¶1concurring: I concur in the result of the majority opinion. I write separately because I believe the majority falls short in its analysis, critique, and application of State v. Hill, 271 Kan. 929, 26 P.3d 1267 (2001), on the multiple acts issue.
¶2The majority is correct that our court had been divided until recently on whether violation of the elect-or-instruct rule for multiple acts cases required reversal for structural error or could be declared harmless error. Compare State v. Wellborn, 27 Kan. App. 2d 393, 4 P.3d 1178, rev. denied 269 Kan. 940 (2000), with State v. Hill, 28 Kan. App. 2d 28, 11 P.3d 506 (2000), aff'd 271 Kan. 929, 26 P.3d 1267 (2001). The Supreme Court addressed that controversy in Hill, categorically rejecting structural error analysis and adopting virtually word-for-word the two steps set forth originally in the Court of Appeals’ opinion in Hill. The second of these steps requires a determination of whether any error is harmless beyond a reasonable doubt. Hill, 271 Kan. at 939. In my view, if the Kansas appellate courts are ever going to straighten out what has become a multiple acts morass, it is vital to consider the holding of Hill and the Supreme Court’s application of it carefully and in appropriate context.
¶3In Hill, the State filed one count of rape among other charges arising out of a series of unwanted sexual advances made in multiple rooms of the victim’s home over the course of a few minutes. The victim testified to the distinct order of each of two digital penetrations of her vagina and to her exact location in the home at the moment of each. The State did not elect which of the two it wished to rely upon to support a conviction on the rape count, and no multiple acts unanimity instruction was requested or given. See PIK Crim. 3d 68.09.
¶4*984Prior to Hill, the Court of Appeals had developed two possible patterns of analysis for facts such as those in Hill. Under the first, a panel could have applied the strict elect-or-instruct rule and reversed the conviction because of structural error. See Wellborn, 27 Kan. App. 2d at 394; State v. Barber, 26 Kan. App. 2d 330, 331, 988 P.2d 250 (1999). Under the second, given the two acts’ relationship to one another as parts of a single series of advances that took place over a few minutes in one home, a panel could have concluded that the acts were not truly “multiple” at all and found no error. See State v. Hazley, 28 Kan. App. 2d 664, 671, 19 P.3d 800 (2001) (simultaneous constructive possession of drugs in numerous rooms of defendant’s house not multiple acts underlying drug possession charge); State v. Staggs, 27 Kan. App. 2d 865, 867-68, 9 P.3d 601, rev. denied 270 Kan. 903 (2000) (punch followed by kick in one continuous fight not multiple acts underlying battery charge).
¶5Under Hill, we are now directed to first determine whether “there is a possibility of juror confusion from the record or if evidence showed either legally or factually separate incidents.” (Emphasis added.) Hill, 271 Kan. at 939. On the surface, this part of Hill’s holding appears to equate the possibility of juror confusion with the disjunctive legal separability or factual separability. Under the surface, however, the Supreme Court actually requires more before it will accept that a jury could have become confused.
“Incidents are legally separate when the defendant presents different defenses to separate sets of facts or when the court’s instructions are ambiguous but tend to shift the legal theory from a single incident to two separate incidents. Incidents are factually separate when independent criminal acts have occurred at different times or when a later criminal act is motivated by ‘a fresh impulse.’ ” Hill, 271 Kan. 929, Syl ¶ 3.
¶6In other words, both a particular type of legal separability and a particular type of factual separability are needed to demonstrate the possibility of jury confusion.
¶7The Supreme Court’s application of its new rule to the facts in Hill also compels us to refine our understanding of this first step in this way. The Hill victim’s ability to identify the order and exact location of the two penetrations at issue unavoidably yielded at least *985one type oí factual separability. Furthermore, the Supreme Court candidly observed that each penetration could have supported a separate count of rape without being vulnerable to a multiplicity argument. This was at least one type of legal separability. Cf. State v. Long, 26 Kan. App. 2d 644, 645-47, 993 P.2d 1237 (1999), rev. denied 268 Kan. 852 (2000) (multiple charges for multiple penetrations during series of attacks in 2-hour period in one home not multiplicitous).
¶8Yet the Supreme Court found no possibility of jury confusion because the defendant “did not present a separate defense or offer materially distinct evidence of impeachment regarding any particular act.” Hill, 271 Kan at 940. Rather, the defendant turned the trial into an up-or-down credibility contest by his “general denial of participation in any wrongful conduct.” 271 Kan. at 940. The jury was either going to believe the victim’s stoiy that the defendant committed all of the alleged illegal acts, thus giving him his unanimous verdict on each component act, or it was going to believe the defendant’s story that he committed none of them. In the Supreme Court’s language,
“[t]he evidence offered no possibility of jury disagreement regarding Hill’s commission of either of these acts. By the jury’s rejection of Hill’s general denial, we-can unequivocally say there was no rational basis by which the jury could have found that Hill committed one rape but did not commit the other.” 271 Kan. at 940.
¶9The Supreme Court’s first step under Hill harkens back to the second possible analysis pattern in the earlier Court of Appeals cases noted above. The difference is in the result that follows from a finding of factual and legal inseparability.
¶10The Court of Appeals’ analysis would have ended there, holding no truly multiple acts existed and thus neither a State election nor multiple acts unanimity instruction was required. This result would have been consistent with that in Simms v. United States, 634 A.2d 442, 445 (D.C. App. 1993) (robbery of van containing money and tools constitutes single unbroken chain of events), which originated the definitions of factual and legal separability relied upon in Hill. The Simmscourt, when it found no factual or legal separability, arrived at the conclusion there were no truly multiple acts and thus *986no possibility of jury confusion, rather than the other way around. Once that conclusion was reached, no harmlessness inquiry was necessary.
¶11Our Supreme Court, however, evidently still regards the dual absences of an election and instruction as error when there is no finding of jury confusion, because it requires the process to continue through Hill’s second harmlessness step. In the words it adopted from the Court of Appeals opinion in Hill: “When jury confusion is not shown under the first step, the second step is to determine if the error in failing to give a unanimity instruction was harmless beyond a reasonable doubt with respect to all acts.” (Emphasis added.) 271 Kan. at 939. This is puzzling, because the first step has already ruled out the existence of jury confusion and the lack of unanimity it can breed. These were the evils the elect-or-instruct rule was designed to eradicate.
¶12Within this analytical context, it is therefore litde wonder that the Supreme Court proceeded to decide that whatever error occurred in Hill was harmless. Indeed, it is difficult to imagine how any failure to elect or instruct could ever be deemed harmful once the result of Hill’s first step was a finding of no possibility of jury confusion.
¶13It is also worth noting that Hill gave rise to several unanswered questions. If we are to engage in a harmlessness analysis when there is no finding of possible jury confusion, how, if at all, should the second step be altered when we find a possibility of such confusion? Structural error analysis is no longer available. Should harmful error be presumed absent a contrary showing from the State? If the burden on appeal does not shift to the State, then is a situation where jury confusion has been demonstrated to be reviewed for harmlessness no different from one where no jury confusion has been found? If so, then what difference does the presence or absence of the possibility of jury confusion actually make? Does the first step have any validity? And finally, if the two steps ultimately make jury confusion irrelevant, then do we not lose sight of the reason we focused on this issue in the first place?
¶14Where does all of this leave M.M. and the alleged multiple acts in this case? If the evidence in Hill left no room for jury confusion *987and lack of unanimity, then the evidence here certainly left no such room. M.M. was simply incapable of testifying to specific characteristics of incidents of repeated abuse sufficient for the prosecution or jury to differentiate among them. As the majority opinion points out, the abuse had “resulted] in an amalgamation of the crimes in the child’s mind.” People v. Luna, 204 Cal. App. 3d 726, 748, 250 Cal. Rptr. 878 (1988).
¶15Other courts have labeled the vague testimony of such child victims who are repeatedly abused “generic” and have fashioned rules specifically designed to protect them and those they accuse. For example, in R.L.G. Jr. v. State, 712 So. 2d 348 (Ala. Crim. App. 1997), one of the many opinions from other jurisdictions cited by the Supreme Court in Hill, the court observed:
“The facts before us do not lend themselves to a straightforward application of the general law of election as it currently stands . . . : the evidence of the appellant’s alleged abuse of [the victim] presents absolutely no ‘distinct and separate transactions’ upon which to apply the general law and require an election.’
“The term ‘elect’ implies a knowledge of facts which go to make up two or more offenses; ... to hold [the prosecutor] to have elected to proceed for a certain offense, he must have learned enough to enable him to individualize the transaction, and then pursue his inquiry with a view of learning the details and particulars of the act or transaction thus individualized.’ Jackson v. State, 95 Ala. 17, 18, 10 So. 657, 657 (Ala. 1892). The very nature of the evidence in this case ... is dictated by the nature of the circumstances of the alleged abuse: abuse upon a young child by an abuser residing with the child, thus an abuser who could perpetuate the abuse so frequently and in so many locations that the young child loses any frame of reference in which to compartmentalize the abuse into ‘distinct and separate transactions.’ Such evidence of abuse has been termed ‘generic evidence.’ ” R.L.G. Jr., 712 So. 2d at 356.
¶16The Alabama court held that, because the statutory scheme at issue did not allow for charging of a continuing offense covering child sexual abuse, a situation Kansas faces as well, see Wellborn, 27 Kan. App. 2d at 395, the strict elect-or-instruct rule must be modified:
“ ‘[W]hen there is no reasonable likelihood of juror disagreement as to particular acts, and the only question is whether or not the defendant in fact committed all of them, the juiy should be given a modified unanimity instruction which, in addition to allowing a conviction if the jurors unanimously agree on specific acts, also allows a conviction if the juiy unanimously agrees the defendant committed *988all the acts described by the victim.’ ” R.L. G. Jr., 712 So. 2d at 361 (quoting People v. Jones, 51 Cal. 3d 294, 322, 270 Cal. Rptr. 611, 792 P.2d 643 [1990]).
¶17In my view, the Alabama court’s idea of a modified unanimity instruction for generic evidence child abuse cases has merit. Such a rule also could be applied in cases such as Hill, when the victim’s testimony is more specific as to multiple acts but the defendant’s general denial reduces the trial to a credibility contest. I believe jurors should be told that they may find the defendant guilty if they unanimously agree he or she committed any one of the acts or all of the acts described. Such an instruction strikes a sound balance between the defendant’s and the victim’s rights while respecting the current statutory scheme. If this rule were adopted, the absence of such an instruction would be regarded as error. Then, in keeping with the Supreme Court’s unequivocal expression of disapproval for structural error analysis, a harmlessness analysis would follow. This pattern, to me, vastly improves upon the internally inconsistent Hill two-step analysis.
¶18A strict elect-or-instruct rule, is inappropriate for the truly generic evidence child abuse case. Some exception is necessary to ensure that perpetrators of the most egregious violations upon the most vulnerable of our children are punished. Given the absence of a continuing offense covering child sexual abuse in our criminal code and the inability of the youngest and most traumatized children to recite specifics of the circumstances surrounding the abuse they have suffered, the rule must be relaxed when credibility is the only genuine issue. It is simply impossible for the prosecutor to elect which among multiple indistinguishable acts he or she relies upon for conviction, and it is likewise impossible for jurors to select one from a string of undifferentiated incidents to arrive at a verdict. Rigid adherence to the elect-or-instruct rule would disable or derail prosecution altogether, removing from the reach of the law those whose depredations were so frequent or so long-term or so aimed at the inarticulate that generic testimony is all that can be marshalled against them.
¶19I would also emphasize, however, that any case in which the defendant had ammunition beyond general attacks on the victim’s *989credibility, or general endorsements of his or her own, would still have to be subject to the elect-or-instract rale. If alibi or identity was in issue, if the victim or other witnesses were able to put any more meat on the bones of the allegations that might be picked off by cross-examination about physical implausibility or other weaknesses, then that requirement would apply. If there was neither an election by the prosecution nor an instruction by the district court, only a convincing demonstration that the omissions were harmless should save a conviction from reversal.
¶20Unless and until my view is adopted, however, I agree Hill’s first step leads inexorably to a finding of no possibility of jury confusion in this case, and, under its second step, any error in failing to elect or instruct was therefore harmless beyond a reasonable doubt as to all acts.