*600DISSENTING OPINION ON APPELLANT’S MOTION FOR REHEARING
CLINTON, Judge.Now that the Court has reexamined the genesis and tracked antecedents of Rule 81(b)(2) in Arnold et al. v. State, 784 S.W.2d 872 (Tex.Cr.App.1990), and denied motions for rehearing in those causes, I would grant rehearing here to reconsider our initial effort “to articulate a coherent standard for determining when an error is harmless” under Rule 81(b)(2), opinion, at 584, to the end that conceptual nuances may be reconciled for benefit of the bench and bar. Because the majority does not, I respectfully dissent.
My dissent is in two parts: first, a modified version of a portion of the opinion of the Court along the lines of its basic views and going with the flow of its “formulation” but revising format and editing content, from the last paragraph on page 584 to first one on page 588; second, an application of the first.
I
Although Rule 81(b)(2) has been cited many times, we have yet to articulate a coherent standard for determining when an error is harmless. The problem is that the rule is expressed in conelusory terms that may be read to admit subjective concerns rather than objective standards which must be explored to reach a legally correct resolution. In this regard it must be emphasized that the function of a harmless error analysis is not to determine how an appellate judge would have decided the facts, but to determine whether an error contributed to the conviction or the punishment. The purpose of the rule dictates that appellate responsibility transcends deciding whether the conviction was correct. See Traynor, supra n. 15, at 35-36.
In performing a harmless error analysis the easiest and consequently the most convenient approach one could employ is to determine whether the correct result was achieved despite the error.16 Or/ notwithstanding the error, in light of all the admissible evidence was the fact finder’s determination of guilt clearly correct? Stated another way, is there “overwhelming evidence” of guilt that was not tarnished by the error? This approach is incorrect because the language of the rule focuses upon the error and not the remaining evidence, and to review and weigh only “untainted” evidence makes an appellate court the factfinder, effectively substituting itself for the trier of fact.
The rule is derived from a passage in Fahy v. Connecticut, 375 U.S. 85, 84 S.Ct. 229, 11 L.Ed.2d 171 (1963). Reviewing a state’s determination of harmless error to admit evidence gathered in an illegal search and seizure, and finding that “the erroneous submission of this unconstitutionally obtained evidence ... was prejudicial [and] therefore, the error was not harmless,” the Supreme Court outlined the correct approach.
We are not concerned here with whether there was sufficient evidence on which the petitioner could have been convicted without the evidence complained of. The question is whether there is a reasonable possibility that the evidence complained of might have contributed to the conviction. To decide this question, it is necessary to review the facts of the case and the evidence adduced at trial.
Id., at 88, 84 S.Ct. at 231.17
As noted, Rule 81(b)(2) mandates that the appellate court focus upon the error and *601determine whether it contributed to the conviction or the punishment. Irrespective of the focus of the inquiry, it is impossible to gauge the significance of the error apart from the remaining properly admitted evidence. This approach obviously implicates a review of the evidence, but the concern is solely to trace the impact of the error. The untainted evidence is not to be weighed in its own right, nor is it to be examined to see if it is cumulative with the tainted evidence; it is to be considered only to uncover the potentially damaging ramification of the error. In other words, the impact of the error cannot be properly evaluated without examining its interaction with the other evidence.18
Those lessons are in Kotteakos and Fahy, and Chapman retaught them, also rejecting an “overwhelming evidence” standard. Chapman v. California, at 23, 87 S.Ct. at 827. Yet in Harrington v. California, 395 U.S. 250, 89 S.Ct. 1726, 23 L.Ed.2d 284 (1967), arguably the Supreme Court departed from their teachings when *602it found that, apart from “cumulative evidence” in erroneously admitted confessions, “the case against Harrington was so overwhelming that this violation of Bruton was harmless beyond a reasonable doubt.” Id., at 254, 89 S.Ct. at 1728. However, speaking through Justice Douglas, the Court made clear that its decision was on “these special facts,” id., at 253, 89 S.Ct. at 1728; that by the test of Chapman it could not “impute reversible weight to the two confessions,” id., at 254, 89 S.Ct. at 1728; that “[w]e do not depart from Chapman; nor do we dilute it. We reaffirm it,” ibid.; that it did not suggest “if evidence bearing on all ingredients of the crime is tendered, the use of cumulative evidence, though tainted, is harmless,” ibid., that its decision “is based on the evidence in this record [which] is so overwhelming that unless we say that no violation of Bruton can constitute harmless error [the conviction must be affirmed]," ibid.
In Schneble v. Florida, 405 U.S. 427, 92 S.Ct. 1056, 31 L.Ed.2d 340 (1972), the majority read all that to mean that where properly admitted evidence of guilt is overwhelming and prejudicial effect of code-fendants’ statements is insignificant by comparison, Bruton error is harmless, id., at 430, 92 S.Ct. at 1059, 31 L.Ed.2d at 344, and following such analysis it found insufficient prejudicial effect of inadmissible evidence of codefendant in light of defendant’s own confession; Justice Douglas, interestingly enough, was among the dissenters. Milton v. Wainwright, 407 U.S. 371, 92 S.Ct. 2174, 33 L.Ed.2d 1 (1972), involved a claimed violation of the Sixth Amendment under Massiah in that an officer pretending to be a fellow inmate testified to certain incriminating statements made by defendant; the majority found the error harmless because defendant himself had made three full confessions to other officers; again, however, Justice Douglas joined dissenters. More recently, in Delaware v. Van Arsdall, 475 U.S. 673, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986), the Supreme Court extended the rule to denial of Sixth Amendment right to crossexamine a prosecution witness about bias, included among five factors to be considered “overall strength” of prosecution’s case, and remanded the cause to state court to make a harmless error determination in the first instance. Contemporaneously, in Rose v. Clark, 478 U.S. 570, 106 S.Ct. 3101, 92 L.Ed.2d 460 (1986), the Court decided that Sandstrom error is subject to a harmless error analysis and an instruction erroneously shifting the burden on malicious intent may be harmless,' not on account of “overwhelming evidence” but because the charge still required the jury to find existence of those facts; it likewise remanded to the state court to make that determination. See Pope v. Illinois, 481 U.S. 497,107 S.Ct. 1918, 95 L.Ed.2d 439 (1988), adhering to Chapman test and applying it to erroneous instruction on an element of offense by ascertaining whether jury was precluded from considering that element and facts jury must have found show guilt beyond reasonable doubt; see also Carella v. California, 491 U.S.-, 109 S.Ct. 2419, 105 L.Ed.2d 218 (1989) (Justice Scalia dissenting) (mode of analysis may depend upon nature of error, e.g., certain jury instructions, even when evidence of guilt is overwhelming, id., 109 S.Ct. at 2421-2422).
Thus in the course of extending the Chapman standard to more and more kinds of constitutional errors, the Supreme Court is finding it cannot be applied like a bright line rule. Relevant considerations vary from problem to problem, and “overwhelming evidence” may resolve one yet have no bearing on solving another. In Harrington, Justice Douglas himself read Chapman to admonish “against giving too much emphasis to ‘overwhelming evidence’ of guilt,” 395 U.S., at 254, 89 S.Ct. at 1728, and other cases have practically ignored that circumstance.
In Satterwhite v. Texas, 486 U.S. 249, 108 S.Ct. 1792, 100 L.Ed.2d 284 (1988), the Supreme Court agreed with this Court that an Estelle v. Smith, 451 U.S. 454,101 S.Ct. 1866, 68 L.Ed.2d 359 (1981), error is subject to the harmless error rule. However, in reversing the sentence of death the Supreme Court disagreed with the conclusion that tainted testimony by Dr. Grigson was harmless beyond a reasonable doubt. *603Without citing any case dealing with harmless error, except Chapman v. California, supra, the Supreme Court noted substantial evidence to support the death penalty but focused on the impact of his testimony and its possible effect on the jury. Without undertaking to determine whether evidence was “overwhelming,” the Court stated:
The question, however, is not whether the legally admitted evidence was sufficient to support the death sentence, which we assume it was, but rather, whether the State has proved ‘beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained.’ Chapman, 386 U.S., at 24, 87 S.Ct., at 828.
Id., 108 S.Ct. at 1798.
Any perceived difficulty in reconciling various cases applying the harmless error rule from Chapman to Van Arsdall and Pope v. Illinois, all supra, is not occasioned by a changing standard of review, but from a failure to recognize that the manner of making an analysis of the facts of the case — not of the offense — will vary according to the error in question.19
Thus the most this Court can do to guide prospective harmless error analyses is provide a formulation of requisites under Rule 81(b)(2), present general considerations that may be relevant, and trust judges to use these observations in assaying what the error meant to other men and women reacting with reason.
An appellate court must not determine harmfulness of an error simply by examining whether there exists “overwhelming evidence” to support a finding of guilt. The impropriety of this standard has been explained:
a court that makes a finding of harmlessness under the overwhelming evidence test is not finding that the ... [error] did not in fact affect the verdict.
[[Image here]]
The court’s affirmance simply indicates its opinion that the untainted evidence is so overwhelming that if the jury had been compelled to rely on it alone, it would have convicted. In so holding, the court is not passing upon what the jury did; it is not determining the propriety of the evidence on which the jury relied. Because it is ruling instead upon what the jury would do if forced to rely on different evidence, it is substituting itself for the jury as factfinder.
Field, “Assessing the Harmfulness of Federal Constitutional Error — A Process in Need of a Rationale,” 125 Univ. of Penn. L.R. 15, 35 (1976).
Rather, the appellate court should calculate as best it can probable impact of the error on jurors in light of other evidence; “overwhelming evidence” may or may not play a role, depending on the particular error and the facts of the case and evidence adduced. The predominant concern must be the error and its likely impact. When the court rules that an error is harmless in *604essence it has found the nature of the error is such that it could not have affected jurors, so the jury must have relied on other evidence. When properly considered overwhelming evidence so dissipates a probable impact on minds of jurors reacting with reason, the error did not contribute to the verdict of guilt, and is harmless. Otherwise, it is not.
In summary, applying the harmless error rule, a reviewing court should not focus on propriety of outcome of trial. Instead, an appellate court should be concerned with impact of the thing done wrong, to determine whether the error might have affected deliberations of the jury. Accordingly, it must identify the source of the error, discern the nature of the error, examine the extent to which it was emphasized by the prosecution, estimate its gravity, and consider its probable consequences. The court is to assume evidence is sufficient to support the verdict, and not be distracted by assaying its weight; concentration must be on the error to judge whether it possibly influenced the jury. Again, it is the impact of the error, not other evidence, that must dictate whether the court can say beyond reasonable doubt that error made no contribution to the verdict.
With those general considerations identified, we now provide a skeleton on which to place them. The methodology is, first, to isolate the error and its effects in light of foregoing considerations and any others suggested by the facts of the particular case and, second, to judge whether the error and its effects might have influenced the minds of rational jurors in reaching their verdict. The Court must be able to declare a belief that it was harmless beyond a reasonable doubt. Chapman, at 24, 87 S.Ct. at 828; Satterwhite, 486 U.S. at 258, 108 S.Ct., at 1798.
II
Of course, for reasons developed in my dissent on original submission, I do not agree with the application of these standards the majority would make to the facts of this cause. That testimony of the teenage accomplice was corroborated, as indeed it must be for evidence to be sufficient, does not constitute “overwhelming evidence.” Even if discernible in the record, that “the State was not attempting to taint the trial process” and only made “passing reference” to them says nothing about impact of two extraneous felony offenses on minds of jurors. That it was “minimal” is still enough to influence the jury.
Accordingly, I would grant appellant's motion for rehearing.
. This is commonly referred to as the “overwhelming evidence" test. Under this test one “does not look to the tainted evidence, but to the untainted evidence, and asks whether it alone compels a verdict of guilty." Field, "Assessing the Harmfulness of Federal Constitutional Error — A Process in Need of a Rationale," 125 Univ. of Penn.L.R. 15 (1976).
. The Fahy analysis is reminiscent of a detailed exposition in Kotteakos v. United States, 328 U.S. 750, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946), construing federal “harmless error statute" in a general conspiracy case.
The statute commanded federal appellate courts in both civil and criminal cases to “give judgment after an examination of the entire record ... without regard to technical errors, defects or exceptions which do not affect the substantial rights of the parties." Id., at 757, 66 S.Ct. at 1244. Noting legislative history that the statute "affects only technical errors,” but if “its natural effect is to prejudice a litigant’s substantial rights, the burden of sustaining a verdict *601will ... rest upon the one who claims under it,” the Supreme Court observed:
Easier was the command to make than it has been always to observe. This, in part because it is general; but in part also because the discrimination it requires is one of judgment transcending confinement by formula or precise rule, [citation omitted]. * * * * Judgment, the play of impression and conviction along with intelligence, varies with judges and also with circumstance. What may be technical for one is substantial for another; what minor and unimportant in one setting crucial in another.
******
In the final analysis judgment in each case must be influenced by conviction resulting from examination of the proceedings in their entirety, tempered but not governed in any rigid sense of stare decisis by what has been done in similar situations, [citation omitted]. Necessarily the character of the proceeding, what is at stake upon its outcome, and the relation of the error asserted to casting the balance for decision on the case as a whole, are material factors in judgment.
Id., at 761-762, 66 S.Ct. at 1246.
In light of a congressional "fear of too easy relaxation of historic securities thrown around the citizen charged with crime,” id., at 762, 66 S.Ct. at 1246, the Supreme Court found help in the matter of making judgments in criminal causes.
Some aids to right judgment may be stated more safely in negative than in affirmative form. Thus, it is not the appellate court’s function to determine guilt or innocence, [citations omitted throughout]. Nor is it to speculate upon probable reconviction and decide according to how the speculation comes out. Appellate judges cannot escape such impressions. But they may not make them sole criteria for reversal or affirmance. Those judgments are exclusively for the jury....
But this does not mean that the appellate court can escape altogether taking account of the outcome. To weigh the error’s effect against the entire setting the record without relation to the verdict or judgment would be almost to work in a vacuum. In criminal causes that outcome is conviction. This is different, or may be, from guilt in fact. It is guilt in law, established by the judgment of laymen. And the question is, not were they right in their judgment, regardless of the error or its effect upon the verdict. It is rather what effect the error had or reasonably may be taken to have had upon the jury's decision. The crucial thing is the impact of the thing done wrong on the minds of other men, not on one’s own, in the total setting.
This must take account of what the error meant to them, not singled out and standing alone, but in relation to all else that happened. And one must judge others’ reactions not by his own, but with allowance for how others might react and not be regarded generally as acting without reason. This is the important difference, but one easy to ignore when the sense of guilt comes strongly from the record.
If, when all is said and done, the conviction is sure that the error did not influence the jury, or had but very slight effect, the verdict and judgment should stand, except perhaps where the departure is from a constitutional norm or a specific command of Congress.... The inquiry cannot be merely whether there was enough to support the result, apart from the phase affected by the error. It is rather, even so, whether the error itself had substantial influence. If so, or if one is left in grave doubt, the conviction cannot stand.
Id., at 763-765, 66 S.Ct. at 1247-48.
. It is important to note that in the context of a harmless error analysis the other evidence is the entire record. Unlike the dictates of Jackson v. Virginia, 443 U.S. 307, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979), in resolving the issue of harmfulness of an error an appellate court is obligated to examine the entire record in a neutral, impartial and even-handed manner and not "in the light most favorable to the prosecution....” Id., at 319, 99 S.Ct. at 2789. See Butler v. State, 769 S.W.2d 234 (Tex.Cr.App.1989). A review of the evidence in this manner is necessary because, for example, an error can be harmful when it has the effect of disparaging a defense, whereas if there is no defense the error could have been harmless, at least in that respect.
. To suggest that any such difficulty follows “from changes in the Court's personnel and the relative emphasis that individual justices have placed on relevant considerations of the extent of harm [that] are a natural reflection of an inherently subjective process” is to misread the Court, to confuse the process. In Kotteakos, supra, at 761, 66 S.Ct. at 1246, the Court did introduce its discussion of proper exercise of judicial judgment with observations that "the play of impression and conviction along with intelligence, varies with judges and also with circumstance [and that] [w]hat may be technical for one is substantial with another; what minor and unimportant in one setting crucial in another." However, the Court went on to develop considerations involved in the process, and ultimately to reject the notion that it is "inherently subjective," viz:
... And the question is not were [jurors] right in their judgment regardless of the error or its effect upon the verdict. It is rather what effect the error had or reasonably may be taken to have had upon the jury’s decision. The crucial thing is the impact of the thing done wrong on the minds of other men, not on one’s own, in the total setting, [citations omitted].
This must take account of what the error meant to them, not singled out and standing alone, but in relation to all else that happened. And one must judge others’ reactions not by his own, but with allowance for how others might react and not be regarded generally as acting without reason. This is the important difference, but one easy to ignore when the sense of guilt comes strongly from the record.
Id., at 764, 66 S.Ct. at 1247.