the plainness of the error is always evaluated at the time of appellate review. See, e.g., United States v. Crosgrove, 637 F.3d 646, 656-57 (6th Cir.2011) ("However, the requirement that the error be plain means `plain under current law.' ... For plain error review, current law `is the law as it exists at the time of review.'"); United States v. Gamez, 577 F.3d 394, 400 (2d Cir.2009) (per curiam) ("A court of appeals cannot correct an error pursuant to Rule 52(b) unless the error is clear under current law .... Whether an error is `plain' is determined by reference to the law as of the time of appeal." (internal quotation marks and citations omitted)); United States v. Ziskind, 491 F.3d 10, 14 (1st Cir.2007) (citing Johnson for the proposition "that error is plain if the law is clear at the time of direct appellate review, even though governing law was unclear at time of trial"); United States v. Underwood, 446 F.3d 1340, 1343 (11th Cir.2006) (noting that "even though the error was not plain at the time of sentencing, the subsequent issuance of [a Supreme Court opinion] establishes that the error is plain at the time of appellate consideration").
We have not previously squarely addressed this issue where the timing of when the "plainness" was judged was critical; however, our decisions are in something of a disarray on this point. Several opinions, including the panel's opinion in this case, have held that the court considers the law at the time of trial when determining whether an error is plain. See, e.g., United States v. Henderson, 646 F.3d 223, 225 (5th Cir.2011) ("[A]n error is plain only if it was `clear under current law at the time of trial.'" (quoting United States v. Jackson, 549 F.3d 963, 977 (5th Cir.2008))); United States v. Garcia-Rodriguez, 415 F.3d 452, 455 (5th Cir.2005) ("`Plain' is synonymous with `clear' or `obvious,' and at a minimum, contemplates an error which was clear under current law at the time of trial."); United States v. Hull, 160 F.3d 265, 272 (5th Cir.1998) (same). Other opinions have concluded that Johnson established that the court considers the error at the time of appeal in deciding whether it is plain. See, e.g., United States v. Bishop, 603 F.3d 279, 281 (5th Cir.) ("We determine whether an alleged error is plain by reference to existing law at the time of appeal."), cert. denied, ___ U.S. ___, 131 S.Ct. 272, 178 L.Ed.2d 180 (2010); United States v. Gonzalez-Terrazas, 529 F.3d 293, 298 (5th Cir.2008) ("[T]he error need only be plain at the time of appellate consideration."). We have not squarely addressed the precise question of whether Johnson applies to the plain error analysis when the law was unclear at the time of trial and later becomes clear.
Our earliest discussion of this issue applying the Olano formulation of plain error (decided prior to Johnson), judged the error at the time of appeal. United States v. Knowles, 29 F.3d 947 (5th Cir.1994). Where "two previous holdings or lines of precedent conflict, the earlier opinion controls and is the binding precedent in this circuit." United States v. Wheeler, 322 F.3d 823, 828 n. 1 (5th Cir.2003) (internal quotation marks omitted). The panel opinion fails to address Knowles in light of this precedent or reconcile our conflicting precedents. For this reason, I recommend en banc consideration of this issue to provide clarity on when plain error should be evaluated.
If our court were to follow the First, Second, Sixth, and Eleventh Circuits and hold that plain error is always evaluated at the time of appeal, the district court's opinion would be reversed. The Supreme Court's decision in Tapia establishes that at the time of appeal, the district court's error was plain. I submit that Henderson easily meets the other requirements for plain error because the district court, in granting a longer sentence, considered a factor the Supreme Court has stated is an impermissible consideration.
Whichever way this court ultimately would come out on the "timing issue," it is worthy of the full court's attention. We deal almost daily with issues of plain error, and it is certainly not an unusual occurrence for a claim of plain error to be made where the law was unclear at the time of the trial court's decision but is clear by the time of appeal. See, e.g., United States v. Newson, No. 11-10073, 2011 WL 5599112, at *1, 2011 U.S.App. LEXIS 23181, at *2-3 (5th Cir. Nov. 17, 2011) (per curiam) (unpublished) (noting that "the lack of a published opinion from this court at the time of the district court proceedings rendered any [error] neither clear nor obvious legal error"); United States v. Gloria, 2011 WL 3966101, at *1-2, 2011 U.S.App. LEXIS 18589, at *4 (5th Cir. Sept. 7, 2011) (per curiam) (unpublished) (addressing a different sentencing issue and judging plain error at the time of sentencing); United States v. Graves, 409 Fed.Appx. 780, 781 (5th Cir.2011) (per curiam) (unpublished) (addressing such a situation on habeas review and noting that "it is enough that the error be plain at the time of appellate consideration" (internal quotation marks omitted)). How we address such a situation should be uniform. Without doubt, Henderson was sentenced based upon an impermissible consideration. Given the discord within our own circuit (and that among our sister circuits), I submit that the full court should resolve this question. Because it fails to do so here, I respectfully dissent.
NOTES
[1] This note addresses former Rule 35(c); however, the substantive provisions of Rule 35(c) were moved to subsection (a) in 2002. Therefore, earlier analyses of subsection (c) now apply to subsection (a).
[2] In discussing this issue, Watkins stated: "The defendants were sentenced before the mandatory provision of the Sentencing Guidelines were modified and rendered advisory by the United States Supreme Court in [Booker].... The defendants first raised their Sixth Amendment claim in a timely Fed.R.Crim.P. 35(a) motion, after the district court had orally pronounced the defendants' sentences. We conclude that they preserved the error." 450 F.3d at 185.
[3] The sequence of events in this case was that the sentencing hearing took place, eight days later Henderson filed his motion, two days later the court entered a written judgment of conviction and sentence containing the sentence announced at the oral hearing. Thereafter, the court ordered briefing on the Rule 35(a) motion before concluding that it could not grant the motion because, by the time it ruled, more than fourteen days had passed. Whether or not that is true, I conclude that the district court erred by not granting the motion within the fourteen days in light of Henderson's timely motion.
[4] The Government's citation to Puckett v. United States, 556 U.S. 129, 129 S.Ct. 1423, 173 L.Ed.2d 266 (2009), adds little to the discussion, as Puckett did not address when plain error is evaluated. The Government concludes that because Puckett states that for an error to be plain, it "must be clear or obvious, rather than subject to reasonable dispute," id. at 1429, that means that "if the law is unclear at the time of trial ... the reviewing court considers the status of the law at the time of trial."