(dissenting):
Were it not for the lack of a counsel rights advisement on the record in this case, I would find the findings and sentence to be correct in law and fact and would, therefore, affirm. However, I do not believe that the record, in its present state, rationally supports an inference that the appellant was fully aware of his counsel rights at the time of trial. Accordingly, absent further proceedings, the record is not complete.
As does the majority I look to United States v. Johnson, 21 M.J. 211 (C.M.A. 1986), as the current statement of military *576law relating to an inadequate counsel rights advisement. As I read Johnson, the Donohew rule remains more or less intact. In fact, the Court expressly declined an invitation by government counsel to overrule Donohew. 21 M.J., at 214. In so doing both participating judges, writing separately, endorsed the concept that an accused who pursues a counsel rights advisement error on appeal might be required to state in what manner he had been prejudiced. As Judge Cox stated: “It is not unreasonable for us to put the burden on the accused to say, through counsel, to the appellate courts that he is dissatisfied with the results in his case; that he wants a new trial or a different result; and that he tell us why.” Id., at 217. I believe that the appellant in this case has clearly satisfied this threshold burden.
The majority, I submit, deals with this issue in entirely too summary a fashion. Nothing I am able to glean from Johnson permits us to resolve the counsel rights issue based on a review of post-trial affidavits. If we follow the suggestion by the Court of Military Appeals in inviting the appellant to state how he was harmed and, then, summarily dismiss his averments as not worthy of belief, we are engaging in “Catch 22” logic rather than dealing seriously with a meaningful appellate remedy. Johnson also suggests, and rather strongly I believe, that development of the facts necessary for a fair adjudication of the counsel rights issue might well be accomplished by a limited DuBay hearing directed by us. 21 M.J., at 215-216. Since the majority has not deemed it necessary to so direct, we should look at what the record tells us to support a conclusion that the appellant was aware of his right to individual military counsel at the time of trial.
What are we to make of the appellant’s averment that had he been aware of his right to request individual military counsel he would surely have done so in view of his dissatisfaction with his civilian counsel and his assigned military counsel? Is the statement self-serving? Of course, it is. Any statement designed to meet the threshold showing of prejudice test suggested by Judge Cox would necessarily be self-serving. By the same token, the counter-affidavits of the appellant’s detailed military counsel and the Article 32 Investigating Officer have self-serving aspects as well. If the detailed counsel had acknowledged that she failed to advise the appellant of his rights to counsel, this would be an admission that she neglected a defense counsel responsibility outlined in the Manual. R.C.M. 502(d)(6), Discussion (A). The Article 32 Investigating Officer, from a reading of his affidavit, had a similar responsibility implicit in a checklist he was using at the time he convened the investigation. I do not make these observations in any attempt to discredit the statements of either party, but merely to demonstrate that an appellate review of affidavits, without benefit of confrontation and cross-examination of live witnesses, is a poor way to resolve contested factual issues. See Chief Judge Everett’s discussion in Johnson, supra, 21 M.J., at 215.
Let us for the sake of argument view this matter in the manner most favorable to the government and accept the counter-affidavits as true. Does this clearly resolve the issue of the appellant’s knowledge of his counsel rights at trial? In my view, it does not. If the two counter-affidavits are true, the appellant has misstated, falsely overstated, or forgotten certain material facts relating to his knowledge of his counsel rights. Whatever the case, the counter-affidavits do not clearly address the issue of what the appellant’s knowledge was with respect to choice of counsel options at the time of trial. Several questions arise. If the appellant was advised at two distinct pretrial junctures as to his rights to counsel, was it reasonable for him to assume by virtue of these prior advisements that his right to state a preference for counsel extended up to the point of actual trial? On the other hand, might the lack of advisement at trial have supported an inference on his part that the time for exercising counsel options had passed? Was the appellant’s mental and emotional state at trial such that he was consciously considering such matters without the reinforcement of a fresh advisement? The ap*577pellant was facing grave charges involving alleged acts of sexual misconduct against his natural daughter when he appeared in court. He had little more than a week earlier been admitted to a military hospital following a suicide attempt or gesture. What can we reasonably infer was going through his mind during this terrible, traumatic period? Was he consciously evaluating his options? Did he even understand what his options were? I do not believe the record answers these questions for us.
May we rely upon any particular assumption of knowledge by the appellant because he was a senior noncommissioned officer who had undoubtedly attended briefings on the Uniform Code of Military Justice? I hesitate to do so. I suspect that all personnel learn quite early in their careers that an accused has a right to retain a civilian counsel if he or she faces a court-martial. Regardless of how many times one is briefed, I would also suspect that the right to request individual military counsel is somewhat less well known in the general military community and is likely to fade from one’s consciousness without reinforcement at an appropriate time. Donohew presupposes that all members brought before a court-martial require an advisement in court to eliminate uncertainty and to make it clear to all concerned that each accused has received proper advice. 39 C.M.R. 149, 152. Surely, given the magnitude of penal and emotional jeopardy which Master Sergeant Jack L. Saul faced on his initial day in court, he was an accused who required the fullest knowledge and understanding of his counsel rights. Can we be certain from this record that he had the benefit of such knowledge and understanding? Apparently the majority is sufficiently satisfied in this regard. I am not.
Is the appellant’s claim of prejudice preposterous on its face? I don’t believe that question can be answered affirmatively with any degree of assurance. The appellant’s claim that he was dissatisfied with the apparent preparatory efforts of his two counsel at trial is not a new averment which suddenly ripened in the light of the discovered counsel rights advisement deficiency. It is entirely consistent with his prior statements in support of his claim of inadequacy of counsel. While not directly applicable to the issue before us, it should be noted that the appellant’s misapprehensions concerning the quality of legal representation he was receiving were not entirely without merit. This much is clear from the majority’s analysis of the adequacy issue. If the appellant is playing the opportunist, he is at least not contradicting that which he had already placed before us before the counsel rights advisement issue arose. The record provides no basis for us to summarily dismiss his claim.
Has appellate insistence in the past for a counsel rights advisement in court been an example of form elevated over substance? See Chief Judge Everett’s discussion in Johnson, supra, 21 M.J., at 214. When we sat for oral argument in this case we posed a question of this general nature to the appellant’s civilian counsel. His response was noteworthy. To paraphrase, there is undeniable value in the military judge’s posing the counsel rights inquiry to an accused in open court. In terms of what the written transcript shows us at the appellate level there is seldom a major variance in such inquiries from one case to another. We can come to view this inquiry as a quaint and somewhat anachronistic ritual. However, an experienced military judge, having eye to eye contact with an accused, can normally determine whether such individual is evincing a true understanding of his counsel options. See Chief Judge Everett’s further discussion along this same line in Johnson, supra, also at 214. Problems can be resolved before they occur and become appellate issues, some of which may by then not be capable of being resolved. Regardless of what we say or what the Court of Military Appeals may eventually say on the precise issue before us, there is no substitute for doing it right the first time at the trial level.
What, then, should this court do at this point? In my view, we should attempt to perfect an incomplete record concerning the appellant’s knowledge of his counsel rights at trial. I am uncomfortable with a DuBay hearing when its purpose is to at*578tempt to recreate trial history as opposed to exploring an issue outside the trial record which might have had an impact on a given case. United States v. DuBay, 17 U.S.C.M.A. 147, 37 C.M.R. 411 (1967) (allegations of command influence which might have affected results of Army courts-martial). I am also concerned when a limited hearing is such as to create a potential conflict for a defense counsel in his roles as advocate for an accused and as an officer of the court. I do not perceive any problem with respect to the second issue in this case. Since I could affirm the findings of guilty in this case but for the counsel rights problem, I would order a DuBay hearing to consider evidence relevant to the question of the appellant’s knowledge of his right to request individual military counsel at the time of trial. This is the type of case which should not be retried unless the ends of justice absolutely require that it be done. I would not be prepared to set aside the findings and sentence and authorize a rehearing without first making an attempt to perfect the record.
I concur in Parts I, II and III of the majority opinion.