concurring.
I agree with the result reached by the majority. I also agree with the majority’s reasoning, with one exception: I do not agree that the trial court’s violation of article 35.23 of the Code of Criminal Procedure (allowing the jury to separate) is not subject to harm analysis under Rule 81(b)(2) of the Rules of Appellate Procedure. I have previously stated my view of the scope of Rule 81(b)(2):
It seems to [me] that the “harm analysis” mandated by Rule 81(b)(2) can be applied to any error without fear of diminishing an appellant’s rights. If the appellate record in a particular case does not contain enough data from which to gauge or quantify the effect of an error, then the reviewing court obviously could not determine beyond a reasonable doubt that the error made no contribution to the conviction or punishment. Accordingly, Rule 81(b)(2) would require that the conviction in such a case be reversed. This approach strikes [me] as far simpler to administer and apply than one in which reviewing courts are left to divine which types of error are subject to a harm analysis and which are not.
Marin v. State, 801 S.W.2d 944, 947 n. 3 (Tex.App.1990, pet. granted). I continue to hold to that view.
Not being confident of their conclusion that the error in the present case is not subject to a harm analysis, the majority proceeds to apply such a harm analysis under Rule 81(b)(2). The ease with which *97they apply Rule 81(b)(2) and reach the correct result demonstrates my point:
We do not find the Harris [v. State, 790 S.W.2d 568 (Tex.Crim.App.1989),] guidelines to be of much assistance in light of the nature of the error and the record in the instant case. The uncertainty would counsel reversal. We cannot determine beyond a reasonable doubt that the jury separation error made no contribution to the conviction.
Opinion at 96. I agree. And I can think of no type of error to which a harm analysis could not be applied with the same ease.
Moreover, although the Court of Criminal Appeals does not seem inclined to apply a harm analysis to all types of error, see, e.g., Sodipo v. State, 815 S.W.2d 551 (Tex.Crim.App.1990) (opinion on rehearing), I see no reason not to apply it in the present case. First, the mere act of allowing the jury to separate after the charge has been given does not, in my opinion, render a trial so fundamentally unfair that it “cannot reliably serve its function as a vehicle for determination of guilt or innocence.” Sodipo, 815 S.W.2d at 554 (quoting from Rose v. Clark, 478 U.S. 570, 577-78, 106 S.Ct. 3101, 3106, 92 L.Ed.2d 460 (1986)). Indeed, article 35.23 apparently permits such separation unless one of the parties specifically moves that the members of the jury be kept together.
Second, to hold that a particular type of error is not subject to a harm analysis is to conclude that there is no possible set of circumstances in which a reviewing court could ever determine beyond a reasonable doubt that that type of error made no contribution to the conviction or punishment. I do not claim to have such prescience as to be able to reach that conclusion with respect to violations of article 35.23.
Nor, apparently, does the majority — they are careful to note that their rejection of a harm analysis in the present case “does not mean that every violation of article 35.23 is to be excluded from a harm analysis under Rule 81(b)(2). Our holding is limited to the violation of article 35.23 as shown by the particular facts and circumstances of the instant case.” Opinion at 96. With all due respect, a violation of article 35.23 either is or is not subject to a harm analysis. The majority’s intimation that some violations of article 35.23 might be subject to a harm analysis sounds suspiciously like a conclusion that, under the particular circumstances of the present case, the error cannot be determined to be harmless. That, however, is the very application of a harm analysis that the majority rejects.