Veteranclaims’s Blog

December 20, 2021

Panel Application; 38 C.F.R. § 3.655; Hyson v. Brown, 5 Vet.App. 262, 265 (1993), which imposed a duty on VA under 38 C.F.R. § 3.655 to show, before suspending benefits, that the veteran lacked good cause to miss a physical examination scheduled in conjunction with a compensation claim;

UNITED STATES COURT OF APPEALS FOR VETERANS CLAIMS
No. 17-3766
ROCCO V. PERCIAVALLE, APPELLANT,
V.
DENIS MCDONOUGH,
SECRETARY OF VETERANS AFFAIRS, APPELLEE.
On Appellee’s Motion for Full Court Review
(Argued September 22, 2020 Decided December 3, 2021)
Kenneth H. Dojaquez, of Topeka, Kansas, for the appellant.
Nathan P. Kirschner, with whom James M. Byrne, General Counsel; Richard J. Hipolit,
Deputy General Counsel; Mary Ann Flynn, Chief Counsel; Christopher W. Wallace, Deputy Chief
Counsel; and Shereen M. Marcus, Appellate Attorney, all of Washington, D.C., were on the brief
for the appellee.
Before BARTLEY, Chief Judge, and PIETSCH, GREENBERG, ALLEN, MEREDITH,
TOTH, FALVEY, LAURER, and JAQUITH, Judges.
LAURER and JAQUITH, Judges, announced the judgment of the Court and filed the
opinion of the Court with respect to Parts I-IV, in which BARTLEY, Chief Judge, and PIETSCH,
GREENBERG, and TOTH, Judges, joined; and filed an opinion in Part V. TOTH, Judge, filed an
opinion concurring with Part V. ALLEN, Judge, filed an opinion concurring in the judgment, in
which MEREDITH and FALVEY, Judges, joined. BARTLEY, Chief Judge, filed an opinion
dissenting in the judgment, in which PIETSCH and GREENBERG, Judges, joined.
LAURER and JAQUITH, Judges: U.S. Army veteran Rocco V. Perciavalle appeals,
through counsel, a September 18, 2017, decision by the Board of Veterans’ Appeals (Board)
finding no clear and unmistakable error (CUE) in a 1971 rating decision and denying, as a matter
of law, his motion to revise the rating decision. In a panel decision issued on September 27, 2019,
the Court reversed the Board’s determination that his CUE motion was prohibited as a matter of
law because it depended on a changed interpretation of law, and the Court remanded for further
adjudication. Perciavalle v. Wilkie, 32 Vet.App. 59 (2019). The Secretary timely moved for
2
reconsideration or, in the alternative, full court review. Based on this motion, the case was
submitted to the Court for en banc consideration.1
The dispute centers on whether the appellant’s CUE motion was based on a change in the
interpretation of law and on whether any Board error in deciding that question was prejudicial. A
majority of the Court agrees that the Board was wrong in denying Mr. Perciavalle’s appeal as a
matter of law. However, a different majority concludes that the proper remedy here is to still affirm
the Board decision.2 In our view, despite Board error in characterizing a later interpretation of an
existing regulation as the only basis on which appellant asserted CUE and denying his CUE
assertion as a matter of law, the only possible outcome was the Board’s ultimate conclusion that
VA did not commit CUE in the 1971 rating decision. So we would affirm the September 2017
Board decision for lack of prejudicial error.
I. BACKGROUND
Mr. Perciavalle served honorably in the Army from 1962 to 1964 as a field artillery unit
commander. R. at 525. He injured his left knee during service playing football at Fort Sill,
Oklahoma. Record (R.) at 514, 535. In May 1966, the veteran submitted a claim for disability
compensation based on his knee injury, which had required surgery. R. at 534-37. He described
the nature of the injury as “removed cartilage anterior cruciate ligament & removed bone chip.” R.
at 535. In August 1966, a physician conducted an orthopedic examination as part of the veteran’s
disability evaluation. The veteran said that his left knee was “not back to normal,” had a “tendency
to give away at times,” and bothered him going up and down stairs. R. at 512. The physician
described the original diagnosis as “a torn left medial meniscus and injury to the left anterior
cruciate ligament and stellate fragmentation of cartilage of [the] left femoral medial condyle,” and
the surgery as a “left medial meniscectomy.” R. at 514. The physician diagnosed postoperative
residuals and moderate anteroposterior (AP) relaxation. Id. He found that Mr. Perciavalle’s left
knee did not present “lateral relaxation, tenderness, crepitus, swelling, effusion, or limitation of
1 The panel decision of September 27, 2019, is vacated, and the appeal is reinstated and resolved by this
opinion.
2 Judges Laurer and Jaquith file this opinion. Chief Judge Bartley and Judges Pietsch and Greenberg join
Parts I-IV only. Judge Toth joins Parts I-IV and concurs with Part V. Accordingly, this opinion states the view of the
Court with respect to Parts I-IV, and it states the nonplurality view of two judges with respect to Part V. Judges Allen,
Meredith, and Falvey concur in the result.
3
motion,” and noted that his leg had normal power and reflexes, he could squat fully, he walked
without a limp, and his left thigh had improved from 1 inch smaller in circumference than the right
to ¼ inch smaller. Id.
In September 1966, VA awarded service connection for the residuals of the veteran’s left
knee injury. R. at 509. The rating decision stated: “Current VA exam shows weakness and feeling
of [instability] of left knee. The operative scar is asymptomatic. There is moderate AP relaxation.
There is slight fullness of the infrapatellar fat pad. No other functional impairment is shown.” Id.
Mr. Perciavalle’s disability was given a 10% rating for a medial meniscectomy under Diagnostic
Code (DC) 5259. Id.
The veteran had another disability examination in July 1971. He reported some discomfort
and weakness ascending and descending steps. R. at 496. An x-ray report reflected: “Left Knee:
The joint space is questionably narrowed medially and there does appear to be some slight blunting
of the tibial spines. On one view there is a question of nodular irregularity of the medial condyle
of the femur.” R. at 492. The physician found that the veteran’s left knee had normal extension,
flexion to 135 degrees, and no swelling or functional deficit, and that he had no quadriceps atrophy
or weakness. R. at 495. However, the physician noted “very slight instability of [the] joint
laterally.” Id. He diagnosed minimal residual effects of the veteran’s injury. R. at 496.
The evaluation report was referred to a rating board, which promptly determined and
advised Mr. Perciavalle that: “The evidence does not warrant any change in the previous
determination. Your knee condition remains 10% disabling and you will continue to receive
compensation for this disability.” R. at 490. The veteran did not appeal this decision, and it became
final.
In February 1994, the Court decided Esteban v. Brown, holding that separate ratings were
warranted for disfiguring facial scars, painful facial scars, and facial muscle damage resulting in
mastication problems because “none of the symptomatology for any one of these three conditions
is duplicative of or overlapping with the symptomatology of the other two conditions.” 6 Vet.App.
259, 262 (1994).
In July 1997, VA’s General Counsel (GC) issued a precedent opinion holding that “[a]
claimant who has arthritis and instability of the knee may be rated separately under diagnostic
4
codes 5003 and 5257.” VA Gen. Couns. Prec. 23-97 (July 1, 1997), https://www.va.gov/ogc
/opinions/1997precedentopinions.asp; see R. at 55-56.
In March 2015, the veteran asked the VA regional office (RO) to revise the July 1971
decision, asserting that it contained CUE. R. at 101-102. He contended that he should have
received two separate 10% ratings for his knee disability, one “for slight instability of his left knee
under Diagnostic Code (DC) 5257 and a separate 10% for limitation of motion of flexion and
discomfort (pain) secondary to arthritis under DC 5003-5260.” R. at 101-02.
In September 2015, the RO denied the CUE motion, determining that no revision of the
July 1971 decision was warranted “because the decision was properly based on the available
evidence of record and the rules in effect at the time the issue was considered.” R. at 94. The
appellant filed his Notice of Disagreement that same month. R. at 76-77. A May 2017 Statement
of the Case (SOC) continued the denial of the CUE motion because the 1971 rating decision
predated the VA GC opinion that allowed VA “to assign separate evaluations for instability and
limitation of motion.” R. at 58. Mr. Perciavalle appealed to the Board 3 days later. R. at 25.
In August 2017, the veteran submitted a letter brief to the Board, reiterating his position
that he should have been awarded separate ratings in 1971 for his left knee. R. at 13-15. He
emphasized that his theory of CUE was not based on a change of law during the 1990s; it was
based on the incorrect application of the law as it existed when the rating decision was made in

  1. Id.
    In September 2017, the Board relied on 38 C.F.R. § 20.1403(e) to deny Mr. Perciavalle’s
    CUE motion, saying:
    The Veteran has not provided any evidence that, in July 1971, VA
    interpreted the rating schedule to allow for separate ratings for limitation of motion
    and instability of the same knee. Instead, the Veteran contends that a more recent
    interpretation of VA regulations should have retroactive effect. . . . Because a later
    interpretation of an existing regulation cannot constitute CUE and that is the only
    basis on which the Veteran asserts CUE, the Veteran’s motion must be denied as a
    matter of law.
    R. at 5-6. This appeal followed.
    On September 27, 2019, a panel of this Court reversed and vacated the Board decision,
    remanding the matter for further adjudication. Perciavalle, 32 Vet.App. at 59. The panel
    determined that neither Esteban nor the VA GC opinion was a change in interpretation because a
    changed interpretation necessarily requires the existence of an antecedent interpretation from
    5
    which a later interpretation departs, and no such prior interpretation existed. The panel also found
    the Board’s error prejudicial because it prevented Mr. Perciavalle from receiving a meaningful
    opportunity to participate in the adjudicative process.
    On October 18, 2019, the Secretary filed a motion for reconsideration or, in the alternative,
    full court review. The panel denied reconsideration, and on May 11, 2020, the case was submitted
    to the Court for en banc consideration.
    II. PARTIES’ ARGUMENTS
    Mr. Perciavalle argues that there was CUE in the 1971 RO decision and the Board’s
    contrary conclusion was “arbitrary, capricious, and not in accordance with the law.” Appellant’s
    Brief (Br.) at 2. He contends that the Board erred in finding that Esteban and the VA GC opinion
    changed the law because the law in 1971 already mandated a second separate rating for his arthritis
    or limitation of motion. Appellant’s Br. at 2-5. In his responsive briefing, he asserts that his CUE
    challenge accords with the plain meaning of the CUE regulation, which rejects CUE motions based
    on changes in the interpretation of the law, not new or novel interpretations. And he maintains that
    his CUE motion based on the law in existence when the rating decision was made is distinct from
    and comports with the relevant caselaw implementing the regulation in cases where there was a
    change from one formal interpretation of the law—in a regulation, court decision, or VA manual—
    to another. Id.
    The Secretary argues that the veteran’s claim is precluded as a matter of law because he is
    seeking “to re-litigate and re-weigh the facts at the time of the prior final rating decision anew and
    hold the 1971 rating decision to a later change in regulatory interpretation,” both of which are
    impermissible in the context of a CUE motion. Secretary’s Br. at 10, 15. He adds that “novelty
    does not require a change, and a ‘change’ is not necessary in the context of a CUE motion” to defeat
    it because the new interpretation cannot be applied retroactively. Secretary’s Motion at 5. The
    Secretary also argues that any error in the Board decision was not prejudicial because even if it
    6
    was possible that the RO could have granted separate ratings in 1971, the evidence of record at the
    time cannot show an undebatable outcome-determinative error. Id. at 6-10.
    III. LEGAL LANDSCAPE
    A. Clear and Unmistakable Error
    By statute and regulation, final decisions by both VA and the Board are subject to reversal
    or revision on the grounds of CUE.3 38 U.S.C. §§ 5109A, 7111; 38 C.F.R. §§ 3.105 (2020),
    20.1403 (2020). Because Mr. Perciavalle asserts that there is CUE in a VA rating decision, his
    motion falls under 38 U.S.C. § 5109A and 38 C.F.R. § 3.105. By statute, a veteran may seek to
    revise or reverse a final decision for CUE at any time and, if successful, receive retroactive benefits
    as if VA granted the claimed benefit in the prior decision. See George v. McDonough (George II),
    991 F.3d 1227, 1233 (Fed. Cir. 2021), aff’g George v. Wilkie (George I), 30 Vet.App. 364 (2019),
    petition for cert. filed, No. 21-234 (U.S. Aug. 13, 2021). To do so, a claimant must allege with
    specificity the error VA committed in its rating decision. Lang v. Wilkie, 971 F.3d 1348, 1352
    (Fed. Cir. 2020); see, e.g., Andre v. Principi, 301 F.3d 1354, 1361 (Fed. Cir. 2002) (“[E]ach
    3 The term “clear and unmistakable error” first appeared in veterans regulations in 1928 that subjected only
    the decisions of rating boards to review for CUE. Smith v. Brown, 35 F.3d 1516, 1524-25 (Fed. Cir. 1994). In 1955,
    the regulation was amended to apply to rating boards and other agencies of original jurisdiction (AOJs). Id. at 1525.
    The CUE regulation was recast as § 3.105 in 1959 and amended and published in the new Part 3 of Title 38 of the
    C.F.R. in 1961. Id.; 26 Fed. Reg. 1561, 1569 (Feb. 24, 1961). Section 3.105 continued to apply to decisions by AOJs—
    VA ROs—but not to decisions by the Board. Smith, 35 F.3d at 1526-27.
    In 1997, § 3.105 was codified at 38 U.S.C. § 5109A, and 38 U.S.C. § 7111 was enacted so that Board
    decisions could be reviewed for CUE. Disabled Am. Veterans (DAV) v. Gober, 234 F.3d 682, 687 (Fed. Cir. 2000),
    overruled on other grounds by Nat’l Org. of Veterans’ Advocs., Inc. v. Sec’y of Veterans Affs., 981 F.3d 1360 (Fed.
    Cir. 2020). In 1999, 38 C.F.R. § 20.1403 was adopted as one of the regulations implementing section 7111. 64 Fed.
    Reg. 2134, 2135-36 (Jan. 13, 1999). The content of § 20.1403 was derived from rulings by the Court because Congress
    intended that “VA adopt the [Court’s] interpretation of the term ‘clear and unmistakable error'” and “not alter the
    standard for evaluation of claims of clear and unmistakable error.” 64 Fed. Reg. at 2136. See DAV, 234 F.3d at 693
    (“Congress intended to extend the principles underlying 38 C.F.R. § 3.105(a) to [section] 7111, which governs CUE
    review of Board decisions.”). In DAV, the Federal Circuit noted that § 20.1403 was a reasonable implementation of
    section 7111 and the Court’s decisions regarding CUE, and it found § 20.1403 to be valid. Id. at 697-98.
    In 2019, § 3.105(a) was updated to mirror the language in § 20.1403(e). 84 Fed. Reg. 138, 142, 167-68 (Jan.
    18, 2019). Based on this history of the development of the two regulations, their coincident principles and provisions,
    and the subsumption that occurs when decisions on issues by an RO have been decided by the Board on appeal, see
    38 C.F.R. § 3.105(a)(5) and Manning v. Principi, 16 Vet.App. 534, 540-41 (2002), aff’d, 85 F. App’x. 216 (Fed. Cir.
    2004), the Court often cites both. Neither party has suggested that §§ 3.105 and 20.1403(e) have been or should be
    interpreted differently, or that either the Board’s citation to § 20.1403 or the 2019 modification of § 3.105(a) impacts
    our analysis today.
    7
    ‘specific’ assertion of CUE constitutes a claim that must be the subject of a decision by the [Board]
    before the Veterans Court can exercise jurisdiction over it.”).
    However, “CUE proceedings are fundamentally different from direct appeals,” as they are
    a limited statutory exception to the rule of finality. Robinson v. Shinseki, 557 F.3d 1355, 1360
    (Fed. Cir. 2009). A CUE motion is a collateral attack of an adjudicated claim that was not appealed
    and became final and binding. Bustos v. West, 179 F.3d 1378, 1380 (Fed. Cir. 1999). When
    challenged, the final decision benefits from a strong presumption of validity. See Fugo v. Brown,
    6 Vet.App. 40, 44 (1993). And the challenger bears an “extra-heavy” burden of persuasion in
    showing error in the final decision. Berger v. Brown, 10 Vet.App. 166, 169 (1997).
    The elements of CUE are well established: (1) “[e]ither the correct facts, as they were
    known at the time, were not before the adjudicator or the statutory or regulatory provisions extant
    at the time were incorrectly applied”; (2) the asserted error must be “undebatable, so that it can be
    said that reasonable minds could only conclude that the original decision was fatally flawed at the
    time it was made,” and not merely “a disagreement as to how the facts were weighed or evaluated”;
    and, (3) to give rise to the need to revise the prior decision, the error must have been one that
    “manifestly changed the outcome at the time it was made.” Russell v. Principi, 3 Vet.App. 310,
    313-14 (1992) (en banc); see Abbott v. O’Rourke, 30 Vet.App. 42, 46–47 (2018). “CUE is a very
    specific and rare kind of ‘error'” of fact or law that makes it “absolutely clear that a different result
    would have ensued” without the error. Fugo, 6 Vet.App. at 43-44.
    Review for CUE in a prior final decision “must be based on the evidentiary record and the
    law that existed when that decision was made.” 38 C.F.R. § 3.105(a)(1)(iii); 38 C.F.R.
    § 20.1403(b)(1). The corollary to that requirement is that CUE “does not include the otherwise
    correct application of a statute or regulation where, subsequent to the decision being challenged,
    there has been a change in the interpretation of the statute or regulation.” 38 C.F.R.
    § 3.105(a)(1)(iv); see 38 C.F.R. § 20.1403(d)(e). Though finality is important to promoting the
    efficient resolution of veterans’ claims, it yields to correcting CUE by collateral attack because
    such corrections conform the result “to the ‘true’ state of the facts or the law that existed at the time
    8
    of the original adjudication.” Russell, 3 Vet.App. at 313. So, the focal point of the CUE landscape
    is the state of the facts and the law when the rating decision was made.
    B. Standard of Review
    “When the Court reviews a Board determination that there was no CUE in a prior final
    decision, the Court ‘cannot conduct a plenary review of the merits of the original decision'” by VA.
    Stallworth v. Nicholson, 20 Vet.App. 482, 487 (2006) (quoting Archer v. Principi, 3 Vet.App. 433,
    437 (1992)). Rather, the Court reviews the Board’s decision on CUE to determine whether it was
    “‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law'” or
    unsupported by adequate reasons or bases. Abbott, 30 Vet.App. at 47 (quoting 38 U.S.C.
    § 7261(a)(3)). That standard of review “contemplates de novo review of questions of law.” Kent
    v. Principi, 389 F.3d 1380, 1384 (Fed. Cir. 2004). Legal questions reviewed anew include whether
    there was a failure to apply pertinent statutes and regulations or a failure to apply them correctly.
    Id.; see Simon v. Wilkie, 30 Vet.App. 403, 408 (2018); Joyce v. Nicholson, 19 Vet.App. 36, 43
    (2005). The Court also reviews de novo a Board determination of whether a veteran presented a
    valid CUE allegation as a matter of law. Andrews v. Principi, 18 Vet.App. 177, 182 (2004), aff’d
    sub nom. Andrews v. Nicholson, 421 F.3d 1278 (Fed. Cir. 2005); Phillips v. Brown, 10 Vet.App.
    25, 30 (1997). So we review the Board’s denial of Mr. Perciavalle’s motion “as a matter of law”
    under the nondeferential de novo standard. See Phillips, 10 Vet.App. at 30.
    C. Rating Separate Disabilities
    Since its inception in 1921, the rating schedule has included separate provisions for
    evaluating knee instability, limitation of leg extension, and meniscal problems
    [and] . . . no portion of the rating schedule pertaining to the knees [has] included an
    express prohibition on separate evaluation of those manifestations of disability,
    despite numerous amendments.
    Lyles v. Shulkin, 29 Vet.App. 107, 115 (2017).4 In contrast, “[t]he rating schedule is replete with
    rules that prohibit separate evaluation of other disabilities.” Id. at 114. Four of the six examples of
    such prohibitions cited in Lyles have been in effect since 1964. See 38 C.F.R. §§ 4.96 (respiratory
    conditions), 4.113 (abdominal conditions), 4.114 (digestive conditions), and 4.115 (heart diseases
    4 Lyles illustrates the superficiality of the judgment concurrence’s claim that citation to post-1971 authority
    taints, corrupts, or is unfaithful to proper plain meaning circa-1971 analysis. See post at 39-40; cf. id. at 41 (“Such
    citations could certainly provide context.”). Later opinions obviously were not available to rating officials in 1971,
    but it does not follow that we must discard the opinions’ historical surveys or replicate them without citation as if they
    are uniquely ours. The information surveyed was there for rating officials in 1971—and for 50 years beforehand—as
    well as for the Board in 2017, though the latter plainly ignored it.
    9
    and nephritis) (1964). Given this history of expressly forbidding separate evaluation of other
    disabilities, “the lack of an express bar to separate evaluation of knee disabilities in [38 C.F.R.]
    § 4.71a must be read as a deliberate decision to permit separate evaluation.” Lyles, 29 Vet.App. at
    115; see Buczynski v. Shinseki, 24 Vet.App. 221, 227 (2011) (explaining that including limitations
    in one DC but not in a related DC indicates that their omission was intentional).
    “[I]f an injury or disease manifests with two different disabilities, then two separate ratings
    should be awarded.” Tropf v. Nicholson, 20 Vet.App. 317, 321 (2006); see 38 C.F.R. § 4.25(b)
    (2020). Section 4.25(b) provides: “Except as otherwise provided in [the rating] schedule, the
    disabilities arising from a single disease entity . . . are to be rated separately as are all other
    disabling conditions, if any.” Subsection (b) was added to § 4.25 in 1989, 54 Fed. Reg. 27,161,
    27,162 (June 28, 1989), but § 4.25 has contemplated combining separate ratings for different
    disabling conditions since its inception. 29 Fed. Reg. 6718, 6720-21 (May 22, 1964). Another
    general policy in rating disabilities—also present since 1964—prescribes “avoidance of
    pyramiding.” 38 C.F.R. § 4.14 (1964). Section 4.14 begins: “The evaluation of the same disability
    under various diagnoses is to be avoided,” and ends: “[T]he evaluation of the same manifestation
    under different diagnoses [is] to be avoided.” 5 Id. Properly read, § 4.14 and § 4.25 are
    complementary provisions instructing VA to rate only distinct manifestations separately. See
    Fanning v. Brown, 4 Vet.App. 225, 230 (1993). “Of course, it is possible for a veteran to have
    separate and distinct manifestations attributable to two different disability ratings, and, in such a
    case, the veteran should be compensated under different diagnoses.” Id. The anti-pyramiding rule
    means “that the rating schedule may not be employed as a vehicle for compensating a claimant
    twice (or more) for the same symptomatology[ because] such a result would overcompensate the
    claimant for the actual impairment of his [or her] earning capacity.” Brady v. Brown, 4 Vet.App.
    203, 206 (1993).
    D. Esteban and the VA General Counsel’s Opinion
    The Court’s 1994 Esteban decision synthesized Brady and Fanning in holding that distinct
    and separate symptomology arising from the same injury should be rated separately.6 Esteban,
    5 Section 4.14 also provides: “Disability from injuries to the muscles, nerves, and joints of an extremity may
    overlap to a great extent, so that special rules are included in the appropriate bodily system for their evaluation.” There
    is no argument that a special rule applies here.
    6 In Esteban, the veteran’s injuries to his face in a motor vehicle accident resulted in disfigurement, painful
    10
    6 Vet.App at 261-62. Nearly 3 decades before Esteban—and before this Court and the U.S. Court
    of Appeals for the Federal Circuit were created—the U.S. Court of Claims7 considered VA’s
    Schedule for Rating Disabilities and its pyramiding prohibition in the context of an Air Force
    colonel’s claim for disability retirement pay at a higher rating. In Wolf v. United States, the Court
    of Claims held that separate ratings for resections of the large and small intestines did not constitute
    improper pyramiding because the manifestations of the resections—the symptoms produced by
    them—were separate and distinct. 168 Ct. Cl. 24, 31 (1964).
    In 1997, the VA GC issued the precedent opinion to address the question whether a
    claimant could receive separate ratings for arthritis and instability of the knee under Diagnostic
    Codes 5003 and 5257, respectively.8 G.C. Prec. 23-97; see R. at 55-56. After advising that she
    knew of “no formal position taken by the Veterans Benefits Administration on this issue,” the GC
    answered the question affirmatively. G.C. Prec. 23-97, at 1. She observed that “DC 5257 provides
    for evaluation of instability of the knee without reference to limitation of motion,” and conversely,
    DC 5003 provides for rating arthritis on the basis of limitation of motion without reference to knee
    instability. Id. Noting that the pyramiding prohibition in § 4.14 applied to the evaluation of the
    same disability or manifestation under different diagnoses, the GC concluded: “[s]ince the plain
    terms of DC 5257 and 5003 suggest that those codes apply either to different disabilities or to
    scars, and facial muscle damage resulting in problems with mastication—symptomatology the Court found separate
    and distinct rather than duplicative or overlapping, and thus warranting separate ratings. 6 Vet.App. at 259-62.
    7 When the Federal Circuit was created in 1982, the judges of the Court of Claims continued in office as
    judges for the Federal Circuit. Federal Courts Improvement Act of 1982, Pub. L. No. 97-164, § 165, 96 Stat. 25, 50.
    8 That question was before the Board on remand from the Court in the nonprecedential case McAdory v.
    Brown, 15 Vet.App. 387 (table), 1997 WL 288699 (May 2, 1997), and other unspecified cases. In McAdory, the Court
    held that the Board “erred in not addressing whether, under Esteban, . . . symptomatology for a rating under DC 5257
    is duplicative or overlapping with the symptomatology for a rating under DC 5010 and DC 5003.” 1997 WL 288699,
    at *2. The Board had considered “all possible diagnostic codes which [might] be applicable,” found that rating the
    veteran’s knee disability under DC 5257 was “most appropriate,” and concluded that an “increased rating [was] not
    supported by the evidence of record.” Id.
    11
    different manifestations of the same disability, the evaluation of knee dysfunction under both codes
    would not amount to pyramiding under section 4.14.” G.C. Prec. 23-97, at 2.
    IV. THE BOARD’S ERRORS
    A. Misstating the Veteran’s CUE Motion
    As previously noted, the VA GC opinion confirmed: “A claimant who has arthritis and
    instability of the knee may be rated separately under diagnostic codes 5003 and 5257.” R. at 56.
    In the case before us, the Board considered that holding a change in interpretation that could not
    be the basis for a CUE challenge, per § 20.1403(e).9 R. at 6. There are several problems with that
    determination. The most obvious error is that the Board, in its penultimate sentence,
    mischaracterized the veteran’s argument. The Board wrote: “Because a later interpretation of an
    existing regulation cannot constitute CUE and that is the only basis on which the Veteran asserts
    CUE, the Veteran’s motion must be denied as a matter of law.” R. at 6. The veteran’s actual
    argument was that the law had always permitted separate evaluations and that the CUE was
    incorrectly applying the law that existed at the time of the July 1971 rating decision. R. at 13-14.
    The veteran’s March 2015 CUE motion, submitted on his behalf by a non-attorney
    representative, contended that the rating specialist committed CUE in 1971 by failing to assign, in
    addition to the veteran’s existing 10% disability rating, a separate 10% rating under DC 5003-5260
    for limitation of motion secondary to “degenerative changes (arthritis)” in his left knee. R. at 101-
  2. He also argued that the DC for his original 10% rating should have been changed from 5259
    (for symptomatic cartilage removal) to 5257 based on instability, but specifically sought corrective
    action based on CUE only to add the separate 10% rating under DC 5003-5260. Id. The motion
    specifically cited 38 C.F.R. § 3.105, Court cases setting forth the test for determining whether there
    was CUE in a rating decision, and the evidence he contended entitled him to the additional rating:
    “X-ray evidence dated July 8, 1971, compared to the X-ray evidence dated August 5, 1966, clearly
    show[ed] degenerative changes in the veteran’s left knee.” R. at 102. In May 2017, the RO denied
    Mr. Perciavalle’s motion based upon the RO’s assertion that VA was not allowed to assign separate
    9 As previously indicated, supra note 3, 38 C.F.R. § 3.105, rather than 38 C.F.R. § 20.1403, is the governing
    regulation for the veteran’s claim of CUE in a VA rating decision, but the distinction is not significant in this appeal.
    12
    evaluations for instability and limitation of motion until the VA GC opinion was issued.10 R. at
    58.
    In August 2017 the same non-attorney representative submitted a letter to the Board that
    said:
    We are not arguing the 1997 General Counsel opinion or any VA rules after
    the 1971 rating decision. We contend that the law has always permitted that a
    separate evaluation can be applied for DC 5257 based on instability and a separate
    evaluation can be applied for loss of range of motion under DC 5003-5260 or due
    to pain/functional loss under 38 CFR §§ 4.40, 4.59. All regulations that applied to
    this claim have not materially changed since 1971. . . . [W]e contend that corrective
    action should be taken under the provisions of 38 CFR § 3.105(a) to correctly apply
    the law that was in existence at the time of the July 22, 1971 rating decision.
    R. at 13-14 (emphasis in original). The Board’s assertion that “a later interpretation of an existing
    regulation” was “the only basis on which the Veteran assert[ed] CUE” was plainly wrong.
    The letter by the veteran’s representative also stated that “there [was] no bar to the
    retroactive effect of Esteban” because it did not change the law or lead to any change in the rating
    schedule or regulations. R. at 14. The judgment concurrence uses that statement to overwrite the
    veteran’s specific contention that the 1971 law required corrective action. Then it assesses the
    veteran’s CUE motion as otherwise inadequately pled. Neither position is supported by our caselaw
    or consonant with the veteran-friendly system Congress created. See Henderson v. Shinseki,
    562 U.S. 428, 440-41 (2011); Cacciola v. Gibson, 27 Vet.App. 45, 58 (2014).
    B. Failing to Sympathetically Construe the Veteran’s Motion
    The Government’s interest in veterans cases is that justice be done, and the systemic
    fairness essential for securing justice includes a duty to construe veterans’ submissions
    sympathetically, especially when the veteran is self-represented or represented by someone who
    is not a licensed attorney. See Comer v. Peake, 552 F.3d 1362, 1367-69 (Fed. Cir. 2009). This
    principle of sympathetic construction applies to CUE motions. Roberson v. Principi, 251 F.3d
    1378, 1384 (Fed. Cir. 2001). Roberson requires the RO and the Board to fully and sympathetically
    develop a veteran’s CUE motion to its optimum before deciding it. Andrews, 421 F.3d at 1282-83.
    That task requires a high degree of care in the context of the requirement that CUE motions be
    10 The RO noted that the VA GC opinion stemmed from an unnamed Court of Appeals for Veterans Claims
    decision from 1997. R. at 58. It is unclear whether the RO was mistakenly referring to Esteban (which was decided in
    1994), or to McAdory, the 1997 nonprecedential decision remanding for the Board to address the impact of Esteban
    on knee symptomatology.
    13
    pled with specificity because denial of a CUE motion may foreclose any further attempt at revision.
    Acciola v. Peake, 22 Vet.App. 320, 326 (2008). In this case, the Board completely failed to fulfill
    its requirement to sympathetically construe the CUE motion submitted on Mr. Perciavalle’s behalf
    by his non-attorney representative.
    By recasting the veteran’s motion as based solely on the retroactive application of a change
    in the law, the Board turned the sympathetic construction requirement upside down. And the
    judgment concurrence endorses the Board’s maneuver by purporting to divine the veteran’s true
    intention and attempting to fill gaps in the Board’s decision, and the Secretary’s defense of it, to
    prop the decision up. As the partial concurrence explains, a threshold problem is that the veteran’s
    sole reference to retroactivity was in his representative’s letter to the Board, responding to the RO’s
    decision; it was not presented to the RO and thus not properly before the Board. See Jarrell v.
    Nicholson, 20 Vet.App. 326, 332 (2006) (en banc). Moreover, even if the Board could reach the
    retroactivity issue, it was not free to liberally construe the veteran’s pleadings unsympathetically
    by subsuming a valid CUE theory in one it finds invalid. Such subsumption violates the basic
    requirement to decide all CUE issues specifically raised. See Comer, 552 F.3d at 1367 (finding
    that VA is obligated to determine claims raised by the evidence).
    Sympathetically construed as written (rather than as reimagined), the veteran’s CUE motion
    was not subject to denial “as a matter of law.” That determination is reserved for allegations that
    could not possibly constitute CUE, such as a CUE motion based on a breach of the duty to assist,
    see Robertson v. Shinseki, 26 Vet.App. 169, 176 (2013); a second CUE challenge for a single
    claim, see Hillyard v. Shinseki, 24 Vet.App. 343, 354 (2011), aff’d, 695 F.3d 1257 (Fed. Cir. 2012);
    a CUE allegation that relies on a later change in the law, see Lamb v. Peake, 22 Vet.App. 227, 235
    (2008); and a motion challenging an RO decision subsumed by a Board decision. See Jones v West,
    250 F.3d 753 (Fed. Cir. 2000); Manning, 16 Vet.App. at 541.
    The judgment concurrence finds no fumble by the Board because it believes that the
    veteran’s CUE allegations are so completely deficient as to be “non-existent.” See post at 42. It
    notes that 38 C.F.R. § 20.1404(b) provides for dismissal of deficient CUE motions “without
    prejudice to refiling.”11 And it faults the Court for precluding a related future CUE motion. But its
    targeting is off, for it is the Board that accomplished what the judgment concurrence decries. The
    11 Since the veteran alleged CUE in the rating decision, the governing regulation actually is 38 C.F.R.
    § 3.105(a)(1)(vii)(B), which sets forth pleading requirements that correspond with those in 38 C.F.R. § 20.1404(b).
    14
    Board preclusively denied Mr. Perciavalle’s CUE motion, it did not dismiss the motion without
    prejudice. The Federal Circuit has made clear that “CUE assertions that fail to satisfy pleading
    requirements must be dismissed without prejudice to refiling, and may not be denied.” Simmons v.
    Principi, 17 Vet.App. 104, 114 (2003) (citing DAV, 234 F.3d at 699). The Court went on to specify
    the Board’s responsibility, holding that “the proper remedy for the Board, when confronted with
    an inadequately [pled] CUE claim, . . . is to dismiss that challenge without prejudice.” Simmons,
    17 Vet.App. at 114. By neither dismissing the veteran’s motion without prejudice nor adjudicating
    it on its merits, the Board improperly imposed the worst of all possible outcomes for the veteran—
    a preclusive denial “as a matter of law.”
    C. Failing To Adjudicate the Veteran’s Challenge to the Application of the 1971 Law
    The Board’s words and deed show that it did not find that the veteran’s CUE motion failed
    the specific pleading requirement, so the judgment concurrence’s addition of this alternative
    explanation is misplaced. Moreover, a sympathetic reading of the actual CUE motion shows that
    it asserts a cognizable challenge that was due a decision on its merits.
    The judgment concurrence’s contrary contention flows from its conflation of the pleading
    requirements and the appellant’s burden of persuasion. See post at 38-42. Our colleagues want
    claimants to develop and supply “esoteric legal arguments” citing cases, using terms of art like
    “implausible,” and analyzing regulations and contemporaneous interpretations of them to explain
    how the CUE elements are satisfied. These points are excellent as suggestions, but they do not
    reflect the lower threshold veterans must clear to get their motions adjudicated and would set too
    stingy a standard for pleadings by veterans and their non-attorney representatives.
    “[A] sympathetic reading of a CUE motion requires the Secretary to fill in omissions and
    gaps that an unsophisticated claimant may leave in describing his or her specific dispute of error
    with the underlying decision.” Acciola, 22 Vet.App. at 326–27. Put another way, “a sympathetic
    reading of the CUE theory may result in clarifying modifications.” Canady v. Nicholson¸
    20 Vet.App. 393, 402 (2006). And CUE allegations that are “rough but recognizable” must be
    adjudicated. Acciola, 22 Vet.App. at 326.
    The discussion of Berger in the judgment concurrence illustrates the conflation problem.
    See post at 38-42. In Berger, neither the Board nor the Court sanctioned the sort of summary action
    imposed on Mr. Perciavalle. Both the Board and the Court went beyond their determination that a
    case that invalidated a portion of a regulation could not be applied retroactively to demonstrate
    15
    CUE. The Berger Board and then the Court each analyzed both the facts of the case and the law
    that applied when the benefits decision was made by the RO and determined, on the facts
    presented, that the adjudicator’s actions accorded with the plain meaning of the statute, regulations,
    and manual in effect at that time. 10 Vet.App. at 167-70. Mr. Perciavalle was not afforded the
    same consideration of his motion. But Mr. Perciavalle’s CUE motion did echo the CUE theory
    Berger endorsed—alleging that “the RO’s interpretation of the plain meaning of the law was clearly
    and unmistakably erroneous,” 10 Vet.App. at 170—when his motion is read sympathetically, as
    required.
    Lamb makes the distinction between pleading and persuasion even clearer. In Lamb, the
    veteran’s CUE motion—prepared by his attorney—rested, in part, on Hyson v. Brown, 5 Vet.App.
    262, 265 (1993), which imposed a duty on VA under 38 C.F.R. § 3.655 to show, before suspending
    benefits, that the veteran lacked good cause to miss a physical examination scheduled in
    conjunction with a compensation claim
    . Lamb, 22 Vet.App. at 233. The regulation in effect in
    1957, when Mr. Lamb’s benefits were suspended, did not expressly impose such a burden on VA.
    Id. The CUE motion also alleged that the veteran had good cause for missing the examination, so
    VA suspended his benefits “without observance of applicable law and regulation.” Id. at 234. The
    Court observed: “Charitably interpreted, appellant’s allegations of CUE are twofold and are stated
    in slightly different terms,” so they “must be addressed separately.” Id. The Board had dismissed
    Mr. Lamb’s CUE motion for insufficiency of pleading, but the Court held that the Board erred by
    not addressing the two aspects of the motion on their merits. Id. However, the Court found that the
    error in dismissing the challenge based on Hyson was not prejudicial because the “correct
    application of a statute or regulation as it was understood at the time does not retroactively become
    CUE where, subsequent to the decision challenged, there has been a change in the interpretation
    of the statute or regulation.” Id. at 235. But the Court reversed “the Board’s decision dismissing a
    claim of CUE based on a failure to correctly apply regulations in effect” at the time of the decision,
    and the Court remanded for adjudication of that CUE contention on its merits. Id. at 237. In doing
    so, the Court observed:
    The appellant also alleges that VA failed to properly apply regulations . . . .
    The Court notes that he does not support his argument about the meaning of “good
    cause” with any authority that existed in 1957 that demonstrates that the regulation
    was understood as he suggests. Nor has the appellant even cited any current
    authority supporting his interpretation of “good cause.” However, at this juncture,
    16
    that is not the Court’s concern. The issue before the Court is whether the allegation
    of CUE was pled with sufficient specificity so that, assuming its truth and legal
    viability, it would satisfy the three elements required to find that the challenged
    final decision was clearly and unmistakably erroneous. Unlike appellant’s reliance
    on Hyson discussed above, there is no legal impediment to revising the 1957 RO
    decision if VA were to find that the RO had failed to properly apply the applicable
    regulations that were in effect at that time. Such a decision, however, is not for the
    Court in the first instance but for VA.
    Id. at 235-36. The Board cited Lamb, R. at 6, but obviously did not follow it. That was clear error.
    And the judgment concurrence also missed Lamb’s requirement that the Board decide, on their
    merits, allegations of CUE based on a failure to correctly apply extant regulations.
    Without addressing the application of Lamb to Mr. Perciavalle’s more detailed contention
    that the RO committed CUE in the application of the regulations to his knee condition, the
    judgment concurrence complains that the appellant’s allegation that the knee regulations permitted
    separate ratings in 1971 was inadequate. But the appellant’s argument was accurately responsive
    to the error that the RO made (and the Board endorsed) in asserting that separate ratings were not
    allowed for instability and limitation of motion of the same knee. R. at 5. The appellant went
    further, contending that the RO’s failure to assign separate ratings was CUE that required corrective
    action R. at 14, 101-02. And he argued that “VA regulations in 1971 mandated that separate ratings
    be assigned for limitation of motion and instability.” Appellant’s Br. at 2.
    In short, Lamb and Berger emphasize that Mr. Perciavalle was entitled to a merits decision
    on his contention that the rating specialist committed CUE in 1971 by incorrectly applying the
    regulatory diagnostic codes in effect at that time to specified medical evidence and failing to assign
    a separate 10% disability rating for degenerative changes and limitation of motion in his left knee.
    D. Incorrectly Barring the Veteran’s Motion Because of a Later Change in Interpretation
    “[C]hanges in the law subsequent to the original adjudication . . . do not provide a basis for
    revising a finally decided case.” Russell, 3 Vet.App. at 313. However, as the panel decision
    detailed,12 the VA GC opinion did not constitute a “change in the interpretation” of the rating
    regulations (within the meaning of § 3.105(a)(1)(iv) and § 20.1403(e)). The statutes and
    regulations governing CUE do not define “change in the interpretation.” Without such a definition,
    we employ the fundamental canon of textual construction, giving words “‘their ordinary,
    12 Perciavalle, 32 Vet.App. at 64-66.
    17
    contemporary, common meaning,'” which we may ascertain from general dictionaries. Sandifer v.
    U.S. Steel Corp., 571 U.S. 220, 227 (2014) (quoting Perrin v. United States, 444 U.S. 37, 42
    (1979)). In its basic definition, “change” means “to make different in some particular; alter;
    transform; to give a different position, course, or direction to; to replace with another; to make a
    shift from one to another; switch . . . .” MERRIAM-WEBSTER’S COLLEGIATE DICTIONARY 206 (11th
    ed. 2014). Another dictionary provides an even simpler formulation, defining “change” as “make
    or become different,” and presents an example of the term used with a direct object: “a proposal
    to change the law.” NEW OXFORD AMERICAN DICTIONARY 289 (3d ed. 2010) (emphasis in
    original). As a third dictionary describes it, “change denotes a making or becoming distinctly
    different and implies either a radical transmutation of character or replacement with something
    else.” WEBSTER’S NEW WORLD DICTIONARY, COLLEGE EDITION 244 (1968) (emphasis in original).
    It follows that a “change in the interpretation” necessarily requires the existence of a prior
    interpretation that is made different, whether modified or replaced in whole; it cannot be the “first
    commentary” on a regulation or statute, as there exists no prior interpretation that the latter
    modifies, alters, or replaces. See Jordan v. Nicholson, 401 F.3d 1296, 1298 (Fed. Cir. 2005) (noting
    that VA’s regulation interpreting 38 U.S.C. § 1111 constituted its “initial interpretation” of that
    statute). Although an initial interpretation constitutes a change from having none at all, this is
    obviously not the same as changing a preexisting interpretation.
    Before analyzing the question presented—whether the rating schedule authorized separate
    ratings for arthritis and instability of the knee—the VA GC opinion readily acknowledged that this
    was the first Agency interpretation, noting that: “We know of no formal position taken by the
    Veterans Benefits Administration on this issue.” G.C. Prec. 23-97, at 1. This point is dispositive:
    there can be no “change” in interpretation absent an existing interpretation from which a later
    interpretation deviates. In glossing over this admission, the Board erred in finding that the VA GC
    opinion constituted a change in interpretation when the opinion was rendered based on the premise
    that no antecedent interpretation existed.
    Moreover, the text of the VA GC opinion makes clear that it arises from and reflects the
    plain terms of the applicable diagnostic codes and § 4.14. The VA GC opinion addresses each
    diagnostic code in detail, highlights the operative language of § 4.14, and then declares: “Since the
    plain terms of DC 5257 and 5003 suggest that those codes apply either to different disabilities or
    to different manifestations of the same disability, the evaluation of knee dysfunction under both
    18
    codes would not amount to pyramiding under section 4.14.” G.C. Prec. 23-97, at 2. The VA GC
    opinion notes that Esteban also “supports the availability of separate ratings.” Id.
    As that notation suggests, our holding in Esteban likewise cannot be deemed a change in
    interpretation. On its own terms, Esteban did not purport to introduce a definitive interpretation of
    § 4.14 but merely held that the Board’s denial of a higher rating did not square with its concession
    that the veteran had three different disabling conditions—which meant that separate ratings did
    not constitute pyramiding under the plain language of the regulation. 6 Vet.App. at 260-61.
    Esteban did not cite any formal interpretation by VA, only the factual findings that showed the
    veteran’s disabilities had no duplicative or overlapping symptomatology, the regulatory language,
    and earlier cases from this Court that read the regulation the same way when applied to similar
    facts, discussing Fanning, 4 Vet.App. at 231, and to dissimilar ones, discussing Brady, 4 Vet.App.
    at 206-07. Esteban, 6 Vet.App. at 261-62. In short, Esteban neither proposed an interpretation of
    a regulation nor cited any existing interpretation; it merely applied the express provisions of
    §§ 4.14 and 4.25 to the facts presented.
    Esteban thus presents a marked contrast to our rulings in Lamb and Berger. In Lamb,
    benefits were suspended based on a regulation, and a later court decision imposed a burden of
    proof not contained in the regulation’s terms. The Court rejected a portion of Mr. Lamb’s CUE
    motion because the requirement imposed by the later case did not reflect the law at the time of the
    benefits decision. 22 Vet.App. at 234-35. Berger likewise involved a CUE motion that could not
    succeed to the extent it was based on a later case that changed the law by invalidating a portion of
    a regulation. 10 Vet.App. at 169-70 (citing Gregory v. Brown, 5 Vet.App. 108, 112 (1993)).
    Esteban was critically different in that it neither added a new burden of proof or other requirement
    to a regulation nor subtracted or invalidated a portion of one.
    The other cases cited by the Secretary likewise involve clear evidence of a change in the
    interpretation of the law. In George I, there were three changes in how the presumption of
    soundness regulation was interpreted between the 1977 Board decision claimed to have been beset
    by CUE and the 2016 Board decision denying that CUE motion: a 2003 VA GC precedent opinion
    invalidating the regulation,13 a 2004 decision by the Federal Circuit finding that the regulation did
    13 See VA Gen. Couns. Prec. 3-2003 (July 16, 2003), https://www.va.gov/ogc/opinions/2003precedent
    opinions.asp.
    19
    not fulfill statutory requirements,14 and the promulgation of an amended regulation in 2005.
    George I, 30 Vet.App. at 367-69. Though the Federal Circuit’s 2004 interpretation in Wagner set
    forth what the statute had always meant, both this Court and the Federal Circuit later held that a
    first judicial interpretation of the statute could not be the basis for finding CUE where the
    regulation in existence at the time of the original decision imposed a different rule that VA
    adjudicators were bound to follow. George I, 30 Vet.App. at 375; George II, 991 F.3d at 1234-35;
    see also Joyce v. Nicholson, 443 F.3d 845, 848 (Fed. Cir. 2006) (“Where the regulations in
    existence at the time of the original decision imposed a different rule, Wagner cannot be the basis
    for a CUE claim.”). The George II court echoed Jordan 15 in declaring that the appellants’
    argument—that their CUE motions were not premised on a change in the law—failed to appreciate
    that the regulation provided the initial interpretation of the statute, regardless of any inaccuracies
    in that interpretation subsequently identified in the later case. George II, 991 F.3d at 1235.
    A recent nonprecedential decision by the Federal Circuit, Steele v. McDonough, 856
    F. App’x 878 (Fed. Cir. 2021), followed George in concluding that a veteran could not rely on its
    holding in Saunders v. Wilkie, 886 F.3d 1356 (Fed. Cir. 2018)—that pain alone may constitute
    functional impairment, and thus a compensable disability—to demonstrate CUE in contrary prior
    Board and RO decisions. The Steele court explained that “[w]e expressly noted in George that our
    precedent does not support the view that a new judicial pronouncement can retroactively apply to
    final decisions of the VA.” 856 F. App’x. at 881.16 Though Steele declined to decide whether
    14 Wagner v. Principi, 370 F.3d 1089 (Fed. Cir. 2004).
    15 Jordan involved the same presumption of soundness regulation—an initial interpretation of the statute—
    the same change in interpretation via the 2003 VA GC opinion, and the same result: the Federal Circuit likewise held
    that applying the regulation in effect at the time of the Board decision could not be CUE. 401 F.3d at 1297-99. The
    Jordan court declared that, notwithstanding the later conclusions about the statutory requirements by VA’s GC and
    Wagner, “the accuracy of the regulation as an interpretation of the governing legal standard does not negate the fact
    that the regulation did provide the first commentary on [the statute], and was therefore the initial interpretation of [that
    statute].” Id. at 1298.
    16 The panel in Steele also said: “We recently clarified in George v. McDonough that whether an interpretation
    is a first interpretation or a change in existing interpretation, it cannot serve as a basis for CUE.” Steele, 856 F. App’x
    at 881. That statement must be read in context. George was talking about the first judicial interpretation of a statute
    that resulted in the court invalidating a regulation that constituted the initial interpretation of the statute and the one
    that was in effect when the rating decision was made. George, 991 F.3d at 1235. George was not referring to the initial
    interpretation of the statute, but a new interpretation that changed the initial one—whether the new interpretation was
    by the Agency or a court. Id. (“Jordan does not differentiate between new agency interpretations and new judicial
    interpretations.”). Moreover, George illustrated that we should exercise caution in applying the Federal Circuit’s
    nonprecedential decisions, completely rejecting the holding by a panel in Patrick v. Nicholson, 242 F. App’x 695 (Fed.
    Cir. 2007), that Wagner could be the basis for CUE in an earlier decision. George, 991 F.3d at 1231.
    20
    Saunders was a change in interpretation, Saunders expressly rejected the holding in Sanchez-
    Benitez v. West (Sanchez-Benitez I), 13 Vet.App. 282, 285 (1999), vacated in part, appeal
    dismissed in part sub nom. Sanchez-Benitez v. Principi (Sanchez-Benitez II), 259 F.3d 1356 (Fed.
    Cir. 2001), that pain alone could not qualify as a disability. Saunders, 886 F.3d at 1365-66.17
    Nothing like that happened in connection with the knee regulations at issue here.
    Most importantly, Mr. Perciavalle’s CUE motion does not depend on the retroactive
    application of Esteban or the VA GC opinion. His case is the flip side of George II, Jordan, and
    Steele, as well as Lamb and Berger. “VA does not commit clear and unmistakable error in a
    benefits claim decision when it faithfully applies a regulation as it existed at the time of decision,
    even if that regulation is later revised or invalidated.” George II, 991 F.3d 1227. But it does commit
    CUE if fidelity to the regulation is lacking and the error is both undebatable and outcome
    determinative, even if that regulation is later validated and affirmed. In addition, however a
    subsequent interpretation of the law is characterized—as new, a change, or an authoritative
    statement—a motion alleging that VA’s interpretation of the plain meaning of the law at the time
    of the rating decision was clearly and unmistakably erroneous states a valid CUE challenge. See
    Berger, 10 Vet.App. at 170. The CUE statutes and regulations protect and promote the faithful
    application of the rules as they existed at the time of the rating decision and the systemic reliance
    on the blend of fairness and finality. See Russell, 3 Vet.App. at 313. It would be nonsensical—and
    antithetical to the CUE regulations—for the later validation of the rating rules to foreclose
    accountability for undebatable error in previously applying them.18
    E. Misunderstanding Separate Disabilities or Manifestations from a Single Injury
    The Board decision and the Secretary’s brief evince a misunderstanding that may explain
    how we got here. The Board declared: “The evaluation of the same disability or manifestation
    17 “In Saunders, the Federal Circuit overruled nearly 20 years of precedent when it held that ‘pain in the
    absence of a presently-diagnosed condition can cause functional impairment,’ which may qualify as a ‘disability’ under
    section 1110. 886 F.3d at 1368 (overruling Sanchez-Benitez [I, 13 Vet.App. at 285]).” Wait v. Wilkie, 33 Vet.App. 8,
    14 (2020); see Saunders, 886 F.3d at 1359 (In the decision being reversed, “The Veterans Court noted that it has
    applied the legal holding of Sanchez–Benitez I more than 100 times since that opinion issued, and that it has relied
    upon or affirmed the Board’s application of this legal principle at least 83 times.”). In 1999, our Court had interpreted
    the statutory use of the word “disability” and the consideration of pain in disability ratings and declared “that pain
    alone, without a diagnosed or identifiable underlying malady or condition, does not in and of itself constitute a
    disability.” Sanchez-Benitez I, 13 Vet.App. at 285; see Spurgeon v. Brown, 10 Vet.App. 194, 196 (1997) (noting that
    the rating schedule does not provide for a separate rating for pain).
    18 There is no call to turn and face the strange when there are no changes. Cf. DAVID BOWIE, Changes, on
    HUNKY DORY (RCA 1971).
    21
    under different diagnoses constitutes pyramiding and must be avoided,” citing § 4.14, and then
    suggests that Esteban and the VA GC opinion changed the interpretation of that regulation. R. at
    5-6. The Secretary argues, in support of his conclusion that the veteran’s appeal “is precluded as a
    matter of law,” Secretary’s Br. at 15, that: “at the time of the 1971 decision, any evaluation of the
    same disability or manifestation under different diagnoses constituted pyramiding,” Secretary’s Br.
    at 4; “38 C.F.R. § 4.14, as it was understood in 1971, precluded the ability to award separate
    evaluations of the same disability or the same manifestation under various diagnoses,” Secretary’s
    Br. at 13; and, “prior to the 1994 decision in Esteban and the 1997 GC opinion, 38 C.F.R. § 4.14
    was interpreted as a bar to separate ratings for the same disability.” Secretary’s Br. at 15.
    The Secretary’s assertions regarding § 4.14’s definition of pyramiding are accurate for both
    1971 and the present day, but they do not address Mr. Perciavalle’s contention that it was CUE not
    to rate distinct disabilities separately. As outlined above, the corollary to § 4.14 is that separate
    ratings for different disabilities or different manifestations are contemplated by § 4.25, and that
    principle is the basis for the veteran’s motion. The distinction between disabilities and injuries and
    the diseases that cause them is part of the foundation of the system for compensation. See 38 U.S.C.
    §§ 1110, 1131 (declaring that the United States will pay compensation to veterans “[f]or disability
    resulting from personal injury or disease contracted in line of duty, or for aggravation of a
    preexisting injury suffered or disease contracted in line of duty.”). As used in these statutes,
    “disability” refers to the veteran’s functional impairment, not the injury or disease that causes the
    disability. Saunders, 886 F.3d at 1363. That distinction is also an essential underpinning for the
    rating schedule. See 38 C.F.R. § 4.1 (“This rating schedule is primarily a guide in the evaluation
    of disability resulting from all types of diseases and injuries encountered as a result of or incident
    to military service.”).
    The Board also found: “The Veteran has not provided any evidence that, in July 1971, VA
    interpreted the rating schedule to allow for separate ratings for limitation of motion and instability
    of the same knee.” R. at 5. However, we know that the rating schedule did in fact allow for separate
    ratings in 1971 (and 50 years earlier), see Lyles, 29 Vet.App. at 115, and that VA’s GC found no
    prior “formal position taken by the Veterans Benefits Administration on this issue” in 1997. VA
    G.C. Prec. 23-97, at 1; R. at 55. Moreover, VA demonstrated no understanding or interpretation
    22
    different from the regulatory language—not even in the rating decision at issue, much less in any
    official pronouncement.
    The judgment concurrence faults the search for a prior formal interpretation, asserting that
    later variance from any interpretation in effect at the time of the disputed rating is sufficient to stop
    a CUE motion pursuant to § 20.1403(e) (and presumably § 3.105(a)(4)). See post at 46-47. As
    previously noted, the statutes and regulations governing CUE do not define “change in the
    interpretation.” Interpretation has been defined as “the ascertainment of a text’s meaning,” and “‘the
    art or process of discovering and expounding the intended signification of the language used.'”
    BLACK’S LAW DICTIONARY 978 (11th ed. 2019) (quoting Henry Campbell Black, HANDBOOK ON
    THE CONSTRUCTION AND INTERPRETATION OF THE LAWS 1 (1896)). In George I & II, Jordan,
    Berger, and Lamb, the Federal Circuit and our Court have looked to interpretations that might be
    called formal—statutes, regulations, and manuals, and VA GC opinions, and judicial
    pronouncements that have precedential effect. And the VA GC precedent opinion at issue, 23-97,
    specifically looked for a “formal position” by VA on the issue of whether the diagnostic codes
    authorized separate ratings for arthritis and instability and found none.
    The judgment concurrence does what the Board and the Secretary did not; it says it “can
    point to evidence of an initial interpretation.” See post at 45-46. But it doesn’t actually deliver.
    Instead, it cites three cases. The first, Fanning, notes that the Board “properly raised the issue of
    pyramiding,” but sets out the rule Mr. Perciavalle relies on: “it is possible for a veteran to have
    separate and distinct manifestations attributable to two different disability ratings, and, in such a
    case, the veteran should be compensated under different diagnoses.” 4 Vet.App. at 230-31. In the
    second, an unpublished order in Sweeney, VA pointed to the VA GC opinion—which VA
    incorrectly said changed regulations—to concede remand. Sweeney v. West, 16 Vet.App. 403
    (1999) (table) (order), 1999 WL 184869, at *1. However, the underlying decision does not mention
    pyramiding as a possible reason for limiting the veteran’s rating but merely notes that he was
    diagnosed with arthritis and knee instability and not yet rated for them. Sweeney v. West,
    16 Vet.App. 47 (1998) (table) (mem. dec.), 1998 WL 223192, at *2. The third case, a
    nonprecedential memorandum decision, cites the antipyramiding regulation but says only:
    The Board noted that the veteran suffered from pain in the left knee, but
    noted also that the recent examinations showed only moderate arthritis and slight
    instability and did not show significant atrophy or motion restriction. The Board
    23
    then rejected a rating higher than the current 30% schedular rating, which
    comprehends “severe” subluxation or lateral instability of the knee.
    Sanders v. Principi, 3 Vet.App. 334, 336 (1992) (mem. dec.) (affirming the Board’s determinations
    of the veteran’s levels of impairment as findings of fact that were not clearly erroneous).
    Even if the cited decisions had interpreted the relevant regulations, invoking them is not
    “faithful to the nature of CUE”—to use the judgment concurrence’s words and logic, see post at
    39-40—because they were rendered more than 20 years after the 1971 rating decision. Crucially,
    unlike Lyles, not one of them says anything about the law that applied in 1971. And only Fanning
    discusses the law at all, with a conclusion that supports Mr. Perciavalle’s CUE motion. At this
    juncture, the 1964 decision of the U.S. Court of Claims in Wolf stands as the only identified
    interpretation before the 1971 rating decision, and it also called for separate ratings for separate
    symptomatology, 168 Ct. Cl. at 31, as Mr. Perciavalle contends were required in his case.
    Considering Fanning, Wolf, and the other salient cases, it seems doubtful that the CUE
    regulations contemplate insulating rating decisions that violate the plain meaning of the law from
    CUE review if they are consistent with a prior rogue decision or two later authoritatively held to
    be wrong. Though the context is different, the conditions for deference to an Agency interpretation
    spelled out in Kisor v. Wilkie seem instructive, such as the requirement that the interpretation be
    the Agency’s official position. 139 S. Ct. 2400, 2415-17 (2019). And the prior and subsequent
    interpretations the Federal Circuit and our Court have pointed to, in affirming the denial of CUE
    challenges to decisions in between, fulfill that test or are precedential judicial decisions. But what
    is most important is that we are not called upon in this case to decide what counts as a prior
    interpretation. The veteran’s CUE motion specifically contended that his rating decision violated
    the plain meaning of the regulation in 1971. And when asked at oral argument whether there is
    evidence of VA’s interpretation of the knee rating regulations in 1971, the Secretary said the plain
    meaning of the regulation is the only guidance we have for how the law was interpreted at that
    time. See Oral Argument (OA) at 1:13:00-1:13:10, 1:15:18-1:15:28,
    http://www.uscourts.cavc.gov/oral_arguments_audio.php. The Secretary did mention that there
    were a few Board decisions through the years invoking the antipyramiding rule to deny separate
    ratings, but the Secretary did not cite or provide any specific examples.19 “[T]he RO is presumed
    19 Of course, “‘statements and arguments of counsel are not evidence.'” United States v. Valois, 915 F.3d 717,
    24
    to have considered all applicable law, absent some showing that it did not.” Hickson v. Shinseki,
    23 Vet.App. 394, 401 (2010). That presumption stands unrebutted, and any implication by the
    Board and the Secretary that the VA adjudicator in 1971 concluded that separate ratings would be
    pyramiding is completely unsupported.
    In sum, separate ratings for different knee disabilities were permissible in 1971, and the
    Board erred in suggesting otherwise. The Board also erred in characterizing “a later interpretation
    of an existing regulation” as “the only basis on which the Veteran asserts CUE,” and erred in
    concluding that the veteran’s CUE assertion “must be denied as a matter of law.” R. at 6. The
    veteran’s motion alleged CUE in how the RO applied the regulations in 1971, and the parties agreed
    that the measure of whether there was CUE was those regulations’ plain language. In that context,
    the veteran was entitled to a decision on the merits of his motion.
    The judgment concurrence’s prediction of dire consequences from finding error in the
    Board’s decision is belied by its “food for thought” explaining that adjudicating the veteran’s CUE
    motion on its merits would not have been burdensome or complicated. See post at 48-49. The
    unbearable burden is restrictive gatekeeping achieved by rewriting cognizable CUE challenges to
    reflexively reject them. That is not doing the right thing in the right way for the right reason.
    Turning Mr. Perciavalle away without consideration of all of his actual CUE motion was patently
    an error.
    V. PREJUDICIAL ERROR AND CUE
    Having found that the Board erred in foreclosing appellant’s CUE allegation about the July
    1971 rating decision as a matter of law, the Court turns to whether that error prejudiced appellant.
    And more specifically, whether the Board’s failure to consider the “manifestly changed outcome”
    prong of a CUE analysis was prejudicial to appellant.
    In our view, the Board’s error is harmless. Because the Board’s denial of appellant’s motion
    to revise the 1971 rating decision would not have been different even if the Board committed no
    error in its application of the CUE regulation, the Board’s error did not affect the ultimate outcome
    of the decision. The appellant has not shown how he might meet the extra heavy burden to establish
    726 (11th Cir. 2019) (quoting United States v. Lopez, 590 F.3d 1238, 1256 (11th Cir. 2009)). And it is doubtful that
    such examples could establish an Agency interpretation as its authoritative or official position, see Kisor, 139 S. Ct.
    at 2416-17, but that determination need not and could not be made without examples to consider.
    25
    undebatable outcome-determinative error in the 1971 rating decision, or how the Board’s legal
    error impacted its conclusion that there was no CUE in that decision. We would also find that the
    Board’s missing manifest-change-in-outcome analysis did not prejudice appellant’s substantial
    rights. Nor was appellant deprived a chance to participate in VA’s processing of his CUE
    allegation. What’s more, on appeal, appellant fails to explain how the Board’s legal error impacted
    its ultimate determination that the 1971 rating decision contained no CUE. And so, we would
    affirm the September 2017 Board decision.
    A. Taking Due Account of Prejudicial Error
    Merely finding error in a Board decision or in VA’s processing of a claim is not enough for
    the Court to remand. In exercising our appellate jurisdiction, the Court must review the record of
    prior proceedings and “take due account of the rule of prejudicial error.” 38 U.S.C. § 7261(b)(2).
    For remand to be the appropriate remedy, an error must prejudice the claimant. Shinseki v. Sanders,
    556 U.S. 396, 411-12 (2009).
    Although the “matter is somewhat more complicated in the context of CUE,” this standard
    remains the same. Simmons v. Wilkie, 30 Vet.App. 267, 277-79 (2018) aff’d, 964 F.3d 1381 (Fed.
    Cir. 2020). “[W]hen the Court is determining whether any Board error is prejudicial to the
    appellant, the Court is considering prejudice as an original matter. . . .[and] the only way to assess
    any prejudice in a Board error is to consider the decision that is the subject of the CUE motion.”
    Id. at 284 n.7; see also George I, 30 Vet.App. at 377-78 (holding in the alternative that affirmance
    was warranted because appellant failed to show that any error in the challenged decision would
    have manifestly changed the outcome). As the Court explained in Simmons, that consideration
    does not defy Archer’s prohibition of plenary review of the facts of the RO decision, 3 Vet.App. at
    437, because Archer applies when the Court is determining whether the Board decision was an
    abuse of discretion, but harmless error analysis arises when the Court has already concluded that
    the Board erred and the Court is assessing whether that error was prejudicial or affected the
    essential fairness of the adjudication. Simmons, 30 Vet.App. at 284.
    How does the Court conduct a prejudicial error analysis? In Sanders, the Supreme Court
    addressed this Court’s statutory prejudicial error mandate and concluded that it requires us “to
    apply the same kind of ‘harmless-error’ rule that courts ordinarily apply in civil cases.” 556 U.S. at
    26
    406;20 see 38 U.S.C. § 7261(b)(2). The Supreme Court rejected the Federal Circuit’s “complex,
    rigid, and mandatory” harmless-error framework that required this Court to presume that a notice
    error was prejudicial. Sanders, 556 U.S. at 407. The Supreme Court reasoned that presumptive
    prejudice is “inconsistent with the statutory requirement” to take due account of prejudicial error.
    Id. at 414. Rather, the Supreme Court said that this Court should use “case-specific applications of
    judgment, based on examination of the record” rather than “mandatory presumptions and rigid
    rules.” Id. at 407; see Simmons, 964 F.3d at 1385 (quoting Sanders, 556 U.S. at 407).
    And so this Court’s harmless error inquiry must be “exceedingly broad.” Simmons,
    30 Vet.App. at 284. Our statutory obligation permits us “to go outside of the facts as found by the
    Board to determine whether an error was prejudicial by reviewing ‘the record of the proceedings
    before the Secretary and the Board.'” Mlechick v. Mansfield, 503 F.3d 1340, 1345 (Fed. Cir. 2007)
    (quoting Newhouse v. Nicholson, 497 F.3d 1298, 1302 (Fed. Cir. 2007)). Indeed, fulfilling that
    obligation “necessarily requires” us to go beyond the Board’s findings. Wait, 33 Vet.App. at 18.
    We must do so because “[t]he Board cannot predict every instance in which it might be found to
    have committed error” and “cannot be expected to make specific factual findings that might
    facilitate a prejudicial error analysis.” Vogan v. Shinseki, 24 Vet.App. 159, 163 (2010). “If the
    Court’s review were restricted to findings made by the Board, the usefulness of Congress’s
    direction that we examine an error for prejudice would be marginalized as a tool for avoidance of
    remands that entail no realistic prospect of an outcome more favorable to a veteran.” Id. at 163-
    64.
    So we conduct a full review of the record to perform our statutory duty to determine
    whether error is prejudicial or harmless—mindful that the undertaking is an important part of our
    20 Criminal cases have a different paradigm, based on the principle that there are “certain basic, constitutional
    guarantees that should define the framework of any criminal trial” and “some errors should not be deemed harmless
    beyond a reasonable doubt.” Weaver v. Massachusetts, 137 S. Ct. 1899, 1907 (2017). Such errors are known as
    structural errors “and require reversal because they cause fundamental unfairness, either to the defendant in the specific
    case or by pervasive undermining of the systemic requirements of a fair and open judicial process.” Id. at 1911. Errors
    found to be structural have violated defendants’ right to conduct their own defense, right to select their own attorney,
    right to have a reasonable doubt jury instruction, right to an unbiased judge, and right to a public trial, see id. at 1908-
    10 (collecting cases), as well as their right to insist that their counsel refrain from admitting their guilt. McCoy v.
    Louisiana, 138 S. Ct. 1500, 1505-11 (2018). Even in the criminal justice system, the judicial duty to find the proper
    balance between the necessity for fair trials and the importance of finality of judgments means that a new trial generally
    will be granted as a matter of right when a structural error is preserved and raised on direct review, but when the error
    is raised in a collateral attack such as a claim of ineffective assistance of counsel, “finality concerns are far more
    pronounced” and may require a showing of prejudice to obtain a new trial. Weaver, 137 S. Ct. at 1913 (denying a new
    trial for defense counsel’s failure to object to the closure of the courtroom during jury selection).
    27
    jurisdiction. See Newhouse, 497 F.3d at 1301-02; see also Pitts v. Shinseki, 700 F.3d 1279, 1286
    (Fed. Cir. 2012) (concluding that an argument asserting prejudicial error “challenges the [Veterans
    Court’s] application of law to fact and therefore falls outside this court’s jurisdiction” (citing
    38 U.S.C. § 7292(d)(2))); Conway v. Principi, 353 F.3d 1369, 1375 (Fed. Cir. 2004) (“[T]he
    ultimate conclusion of the effect of the rule of prejudicial error on this case is beyond our
    jurisdiction.”). But the Court remains equally mindful to not exceed its jurisdiction. If the record
    contains no VA findings to support affirmance, the Court will affirm only when “the entire record
    makes evident that the Board could not have reached any other decision.” Tadlock v. McDonough,
    5 F.4th 1327, 1337 (Fed. Cir. 2021). And when “additional findings of fact are necessary regarding
    matters open to debate,” it is the Board, and not the Court, which must weigh those facts in the
    first instance. Id.
    Against this backdrop, the Court turns to appellant’s case. As explained above, the Court
    holds that the Board erred in its determination that appellant’s CUE allegation must be denied as a
    matter of law. So the Court must determine whether the Board’s ultimate conclusion that the
    challenged 1971 RO decision contained no CUE could have been different had the error not
    occurred.
    In its September 2017 decision, the Board found that the 1971 RO did not commit CUE
    because VA’s ability to award separate ratings for different disabilities of the same knee was a new
    interpretation of law. R. at 5-6. But the Board addressed none of the CUE elements, specifically
    whether (1) VA incorrectly applied the rating schedule in existence at the time, (2) any such error
    was “undebatable” and not merely a disagreement over how VA weighed or evaluated the facts in
    the final decision, or (3) any error “manifestly changed the outcome” of the final decision. R. at 5-
  3. And so the Board made no alternative finding to support an affirmance of its decision.21
    Appellant asserts that remand is necessary because the Board “never reached the merits”
    of his CUE motion. Reply Br. at 1. But he provides no reasoning for this requirement in any of his
    pleadings or at oral argument. The Secretary responds that appellant has an “extra heavy” burden
    21 Our judgment-concurring colleagues highlight the absurdity of the Board’s approach in adjudicating
    appellant’s CUE allegation. Post at 48-49. The full Court would not be here today had the Board merely done the
    work of assessing whether the facts before the RO in 1971 clearly and unmistakably showed that appellant was entitled
    to another rating for his right knee disability. But if VA’s routine misapplication of its own regulation as written opens
    a floodgate, then perhaps there was a storm surge with the potential to cause greater harm that required its opening.
    And, as here, should the gates remain locked, then it is the facts existing in the record at the time of the challenged
    decision that drive that choice.
    28
    of proof to establish CUE and that the Board’s ultimate conclusion on CUE in the 1971 rating
    decision was not prejudicial because appellant “did not meet his burden of proving an outcomedeterminant
    error.” Secretary’s Motion at 10.
    B. The Only Possible Outcome
    As the terms make evident, deciding whether error is prejudicial or harmless hinges on the
    effect of the error—whether it harmed the veteran by prejudicing the outcome of the Board’s
    ultimate conclusion. Put simply, for an error in a Board decision to warrant remand, an appellant
    must persuade the Court that the error could have made a difference in the Board’s decision. See
    Sanders, 556 U.S. at 413.
    For remand to be appropriate, the Court need not conclude that an outcome would be
    different had VA not committed an error, only that it could be. See Mayfield v. Nicholson,
    19 Vet.App. 103, 115 (2005) (rejecting the Secretary’s argument that prejudice requires a showing
    that “the correction of the error would yield a different result”), rev’d on other grounds, 444 F.3d
    1328 (Fed. Cir. 2006). As this Court spelled out in Simmons:
    [P]rejudice is established by demonstrating a disruption of the essential fairness of
    the adjudication, which can be shown by demonstrating that the error (1) prevented
    the claimant from effectively participating in the adjudicative process, or (2)
    affected or could have affected the outcome of the determination.
    30 Vet.App. at 279. On the other hand, “a demonstration that an outcome would not have been
    different even if the error had not been committed would demonstrate that there was no prejudice.”
    Mayfield, 19 Vet.App at 115 (emphasis in original). Put another away, the Board could only deny
    the claim.
    For a CUE allegation, this includes showing that there could be a manifest change in the
    outcome of the prior decision. “A manifest change in the outcome of the adjudication means that,
    absent the alleged clear and unmistakable error, the benefit sought would have been granted at the
    outset.” King v. Shinseki, 26 Vet.App. 433, 441 (2014), aff’d sub nom. King v. McDonald,
    599 F. App’x 957 (Fed. Cir. 2015).
    In a March 2015 motion to revise based on CUE, appellant alleged that, in 1971, he was
    undebatably entitled to separate ratings for left knee disabilities under DCs 5003 and 5257 for
    arthritis and instability. R. at 102; see R. at 492. For appellant to prevail on his CUE motion before
    the Board, he must meet the “extra-heavy burden” of showing that it was error for the 1971 RO
    not to apply the rating criteria in DCs 5003 or 5257 and that the evidence undebatably established
    29
    the existence of arthritis and knee instability. See King, 26 Vet.App. at 441; Fugo, 6 Vet.App. at
  4. To assess prejudice, the Court must determine whether the Board might find that the record in
    1971 compelled the conclusion that VA would have awarded a separate evaluation for a left knee
    disability. Berger, 10 Vet.App. at 169; see Simmons, 30 Vet.App. at 285; George I, 30 Vet.App.
    at 377-78.
    In 1971, as now, VA had eight DCs for rating knee disabilities. Included were DC 5257
    for “recurrent subluxation or lateral instability”; DC 5259 for “cartilage, semilunar, removal of,
    symptomatic”; and DC 5260 for “limitation of flexion.” 38 C.F.R. § 4.71a, DCs 5256-5263 (1971).
    VA rates degenerative arthritis under DC 5003 and awards a minimal compensable rating for
    limitation of motion for degenerative arthritis established by x-ray findings. See Lichtenfels v.
    Derwinski, 1 Vet.App. 484, 488 (1991) (holding that, when read together, DC 5003 and 38 C.F.R.
    § 4.59 entitle a veteran to a minimum 10% rating “where the arthritis is established by x-ray . . .
    even though there is no actual limitation of motion”). In 1971, VA rated appellant only under DC
    5259 for medial meniscectomy, evaluated at 10%. R. at 509.
    We would find that, because appellant could not meet his burden to show a manifest change
    of outcome in the 1971 rating decision, the Board error did not affect its ultimate conclusion.
    Specifically, the evidence in 1971 does not compel the conclusion that VA clearly and
    unmistakably erred in not awarding a separate evaluation for left knee arthritis.22 Appellant points
    22 The veteran’s initial CUE motion included a statement that “the Rating Specialist did not change the DC
    from 5259 to 5257 based on instability[, nor] was a separate evaluation considered for the degenerative changes
    (arthritis) and limitation of motion flexion under DC 5003-5260.” R. at 102. By its terms, this statement could be
    construed as asserting first a failure to assign a rating under DC 5257 for instability rather than his original rating
    under DC 5259, and then to add a separate other rating under DC 5003-5260. But neither the veteran’s initial motion
    nor his later arguments assert that he should have received separate ratings for his meniscectomy residuals and
    instability under DCs 5259 and 5257. He has consistently maintained that, along with his initial 10% rating, he is
    entitled to a second, separate 10% rating for arthritis, and that is how his argument was characterized and addressed
    in the proceedings below. For example, the September 2015 rating decisions lists DC 5259 and DC 5257 together. R.
    at 95. The May 2017 SOC displaces DC 5259 with DC 5257 but still characterizes the claim for a separate rating as
    one for “a separate 10% for limitation of motion for flexion and discomfort secondary to arthritis.” R. at 58. The
    veteran’s August 2017 letter echoes that formulation, seeking a substituted “10% evaluation for slight instability under
    DC 5257 and a separate 10% evaluation for loss of range of motion due to pain/discomfort under DC 5003-5260.” R.
    at 14.
    Although the record does not reflect that VA or the Board ever interpreted his argument this way, for the
    sake of completeness we note that the failure to switch his rating from DC 5259 to DC 5257 alone could not constitute
    CUE. The veteran did not seek separate ratings under DCs 5259 and 5257, so neither VA nor the Board addressed that
    issue, but an argument for separate ratings under those diagnostic codes would have likewise been unavailing in this
    context. Under DC 5257’s rating criteria in 1971, VA rated “slight” instability at 10%, “moderate” instability at 20%,
    and “severe” instability at 30%. 38 C.F.R. § 4.71a, DC 5257 (1971). The initial September 1966 rating decision
    awarding a 10% disability under DC 5259 for appellant’s medial meniscectomy was based on an examination showing
    30
    to a July 1971 radiographic report that revealed “questionably narrowed” joint space, “slight
    blunting” of the tibial spines, and “a question of nodular irregularity” as undebatable evidence of
    arthritis. R. at 102; see R. at 493-96. But notation of questionable joint irregularities in an x-ray
    report is not an arthritis diagnosis. See Lichtenfels, 1 Vet.App. at 488; 38 C.F.R. § 4.71a, DC 5003.
    Nor does appellant explain how the x-ray evidence from 1971 supports an arthritis diagnosis. Thus,
    absent evidence of arthritis confirmed by x-ray, appellant can’t meet his burden of showing that
    VA would have awarded a 10% rating for arthritis under DC 5003.
    Even if the Court, or the Board, may have reached a different conclusion than the 1971 RO
    and determined that questionable joint irregularities show arthritis warranting a separate 10%
    rating, that retrospection does not establish CUE. George I, 30 Vet.App. at 378; see King,
    26 Vet.App. at 442 (“[T]here will be times when the Court arrives at a different conclusion when
    reviewing a motion to revise a prior, final decision than it would have had the matter been reviewed
    under the standards applicable on direct appeal.”). The evidence must be absolutely clear that VA
    erred in not granting a separate rating for arthritis. Fugo, 6 Vet.App. at 44. And that simply is not
    the case. After all, disagreeing with the weighing of evidence does not support CUE. So the Court
    finds that appellant fails to show that the Board erred when it found that any error in the 1971
    rating decision did not manifestly change the outcome. And so despite the Board’s error in rejecting
    appellant’s CUE motion as a matter of law, the only possible outcome was the Board’s ultimate
    determination that the 1971 rating decision contained no CUE.
    C. Harm to Appellant’s Substantial Rights
    In Sanders, the Supreme Court identified some case-specific factors a court may consider
    in assessing prejudice: “an estimation of the likelihood that the result would have been different”
    but for the error; an awareness of what body “has the authority to reach that result”; and the “error’s
    likely effects on the perceived fairness, integrity, or public reputation of judicial proceedings.”
    Sanders, 556 U.S. at 411-12.
    “weakness and feeling of instability of left knee.” R. at 509. Though a rating under DC 5259 does not preclude a
    separate rating under DC 5257, the anti-pyramiding rule means that a veteran cannot be compensated more than once
    for the same manifestation of a knee disability or for overlapping symptoms that do not cause more functional
    impairment. See Lyles, 29 Vet.App. at 112-19. Since the 1971 decision continued the 10% rating based on an
    examination that found no weakness or functional defect but “very slight instability of joint laterally,” R. at 490, 495,
    it cannot be undebatable that separate ratings under DC 5259 and 5257 were warranted. In other words, it is far from
    clear that the evidence compels the conclusion that VA clearly and unmistakably erred in not awarding a separate
    rating under DC 5257 in 1971.
    31
    The Supreme Court also noted that this Court’s special expertise in reviewing VA decisions
    enables it to make “empirically based, nonbinding generalizations” about errors which have the
    “natural effect” of prejudicing a claimant’s “substantial rights.” Sanders, 556 U.S. at 411-12; see
    Kotteakos v. United States, 328 U.S. 750, 760 (1946) (reviewing courts may learn over time that
    the “natural effect [of certain errors] is to prejudice a litigant’s substantial rights”) (quoting H.R.
    REP. NO. 913, 65th Cong. (3d Sess. 1919) at 1).
    The Supreme Court also recognized that some errors that could be considered harmless in
    other adjudicatory proceedings may be considered harmful in a veteran’s case. Sanders, 556 U.S.
    at 412. This is because of the “special solicitude” Congress has expressed for veterans cases, VA’s
    statutory duty to help veterans develop their claims, and the nonadversarial nature of VA
    proceedings. Sanders, 556 U.S. at 412. The VA claims system requires “notice and an opportunity
    to be heard at virtually every step.” Thurber v. Brown, 5 Vet.App. 119, 123 (1993); Bernard v.
    Brown, 4 Vet.App. 384, 392-94 (1993) (holding that statutory and regulatory provisions provide
    “procedural safeguards” to “ensure a claimant’s right to full and fair assistance and adjudication in
    the VA adjudication process”).
    Following Sanders, this Court identified some procedural errors that had the “natural
    effect” of harming a claimant’s “substantial rights.” Sanders, 556 U.S. at 412; see Kotteakos,
    328 U.S. at 759. This is because some procedural errors can have an “unquantifiable” effect on the
    outcome of a proceeding and thus must be returned to the Board. See Quinn v. Wilkie, 31 Vet.App.
    284, 292 (2019) (“‘Where the effect of an error on the outcome of a proceeding is unquantifiable,
    however, we will not speculate as to what the outcome might have been had the error not
    occurred.'” (quoting Wagner v. United States, 365 F.3d 1358, 1365 (Fed. Cir. 2004))); Arneson v.
    Shinseki, 24 Vet.App. 379, 389 (2011) (finding prejudice when error “could have altered” the
    Board’s determinations).
    Prejudice to a claimant’s substantial rights naturally stems from the deprivation of
    meaningful opportunities to participate in the adjudicative process, which may result from a VA
    failure to provide notice to a veteran of evidence necessary to substantiate a claim, see Vazquez-
    Flores v. Shinseki, 24 Vet.App. 94, 105-07 (2010), or from denial of the veteran’s right to a hearing
    before the Board, see Arneson, 24 Vet.App. at 389. “An error is prejudicial when it ‘affects a
    substantive right that a statutory or regulatory provision was designed to protect’ and ‘affects the
    32
    essential fairness of the adjudication.'” Sapp v. Wilkie, 32 Vet.App. 125, 143 (2019) (quoting
    Overton v. Nicholson, 20 Vet.App. 427, 434-35 (2006)).
    We would also find that the Board’s error did not have the natural effect of harming
    appellant’s substantial rights or an “unquantifiable” effect on the outcome of the proceeding.
    Wagner, 365 F.3d at 1365. Recall that VA’s processing and decision-making on CUE allegations
    are much different than direct appeals of nonfinal decisions. In the adjudication of a CUE
    allegation, VA has no duty to assist a claimant with developing a claim. Cook v. Principi, 318 F.3d
    1334, 1344 (Fed. Cir. 2002) (holding that a breach of the duty to assist cannot form the predicate
    for a motion for revision of a finally decided claim based on clear and unmistakable error); see
    also Szemraj v. Principi, 357 F.3d 1370, 1376 (Fed. Cir. 2004) (“[T]he failure to assist in
    developing the evidentiary record cannot constitute CUE.”); King, 26 Vet.App. at 438 (noting that
    the breach of the duty to assist creates an “‘incomplete rather than an incorrect record'” (quoting
    Caffrey v. Brown, 6 Vet.App. 377, 384 (1994))). Indeed, review for CUE must be based on the
    record at the time the prior decision was made. May v. Nicholson, 19 Vet.App. 310, 313 (2005);
    38 C.F.R. § 3.105(a)(iii). There is no factfinding or weighing of evidence—VA and the Court are
    explicitly prohibited from “reweighing” evidence. Damrel v. Brown, 6 Vet.App. 242, 245-46
    (1994) (holding that appellant’s contention that the RO “misevaluated and misinterpreted evidence
    available at the time” of the prior decision is not CUE because it merely requires reweighing the
    evidence).
    VA’s adjudication of a CUE allegation is limited to determining whether a claimant’s
    specifically pled allegation of error within a final decision was CUE. See Lang, 971 F.3d at 1352.
    Because a CUE allegation is limited to the record as it existed at the time of the final decision and
    the error as pled before VA, the Court can evaluate a Board decision on a CUE allegation that
    lacks a manifest-change-of-outcome analysis based on the record before it. Here, the Court has
    everything required to determine whether appellant was harmed by the Board’s ultimate conclusion
    that the 1971 RO did not commit CUE in its rating decision.
    What’s more, VA provided appellant a meaningful chance to participate in the processing
    of his claim at every step and discerns no procedural error. See Simmons, 30 Vet.App. at 285.
    Through his chosen representative, Veterans Advocates Group, LLC, appellant submitted a CUE
    motion to the RO. R. at 101-02. He then participated in an informal hearing at the RO with his
    representative present. R. at 32 (noting that VA held an informal hearing with the veteran’s
    33
    attorney). And, on appeal to the Board, appellant presented more argument in support of his CUE
    allegation. R. at 13-15. At every step of the adjudication process, VA provided appellant with the
    chance to meet his burden to clearly and unmistakably show that the 1971 rating decision, which
    was presumed valid, contained an outcome-determinative error.
    We also conclude that VA provided appellant with all required notice in the CUE
    adjudication process. In a rating decision, the RO addressed appellant’s allegation that the 1971
    RO’s “failure to grant a 10% rating for slight instability of the left knee under DC 5257 and a
    separate 10% rating for limitation of motion for flexion and discomfort secondary to arthritis” was
    CUE. R. at 94. In an SOC, a decision review officer addressed the same CUE allegation. R. at 32-
  5. And the Board addressed appellant’s contention that the 1971 “RO committed CUE by not
    granting separate compensable evaluations for limitation of motion and for instability.” R. at 5. At
    every point in its processing of appellant’s CUE allegation, VA addressed appellant’s CUE
    allegation and provided written notice on why it found no CUE in the 1971 rating decision.
    Our dissenting colleagues would hold that VA’s mischaracterization of appellant’s CUE
    allegation deprived him of the right to a decision on his CUE motion, which disrupted the essential
    fairness of the process. Post at 54. But what happened here is not the kind of situation when the
    error could somehow shake faith in the adjudication process. Sanders, 556 U.S. at 410. VA can,
    as it did here, reach an incorrect conclusion without preventing a claimant from participating in
    the proceedings or otherwise upsetting fundamental fairness. An appellant’s lack of meaningful
    opportunity to participate in the processing of his or her claim may cause remandable harm. But
    an appeal to the Court is the correct recourse for an appellant to resolve harm by an error in VA’s
    adjudication of a claim. And on appeal to the Court, for an adjudication error to warrant remand,
    an appellant bears the burden of showing how the outcome of a new Board decision on the claim
    could be different.
    And now we address whether the Board’s lack of manifest-change-of-outcome analysis in
    its adjudication of appellant’s CUE allegation prejudiced appellant. This was not a procedural error
    that had the natural effect of harming appellant’s “substantial rights” or some “unquantifiable”
    effect on the outcome. See Sanders, 556 U.S. at 412; Wagner, 365 F.3d at 1365. As discussed
    above, the Board erred in its adjudication of appellant’s CUE allegation. The Board made an
    erroneous conclusion that appellant’s CUE allegation hinged on a new interpretation of law and
    was thus precluded by law. But adjudication error does not mean procedural error. All the post34
    Sanders cases when the Court found prejudice to a claimant’s substantial rights naturally stemmed
    from VA error that denied the veteran some whole step in the adjudication process. See Simmons,
    30 Vet.App. at 281-82 (compiling cases). Here, once appellant submitted his argument to the
    Board, there was no more participation available to him. He had seized every opportunity the
    process provides.
    Here if the Court determined, after a full review of the record, that the Board could have
    found CUE in the 1971 decision, then appellant was prejudiced, and the Court would have to
    remand. But absent a showing of effect on the Board’s ultimate CUE determination, there is no
    prejudicial error. In this context, the effect of error is assessed (or quantified) by applying set
    standards to facts and law fixed by the date of the decision being reviewed. In other words, an
    appeal to our Court and its obligation to provide a broad scope prejudicial error analysis cures any
    Board error here in not providing a full CUE analysis. See Mayfield, 19 Vet.App. at 129 (holding
    that the Court’s review was “not hindered by any reasons-or-bases deficiency in the Board decision,
    and a remand for a reasons-or-bases error would be of no benefit to [appellant] and would therefore
    be pointless”); Soyini v. Derwinski, 1 Vet.App. 540, 546 (1991) (concluding that where evidence
    is overwhelmingly against claim, remand for a reasons-or-bases deficiency would be superfluous).
    Thus, we discern no prejudicial error in VA’s full and fair processing of appellant’s CUE
    allegation and find that appellant had a meaningful chance to participate effectively in the
    processing of his CUE motion. And so the Board’s lack of manifest-change-of-outcome analysis
    in its adjudication of appellant’s CUE allegation did not harm appellant’s substantial rights. See
    Sanders, 556 U.S. at 410.
    D. Burden to Show Harm in the Board’s Error
    Finally, the appellant failed to show how the Board’s error affected the outcome of the
    Board’s CUE determination.
    Though the Court has an independent duty to assess harm to an appellant based on an error
    in a Board decision, the appellant has the burden to show that prejudice. See Sanders, 556 U.S. at
    409 (“[T]he burden of showing that an error is harmful normally falls upon the party attacking the
    agency’s determination.”); Hilkert v. West, 12 Vet.App. 145, 151 (1999) (en banc) (“An appellant
    bears the burden of persuasion on appeals to this Court.”), aff’d per curiam, 232 F.3d 908 (Fed.
    Cir. 2000) (table). An appellant must show how a claimed error in a Board decision harmed the
    outcome of her or his claim. Sanders, 556 U.S. at 410 (“[T]he party seeking reversal normally
    35
    must explain why the erroneous ruling caused harm[,] . . . by marshaling the facts and evidence . .
    . .”).
    Appellant, through counsel, chose to raise only legal challenges to the Board’s decision and
    made the, presumably strategic, choice not to present any argument on prejudice to this Court. See
    Robinson v. Peake, 21 Vet.App. 545, 554 (2008) (presuming that “an experienced attorney in
    veteran’s law[] says what he means and means what he says”), aff’d sub nom. Robinson v. Shinseki,
    557 F.3d 1355 (Fed. Cir. 2009). This is despite a Court order for appellant to respond to the
    Secretary’s motion for full-Court consideration that specifically raised a challenge to the threejudge
    panel’s prejudicial error analysis. March 18, 2020, Court order (“[F]ile a response to the
    Secretary’s motion for full-Court consideration.”); Secretary’s Motion at 7 (alleging that “the Court
    overlooked relevant evidence and legal authority in its prejudicial error analysis”). Yet appellant
    declined to articulate to the Court how the Board’s error on what could constitute CUE prejudiced
    his motion to revise the 1971 rating decision. Without more, the Court is left to divine argument
    for an appellant represented by experienced counsel at this Court. The Court will not fill in the
    blanks and identify possible prejudice in a Board decision for itself when it is the responsibility of
    counsel to do so. See Coker v. Nicholson, 19 Vet.App. 439, 442 (2006) (per curiam), vacated on
    other grounds sub nom. Coker v. Peake, 310 F. App’x 371 (Fed. Cir. 2008) (per curiam order).
    In sum, we find no prejudice in the Board’s erroneous finding that appellant’s CUE
    allegation was precluded by law. As appellant failed to show harm in the Board’s legal error, the
    Court should affirm the September 2017 Board decision. See Sanders, 556 U.S. at 411-14; see also
    King, 26 Vet.App. at 441; Fugo, 6 Vet.App. at 43-44.
    VI. CONCLUSION
    Although a majority of the Court agrees that the Board erred, a different majority concludes
    that the proper remedy in this case is to nevertheless affirm the Board decision. However, the Court
    is divided as to the reasons for affirmance. Accordingly, the Court VACATES the Court’s
    September 27, 2019, panel decision and AFFIRMS the September 18, 2017, Board decision
    finding no CUE in the July 1971 rating decision.
    36
    TOTH, Judge, concurring: I agree that 38 C.F.R. § 20.1403(e) doesn’t apply here because
    there has been no actual change in interpretation for Mr. Perciavalle to apply retroactively. So, the
    Board’s rejection of his CUE motion on that basis was erroneous. I also agree that such error was
    harmless, as I don’t see a way for Mr. Perciavalle to prevail on his CUE motion under the reasoning
    spelled out in George v. McDonough, 991 F.3d 1227 (Fed. Cir. 2021), and Steele v. McDonough,
    856 F. App’x 878 (Fed. Cir. 2021).
    (Before proceeding, it’s necessary to address the proverbial elephant in the room: the
    peculiar circumstance presented by Steele, a nonprecedential opinion in which the Federal Circuit
    weighed in on this pending case and previewed both its ruling and reasoning were the matter to
    reach that court. Although not exactly the law of the case, Steele appears much the same as if we
    had certified the legal question to the Federal Circuit rather than hearing the matter en banc. Suffice
    it to say, a sneak preview of a potential appellate decision is an odd thing for a lower court to factor
    into its decisionmaking.)
    That said, I read George and Steele as forming a one-two punch prohibiting CUE motions
    wherever the alleged legal error or disputable question of law was resolved by a court decision or
    official Agency publication (such as a General Counsel precedential decision) issued after the
    decision the veteran seeks to collaterally attack became final. Stated differently, George and Steele
    require that, in CUE motions claiming legal error, the error be undebatable at the time of the
    disputed decision; otherwise, the error can only be proven via the retroactive application of a later
    court ruling or Agency publication. And because retroactive application is forbidden, only legal
    errors known at the time of the underlying decision can support a collateral challenge for CUE.
    This reading explains the rejection of counsel’s argument in George that later court
    decisions may be relied on in the CUE context because they merely state what the law has always
    meant. 991 F.3d at 1236. Steele offers a succinct repudiation of this argument, concluding that,
    “whether an interpretation is a first interpretation or a change in existing interpretation, it cannot
    serve as the basis for CUE.” 856 F. App’x at 881. This prohibition against an initial legal
    interpretation forming the basis of a CUE motion recognizes that the only way to prove CUE in
    such situations would be via retroactive application of a later authoritative pronouncement to a
    finally decided VA decision, which is not permitted. Id.
    But instead of reading § 20.1403(e) to cover cases where there is no clear change in the
    interpretation of the law (e.g., from interpretation A to interpretation B), I read George and Steele
    37
    as effecting the broader point that an alleged legal error cannot satisfy the second CUE element—
    that the error is “undebatable”—where it has yet to be identified as erroneous by a court decision
    or VA publication. See Russell v. Principi, 3 Vet.App. 310, 313-14 (1992) (en banc) (“The words
    ‘clear and unmistakable error’ are self-defining. They are errors that are undebatable, so that it can
    be said that reasonable minds could only conclude that the original decision was fatally flawed at
    the time it was made.”).
    Under this reading, it’s irrelevant whether or not an interpretation presents a change from
    any predecessor; what matters is whether the claimed legal error was known at the time of the
    disputed underlying decision. And if nothing else, shifting the focus from retroactivity to what
    constitutes an undebatable error will relieve adjudicators from having to engage in abstruse
    speculations about whether a court decision interpreting a statute amounts to a change in law or
    whether it merely elucidates what the law has always meant, or whether one can, in fact,
    demonstrate legal error without consideration of the later decision where the error was identified.
    As the author of the original panel opinion in this case, I still believe that the Board erred
    in characterizing Mr. Perciavalle’s CUE motion as resting on a change in interpretation. However,
    George and Steele have since made clear that CUE is not available on the theory that Mr.
    Perciavalle presents, and so any Board error here must be harmless. For these reasons, I concur in
    the Court’s decision.
    ALLEN, Judge, with whom MEREDITH and FALVEY, Judges, join, concurring in the
    judgment: Although we concur in the Court’s judgment that the Board’s decision should be
    affirmed, we write separately because our path to affirmance bears no resemblance to that of the
    opinion co-authored by Judges Laurer and Jaquith. In short, we conclude that affirmance is
    warranted because the Board did not commit error. The opinion of the Court (in particular Part IV)
    concludes that the Board legally erred when it denied appellant’s motion to revise a July 1971
    rating decision based on CUE. However, the opinion determines that affirmance is nevertheless
    appropriate because the Board’s error was harmless. As we will explain, we disagree with the
    Court’s determination that the Board erred. It did not. Because we believe the Board reached the
    correct result as a matter of law, we cannot join anything more than the judgment of affirmance.
    Because we would affirm on the ground that the Board did not err, we do not reach the question
    whether any supposed error was harmless. For ease of reference, in the balance of this opinion we
    38
    will refer to the portion of the Court’s opinion dealing with what it determines to be the Board’s
    legal error as the “majority” because it represents the views of six Judges.23
    The majority holds that “separate ratings for different knee disabilities were permissible in
    1971, and the Board erred in suggesting otherwise.”24 Further, the majority holds that “[t]he Board
    also erred in characterizing ‘a later interpretation of an existing regulation’ as ‘the only basis on
    which the Veteran asserts CUE,’ and erred in concluding that the veteran’s CUE assertion ‘must be
    denied as a matter of law.'” 25 To get there, the majority reasons that, first, the Board
    mischaracterized appellant’s CUE theory.26 According to the majority, appellant’s CUE motion27
    “does not depend on the retroactive application of Esteban or the OGC opinion”; rather, his CUE
    motion depends on the allegation that “VA’s interpretation of the plain meaning of the law at the
    time of the rating decision was clearly and unmistakably erroneous.”28
    Next, the majority reasons that neither the 1997 precedential VA GC opinion nor Esteban
    v. Brown was a change in the interpretation of the rating-schedule and anti-pyramiding regulations.
    The 1997 precedential VA GC opinion formally stated that a claimant could receive separate
    ratings for knee arthritis and instability under their respective diagnostic codes and that those
    separate ratings do not constitute pyramiding.29 According to the majority, the VA GC opinion did
    not constitute a change in interpretation because no interpretation preexisted the VA GC opinion
    and the opinion was based on the regulations’ plain terms.30 Nor was a change rendered by Esteban
    v. Brown, in which this Court held in 1994 that “distinct and separate symptom[at]ology arising
    from the same injury should be rated separately.”31 The majority reasons that Esteban “cannot be
    deemed a change in interpretation” because “Esteban neither proposed an interpretation of a
    23 Chief Judge Bartley and Judges Pietsch, Greenberg, Toth, Laurer, and Jaquith agree on Parts I through IV of the
    Court’s opinion concerning the purported Board error.
    24 Ante at 23.
    25 Id. (quoting R. at 6).
    26 Id. at 10.
    27 “[A]n assertion of [CUE] is a motion or a request, rather than a claim.” Hillyard v. Shinseki, 24 Vet.App. 343, 355
    (2011) (emphasis omitted), aff’d, 695 F.3d 1257 (Fed. Cir. 2012).
    28 Ante at 20 (citing Berger v. Brown, 10 Vet.App. 166, 170 (1997)).
    29 Id. at 10-11, 16-17.
    30 Id. at 16.
    31 Id. at 9, 17-20 (citing 6 Vet.App. 259, 261-62 (1994)).
    39
    regulation nor cited any existing interpretation; it merely applied the express provisions of §§ 4.14
    and 4.25 to the facts presented.”32 In other words, according to the majority, Esteban merely stated
    what the law has always meant.
    A. Characterization of Appellant’s CUE Theory
    We believe the majority is wrong no matter how one views this appeal. All else (for the
    majority) flows from the majority’s conclusion that the Board mischaracterized appellant’s CUE
    theory, so that is where we will begin. The only way that the majority can conclude that the Board
    mischaracterized appellant’s CUE theory is by glossing over appellant’s assertion before the Board
    that Esteban merely “clarified” what the law has always been, and, “[t]herefore, there is no bar to
    the retroactive effect of Esteban,” citing a Federal Circuit decision regarding retroactivity.33 We
    cannot fault the Board for taking cues from appellant’s use of legal terms of art.34 The Board did
    not mischaracterize appellant’s CUE theory in form or, as we explain next, in substance.
    Though the characterization of the CUE motion matters greatly to the majority, it matters
    not at all to us. Even if we ignore appellant’s explicit references to retroactivity, we would hold
    that the Board did not err for two reasons: (1) Appellant failed to state what we’ll call a Berger
    CUE motion; and (2) the Board correctly recognized and denied as a matter of law appellant’s CUE
    theory as, if anything, retroactivity.
    In Berger, this Court reiterated that “the regulation that establishes CUE . . . preclud[es]
    new interpretations of the law from being the basis of revising decisions.”35 And the Court took
    the occasion “to specifically hold that opinions from this Court that formulate new interpretations
    of the law subsequent to an RO decision [that is the subject of a CUE challenge] cannot be the
    basis of a valid CUE [motion].”36 But the Court concluded by articulating
    one theory upon which CUE could be based: If the RO’s interpretation of the plain
    meaning of the law was clearly and unmistakably erroneous . . . , then there may be
    a basis for a CUE motion. On the other hand, if it was a plausible interpretation, in
    32 Id. at 17-20.
    33 R. at 13-14 (emphasis added) (citing Princess Cruises, Inc. v. United States, 397 F.3d 1358 (Fed. Cir. 2005)).
    34 See R. at 5-6 (“The Veteran argues that Esteban and the General Counsel opinion clarified existing regulations as
    opposed to changing them and so their interpretation should have retroactive effect.”).
    35 Berger, 10 Vet.App. at 170 (citing 38 C.F.R. § 3.105 (1996)).
    36 Id.
    40
    the context of the body of law that existed [at the time of the decision allegedly
    containing CUE], then there is no basis for such a claim.[37]
    This cognizable, incorrect-application CUE theory is what we call, for short, a Berger CUE motion
    and what the majority says appellant alleged before the Board.38
    But appellant did not assert a Berger CUE motion—or at least the Board did not err in
    failing to recognize such a request. Before the Board, contrary to the majority’s holding, appellant
    failed to state a Berger CUE motion because he failed to make any semblance of an allegation
    about how, in the context of the body of law that existed in 1971, the 1971 RO could not have
    plausibly interpreted the rating schedule and anti-pyramiding regulations in his case to preclude
    the separate ratings appellant sought.39 CUE will cease to be a specific and rare kind of error if we
    hold claimants to such a lesser pleading standard and require VA not only to adjudicate on the
    merits undeveloped theories but to supply in its entirety the missing legal argument.40
    To state and consider a Berger CUE motion—without corrupting the inquiry into direct or
    plenary review, which is not a proper CUE inquiry—a claimant and the Board must pretend that
    they exist in the time the decision allegedly containing CUE was rendered, without the benefit of
    any later law or interpretation.41 That is the context in which a claimant must allege how, and the
    Board must decide whether, the Agency decisionmaker made a clear and unmistakable error.42 To
    reason in such a mindset is difficult, to be sure; it is all too easy to draw on later, “clarify[ing]”43
    law or interpretations. But to stay faithful to the nature of CUE, one must fight the urge. And it is
    certainly not impossible. In fact, trial judges do something similar all the time, when they hold
    37 Id.
    38 See ante at 20 (citing Berger, 10 Vet.App. at 170) (holding that “a motion alleging that VA’s interpretation of the
    plain meaning of the law at the time of the rating decision was clearly and unmistakably erroneous states a valid CUE
    [motion].”).
    39 Berger, 10 Vet.App. at 170.
    40 Acciola v. Peake, 22 Vet.App. 320, 328 (2008) (holding that merely citing a general entitlement statute or regulation
    is insufficient to allege a specific CUE theory and explaining that to accept such broad pleadings would create a
    “pleading gimmick” placing undue burden on the Secretary to invent CUE theories to adjudicate).
    41 See Berger, 10 Vet.App. at 170.
    42 See id.
    43 R. at 13-14.
    41
    bench trials. In bench trials, judges hear objections, exclude evidence, and must make decisions
    based on admissible evidence only, pretending they never heard or saw the inadmissible evidence.
    Relatedly, and also under Berger, it is simply not enough to state, as the majority does and
    appellant did in a conclusory way before the Board, that in 1971 the regulations “permitted”
    separate ratings.44 Rather, at the bare minimum, appellant must have somehow alleged that, in the
    context of the body of law that existed in 1971, the 1971 RO’s interpretation and application of
    controlling law were implausible given the circumstances of his case.45 Specifically, here, that a
    purported, implicit interpretation and application of the rating schedule and anti-pyramiding
    regulations, under which the 1971 RO failed to award separate ratings, was implausible. Even if
    he did not speak in terms of plausibility, he could have adequately pled the Berger CUE motion in
    substance in one of two ways.
    One way was to attempt to analyze the regulations’ plain language to show how the RO’s
    interpretation was clearly and unmistakably wrong.46 But appellant did not do that. He did not
    explain how in 1971 the regulations’ plain language was not “susceptible of differing
    interpretations depending, as is true in most cases, on the facts that [were] to be applied to the
    law”;47 or, in other words, how in 1971, “reasonable minds could [not have] differ[ed] concerning
    the ‘correct’ interpretation” or reconciliation of the rating schedule and the anti-pyramiding
    regulations. 48 Instead he was nothing but conclusory in saying that he should have received
    separate ratings.49 And the majority’s analysis does him no favors; the majority’s attempt to analyze
    the regulations’ plain language on appellant’s behalf involves legion citations to post-1971 law and
    interpretations.50 Such citations could certainly provide context in terms of determining what the
    44 Ante at 23; R. at 13.
    45 See Berger, 10 Vet.App. at 170; cf. Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555 (2007) (holding that a plaintiff’s
    obligation to provide the grounds of his entitlement to relief “requires more than labels and conclusions, and a
    formulaic recitation of the elements of a cause of action will not do”). Twombly is an apt comparison because,
    ultimately, Federal complaint pleading standards are all about notice, see Twombly, 550 U.S. at 555, and the majority’s
    entire position depends on the assumption that the Board had adequate notice that appellant was alleging a Berger
    CUE motion, not retroactive application of Esteban or the VA GC opinion, and thus should have addressed that claim.
    46 See Berger, 10 Vet.App. at 170.
    47 Id.
    48 Id.
    49 See R. at 13-14, 101-02.
    50 See ante at 8-9 (citing cases from 1993, 2006, 2011, and 2017 to explain how to reconcile the regulations as of
    1971), 20-21 (citing Saunders v. Wilkie, 886 F.3d 1356, 1363 (Fed. Cir. 2018) (explaining the definitions of and
    42
    law was at an earlier time. But the majority appears to use them here to show what the law actually
    was in 1971.
    Alternatively, to state a Berger CUE motion in substance, appellant could have produced
    a contemporaneous interpretation that reconciled the regulations in his favor. But he did not do
    that either. The Board found that “[appellant] has not provided any evidence that, in July 1971,
    VA interpreted the rating schedule to allow for separate ratings for limitation of motion and
    instability of the same knee.”51
    Because appellant did not evoke Berger in form or in substance before the Board—he did
    not cite Berger, use terms of art evocative of Berger’s standard, analyze the regulations’ plain
    language, or produce a contemporaneous interpretation—as a matter of law, appellant did not
    allege a Berger CUE motion. And the majority went too far in holding that he did.52 Therefore, the
    Board did not err in failing to recognize a Berger CUE motion—because there wasn’t one.
    Mind you, our position is not that appellant must have used magic words to state a Berger
    CUE motion. But this Court cannot hold the Board to the impossible standard of recognizing a
    CUE allegation that was not properly pled in form or in substance. To the extent appellant intended
    to allege a Berger CUE motion, his request was sorely underdeveloped to the point of being
    unrecognizable. No amount of more sympathetic reading or development would have changed
    that.53 Neither the Board nor this Court is required to create esoteric legal arguments for claimants.
    relationships among regulatory terms by relying at times on noncontemporaneous authorities)).
    51 R. at 5. The closest anyone comes to making such a showing is the majority citation to a U.S. Court of Claims case
    from 1964 that concerns actions of the Secretary of the Air Force and the Air Force Board for the Correction of
    Military Records. See ante at 9-10 (citing Wolf v. United States, 168 Ct. Cl. 24, 31 (1964)). To the extent this decision
    is relevant at all—and recall appellant did not cite it—it does not appear the majority argues that this Court of Claims
    decision established the operative rule VA followed at the time of the 1971 decision. And, in any event, as we discuss
    below, it does not appear that Wolf’s 1964 interpretation of rating intestine symptoms was VA’s interpretation of rating
    knee symptoms before 1997. See infra note 55.
    52 See ante at 20.
    53 See Acciola, 22 Vet.App. at 326 (holding that though “a sympathetic reading of a CUE motion requires the Secretary
    to fill in omissions and gaps that an unsophisticated claimant may leave in describing his . . . specific dispute of error
    with the underlying decision,” a sympathetic reading does not require “the Secretary to imagine ways in which the
    original decision might be defective”); see also Andrews v. Nicholson, 421 F.3d 1278, 1282-83 (Fed. Cir. 2005)
    (“Although CUE [motions] must be pled with specificity, 38 C.F.R. § 20.1404(b) (2004), Roberson [v. Principi]
    requires the RO and the Board to ‘fully and sympathetically’ develop a veteran’s pro se CUE motion ‘to its optimum
    before deciding it on the merits.’ Roberson, 251 F.3d [1378, 1384 (Fed. Cir. 2001)]. Quite simply, the VA’s duty to
    sympathetically read a veteran’s pro se CUE motion to discern all potential claims is antecedent to a determination of
    whether a CUE claim has been pled with specificity.”).
    43
    Nor is it prudent to do so. Unlike compensation claims, which can be readjudicated or
    reconsidered,54 the preclusive effect of res judicata bars relitigation of previously adjudicated CUE
    allegations.55 For this reason, a VA regulation requires that non-specific allegations of CUE be
    dismissed without prejudice.56 By creating, reviewing, and affirming as to appellant’s non-existent
    Berger CUE motion, this opinion may “preclude[] VA from hearing any related future argument,
    because there is now a final denial of this portion of his CUE motion by the Court.”57
    B. Retroactivity
    Which brings us to our second reason (really, the other side of the coin) why, even if we
    ignore appellant’s explicit references to retroactivity, the Board did not err in denying the motion
    as a matter of law: In substance, “the only basis on which the Veteran asserts CUE” is retroactivity
    masquerading as a Berger CUE motion, to put it generously.58 Retroactivity by any other name is
    still retroactivity.59
    Appellant’s CUE motion was retroactivity in all but name because, as the Board said,60 the
    request depended on the retroactive application of a “later-in-time”61 interpretation, as evidenced
    by either Esteban or the VA GC opinion. And that is not a cognizable CUE theory. Esteban and
    54 See 38 C.F.R. § 3.156 (2021).
    55 See Link v. West, 12 Vet.App. 39, 44 (1998) (“Under the principle of res judicata, ‘[o]nce there is a final decision on
    the issue of [CUE] . . . that particular [request for revision on the basis of CUE] may not be raised again.'” (quoting
    Russell v. Principi, 3 Vet.App. 310, 315 (1992) (en banc))).
    56 38 C.F.R. § 20.1404(b) (2021) (providing that the Board must dismiss without prejudice when a CUE motion lacks
    the requisite pleading specificity); see Canady v. Nicholson, 20 Vet.App. 393, 400 (2006) (noting that a request for
    revision that fails to comply with the pleading requirements “‘shall be dismissed without prejudice to refiling'” (quoting
    38 C.F.R. § 20.1404(b))). We recognize that 38 C.F.R. § 3.105(a)(1)(vii)(B) is the governing regulation in this matter
    because the Board is reviewing appellant’s assertion of CUE in an RO decision. We cite § 20.1404(b) to provide more
    general context concerning CUE motions and the requirement that CUE be alleged with specificity. See 38 C.F.R.
    § 3.105(a)(1)(vii)(B) (2021) (requiring specific allegations of CUE); 38 C.F.R. § 20.1404(b) (same). The difference
    between these regulations is immaterial for purposes of the resolution before the Court.
    57 Lamb v. Peake, 22 Vet.App. 227, 238 (2008) (Schoelen, J., concurring in part and dissenting in part).
    58 R. at 5-6.
    59 “What’s in a name? that which we call a rose [b]y any other name would smell as sweet[.]” WILLIAM SHAKESPEARE,
    ROMEO AND JULIET act II, sc. II, l. 43-44, accessible at http://shakespeare.mit.edu/romeo_juliet/romeo_juliet.2.2.html
    (last accessed June 8, 2021).
    60 R. at 5-6.
    61 George v. McDonough, 991 F.3d 1227, 1234 (Fed. Cir. 2021) (“Jordan[ v. Nicholson, 401 F.3d 1296 (Fed. Cir.
    2005)], in view of [Disabled American Veterans v. Gober, 234 F.3d 682 (Fed. Cir. 2000)], squarely forecloses
    Appellants’ argument that Wagner[ v. Principi, 370 F.3d 1089 (Fed. Cir. 2004),]’s later-in-time interpretation of
    [38 U.S.C.] § 1111 can serve as the basis for CUE.”), petition for cert. filed, No. 21-234 (U.S. Aug. 13, 2021).
    44
    the VA GC opinion are changes in the law, novel interpretations of the law, or authoritative
    statements—no matter which way one views them, they cannot be applied retroactively to establish
    CUE.62 Therefore, and as we’ll explain further, based on the Federal Circuit’s decisions in George
    v. McDonough, Steele v. McDonough, and Jordan v. Nicholson, as well as this Court’s decisions
    in Lamb v. Peake and Berger v. Brown, we would hold that the Board’s denial of appellant’s CUE
    motion as a matter of law under 38 C.F.R. § 20.1403(e)63 was not “arbitrary, capricious, an abuse
    of discretion, or otherwise not in accordance with law.”64
    “[CUE] does not include the otherwise correct application of a statute or regulation where,
    subsequent to the Board decision challenged, there has been a change in the interpretation of the
    statute or regulation.”65 “CUE must be analyzed based on the law as it was understood at the time
    of the original decision and cannot arise from a subsequent change in the law or interpretation
    thereof to attack a final VA decision.”66
    Those CUE principles complement each other, and the Federal Circuit’s recent decisions
    in George and Steele instruct us on how they come to bear on this appeal. In George, the Federal
    Circuit affirmed a decision of this Court “[b]ecause Jordan v. Nicholson and Disabled American
    Veterans v. Gober [(DAV)] establish that a legal-based CUE requires a misapplication of the law
    as it was understood at that time, and cannot arise from a subsequent change in interpretation of
    law by the agency or judiciary.”67 That holding depended on the Federal Circuit’s rejection of the
    claimant’s attempt to circumvent the bar on retroactivity by relying on a later judicial interpretation
    that, according to the claimant and the dissenting judge from this Court, merely provided an
    “authoritative statement” of what a law has “always meant.”68
    Less than a month later, in an unpublished decision, Steele, the Federal Circuit again
    rejected a claimant’s argument that was nearly identical to the one made in George—and, this time,
    62 See George, 991 F.3d 1227; Jordan, 401 F.3d 1296; Steele v. McDonough, 856 F. App’x 878 (Fed. Cir. 2021);
    Lamb, 22 Vet.App. 227; Berger, 10 Vet.App. at 170.
    63 R. at 5-6.
    64 38 U.S.C. § 7261(a)(3)(A).
    65 38 C.F.R. § 20.1403(e) (2021).
    66 George, 991 F.3d at 1234 (emphasis in original); see Lamb, 22 Vet.App. at 234-35.
    67 991 F.3d at 1229.
    68 Id. at 1231-32, 1234, 1236.
    45
    that was explicitly reliant on the reasoning in this Court’s panel decision in Perciavalle v. Wilkie,
    32 Vet.App. 59, 65 (2019), that “a changed interpretation cannot be the ‘first commentary’ on a
    regulation or statute as there exists no precursor interpretation that the latter modifies, alters, or
    replaces.”69 The Federal Circuit’s reasoning was short and sweet:
    We recently clarified in George v. McDonough that whether an interpretation is a
    first interpretation or a change in existing interpretation, it cannot serve as a basis
    for CUE. 991 F.3d 1227, 1235 (Fed. Cir. 2021) (“That Wagner was the first judicial
    interpretation of § 111[1] by this court does not lead to a contrary result.”). We
    expressly noted in George that our precedent does not support the view that a new
    judicial pronouncement can retroactively apply to final decisions of the VA. Id. at
    1236-37. Consequently, we need not decide whether Saunders was an authoritative
    statement or a change in interpretation because neither can form the basis for
    CUE.[70]
    Labels aside, we see no daylight between the substance of the losing arguments in George
    and Steele and the substance of the argument that appellant and the majority make in this case. If
    there is a difference, it is a product of wordsmithing. Essentially, appellant and the majority want
    to have their cake and eat it too: They make pronouncements about what the law meant in 1971
    that derive from Esteban and the VA GC opinion, but they refuse to credit Esteban or the VA GC
    opinion for what now seems self-evident to them thanks to Esteban and the VA GC opinion. Never
    mind that the 1971 RO did not have the benefit of Esteban’s or the VA GC opinion’s interpretation
    when it made its decision. Such reasoning flies in the face of the retroactivity principles that the
    Federal Circuit reiterated in George and Steele, as well as the exact reasoning that the Federal
    69 The Steele court further provided:
    On appeal, Mr. Steele argues that the RO decision was CUE because it was not based on the law
    that applied at the time the decision was reached. Specifically, Mr. Steele contends that our holding
    in Saunders [v. Wilkie, 886 F.3d 1356 (Fed. Cir. 2018),] has retroactive application because it is not
    a statutory interpretation, but rather an authoritative statement of the law as it existed when the RO
    rendered its decision. According to Mr. Steele, unlike a statutory interpretation that provides a new
    understanding or meaning to the law, an authoritative statement has retroactive-like effect because
    it explains the meaning of the law existing before any subsequent interpretations. In support, Mr.
    Steele cites [this Court’s] decision in Perciavalle v. Wilkie, 32 Vet. App. 59 (2019), arguing that the
    decision distinguishes between an authoritative statement and a subsequent statutory interpretation.
    Appellant [Brief at] 7; see also Perciavalle, 32 Vet. App. at 65 (“[A] changed interpretation cannot
    be the ‘first commentary’ on a regulation or statute as there exists no precursor interpretation that the
    latter modifies, alters, or replaces[.]”). Mr. Steele therefore contends that the RO’s failure to apply
    Saunders resulted in a misapplication of the statute or regulation then in effect. We disagree.
    856 F. App’x at 881.
    70 Id.
    46
    Circuit used to reject the George and Steele claimants’ arguments. It does not matter how one
    characterizes the interpretation that appellant and the majority derive from Esteban or the VA GC
    opinion—that interpretation cannot form the basis of a CUE motion.
    C. Changed or Novel Interpretation
    But supposing we misunderstand the impact of George and Steele’s one-two punch, we
    would still hold that Esteban or the VA GC opinion constituted a change in interpretation under
    Jordan or a recent or novel interpretation under Lamb. And neither can support a CUE motion.
    Let’s begin with the change in the law Esteban wrought. Here, as in Jordan, “the
    interpretation of [38 C.F.R. § 4.14] has indeed changed, but because the [1971 rating] decision
    was final, Mr. [Perciavalle] has no recourse for appeal through a CUE [motion]; 38 C.F.R.
    § 20.1403(e) clearly proscribes Mr. [Perciavalle]’s claim.”71 In Jordan, the Federal Circuit held
    that § 20.1403(e) applied because a change in the interpretation of a statute had occurred.72 There
    a “regulation . . . provide[d] the first commentary on [the statute], and was therefore the initial
    interpretation.”73 The change in interpretation occurred when VA’s General Counsel issued an
    opinion “stating that [the regulation] . . . conflicted with the language of [the statute].”74 In other
    words, through the regulation, “VA [had] interpreted [the statute] contrary to its facially apparent
    meaning.” 75 Even so, the Federal Circuit held that a facially incorrect interpretation of the
    governing legal standard is still an interpretation for purposes of § 20.1403(e).76
    Here too, though our frame of reference is a regulation, not a statute as in Jordan, we can
    point to evidence of an initial interpretation: VA was denying separate ratings for different
    conditions seemingly related to the same injury or body part because it understood § 4.14 to
    preclude additional ratings as pyramiding.77 That informal interpretation of § 4.14 does not cease
    71 401 F.3d at 1299.
    72 Id. at 1298.
    73 Id.
    74 Id.
    75 Id. at 1297.
    76 Id. at 1298.
    77 E.g., Fanning v. Brown, 4 Vet.App. 225, 228, 230-31 (1993) (recognizing that “[t]he [Board] properly raised the
    issue of pyramiding” under § 4.14 when it denied separate ratings for a tender scar and the residuals of a right inguinal
    hernia, but remanding the case for an explanation of reasons or bases); Sweeney v. West, 16 Vet.App. 403 (table), 1999
    WL 184869, at *2 (denying an Equal Access to Justice Act (EAJA) application when “the Court [could not] find that
    the Secretary’s position at the administrative level was not substantially justified” because the VA General Counsel
    47
    to be the initial one just because, appellant would say, in those cases “VA interpreted [§ 4.14]
    contrary to its facially apparent meaning.”78 A change in the Agency’s interpretation of § 4.14
    (focusing on the injury or disability) occurred in 1994 when this Court in Esteban held that § 4.14
    permitted separate ratings when “none of the symptomatology for any one of [the] conditions [at
    issue] is duplicative of or overlapping with the symptomatology of the other . . . conditions.”79 VA
    lacked the benefit of that interpretation before Esteban. So, under § 20.1403(e) and as in Jordan,
    appellant cannot request CUE as a matter of law because the RO’s 1971 denial of appellant’s claim
    was otherwise correct under § 4.14 as VA seemingly initially understood and interpreted it and
    seems incorrect only based on a later and different interpretation of § 4.14.80
    The majority avoids this logic largely because it assumes that only a “formal” interpretation
    counts for purposes of § 20.1403(e).81 But nothing in § 20.1403(e) says so, and we would not read
    Precedent Opinion 23-97 “change in regulatory interpretation after the [Board] decision on appeal was the sole reason
    for the remand”); Sanders v. Principi, 3 Vet.App. 334, 336-37 (1992) (mem. dec.) (affirming the Board’s application
    of § 4.14 to rate multiple same-knee conditions under a single DC).
    78 Jordan, 401 F.3d at 1297-98.
    79 6 Vet.App. at 261-62 (emphasis in original), superseding Esteban v. Brown, 1993 U.S. Vet. App. LEXIS 545, at *1
    (Oct. 5, 1993) (affirming the Board’s finding that “the applicable regulations do not allow a veteran to obtain an
    increased rating by aggregating ratings for separate symptoms which collectively constitute one injury; therefore,
    appellant may not receive separate ratings for each of his symptoms”).
    Both appellant and the majority put a lot of weight on the VA GC 23-97 opinion’s statement that the General Counsel
    “kn[e]w of no formal position taken by the Veterans Benefits Administration on [the] issue” “whether symptomatology
    [sic] for a rating under DC 5257 is duplicative or overlap[]ping with the symptomatology for a rating under
    DC 5010 and 5003.” VA Gen. Coun. Prec. 23-97, at 1 (July 1, 1997) [hereinafter G.C. Prec. 23-97]. They understand
    that statement to mean that the General Counsel knew of no formal, initial interpretation of § 4.14. Appellant’s Br. at
    4; ante at 21. But theirs is not the only possible reading; one could alternatively read G.C. Prec. 23-97 to say that the
    General Counsel knew of no formal position taken on how Esteban’s interpretation of § 4.14 applied to DCs 5257,
    5010, and 5003. This reading is possible because G.C. Prec. 23-97 cites McAdory v. Brown as the case directing the
    Board to consider the issue. G.C. Prec. 23-97, at 1. And McAdory cited Esteban in its statement of the issue to be
    addressed on remand. No. 96-1172, 1997 WL 288699, at *2 (Vet. App. May 2, 1997). So then, G.C. Prec. 23-97 does
    not necessarily do the work that appellant and the majority would have it do.
    80 We do not maintain that one could never establish CUE based on the plain language of a regulation or statute even
    if a later judicial decision (or other authoritative source) adopted a given meaning of the regulation or statute based
    on its plain language. See Berger, 10 Vet.App. at 170. For example, one might show that the overwhelming Agency
    practice at the time of the decision in which CUE is alleged was in accord with the plain language. In other words,
    one could show that an individual adjudicator in a particular veteran’s case did not follow what the Agency generally
    understood the regulation or statute at issue to mean at the time, even though there was no judicial or authoritative
    Agency position on the matter. However, that is not the case here.
    81 See, e.g., ante at 17 (“Before analyzing the question presented—whether the rating schedule authorized separate
    ratings for arthritis and instability of the knee—the VA GC opinion readily acknowledged that this was the first agency
    interpretation, noting that: ‘We know of no formal position taken by the Veterans Benefits Administration on this
    issue.’ VA Gen. Coun. Prec. 23-97. This point is dispositive: there can be no ‘change’ in interpretation absent an
    existing interpretation from which a later interpretation deviates.”).
    48
    such a requirement into the regulation. In addition to the interpretive principle that courts should
    not add terms to a law, 82 it doesn’t make sense that merely because there is no “formal”
    interpretation there is no interpretation at all. And, in light of memorandum decisions we
    reviewed,83 along with the VA GC opinion and the outcome in appellant’s case, it appears that VA
    generally did not rate such conditions separately before the mid-to-late 1990s. Given the limited
    evidence on the matter, though, VA adjudicators may have applied the regulation inconsistently.
    Either way, Esteban or the VA GC opinion was a change by establishing a consistent Agency
    position.
    We would reach the same conclusion, that the Board did not err, even if we assumed that
    the majority was correct that there must have been a formal Agency position for there to have been
    a “change” in the law. The reason is our decision in Lamb, on which the Board relied.84 In Lamb
    we held in part, based on Jordan and § 20.1403(e), that “a recent or novel interpretation of law
    cannot be used to support a CUE motion, because any interpretation would not change what the
    law was understood to mean at the time of the original decision.”85 (Lamb’s holding is nothing
    revolutionary; it is a corollary of the well-established principle that “CUE must be analyzed based
    on the law as it was understood at the time of the original decision”86) In other words, under Lamb,
    appellant cannot support his CUE motion with his purportedly own present-day “interpretation” of
    the law that supports his entitlement to separate ratings—whether derived from Esteban or the VA
    General Counsel opinion or the regulations’ plain language—because if there was no preexisting
    interpretation, these are all recent or novel interpretations under Lamb. 87 And without those
    borrowed interpretations, appellant is left empty-handed. As we’ve already noted, and as the Board
    82 See Hardt v. Reliance Standard Life Ins. Co., 560 U.S. 242, 252 (2010) (declining to read into a statute an express
    limitation that Congress did not include in the statute’s plain language and holding that the Court of Appeals’ decision
    to add one “more closely resembles ‘invent[ing] a statute rather than interpret[ing] one'” (quoting
    Pasquantino v. United States, 544 U.S. 349, 359 (2005)).
    83 See supra note 55.
    84 R. at 6.
    85 Lamb, 22 Vet.App. at 235; see Berger, 10 Vet.App. at 170.
    86 George, 991 F.3d at 1234 (emphasis in original); see Lamb, 22 Vet.App. at 234-35.
    87 In Berger, this Court similarly held that Gregory v. Brown’s new, later-in-time interpretation of the law at issue was
    legally irrelevant to the Court’s consideration of the incorrect-application CUE motion because Gregory postdated the
    challenged RO interpretation, even though “[t]here was no exegesis or citation of authority for [Gregory’s]
    interpretation,” suggesting that this Court did not entertain a subsequent interpretation even if it purported to be based
    on a law’s plain language and meaning. 10 Vet.App. at 170 (citing Gregory, 5 Vet.App. 108, 112 (1993)).
    49
    noted, appellant points to no contemporaneous interpretation to show that in 1971 VA—or anyone
    else for that matter—understood the law to entitle him to separate ratings for arthritis and
    instability of the same knee.88 That was his burden to bear and he failed to carry it. All in all, the
    Board correctly applied Lamb and § 20.1403(e).89
    And so, even after a thorough review of the law and record, even after oral argument, even
    after reading the majority opinion, we still fail to see how appellant’s CUE theory is any different
    from a request for “a retroactive remedy for [a] final judgment[].”90 If appellant can proceed with
    his incorrect-application CUE theory without showing that in 1971 the law was understood to
    entitle him to separate ratings, the only remaining standard of comparison is the law as it is
    understood today. And to apply a present-day standard to a decades-old final decision is
    antithetical to CUE and the distinction between direct and collateral review. So, we cannot fault
    the Board for using the term “retroactive” in reference to appellant’s theory. And, let’s not forget:
    Before the Board, appellant argued that Esteban merely “clarified” what the law has always been
    and, “[t]herefore, there is no bar to the retroactive effect of Esteban,” citing a Federal Circuit
    decision regarding retroactivity.91 But no matter—call it retroactivity or not, the bottom line is that
    neither appellant’s CUE theory nor the Court’s holding of legal error “accord[s] sufficient respect
    to a final judgment.”92
    D. Board’s Reasoning
    One last note: some food for thought for the Board. There are times when a case requires
    a complex analysis. As much as decisionmakers might like, we can’t avoid wrestling with esoteric
    legal reasoning on occasion. But there are equally times when a case is straightforward as it appears
    at first blush. We think this is one of those cases. But the Board missed the simple approach, and
    that in many respects led to today’s en banc decision.
    88 R. at 5.
    89 R. at 5-6. The majority does not purport to overrule Lamb, so we take it that it still states what the law is.
    90 Jordan, 401 F.3d at 1299.
    91 R. at 14 (citing Princess Cruises, Inc., 397 F.3d 1358).
    92 Jordan, 401 F.3d at 1299.
    50
    In its decision on appeal, the Board relied on esoteric legal reasoning to deny appellant’s
    CUE motion as a matter of law.93 That reasoning led to an extensive legal debate in this Court,
    which yielded a holding that opens the floodgates to what we believe to be an entirely new class
    of CUE motions.
    And for what? Appellant cannot show that in 1971 his own record undebatably raised an
    issue of separate ratings—that in 1971 he had two conditions that could have been rated
    separately.94 The 1971 rating decision says nothing about the rating schedule or separate ratings
    or pyramiding. What it does say is, “The evidence does not warrant any change in the previous
    determination. Your left knee condition remains 10% disabling and you will continue to receive
    compensation for this disability.”95
    We cannot see how anyone could claim, let alone find clearly and unmistakably, that the
    1971 RO incorrectly applied the law extant at the time based on such a sparse explanation. As we
    see it, appellant’s CUE motion at its heart was nothing more than a dispute over how the 1971 RO
    weighed the evidence, which cannot be CUE.96 Had the Board said that or assumed, without
    deciding, that appellant’s CUE theory was legally tenable and concluded (or reasoned in the
    alternative) that appellant simply failed to show a clear and unmistakable error because the facts
    weren’t there, this en banc Court would have had no occasion to announce the legal rule it does
    today.97
    We are not criticizing the Board’s decision for the sake of criticizing. Rather, we are
    highlighting a dynamic that could go overlooked, one that could recur: The majority may have
    opened the CUE floodgates, but the Board unnecessarily and avoidably unlocked them.
    With all that said, we believe the Board did not err and we would affirm the decision on
    appeal on that basis. So, we concur only in the judgment of affirmance.
    93 R. at 6.
    94 Ante at 28-29.
    95 R. at 490 (emphasis added).
    96 Berger, 10 Vet.App. at 169 (“To the extent that the attack on a final decision is based on a disagreement with how
    the adjudicator weighed the facts, then it cannot constitute CUE.”).
    97 Judges Lauer and Jaquith recognize this fundamental point, but consider it only as a reason to find any error in the
    Board’s decision harmless. But we see the issue as more fundamental, going to the merits of the CUE assertion in the
    first place.
    51
    BARTLEY, Chief Judge, with whom PIETSCH and GREENBERG, Judges, join,
    dissenting in the judgment: I agree with Part IV of the decision, explaining that the Board erred in
    denying Mr. Perciavalle’s CUE motion as not legally valid. I disagree with Part V however, and
    write separately to explain why the Board error prejudiced the veteran and why this matter should
    be returned for the Board to conduct a proper CUE adjudication that addresses the specific CUE
    motion filed in March 2015—without erroneously characterizing it as a request for retroactive
    application of new law.
    The veteran’s non-attorney representative argued in the March 2015 CUE motion that the
    July 1971 RO decision contained CUE in failing to revise Mr. Perciavalle’s 10% disability
    evaluation under DC 5259, contemplating symptomatic removal of semilunar cartilage, to a 10%
    evaluation under DC 5257 for lateral knee instability. R. at 101-02. Mr. Perciavalle based this
    argument on the fact that the RO in 1971 noted that he experienced knee instability due to an inservice
    meniscectomy but failed to compensate him for it. Id. The non-attorney representative also
    argued that the July 1971 decision erred in failing to award Mr. Perciavalle a separate 10%
    evaluation under DC 5003-5260 for degenerative changes (arthritis) and limitation of flexion,
    citing 38 C.F.R. §§ 4.40, 4.45, 4.59, and 4.71a.98 Id. He argued that an x-ray report dated July 8,
    1971, which was before the RO prior to the 1971 decision, “clearly” showed degenerative changes
    in the veteran’s left knee. R. at 102; see R. at 492. The veteran has not as of this date received a
    decision responsive to this two-pronged CUE motion.
    In the September 2015 rating decision and May 2017 SOC, the RO found no CUE because
    it determined that a 1997 VA GC precedent opinion 99 concerning assignment of separate
    evaluations for instability and limitation of motion was not in effect in 1971. R. at 94, 58. Those
    adjudications did not engage the veteran’s arguments that at the time of the 1971 RO decision it
    was CUE under DC 5257 to fail to award a 10% evaluation for instability and also that it was CUE
    under DC 5003-5260 to fail to award a separate 10% evaluation for degenerative changes and
    limitation of flexion pursuant to §§ 4.40, 4.45, 4.59, and 4.71a. Instead of adjudicating the CUE
    motion based on the regulations and DCs cited by Mr. Perciavalle, which were in effect in 1971,
    98 Sections 4.40, 4.45, and 4.59 have not changed since 1971. Nor have the operative knee DCs.
    99 In the September 2015 decision, the RO mistakenly characterized this authority as a VA Office of Inspector
    General decision. R. at 94.
    52
    VA looked ahead 26 years to the issuance of the 1997 VA GC precedent opinion and made its
    decision to deny CUE on that basis.
    Before the Board, the veteran’s lay representative again argued as follows:
    We contend that the law has always permitted that a separate evaluation can be
    applied for DC 5257 based on instability and a separate evaluation can be applied
    for loss of range of motion under DC 5003-5260 or due to pain/functional loss
    under 38 CFR §§ 4.40 [and] 4.59. All regulations that applied to this claim have
    not materially changed since 1971.
    R. at 13-14. He argued that “corrective action should be taken . . . to correctly apply the law that
    was in existence at the time of the July 12, 1971, rating decision.” R. at 14.
    The September 2017 Board decision under consideration continued to misrepresent Mr.
    Perciavalle’s CUE argument, stating that “the Veteran contends that a more recent interpretation
    of VA regulations should have retroactive effect.” R. at 5. This is incorrect. The veteran in his
    CUE motion filed with the RO did not argue retroactive effect, R. at 101-02, nor did he make that
    argument in his NOD, R. at 76-77, or his Substantive Appeal, R. at 25. And in his informal brief
    to the Board, the veteran argued that, had the law in existence in 1971 been correctly applied, he
    would have received separate 10% evaluations. R. at 13. Nevertheless, the Board erroneously
    concluded that “[b]ecause a later interpretation of an existing regulation cannot constitute CUE[,]
    and that is the only basis on which the Veteran asserts CUE, the Veteran’s motion must be denied
    as a matter of law.” R. at 6 (emphasis added).
    Some of my colleagues assert that the Board cannot be faulted for characterizing Mr.
    Perciavalle’s argument as based on retroactivity. Ante at 39. They contend that no
    mischaracterization occurred because the veteran mentioned retroactivity in his informal brief to
    the Board. But Mr. Perciavalle’s single reference to retroactivity, in his informal Board brief, was
    preceded by this sentence: “We are not arguing the 1997 General Counsel opinion or any VA rules
    after the 1971 rating decision.” R. at 13 (emphasis removed). Thus, Mr. Perciavalle’s lone reference
    to retroactivity occurred when he attempted to counter the RO’s rationale for denying the CUE
    motion.
    And more important, even were we to consider this single reference to retroactivity as a
    CUE allegation, caselaw clearly prevented the Board from considering it. No retroactivity
    argument was presented to the RO and, since each theory of CUE is a separate and distinct matter,
    a veteran is not allowed to make a “new” CUE argument at the Board when challenging an RO
    decision. Andre v. Principi, 301 F.3d 1354, 1361 (Fed. Cir. 2002); Jarrell v. Nicholson,
    53
    20 Vet.App. 326, 332 (2006) (en banc). Because Mr. Perciavalle’s CUE motion, NOD, and VA
    Form 9 made absolutely no mention of retroactivity or argument concerning it, the Board erred in
    characterizing his CUE motion as based on retroactivity. Although a CUE movant may flesh out
    or recharacterize the argument, see Jordan v. Principi, 17 Vet.App. 261, 271 (2003), aff’d sub nom.
    Jordan v. Nicholson, 401 F.3d 1296 (Fed. Cir. 2005), the Board is not entitled to recharacterize a
    veteran’s CUE arguments to suit their own needs and shoehorn his arguments into a theory that is
    legally nonviable. This Court should not sanction such Board action.
    Some of my colleagues contend that regardless of the specific CUE arguments Mr.
    Perciavalle raised, his CUE motion was based on retroactivity in all but name. Ante at 43-46. They
    accuse the veteran of failing to explain why the regulations were susceptible only of a single
    interpretation, failing to explain why reasonable minds could not differ as to their correct
    interpretation, and failing to proffer a contemporaneous interpretation that reconciled the
    regulations in his favor. Ante at 41-42. But requiring a veteran who filed his CUE motion without
    assistance of counsel to include this degree of detail and minute specificity is antithetical to VA’s
    duty to sympathetically read pro se pleadings, including CUE motions. See Andrews v. Nicholson,
    421 F.3d 1278, 1282-83 (citing Roberson v. Principi, 251 F.3d 1378, 1384 (Fed. Cir. 2001);
    Szemraj v. Principi, 357 F.3d 1370, 1373 (Fed. Cir. 2004)).
    In addition, whatever my colleagues believe about the sufficiency of Mr. Perciavalle’s pro
    se CUE motion, the Board didn’t dismiss the motion as insufficiently pled. Instead, the Board, in
    contravention of its duty to “‘fully and sympathetically’ develop a veteran’s pro se CUE motion ‘to
    its optimum before deciding it on the merits,'” id. at 1282-83 (quoting Roberson, 251 F.3d at 1384),
    recharacterized the CUE motion as predicated on a legally nonviable basis and then denied the
    motion as a matter of law. Even if the lone reference to retroactivity created ambiguity as to the
    specific theory of CUE raised, the Board does not comport with “VA’s duty to sympathetically
    read a veteran’s pro se CUE motion to discern all potential claims,” id. at 1283, when they resolve
    any ambiguity that exists in a manner unfavorable to the veteran, and in a manner that precludes
    its adjudication on the merits.
    Moreover, my colleagues’ view is particularly dangerous and runs counter to the wellestablished
    CUE rule that, once a veteran alleges a specific CUE, VA will address only that CUE
    argument. In Acciola v. Peake, we warned that reading a CUE motion expansively prejudices a
    veteran because it could result in a broad res judicata effect that would hinder a further attempt at
    54
    revising a purportedly erroneous decision. 22 Vet.App. 320, 326 (2008). It is incomprehensible
    that VA would require sufficient specificity in a CUE motion and in fairness to the veteran be
    confined to deciding solely the specific CUE argument raised, only to have the RO and the Board
    recharacterize the veteran’s CUE argument to suit itself. There are untold risks ahead if VA is
    permitted to proceed in this manner. These risks include that a veteran’s CUE contention will fail
    to be heard and adjudicated, as occurred in Mr. Perciavalle’s case.
    Given the above circumstances, any conclusion that Board error did not harm Mr.
    Perciavalle cannot stand. Part IV of the decision concludes that the Board erred by misstating and
    failing to sympathetically construe the veteran’s motion, resulting in the Board failing to adjudicate
    the substance of the CUE motion. Despite this conclusion, my colleagues conclude in Part V that
    it is acceptable for this Court to undertake a merits analysis of the CUE motion in the first instance,
    under the guise of determining whether the Board error was harmless. But because the Board failed
    to adjudicate the CUE motion the veteran advanced, it is improper for us to proceed to address the
    merits of his motion, even under a harmless error analysis. “‘The rule of harmless error cannot be
    invoked to allow the [Court] to decide a matter that is assigned by statute to the [VA] for the initial
    determination.'” Tadlock v. McDonough, 5 F.4th 1327, 1337 (Fed. Cir. 2021) (quoting Winters v.
    Gober, 219 F.3d 1375, 1380 (Fed. Cir. 2000)). “[N]or can the rule be invoked to support an
    affirmance that ‘may [] have required it to make improper de novo findings of fact.'” Id. (quoting
    same). Yet in Part V my colleagues undertake the initial CUE analysis that is reserved to VA and
    do so despite acknowledging that the Board did not make factual findings to support an affirmance.
    Ante at 27.
    Aside from it being improper for the Court to address the substance of Mr. Perciavalle’s
    CUE motion in the first instance under the pretense of a harmless error analysis, see Tadlock,
    5 F.4th at 1337, there is a separate reason that my colleagues’ harmless error analysis fails. In
    Simmons v. Wilkie, this Court recognized that VA errors that affect essential fairness automatically
    prejudice veterans and claimants. These are errors that deprive veterans of a meaningful
    opportunity to participate and be heard in the adjudication process. 30 Vet.App. 267, 281-82
    (2018), aff’d, 964 F.3d 1381 (Fed. Cir. 2020). Simmons provided examples of these types of errors,
    including VA’s failure to provide notice of evidence necessary to substantiate a claim, see Vazquez-
    Flores v. Shinseki, 24 Vet.App. 94, 105-07 (2010), and VA’s failure to provide a required hearing,
    see Arneson v. Shinseki, 24 Vet.App. 379, 388-89 (2011). We determined that these types of errors
    55
    have the natural effect of depriving claimants of meaningful participation and the opportunity to
    be heard during the VA adjudicatory process. See Simmons, 30 Vet.App. at 281-82 (citing Shinseki
    v. Sanders, 556 U.S. 396, 411 (2009)); see also McDonough Power Equipment, Inc. v. Greenwood,
    464 U.S. 548, 553-54 (1984) (characterizing an error as prejudicial when it affects the essential
    fairness of the adjudication).
    In Mr. Perciavalle’s case we likewise confront an error that affects essential fairness—
    wrongly foreclosing a CUE motion by recharacterizing it as a request for retroactive application
    of law, which is barred by 38 C.F.R. § 20.1403(e) as a matter of law. That error automatically
    prejudiced the veteran because under that framework the Board was prevented from addressing
    and deciding the merits of Mr. Perciavalle’s March 2015 two-pronged CUE motion or
    substantively reviewing the DCs and VA regulations in existence in 1971 that were referenced
    therein. Instead, the Board concluded that Mr. Perciavalle was legally barred from making his
    allegations of error because, by definition, they could never amount to CUE. But contrary to the
    Board’s conclusion, Mr. Perciavalle’s CUE motion was not based on the retroactive application of
    a change in law—in fact, he specifically disavowed such argument. R. at 13.
    Veterans have a statutory right to a decision with respect to the claims and motions that
    they file. 38 U.S.C. § 511(a) (“The Secretary shall decide all questions of law and fact necessary
    to a decision by the Secretary under a law that affects the provision of benefits by the Secretary to
    veterans or the survivors or dependents of veterans.”). Veterans also have a right to one review on
    appeal to the Secretary. 38 U.S.C. § 7104(a). With the RO and the Board mischaracterizing his
    CUE motion, Mr. Perciavalle was deprived of his rights under sections 511(a) and 7104(a) and has
    yet to receive an adjudication of the two-pronged CUE motion that he filed and the one review on
    appeal to which he is entitled.
    These circumstances demonstrate that VA’s actions prejudiced Mr. Perciavalle and that
    essential fairness, an indispensable element of the VA adjudication process, was absent. Because
    VA’s process prejudiced the veteran, I do not agree with the Part V prejudicial error analysis,
    particularly as VA has yet to substantively address the merits of the CUE motion. Thus, I would
    remand the appeal for the Board to properly adjudicate the veteran’s CUE motion.

Powered by WordPress.com.