Full opinion text
OPINION ON PETITIONS FOR DISCRETIONARY REVIEW CLINTON, Judge. These causes present various aspects of application of Tex.R.App.Pro.Rule 81(b)(2) to “statutory” error at the punishment stage in charging a jury on parole law pursuant to Article 37.07, § 4, V.A.C.C.P. Rose v. State, 752 S.W.2d 529 (Tex.Cr.App.1987-1988). “Rose error” is giving an instruction in terms of the constitutionally infirm statute. Id., at 535, 537 and 553-554. Finding that courts of appeals are conducting disparate harm analyses of Rose error, we granted review in these causes and consolidated them for submission on briefs and oral argument in order for this Court to address recurring problems in appellate review of Rose error. Tex.R.App. Pro.Rule 200(c)(1), (2) and (6). We begin with pertinent basic underlying propositions, then discuss various germane considerations and finally apply them to specific situations at issue. I Unlike the rule governing reversal of judgment in a civil cause, Rule 81(b)(2) provides: “If the appellate record in a criminal case reveals error in the proceedings below, the appellate court SHALL reverse the judgment under review, unless the appellate court determines beyond a reasonable doubt that the error made no contribution to the conviction or to the punishment.” To the fullest extent of legislatively granted authority, this Court created the rule in interest of consistency and our formulation of the “unless” clause is taken practically verbatim from language in Fahy v. Connecticut, 375 U.S. 85, 84 S.Ct. 229, 11 L.Ed.2d 171 (1963), that the Supreme Court isolated and iterated in Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967), in fashioning its rule for determining when “a federal constitutional error can be held harmless,” viz: There is little, if any, difference between our statement in Fahy v. Connecticut about ‘whether there is a reasonable possibility that the evidence complained of might have contributed to the conviction’ and requiring the beneficiary of constitutional error to prove beyond a reasonable doubt that the error com plained of did not contribute to the verdict obtained. We, therefore, adhere to the meaning of our Fahy case when we hold, as we do now, that before a federal constitutional error can be held harmless, the court must be able to declare a belief that it was harmless beyond a reasonable doubt." Id., at 24-26, 87 S.Ct., at 828-829, 17 L.Ed.2d, at 710-711. See Mallory v. State, 752 S.W.2d 566, at 569 (Tex.Cr.App.1988) (Rule 81(b)(2) eliminated elective propriety of expressing test for harmless error in a less than uniform fashion); Harris v. State (Tex.Cr.App. No. 69,366, delivered June 28, 1989, motion for rehearing pending) (Rule 81(b)(2) is “rhetorical and semantic equivalent of the harmless error standard announced by the Supreme Court for constitutional errors in Chapman v. California," id., majority slip opinion at 29); Bennett v. State, 766 S.W.2d 227, at 229, n. 7 (Tex.Cr.App.1989) (Rule 81(b)(2) is codified progeny of Chapman v. California harmless error analysis). Applying the federal rule to Texas punishment proceedings are Burgett v. Texas, 389 U.S. 109, 115, 88 S.Ct. 258, 262, 19 L.Ed.2d 319 (1967) (admission of presumptively void prior conviction not cured by instruction to disregard, nor harmless within meaning of Chapman v. California, supra), and the opinion in Satterwhite v. Texas, 486 U.S. 249, 108 S.Ct. 1792, 100 L.Ed.2d 284 (1988); see also our own decisions preceding Rule 81(b)(2) in, e.g., Maynard v. State, 685 S.W.2d 60, at 67-68 (Tex.Cr.App.1985), Clemons v. State, 605 S.W.2d 567, at 571-572 (Tex.Cr.App.1980), and Jordan v. State, 576 S.W.2d 825, at 829-830 (Tex.Cr.App.1978). Thus while previously this Court has often dealt with harm in jury misconduct implicating parole law, insofar as making an analysis under Rule 81(b)(2) to determine likelihood of harm resulting from an unconstitutional instruction pursuant to Article 37.07, § 4, appellate courts confront a task with little precedent. We do know the rule is applicable and that as beneficiary of the error the State has the burden to show beyond a reasonable doubt that the error did not contribute to the verdict on punishment. Chapman v. California, supra, 386 U.S. at 24, 87 S.Ct. at 828, 17 L.Ed.2d, at 710; Satterwhite v. Texas, supra, 486 U.S. at 259, 108 S.Ct., at 1798, 100 L.Ed.2d, at 295; Harris v. State, supra, (Clinton, J., dissenting at 10); see Hargraves v. State, 738 S.W.2d 743, at 749 (Tex.App.—Dallas 1987) PDR refused. Thus an appellate court must be able to find an error harmless beyond a reasonable doubt. Rule 81(b)(2); Chapman v. California, supra. II A Where the evil produced by a constitutional violation at trial is limited in scope to erroneous admission of particular evidence, an impermissible comment or a flawed instruction on guilt, usually a reviewing court may undertake with some confidence its task of assessing the likelihood that the error materially affected deliberations of the jury. Satterwhite, supra, 486 U.S. at 257, 108 S.Ct., at 1797-1798, 100 L.Ed.2d, at 294-295; Holloway v. Arkansas, 435 U.S. 475, at 490-491, 98 S.Ct. 1173, at 1182, 55 L.Ed.2d 426, at 438 (1978); see Carella v. California, 491 U.S. -, 109 S.Ct. 2419, 105 L.Ed.2d 218 (1989). However, the scope of Rose error may not be so readily discerned, or its consequences so easily assessed. Rose, supra, at 537 and 554; cf. Satterwhite, supra, 486 U.S. at 256, 108 S.Ct., at 1797, 100 L.Ed.2d, at 1798. In all felony cases where a jury determines punishment, there is not an “issue” as such for it to decide; indeed common practice is to admonish the jury that “it will not be proper for you in determining the penalty to be assessed to fix the same by ... any method other than a full, fair, and free exercise of the opinion of individual jurors under the evidence admitted before you.” Texas Criminal Pattern Jury Charges (State Bar of Texas 1975), § CPJC 12.42((a)-(c), at 84, and § CPJC 12.42(d), at 87; McCormick & Blackwell, Texas Criminal Forms and Trial Manual, § 81.05, 8 Texas Practice 270; McClung, Jury Charges for Texas Criminal Practice (Rev. Ed.1988) 242. Intruding into its wide area of discretion, the Legislature commanded that the jury be instructed about the parole law — “the court shall charge the jury in writing as follows” — as applicable to the status of defendant depending upon its verdict of guilt, an affirmative finding of a deadly weapon and allegations of prior conviction for enhancement. Article 37.07, § 4(a), (b) and (c). “The evil to be avoided is the consideration by the jury of parole in assessing punishment.” Rose v. State, 752 S.W.2d 529, at 535, quoting Clark v. State, 643 S.W.2d 723, 725 (Tex.Cr.App.1982). Thus the task of a reviewing court is to make “an intelligent judgment” about whether the unconstitutional instruction “might have affected [or influenced]” deliberations of the jury on punishment. Satterwhite, supra, 486 U.S. at 258, 108 S.Ct., at 1798, 1799, 100 L.Ed.2d, at 295, 296; see Maynard v. State, supra, at 67-68; Clemons v. State, supra, at 571-572; Jordan v. State, supra, at 830; Spelling v. State, 768 S.W.2d 949 (Tex.App.—Fort Worth 1989) PDR pending (Keltner, J., dissenting) (“whether the trial court’s instruction on parole subjectively influenced the jury’s verdict”). Again, under Rule 81(b)(2) the burden of negativing any such influence beyond a reasonable doubt is on the State. See ante, at 298. As we analyzed § 4 in Rose, in each subsection the statute mandates the trial court to inform the jury: first: “Under the law applicable to this case, the defendant... may earn time off the sentence imposed through the award of good conduct time;” second: that length of imprisonment “might be reduced by the award of parole;” third: “Under the law applicable to this case,” there is a precise formula (with or without good conduct time) to determine when “the defendant” will become eligible for parole; fourth: While one cannot accurately predict “how the parole law and good conduct time might be applied to this defendant,” because that “will depend on decisions made by prison and parole authorities;” still, fifth: jurors are instructed: “You may CONSIDER the existence of the parole law and good conduct time.” Rose, at 535. That is to say, when it comes to assess punishment the jury may deliberate on the content of what the trial court has just explained in the preceding four paragraphs in making its decision as to the number of years it will assess as punishment. Ibid; Rogriguez v. State, 762 S.W.2d 727, at 733 (San Antonio 1988) no PDR (instruction encourages jury to consider existence of parole law and good conduct time); Olivarez v. State, 756 S.W.2d 113, at 115-116 (Tex.App.—San Antonio 1988) no PDR (jury allowed to deliberate on content of first four paragraphs of statutory parole law charge). That the jury is thereafter told not to consider “the extent to which good conduct time may be awarded or forfeited” or “the manner in which the parole law may be applied” to the defendant, we held to be of “no constitutional consequence,” because jurors “had already been instructed that they may CONSIDER the stated explanation of parole law and good conduct time.” Ibid. Obviously, the only purpose in allowing jurors to consider “existence” of parole law and good conduct time is to inform their assessment of punishment; the instruction is “clearly designed to increase [the] sentence,” Gabriel v. State, 756 S.W.2d 68, at 70 (Tex.App.—Houston [1st] 1988) no PDR. In considering such “existence,” however, jurors must not attempt to “predict” how prison and parole authorities, respectively, might treat good conduct time and apply the parole law. In short, the jury can take the declarations into account in considering punishment, but without regard to what authorities may later do with the defendant. The vice in a § 4 instruction is creating an environment for trial participants to induce harm. Because it purports to convey statements of law (inapplicable to any fact in the case, Article 36.14. V.A.C.C.P.), evidence relevant to its declarations is barred by decisional rules and not permitted by § 4(d), and consequently counsel seldom have a real factual basis for arguing those matters to the jury, Rose, supra, at 537, the impact of a § 4 instruction is virtually immune from normal appellate review: that jurors actually did consider parole law and good conduct time is rarely evident in the record. See Rose, at 537, 554; Gabriel v. State, supra, at 69-70; see Brooks v. State, 768 S.W.2d 481 (Tex.App.—Houston [1st] 1989 (O’Connor, J., dissenting), and Satterwhite v. Texas, supra (Marshall, J., dissenting, 486 U.S. at 262, 108 S.Ct., at 1800, 100 L.Ed.2d, at 297-298 (because few “intangibles” jury might consider can be gleaned from record, appellate court “ill-equipped to evaluate effect of constitutional error on sentencing determination”). Yet it is within that framework that a harmless error analysis must be made. Otherwise any inquiry would be purely speculative such that a harm analysis would be inappropriate. Satterwhite v. Texas, supra, at 256, 108 S.Ct., at 1797, 100 L.Ed.2d, at 294; Sorrell v. State, 74 Tex.Cr.R. 505, 169 S.W. 299, at 303 (1914). But as experience demonstrates, best likelihood is that a jury will consider the “existence,” and thereby assess a term of years it believes may ensure the defendant serves more than the minimum term prescribed regardless of what prison and parole authorities may later decide. See Rose, at 536; Blackwell v. State, 768 S.W.2d 9, at 12 (Tex.App.—Fort Worth 1989) no PDR; Martin v. State, 768 S.W.2d 12, at 13 (Tex.App.—Fort Worth 1989) no PDR; Escobar v. State, 770 S.W.2d 24 (Tex.App.—Dallas 1989) PDR pending; Rodriguez v. State, 762 S.W.2d 727, at 732-733) PDR granted on other grounds; Zwack v. State, 757 S.W.2d 66 at 71-72 (Tex.App.—Houston [14th] 1988) no PDR; Gil v. State, 756 S.W.2d 108, at 109 (Tex.App.—Austin 1988) no PDR; Austin v. State, 748 S.W.2d 546 (Tex.App.—Beaumont 1988) PDR refused. A harm analysis should be based on that premise. See, e.g., Olivarez v. State, 756 S.W.2d 113, at 114-115 (Tex.App.—San Antonio 1988) no PDR. With that premise, a reviewing court must examine the record for indicia of factors reasonably conducing to affect minds of average rational jurors in their determination of punishment, the ultimate inquiry being whether it is impossible to say beyond a reasonable doubt that considering declarations made by the trial court in its § 4 instruction law did not influence the jury adversely to appellant in assessing punishment. B The leading opinion on rehearing in Rose conducted its analysis on inability “to know what process the jury underwent” and looked to “factors which indicate that the error was harmless.” Id., at 554-555. But we do know from other cases there are other revelations for examination pertinent to the inquiry. We will discuss them in order of trial. 1 First is voir dire examination; when reported in the statement of facts a few courts of appeals routinely review it, only to find either that parole was not mentioned or that certain references to it are not all that significant. But in our view, endeavoring to “qualify” prospective jurors on a § 4 instruction introduces them to a constitutionally forbidden exercise, tends to lay a predicate for further discussion of such matters during the proceedings and carries a high risk of inducing jurors to engage in it during deliberations on punishment. For example, see Johnson v. State, 768 S.W.2d 788 (Tex.App.—Dallas) PDR refused. 2 Since Rose an objection is not required to preserve error, id., at 552-553; that § 4 mandates a prescribed instruction is calculated to render it unobjectionable, and thus might have dissuaded defense counsel from making an objection. (To request it is another matter.) But nevertheless that an instruction was submitted over objection serves to dispel any speculative notion that counsel might be seeking to gain some advantage from the instruction. See Gabriel v. State, supra, at 70. That an objection is overruled practically invites legitimate comment, which brings us to argument of counsel. 3 Although the leading opinion in Rose said that counsel are prohibited from arguing the matter, id,, at 537, the cases demonstrate that it is not uncommon for counsel, usually a prosecutor, to address the subject. Moreover, we now agree that “it was not error for the State or the defendant to argue the law of parole to the jury,” Blackwell v. State, supra, at 11. Therefore, it matters not which counsel first “opens” the subject of parole, and consequently an “opener” does not necessarily “invite” opposing counsel to respond in kind or otherwise, such that a reviewing court can say “a defendant may not create reversible error by his own manipulations,” Escobar v. State, at 27; it is not error and, if not improper, it is not a manipulative invitation. However, there is an important caveat: while an argument may not be error per se, one made in terms tending to induce consideration of the eligibility formula and other teachings of a § 4 instruction compounds Rose error and may influence the jury in its deliberations on punishment. Rodriguez v. State, 759 S.W.2d 141 (Tex.Cr.App.1988), remanding for reconsideration Rodriguez v. State, 721 S.W.2d 504, at 507-508 (Tex.App.—Houston [14th] 1986); Jones v. State, 725 S.W.2d 770, at 772 (Tex.App.—Dallas 1987); see Satterwhite v. Texas, supra, 486 U.S. at 260, 108 S.Ct., at 1799, 100 L.Ed.2d, at 296 (“district attorney highlighted Dr. Grigson’s credentials and conclusions in his closing argument”). A classic example of an offending argument by a prosecutor is found in Moore v. State (Tex.App.—Dallas No. 05-86-00664-CR, delivered April 10, 1989) PDR refused, (edited somewhat here for brevity, and by deleting emphases and other notations) viz: “The first thing I want to get through right away is this parole charge.... Now, we can’t say what might happen ... to the defendant, because we can’t read the crystal ball. But it says that the defendant ... or person in a case like this, of first-degree murder, with a deadly weapon, could be released on parole when his time served equals a third or twenty, whichever comes first.... Let’s take an example of a sixty-year sentence. Its a third or twenty, whichever comes first. Well, a third of sixty is twenty. Twenty years in the penitentiary before he becomes eligible for parole. I bet you thought, before you heard that, that a sixty-year sentence meant sixty calendar years in the Texas Department of Corrections. As the Judge has given you in his instructions, he’s entitled to be — a person charged in this type of case is entitled to be paroled when they have served twenty years. So, really, those sentences are a joke. [Judge injects, “He’s eligible to be paroled [not entitled];” counsel corrects himself: “I’m sorry, eligible.... No one is entitled to parole, but eventually everyone serves out their sentence, one way or the other.”] So, bear that in mind. Whatever sentence you give, for instance, if its a thirty-year sentence, then that means ten before they become eligible for parole. So that’s why I’m asking for a maximum sentence in this case, because it will take a sixty-year sentence or better to keep this man in the penitentiary for at least twenty calendar years. Consider ninety-five or ninety, if you think he’s entitled to more consideration than that, maybe you’ll consider eighty-five years. But always bear in mind what that one third sentence rule means, because that is where the real impact of your verdict will be felt. Thank you.” Id., slip opinion at 4-5. The jury assessed life; on the strength of that argument the Dallas Court reversed as to punishment. Accord: Woods v. State, 766 S.W.2d 328, at 329-340 (Tex.App.—Houston [14th] 1989) no PDR; Howell v. State, 757 S.W.2d 513, at 518 (Tex.App.—Houston [1st] 1988) PDR refused; Morris v. State, 755 S.W.2d 505, at 511 (Tex.App.—Houston [1st] 1988) PDR refused. At the other end of the spectrum is, e.g., Barger v. State (Tex.App.—Dallas No. 05-85-01242, delivered April 28, 1989) PDR pending (offense is murder; parole not specifically mentioned in argument; prosecutor told jury to look at page four of charge (containing § 4 instruction), that it was “important”); Pope v. State (Tex.App.—Dallas No. 05-86-00065-CR, delivered March 27, 1989 (defense counsel mentioned life means serving at least twenty years in argument to lessen punishment). Somewhere in between are such cases such as Bonner v. State, 779 S.W.2d 81 (Tex.App.—Houston [1st] 1989) PDR pending. The offense is murder; defense counsel “emphasized to the jury that they were not permitted to discuss parole law in reaching their decision on punishment;” in his closing argument the prosecutor seemed ambivalent and commented somewhat ambiguously, viz: “... I’m not going to say a word on the parole laws. If I knew what they were, which I don’t — the charge controls it. He’s told you about that. But I want you to consider what the charge has to say. And that’s it. But I want you to realize that parole is there. That’s it. That’s all there is to it. I don’t have any control over it. And he doesn’t have any control over it. The legislature doesn’t have any control over it. Really, you can send a message to them. You can send a message to the community; that is that the penalty for crime such as this is grievous, grievous, as it should be.” The jury assessed a term of fifteen years. Compare, however, Williams v. State (Tex.App.—Dallas No. 05-86-00439-CR, delivered February 1, 1989) PDR pending (former “curative instruction” not given; prosecutor mentioned parole and good conduct time under § 4(b), told jury to “think about” how long defendant would serve, but argument “confusing,” so in light of “mid-range sentence” no harm). The conviction is for indecency with a child enhanced by a prior conviction; although not noted by the Dallas Court, at punishment the prosecution waived opening argument and, ending his, defense counsel alluded to parole before suggesting a term “somewhere between five and ten years;” in closing the prosecutor traced the prior criminal record of defendant, reprised the facts of the case and concluded by picking up on defense counsel’s submission to ask for life. (Jurors sent out a note, and for more on that incident, see post). Then it fixed punishment at fifty years and no fine. Id., slip opinion at 4-5. A § 4 instruction is sui generis, and we do not see a common statewide pattern for jury argument about it. Counsel who choose to address its declarations and their ramifications can create implications of harm, either by their remarks alone or coupled with other indicia in the record. See, e.g., Miller v. State, 772 S.W.2d 491 (Tex.App.—Dallas 1989) PDR refused; Shaw v. State, (Tex.App.—Fort Worth Nos. 2-86-036-CR & 2-86-073-CR, delivered April 27, 1989) PDR pending; Johnson v. State, 768 S.W.2d 788 (Tex.App.—Dallas 1989) PDR refused; Woods v. State, 766 S.W.2d 328 (Tex.App.—Houston [14th] 1989), no PDR. 4 Regardless of admonitions, argument or other incidents may induce jurors to consider and apply parole law and good conduct time in assessing punishment, and again the cases are replete with proof of violations. See Rose, at 356. Which brings us to another factor indicative of harm not extant in Rose, namely, the “jury note.” Our appellate courts appear to be reacting variously to a “jury note” and evaluating it in different lights. Compare, e.g., Martin v. State, supra, (note “clear indication” jury considered parole instruction) with, e.g., Escobar v. State, supra, (presumption that jury followed instructions not rebutted). In Reyes v. State (Tex.App.—Dallas No. 05-86-00037-CR, delivered April 26, 1989), the prosecutor barely alluded to the § 4 instruction, viz: “I’m going to ask you to return a life sentence. You have got the parole law in that charge to look at. He deserves all that you can give him. Unfortunately, under the law, the most you can do is a life sentence. And you know that you can do whatever you want to do with it, but that’s what I’m going to ask you to do in this case.” Id., slip opinion, at 6. Nonetheless the jury sent a note, viz: “We need clarification of the parole options as applied to life, 99 years, 60 years sentence.” See also, e.g., Rodriguez v. State, 762 S.W.2d 727, at 732 (Tex.App.—San Antonio 1988) (“Please clarify the para, that begins ‘you may consider the existence of the parole law and good conduct time.’ ”) punishment set aside, PDR granted on other grounds; Hawkins v. State, 766 S.W.2d 840, at 842 (Tex.App.—Beaumont 1989) (“Any possibility of parole with a life sentence?”) punishment vacated, PDR refused. Similarly, juries want definition of “punishment for life,” Garcia v. State, 1988 WL 131815 (Tex.App.—Houston [1st] No. 01-87-00708-CR, delivered December 8, 1988) (slip opinion at 6) PDR pending; explanation of “the differences between a ‘life’ sentence and ‘99 years,’ ” Williams v. State, supra; whether “terms 99 years and life [are] considered to be one and the same,” Weatherall v. State (Tex.App.—Dallas No. 05-86-00003-CR, delivered March 20, 1989) (slip opinion at 5) PDR pending. Patently when a jury sends out a note making an inquiry of some sort related to parole, it reveals that jurors are then and there “discussing” and “considering” the subject. Hawkins v. State, supra, at 842, 843, 844; Rodriguez v. State, supra, at 732-733 (apparent jury “did just that and sought further clarification”). Such an important factor in the harm analysis cannot be easily dismissed. Whether a jury has progressed to the point of no return, so to speak, in the sense of “risk that punishment will be based on extraneous considerations,” Rose, at 537, depends on circumstances of a given case. But when it appears the jury passed the point, reasons for finding no risk and to justify the punishment must be more than a subjective view of the facts of the offense and a prior criminal record. Satterwhite, supra. 5 Also to be factored in, but not necessarily related to argument or a “jury note,” is the term of years finally assessed. Both features are found in Williams v. State, supra. The offense in Williams is indecency with a child (touching her genitals); a prior conviction for burglary of a building was alleged for enhancement; the court instructed that upon finding the prior conviction the jury “will assess the defendant’s punishment somewhere in that range pro vided by law; that is, confinement ... for life or for any term of not more than 99 years or less than five years[;]” see ante, at 304, nn. 9 & 10, for summary of argument under § 4(b), and at 305 for jury note. The jury found enhancement and assessed a term of fifty years, which the Dallas Court characterized as a “mid-range sentence;” it gave that fact considerable weight in finding a § 4 instruction did not contribute to punishment. Id., at 298-99. But see Barreda v. State, 760 S.W.2d 1 (Tex.App.—Corpus Christi 1987) (nine years for sexual assault “roughly in middle of range” does indicate possibility of contribution) PDR dismissed. As we later discuss, given a current context of a § 4 instruction and sophisticated jurors, any notion that a term of years somewhat less than maximum permitted amounts to a “break” for defendant is passe’. See post, note 24, at 312. When early release from confinement on parole is implicated and jurors may comprehend that possibility, it is not enough to say that a § 4 instruction made no contribution to punishment merely because the term assessed is “mid-range” relative to the potential maximum. To the contrary, once a jury comes to understand that any term of years beyond the formula number has no effect on eligibility for parole, cases cited in note 24 post demonstrate that jurors are willing to settle on a term higher than the minimum but within limits of the formula. See also, e.g., Jones v. State, 725 S.W.2d 770 (Tex.App.—Dallas 1987) PDR refused (prosecutor literally asked the jury to assess a sixty year sentence to assure that defendant would serve at least twenty years compensating for the possibility of parole, and the jury did exactly that). The Houston [1st] Court makes the point that even without an explanation from counsel jurors are capable of making that calculation themselves, and fixing a term of years to compensate for parole eligibility — what it calls the “one-third rule” in Wheatly v. State, 764 S.W.2d 271, at 273 (Tex.App.—Houston 1988) no PDR. When the punishment is higher than an outer limit on parole eligibility, however, speaking through Justice Cohen, the Houston [1st] has attributed the excess to an aggravating circumstance rather than to the § 4 instruction, particularly where counsel never mentioned the latter. See, e.g., Villanueva v. State, supra (indicates abhorrence for perpetrators of the crime); Black v. State, 1989 WL 28388 (No. 01-85-00869, delivered March 30, 1989) (presumably defendant seen as “career criminal”) PDR pending. Thus a verdict on punishment alone is not a gauge for harm; rather, it serves somewhat as a barometric measure of other pressures we have found are likely to influence the jury in assessing punishment. There is no “bright line” rule. 6 There may be, of course, still other considerations, found so rarely, however, that we have not categorized them. One is a jury finding of a deadly weapon. Aside from determining which instruction under § 4 is appropriate, it may seem to have some bearing on deliberations of jurors in determining punishment. But any bearing should be reasonably rather than erroneously inferred. For example, in Nunez v. State, 769 S.W.2d 599 (Tex.App.—El Paso 1989) PDR pending, the offense is aggravated assault, the jury found use of a deadly weapon and assessed punishment at two and one half years. In finding harmless error, the El Paso Court was “further persuaded by the fact that the judgment in this case ... carries a finding of use of a deadly weapon, requiring service of at least one third of the sentence assessed,” so that “potential harm” was only the “two month difference between a mandatory eight months service ... and a mandatory ten months service.” Id., at 601. Another is whether there is significance in the fact that the jury rejected an application for probation properly proved up and submitted to it by the court’s charge on punishment. In Garcia v. State (Tex.App.—Amarillo No. 07-85-0288-CR, delivered January 30, 1989) PDR pending, although appellant argued that he was eligible for probation and the facts surrounding the offense of murder were “hotly contested” and “not particularly heinous,” because jury assessed punishment at thirty years there was a reasonble doubt as to effect of the parole instruction; in its opinion, the Amarillo Court did not even address that contention, and concluded that the erroneous instruction did not contribute to the punishment. Slip opinion, at 3-4. Similarly, see Sanders v. State (Tex.App.—Texarkana Nos. 6 — 86—005—CR & 6 — 86—006—CR, delivered January 18, 1989) PDR pending on motion for rehearing after PDR refused; Patton v. State (Tex.App.—Fort Worth No. 2-85-290-CR, delivered March 30, 1989). However, in Jarmon v. State (Tex.App.—Fort Worth No. 2-86-140-CR, delivered April 27, 1989) PDR pending, the court concluded the fact that the defendant was “technically eligible for probation has no correlation to the [§ 4] instruction,” viz: “If the jurors had desired Jarmon be granted probation, they knew they could impose a sentence of not more than ten years; the parole instruction in no way affected Jarmon’s eligibility for probation. The jury was obviously persuaded by the State’s argument that one who participated in a planned armed robbery resulting in death should not be granted probation.” Id., slip opinion at 4-5. Comments in the first sentence are true but not dispositive; the deduction in the second may be. Our view is that what is not seen as correlative in one case may be more apparent in another. Thus in Fonseca v. State (Tex.App.—Corpus Christi No. 13-86-011-CR, delivered February 9, 1989), Appellant’s PDR refused, where the offense was sexual assault, a second degree felony, application for probation, along with a § 4(a) instruction, was submitted to the jury, but it rejected probation and assessed punishment at twelve years, the prosecutor discussed the instruction and “asked the jury to ‘consider the manner in which the parole law may be applied to this particular defendant.’ ” The Corpus Christi Court was unable to find beyond a reasonable doubt that the erroneous charge made no contribution to the punishment assessed. Id., slip opinion, at 3. See also Early v. State, 779 S.W.2d 79 (Tex.App.—Houston [1st] 1989) (application for probation supported by ten "quality” witnesses; jury assessed six years confinement, making defendant eligible for parole after two years) PDR pending; Bonner v. State, 779 S.W.2d 81 (Tex.App.—Houston [1st] 1989) (“candidate for probation or minimum sentence (five years) under these facts” given a relatively light sentence “evenly divisible by three and exactly triple the minimum sentence of five years”) PDR pending. We conclude that in a given case a correlation between probation and a § 4 instruction may appear. Finally, there is that unusual situation in which an accused is convicted in one trial of two or more violent offenses involving multiple victims, and the jury assesses various punishments. See, e.g., Johnson v. State, 768 S.W.2d 788 (Tex.App.—Dallas 1989) (reversed) PDR refused; Hartley v. State, 765 S.W.2d 883 (Tex.App.—Dallas 1989) (affirmed) PDR refused; see also Shaw v. State (Tex.App.—Fort Worth Nos. 2-86-036-CR & 2-86-073-CR, delivered April 27, 1989) (reversed) PDR pending, on remand of Shaw v. State, 729 S.W.2d 134 (Tex.App.—Fort Worth 1987), per Shaw v. State, 761 S.W.2d 9 (Tex.Cr.App.1988). We observe, however, that feature seems to be dismissed as having no particular bearing on the outcome. Yet, jurors must have something on their collective mind to bring the jury to make those kinds of determinations. The question is whether it is the § 4 instruction. It occurs to us that an analysis of punishments meted out may suggest further examination of the record to gain an insight for answering that question. In Johnson v. State, supra, each punishment is the maximum. In Hartley v. State, supra, a jury assessed the limit for but one of three offenses, and it is the least punishment. In Shaw v. State, supra, both punishments are less than the maximum. Degrees of offenses aside, what induces jurors to produce such results? Defendant in Johnson accosted in turn several strangers and without provocation or justification shot and killed one person, shot and wounded another and shot at others; he was convicted of murder, attempted murder and aggravated assault. There were many references to the § 4(a) instruction, defense asking for no more than the minimum to give defendant hope for early release on parole, the State arguing on the facts for maximum punishment, explaining all sentences would run concurrently and only a life sentence would give any assurance the defendant would be confined in prison for any length of time. Johnson, at 789-790. More to its point, the prosecutor declared: “... There has been no evidence on the stand that he would be rehabilitated, in fact, down at the Department of Corrections. What this tells you about is a bunch of silliness that goes on down there about good time. When five years comes up and they think he’s got some good time, they’re going to pop him back out.” Id., at 790. The jury assessed maximum of life, twenty years and ten years, respectively- In Hartley defendant and a companion encountered two women in a nightclub, and what was supposed to be “partying” with cocaine at his apartment turned into an ordeal of sexual assaults, aggravated kidnappings and aggravated sexual assault, defendant inflicting substantial bodily injuries upon his victims. The prosecutor told the jury: “Every year you come down off a life sentence, you’re gambling with your safety and that of your families. The judge read to you what the parole law is. All you can do is sentence him.” Hartley, at 885. The jury found appellant did not voluntarily release victims in a safe place, and assessed punishment at sixty years for each aggravated kidnapping, forty years for an aggravated sexual assault and twenty years for a sexual assault — only the latter being the maximum. Finally, looking at Shaw, we find a twen-tythree year old defendant convicted of murder, attempted murder and criminal mischief, and both parties approaching the § 4(a) instruction from their respective perceptions. Thus for reasons given the defense urged less than maximum punishment, while the State pleaded for maximum for each offense. The jury assessed much less, viz: fortyfour years for murder, fifteen years for attempted murder and five years for criminal mischief. In our judgment, in each case jurors did consider the § 4 instruction in assessing punishment in light of arguments made to them. Moreover, in such situations they seem to focus on the offense carrying the highest penalty, the others more or less falling in line below it. Harm thus attaches primarily to a greater offense and spreads to the lesser ones. In all events, we cannot declare that the § 4 instruction made no contribution to the punishments. C The leading opinion on rehearing in Rose did not lay down a methodical approach to, nor pretend to formulate guidelines for, making that determination in other cases presenting “Rose error.” See separate opinion of Teague, J., at 557. We are now undertaking to do that which on rehearing Rose the Court did not purport to do. Examining the plurality review in Rose, we note that at the outset it acknowledged, “We are unable to know what process the jury underwent in assessing punishment[,]” id., at 554. Such an inability, of course, is one reason the rule places a burden on the beneficiary to show beyond a reasonable doubt that the error did not contribute to the verdict on punishment, and renders any harm analysis mostly speculative unless and until the State is able to discharge its heavy burden. But the plurality announced that “the record in this case presents factors which indicate that the error was harmless.” Id., at 554. We address them seriatim. 1 First is an appellate “rebuttable presumption” that “a jury follows the instructions given by the trial judge in the manner presented.” Rose, at 554, citing Cobarrubio v. State, 675 S.W.2d 749, 752 (Tex.Cr.App.1983). Analogizing “instructions to disregard,” the plurality deemed it “particularly significant” that the “last word” was the “curative instruction” traditionally given before advent of § 4. Rose, at 554. But to apply such a “rebuttable presumption” to Rose error presents knotty problems. Rule 81(b)(2) itself mandates an appellate “presumption” of harm from error in giving a § 4 instruction. Mallory v. State, 752 S.W.2d 566, at 569-570 (Tex.Cr.App.1988); see Hargraves v. State, 738 S.W.2d 743, at 749 (Tex.App.—Dallas 1987) PDR refused. From that mandate it follows that the jury is presumed to understand that it “may consider the existence of the parole law and good conduct time” as declared in the preceding four paragraphs of § 4, yet also understand it is not to consider “the extent” good conduct time may be awarded or “the manner” parole law may be actually applied to defendant. The State must demonstrate that the error is harmless. To say that the appellate “rebuttable presumption” may be utilized at once sets up opposing “presumptions.” An appellate “rebuttable presumption” alone does not demonstrate beyond a reasonable doubt that an error made no contribution to punishment. But, if used, the effect is to shift to appellant the burden of demonstrating harm — contrary to the rule. See, e.g., Francis v. Franklin, 471 U.S. 307, at 319-325, 105 S.Ct. 1965, at 1973-1977, 85 L.Ed.2d 344, at 356-360 (1985). In our judgment, the presumption of harm in Rule 81(b)(2) must prevail over an appellate “rebuttable presumption” that experience with parole law teaches is more fiction than fact. 2 The “curative instruction” the Rose plurality found “particularly significant” is a standard admonishment trial courts traditionally submitted to a jury that the Court accepted over the years, viz: “You are further instructed that in determining the punishment in this case, you are not to discuss among yourselves how long the defendant will be required to serve any sentence you decide to impose. Such matters come within the exclusive jurisdiction of the Board of Pardons and Paroles and the Governor of the State of Texas and are no concern of yours.” Id., at 554, and generally, at 532. While we have rejected its presumptive effect, that the jury was given that kind of “curative instruction” is a proper factor, among others, in a harm analysis for “Rose error.” As is often shown in our records, however, “jurors cannot resist the temptation to discuss parole laws,” Rose, at 536; Ready v. State, 687 S.W.2d 757 (Tex.Cr.App.1985), and cases cited ante at 298. When the record suggests that jurors did indeed succumb to that temptation, manifestly the “curative instruction” failed to accomplish its purpose, and loses its value as a probative factor. See Hernandez v. State, 774 S.W.2d 319 (Tex.App.—Dallas 1989) (reversed) PDR pending. 3 Another matter noticed in Rose is “particularly heinous facts of this case,” id., at 554. Of course, “heinousness,” like beauty, is in the eye of the beholder. See Godfrey v. Georgia, 446 U.S. 420, at 429-430, 100 S.Ct. 1759, at 1765, 64 L.Ed.2d 398, at 406 (1980) (interpretation of like terms can only be subject of sheer speculation). With or without other facts of the case some appellate courts are treating what they regard as facts “militating in favor of a harsh sentence” as “ample evidence to support the jury’s findings on punishment,” e.g., Lancaster v. State 754 S.W.2d 493, at 496 (Tex.App.—Dallas 1988) PDR refused; “circumstances [to] support the sixty-year sentence assessed by the 'jury,” Baker v. State, 752 S.W.2d 237, at 239 (Tex.App.—Fort Worth 1988) PDR refused; “overwhelming proof of guilt,” Diaz v. State, 769 S.W.2d 307, at 308, 309 (Tex.App.—San Antonio 1989) PDR pending. Yet, sufficiency of evidence is a matter of little consequence in a harmless error analysis. Satterwhite, supra, 486 U.S. at 258, 108 S.Ct. at 1798, 100 L.Ed.2d at 295 (question not whether evidence sufficient to support jury’s finding but whether State proved error did not contribute to verdict); Olivarez v. State, supra, at 115; see generally Harris v. State, supra. In Diaz v. State, supra, dissenting Justice Carr reasoned, viz: “... If a jury considers ‘A’ (evidence of guilt) and ‘B’ (existence of parole and good conduct time) in assessing punishment, that does not show beyond a reasonable doubt that ‘B’ made no contribution to the punishment.” Id., at 311 (his emphasis). We agree. Beyond that, in context of a § 4 instruction “heinousness” is a slippery indicator for gauging how a jury evaluated conduct of appellant in assessing punishment. In other words, to find the facts are so aggravating that the punishment is appropriate is simply not a satisfactory conclusion. 4 In like vein, the Rose plurality opined, “Compounding these facts surrounding the offense and subsequent arrest was appellant's criminal record,” id., at 554. “These convictions,” according to the opinion, “most certainly ... contributed to the jury’s assessment of punishment,” ibid. However, that contribution to punishment “does not establish beyond a reasonable doubt that the erroneously given instruction made no contribution to the punishment.” Olivarez v. State, supra, at 115. Arguably, in Rose consideration of “the existence of the parole law and good conduct time” could have just as well contributed to motivating the jury to assess “the maximum sentence, life in the Texas Department of Corrections,” ibid. Exhibits of “prior criminal record” are available for scrutiny by jurors. 5 As we have seen, then, on rehearing the Rose plurality found from the factors it addressed that “the error made no contribution to the punishment assessed,” id., at 554-555. It did not look for other germane factors; it addressed only the three and failed to rule out others. Therefore, while the plurality and concurrences were enough to affirm the judgment of the court of appeals, none purports to present a “pattern” harmless error analysis of Rose error — as done here. Ill Although harm analyses of Rose error run the gamut, there emerge acceptable rationales for relevant considerations in applying Rule 81(b)(2). While we have prescribed an analysis herein, no court can lay down a “bright line rule” to measure the contribution a § 4 instruction will make to every result. In sum, the appellate undertaking is to assay extant factors and circumstances germane to punishment, as we have evaluated them, for a likelihood that constitutional error conducive to introduction of offending parole matters into environment of a punishment proceeding affected jury deliberations, and thereby influenced jurors in assessing the terms of punishment reflected in their verdict. The reviewing court must be able to say beyond a reasonable doubt that the error made no contribution to the verdict obtained. Rule 81(b)(2); Chapman v. California, supra. With that in mind, we now turn to the particular causes before the Court. A Arnold v. State These are two convictions for murder. The judgments were affirmed in an unpublished opinion in Arnold v. State (Tex.App.—Dallas Nos. 05-85-01245-CR & 05-85-01246-CR, delivered December 4, 1986). On remand for reconsideration in light of Rose, the Dallas Court, raising but, over a dissent, not deciding the question of burden of proof under Rule 81(b)(2), and relying exclusively on the Rose “curative instruction,” again affirmed. Arnold v. State, 766 S.W.2d 852 (Tex.App.—Dallas 1989). Accordingly, we granted appellant’s petition for discretionary review, and for reasons about to be given we will affirm its judgment. 1 Jeffrey Allen Licker, a practicing civil attorney and a heavy user of cocaine, determined that his former supplier, Koby Sandovsky, stole a safe from his home containing controlled substances, jewelry and cash money; Licker decided to confront Sandovsky and his girlfriend, Lesia Kahl, about the theft, and he hired appellant and another suspect to kill them after the confrontation. His plan went awry in execution, but appellant and his confederate managed to kill the victims by bludgeoning each with a hammer. Appellant and Licker were jointly tried for the two murders. Both were eligible and applied for probation. At punishment the State pretermitted evidence and rested; appellant and Licker presented relatives and other witnesses to prove up eligibility and to support their respective applications with general “character” evidence. After all sides closed the judge heard objections to the proposed charge of the court, particularly pertaining to alternative § 4 instructions depending on a deadly weapon finding. Overruling those objections, the court submitted the applications to the jury, gave the § 4 instructions and added a traditional Rose “curative instruction.” In making their arguments all counsel seemed wary of limine instructions from the court. First prosecutor cautioned the jury about the deadly weapon verdict; suggested that though defendants were different men engaging in different types of criminal conduct, each committed murder in his own way; discussed condition of the victims and reasons for punishments; he concluded, “the only thing you can do to assess a meaningful punishment to both these defendants in all four of these cases is to assess life punishments to protect Dallas County as much as it’s humanly possible from these two and people like them.” Counsel for Licker discounted “an eye for an eye, a tooth for a tooth” in favor of “forgiveness and mercy,” and concentrated on probation for his client; along the way he alluded to “availability of needed treatment sources in the community,” and argued: “... Send him to the penitentiary, you send him to the penitentiary, life, he’s not going to be there for life. Okay. So the question is what are you going to do with a person that you really can’t get rid of. Now, the prosecutor’s just going to want y’all just throw him away, get rid of him, but these people ultimately come out; and the question is do you want them to come out the way they come out from the penitentiary or do you want them to come out so that they’re going to be able to fit in society, so we don’t have any problems like this any more. That’s the hard issue; that’s what you’ve got to determine.” For his part, attorney for appellant also talked about objectives of punishment, but candidly told the jury that in his opinion, “for what it’s worth, [appellant] should go to the penitentiary.” He suggested a formula, viz: “If you sent [appellant] to the penitentiary for 5 years, he would go down there. You know, there’s something on parole. Y’all read that so you will understand it better, but he would go down for 5 years ... — let’s send him down ... for the girl, because it just goes against the grain for somebody to hurt a girl, I don’t care if they are drug dealers; but on the man, let’s give him a ten-year sentence in the penitentiary and then probate it, so when he gets out of the penitentiary, whether he serves his full 5 years or what, he’s not loose. We’ll watch him. He’s still on probation ... and if he ain’t straightened his act up, that probation can be revoked and he can go down for another ten.” In closing argument the last prosecutor doubted that rehabilitation would work for these kinds of killers, disputed the notion that because victims may be “dope dealers” different considerations apply, urged jurors to tell them to get out and stay out of Dallas by rejecting appellant’s proposed formula and assessing life sentences in these cases, and add on for Licker fines of $10,000. During its deliberation the jury asked the judge to identify the case numbers pertaining to “the female,” and the court complied. Against appellant, the jury found he used a deadly weapon and assessed punishment at thirty years for murder of Sandov-sky, and life for murder of Kahl. For Licker, it assessed punishment of ten years confinement and a fine of $10,000 for death of Sandovsky, and ten years probated and a fine of $10,000 for death of Kahl. 2 We granted review in these causes to determine whether the appellate court correctly construed and properly applied Rule 81(b(2) in finding that the unconstitutional § 4 instructions made no contribution to the verdicts on punishment. In that it was content to rely solely on a “presumption” that the jury followed traditional Rose “curative instruction,” for reasons stated in Part II ante, we conclude the court fell into error. Because Rule 81(b)(2) mandates an appellate presumption of harm from error in giving a § 4 instruction, and because the State must demonstrate that the error is harmless, the presumption of harm in the former prevails over an appellate “rebut-table presumption” that “a jury follows the instructions given by the trial judge in the manner presented.” Ante, at C 1. That a “curative instruction” was given is, however, a proper factor in a harm analysis unless the record suggests that the jury did indeed consider the parole law in assessing punishment. Ante, at C 2. Thus we must examine the record for indicia of factors reasonably conducing to affect minds of average rational jurors in their determination of punishment. Ante, at 300. In the instant causes, the Court confronts a rare situation: a jury assessed disparate punishments against two defendants for the part each played in murdering two victims. Manifestly the jury determined the degree of their respective culpability was unequal. We have reviewed every germane revelation in this record and are satisfied that they did not influence the jury adversely to appellant in assessing punishment. The trial judge imposed his own limitations on commenting about parole law; defense counsel objected to any § 4 instruction and stayed within the bounds of comment; the State persistently pleaded for life sentences in all four cases, while counsel for Licker focused on probation and counsel for appellant conceded his client was due some time in the penitentiary for killing Kahl but requested the jury follow that with probation in the other case; in its only note the jury asked for and got case numbers relating to Kahl; its assessment of life against appellant and ten years probated against Licker for the death of Kahl appears to reflect the jury made a reasonably consistent assay of criminal responsibility of each, see n. 30, ante, as does ten years and thirty years confinement, respectively, in the Sandovsky cases; unlike a “pattern” indicative of consideration of parole, cf. cases at 308-310, ante, nothing here suggests the life and thirty year terms have any correlation. Thus we determine and are able to declare beyond a reasonable doubt that § 4 instructions made no contribution to the punishments. The judgment will be affirmed. B Gaines v. State This a conviction for aggravated sexual assault; the jury assessed punishment at twentyfive years. The judgment was affirmed in a published opinion. Gaines v. State, 723 S.W.2d 302 (Tex.App.—San Antonio 1987). On remand for reconsideration in light of Rose, the San Antonio Court, with one justice concurring, concluded the jury was not “influenced by the parole law instruction” because the trial submitted the “statutory curative instruction” (last two paragraphs of § 4), which the concurring justice discredited), the offense was a traumatic crime of violence, the jury assessed “less than one-third of the possible range of punishment,” and again affirmed the judgment. Gaines v. State (Tex.App.—San Antonio No. 04-86-00199-CR, delivered January 18, 1989). Accordingly, we granted appellant’s pro se petition for discretionary review, and the cause was submitted on brief of his court appointed counsel; the State relied on its brief below. For reasons discussed post we will affirm the judgment. 1 The ten year old female victim, Crystal, ran away from home about eight o’clock at night and began hitchhiking in a residential section of San Antonio; after she was picked up by adult male, one Paez, and agreed to “make sex to have a place to stay,” he pulled into a driveway, got out and talked to appellant; shortly all three were in a vacant house formerly occupied by Paez. First Paez had deviate sexual intercourse and sexual intercourse with her and left; then appellant engaged in sexual intercourse and left, but later returned with pillow, blanket and flashlight, and had intercourse with her again; he left and returned with some food. The victim fed the food to a dog, departed the premises and was on her way back home when a police officer on the lookout for her stopped and took her home around three o’clock in the morning. She made prompt outcry to her mother. Appellant was tried alone. Paez was a codefendant, but entered into a plea bargain on the eve of trial. Voir dire of prospective jurors is not in our record. Appellant presented witnesses to set up a sort of alibi defense and, testifying on his own behalf, denied committing any offense. However, on rebuttal the State called Paez further to implicate appellant; during crossexamination he said that he had been promised ten years for his part in the episode. Appellant was eligible and applied for probation. At punishment the State reof-fered all prior evidence and rested; appellant offered witnesses as to his character for being peaceable and law abiding. The trial court submitted his application for probation to the jury and, without objection, included a § 4(a) instruction. It did not contain a traditional Rose “curative instruction.” The State waived opening argument. At the outset counsel for appellant told the jury he had elected to have it assess punishment, and requested jurors “to grant his motion for probated [sentence],” asking that they “give him no more than ten years, anywhere from five to ten years.” He went on to justify that request, assuring the jury that if probation is revoked appellant will be committed to TDC for the period of punishment imposed. Counsel never alluded to § 4(a) instruction or parole. For its part the State asked jurors to reconsider the evidence previously admitted during guilt phase and told the jury punishment is “entirely up to you.” The prosecutor restated range of punishment for a first degree felony, explaining that a wide range enables them “to take the facts of the case and find an appropriate slot in between.” She knew that “all of you are going to consider what is going to happen to the Defendant,” but asked they not forget that “he’s here not for what is going to happen to him but for what he did,” and then proceeded to reprise the facts and circumstances of appellant. She rejected probation, pleaded for “time to serve ... so that he can have time to reflect on what he did,” and made a bit of the “they” argument. Concluding, she recommended fifteen to twenty years, acknowledged that “whatever you give is up to you” and again urged them to “make him serve time.” The State never alluded to § 4(a) instruction or parole. The jury deliberated an hour and twenty minutes without seeking additional information before returning its verdict of confinement for term of twentyfive years. 2 We granted review in this cause principally to determine whether the appellate court correctly regarded the last two paragraphs of a § 4(a) instruction as a “statutory curative instruction.” That consideration and that “aggravated sexual assault is a crime of violence,” that the victim expressed “fear for her safety as to the co-defendant” and “continues to suffer from the trauma,” brought the appellate court to conclude that “assessment of punishment at less than one-third of the possible range of punishment with no fine does not reflect [jurors] were influenced by the parole law instruction.” Slip opinion, at 3-4. The concurring justice disagreed that “the statutory instruction submitted here ... is the same ‘curative’ instruction the Court of Criminal Appeals found ‘particularly significant’ in Rose,” and referred to the earlier opinion of the court in Olivarez v. State, 756 S.W.2d 113, at 115 (Tex.App.—San Antonio 1988) no PDR. For reasons stated ante under II C 2, n. 23 and accompanying text at 311-312, the concurring view is substantially correct; we have determined that “no part of a § 4 instruction can be reasonably characterized and fairly regarded as ‘curative.’ ” Ibid. The jury may well have taken into account the other matters mentioned in the opinion below, but this Court now proceeds on the premise that a jury may just as likely take into account the § 4 instruction and, therefore, examines the record “for indicia of factors reasonably conducing to affect minds of average rational jurors in their determination of punishment.” Ante, at 300-301. In this connection, appellant pro se recognizes and acknowledges that “both from the argument of the State’s attorney and the sentence assessed by the jury, that there was never an intent to give a maximum sentence.” PDR, at 5. Nevertheless, he contends that the instruction influenced the jury because “two of the three factors suggested by the [Rose court on rehearing] in determining ‘harm’ [are] absent in the present case,” to wit: no prior criminal record and no “curative instruction.” Ibid. His attorney makes much the same argument. Appellant’s Brief on the Merits, at 4. But as we have made clear in this consolidated opinion, “there are other revelations for examination pertinent to the [ultimate] inquiry.” Ante, at 300 and see Part II B. So we have reviewed this record to find that the only mention of parole is in the § 4(a) instruction; otherwise there is not even an allusion to it by any trial participant, and perforce no indicia of factors inducing jurors to consider parole law. Thus that appellant had no prior criminal record and that the trial court did not give a “curative instruction” never becomes germane to the inquiry. Accordingly, we are satisfied that the fact the jury rejected appellant’s application for probation and assessed punishment somewhat higher than sought by the State does not indicate it was operating under influence of the § 4(a) instruction. Thus we determine and are able to declare beyond a reasonable doubt that § 4 instruction made no contribution to the verdict on punishment. The judgment will be affirmed. C Hooper v. State This is a conviction for aggravated assault; finding a repeater allegation “true,” the jury assessed punishment at twenty years plus a $10,000 fine. The judgment was affirmed in an unpublished opinion. Hooper v. State, 1987 WL 11568 (Tex.App.—Houston [1st] No. 01-86-00487-CR, delivered May 28, 1987). On remand for reconsideration in light of Rose, noting a “curative instruction” had not been given and that the State discussed the parole laws, the Houston (1st) Court concluded it was unable to find beyond a reasonable doubt that the erroneous charge made no contribution to punishment, and set aside the punishment. Hooper v. State, 1989 WL 27549 (Tex.App.—Houston [1st] No. 01-86-00487-CR, delivered March 23, 1989). Accordingly, the Court granted the State’s petition for discretionary review, and for reasons hereafter given, we will affir