Opinion of the Court by
Justice NOBLE.
This case presents the issue of whether proof at trial showing only an opportunity to commit first-degree criminal abuse against a child is sufficient to support a jury instruction and resulting verdict for directly abusing the child, as opposed to the alternative theory of permitting the abuse of the child. This Court holds that such proof is not sufficient.
I. Background
Cecilia Alvarado was only six months old when she died on August 22, 2005. Her death and subsequent autopsy sparked an investigation that ultimately led to the arrest and prosecution of her mother, Samantha Monahan Acosta,1 and Acosta’s boyfriend, Roy Rankin.
Rankin was charged with and convicted of murder and first-degree criminal abuse. This Court affirmed his conviction. See Rankin v. Commonwealth, 327 S.W.3d 492, 500 (Ky.2010).
Acosta was charged only with first-degree criminal abuse. Her conviction for that crime is the subject of this case.
Acosta had known Rankin from childhood. In early 2005, she lived outside Kentucky but struck up a long-distance relationship with Rankin. She visited him frequently starting in February or March of 2005, near the time Cecilia was born. In July 2005, she moved to Kentucky and began living with Rankin. She got a job, and Rankin watched her children when she worked. Acosta claimed at trial that she [812]*812never saw Rankin mistreat the children or even lose his temper. Unbeknownst to Acosta, Rankin was borderline mildly mentally retarded, though his social skills hid this.
Medical records from when Acosta lived outside Kentucky showed no evidence of abuse of Cecilia. It appears that Acosta got good, consistent medical care for her children at that time, and that Cecilia had only a few mild, common health problems (e.g., neo-natal jaundice, vomiting and diarrhea caused by acid reflux, and some wheezing). The last of these medical visits was in early June, shortly before the move to Kentucky.
Acosta and Amanda Huff, Rankin’s niece, testified that soon after Acosta’s move to Kentucky, in mid-July, they were in a car accident with Cecilia as a passenger in a car seat. No police report associated with Acosta could be found for this accident. Acosta and Huff said this was because Acosta lied to the police who responded, claiming instead that Huff had been driving the car, because Acosta had an outstanding warrant from Tennessee.
At least one victim from the other car was taken to the hospital, but Acosta and Huff did not ask for medical attention at the scene for themselves or their children. Instead, they claimed to have gone to the hospital on their own. According to Acosta, she had Cecilia examined by a doctor, who said the child was fíne. No medical record of this visit was introduced.
According to Acosta, on August 11, she called to schedule a six-month well-baby medical check-up for Cecilia. This appointment, scheduled for August 18, would be the first since the move to Kentucky.
Also according to Acosta, around August 15, Rankin told her that Cecilia had fallen off the bed while Acosta was in the shower. A few days later, around August 17, Acosta noticed what she described as a soft “jelly spot” on Cecilia’s head.
Several of Rankin’s relatives and acquaintances, including his mother and sister, testified about how the baby acted before her death. None of them, however, ever testified to seeing Rankin or Acosta abuse Cecilia.
Rankin’s mother, Rebecca Rankin, testified about the “jelly spot,” confirming that Acosta had shown it to her around August 17. She also testified that the baby would not cry but would whine, as though something was hurting her, when she was picked up. She testified that she thought something was wrong with the baby, that well babies did not act like that, that something was hurting the baby, and that she told her son and Acosta this. She also testified that the child had a fever for several days at one point and that her daughter-in-law threatened to report Acosta to the police if she did not take the baby to see a doctor. She saw Acosta and her son give the child Tylenol.
On cross-examination, she said that she never saw any signs of broken bones or any emergent condition (before the day of the child’s death), though she could tell there was something wrong with the baby. She also stated that Acosta had been the one to mention the “jelly spot” and that she had not tried to hide it. She described the spot as a painful-looking soft knot on the back of the head, which Acosta had told her had been there “quite a while.” She said it was not the “soft spot” on a baby’s head. She also testified to seeing a bruise on the baby’s head, which she had been told was from a fall from a bed.
Rankin’s sister, Wanda Goodlet, also testified that Cecilia appeared to be in pain when she was picked up, and that Acosta, when asked about this, said the child “did that sometimes.” According to Goodlet, Cecilia would groan but not cry. A few [813]*813days before the scheduled check-up, Good-let noticed that Cecilia was running a fever, which she mentioned to Acosta. According to Goodlet, the next day she asked why Acosta had not taken the child to a doctor for the fever; Acosta said she had not wanted to do so because of a cigarette burn on the child’s legs and a bruise on her head, which Acosta feared would lead to a call to social services. The next day, after again talking to Goodlet, Acosta agreed to take the child to a doctor, and did so the next day. Goodlet claimed that she and her family had to pressure Acosta to take the child to the doctor, having gone so far as to threaten to call social services if she did not.
On cross-examination, she stated that she witnessed the cigarette burn happen but that it was an accident, though she did not say when it happened. She also said that she had had no reason to think that the child had any broken bones.
David Goodlet, Wanda Goodlet’s father-in-law, testified primarily about the day Cecilia died. But he also testified about how the child looked and acted a few days before. Specifically, he said that something had been wrong with the baby and that she would say “waah,” as though in pain whenever held or touched, but she would not cry. She would quit saying “waah” when put in her car seat. When asked whether he saw any signs of broken bones, he said that he was hesitant to hold the baby after seeing her appear to be in pain and that she was a sickly baby. He also testified that his wife had threatened to call the police when the child had a sustained high fever and Acosta seemed unwilling to take the child to see a doctor. When asked again about signs of broken bones, he said he had seen none.
In addition to describing the car accident, Amanda Huff testified that she had no reason to think the child had broken bones. She stated, however, that the child was kept in her car seat “twenty-four seven,” and that she was concerned the child could not hold a bottle.
On August 18, Acosta took Cecilia to her six-month well-baby visit, which was her first medical check-up since the move to Kentucky. Cecilia was examined by a nurse practitioner named Susan Hays. In her examination, she noted a faint bruise on Cecilia’s head and two small symmetrical scabs with some bruising on her lower legs that appeared to have been there a while. When asked about the head bruise, Acosta told her that the baby had fallen off the bed a few days before. Hays did not recall what Acosta said about the scabs on the baby’s legs. Her examination included feeling the baby’s abdomen, clavicle, and hip. She testified that Cecilia’s muscle tone was poor and that she was lax and slumped, but that she “did good” in a pull-to-sit response test: The lax muscle tone was concerning, as were the head bruise and the symmetrical marks on the legs, and she asked Acosta to set up a follow-up appointment.
She did not recall Acosta saying anything about a “jelly spot” on the child. She saw no evidence of bone fractures, and did not suspect abuse. When cross-examined about the pull-to-sit test, she stated that she would not have done the test had she known of a fractured arm because it would have caused pain to the baby. She also téstifíed that she regretted not reporting the marks on Cecilia’s legs to Social Services.
When Acosta testified, she admitted that Cecilia acted sore when lifted and was not smiling or kicking her legs in the time leading up to the six-month medical visit. She also admitted that the child had stopped smiling, and would grip objects but could not hold onto them. She claimed that Cecilia fell off a bed a second time [814]*814between the August 18 medical visit and her death.
The day of the killing, Rankin watched Cecilia while Acosta went to work. After Acosta left, Rankin took Cecilia and her two-year-old brother M.A. to his parents’ house. Rankin claimed at trial that around 7:00 p.m., he left the children briefly while he went to another room, and that when he returned 15 to 20 minutes later, he found M.A. with his knees on Cecilia’s neck. According to Rankin, Cecilia had been in her car seat and had been tipped out of it onto the floor by M.A. He claimed that he pulled M.A. off and picked up Gecilia. He testified that her head was limp, so he took her to the front porch where his parents were; 911 was called at that time.
Emergency responders took Cecilia to the hospital. At that time, she had no pulse and was not breathing. She had dark bruises on both sides of her neck that looked like handprints. The emergency physician who saw Cecilia testified that this bruising was “evolving” and increasing. Cecilia also had bruises on her legs and scalp. Resuscitation attempts failed, unfortunately, and Cecilia was pronounced dead.
Because of the circumstances of the death, an autopsy was performed. Evidence related to it was presented through the testimony of four expert witnesses: the forensic pathologist who performed the autopsy; a forensic pediatrician and a forensic anthropologist who were present and assisted in the autopsy; and a second forensic anthropologist who examined bones after the autopsy.
The autopsy revealed that Cecilia’s skull was fractured all along the right side and that she had multiple subdural and epidural hematomas. Her neck was bruised, which could not have been caused by CPR. According to the Commonwealth’s expert witnesses, Cecilia’s death resulted from the head trauma, which was caused by a substantial impact from her having been thrown or swung against an object. In addition to the skull fracture, the autopsy showed bruises and contusions to the child’s head, and a torn frenulum of her upper lip.
The autopsy also revealed several preexisting injuries, some of which must have resulted from abuse. Cecilia had multiple fractures to her arms, legs, and ribs; a dislocated shoulder; and a pair of round, cigarette-type burns on her legs. The proof showed that the arm, leg, and rib breaks occurred in the weeks or months leading to her death, with the earliest likely occurring in May 2005. One of the forensic anthropologists said one of the breaks could have happened as much as four months back (before the mother moved to Kentucky) but that it was very unlikely.
According to the experts, these injuries were not the cause of death. They were the result of mechanical trauma, such as by pulling, bending or twisting, not disease. In fact, the doctor who performed the autopsy testified that the injuries were very probably caused by abuse, not accident, and that she had never seen such a pattern of injuries without abuse.
At least one of the physicians, the forensic pediatrician, testified that the fractures would have resulted in noticeable swelling and pain. She also stated that the physical examination by Nurse Hays likely would have shown pain. She noted specifically a test wherein the nurse pulled the child into a sitting position by her arms, at which time the child had at least one arm fracture and a dislocated shoulder. She also stated that the muscular “floppiness” seen by Nurse Hays was likely the result of bone fractures that were present at the [815]*815time of the six-month well-baby visit because any movement or tensing of muscles would have been painful for the child.
The emergency physician who saw the baby at the time of death testified that these bone fractures were not noticeable from her visual examination, though her examination was not focused on that and was instead focused on the resuscitation efforts. She said if the child had survived, she would have suspected fractures and would have then done a physical exam and other studies to determine whether they were present.
Rankin presented an expert witness, also a forensic pathologist, who testified that the baby was either hit in the head multiple times with an object or was slammed into something multiple times. He also testified that one or more persons caused the head injuries, which were not accidental. When asked about the other broken bones, he stated that it was very unusual that a broken bone would not cause pain when the child was manipulated, as happens in a medical exam, and agreed that if a medical professional could not detect broken bones, then it was highly likely that a lay person would have less awareness of it.
After all the. proof, the trial court instructed the jury on one count of first-degree criminal abuse as to Acosta. The court’s instructions were somewhat unusual, however, in that they covered two different theories of how Acosta allegedly committed the crime in separate, alternative instructions. The first instruction allowed the jury to find Acosta guilty if she intentionally abused Cecilia. The second instruction, to which the jury was to proceed only if it found her not guilty under the first, allowed the jury to find her guilty if she had intentionally permitted Roy Rankin to abuse Cecilia.
The jury found Acosta guilty under the first instruction, and she was sentenced to ten years in prison.
The Court of Appeals affirmed. While the court addressed several issues, the only one relevant to the case at this point was Acosta’s claim that she was entitled to a directed verdict of acquittal. The court found that there was sufficient evidence under both theories presented to the jury to allow conviction.
As to the second theory, the court noted that the Commonwealth had presented evidence that Acosta had reason to know that Cecilia suffered from serious injuries before her death. Specifically, it noted that Cecilia had cried in an unusual manner and had been unable to hold a bottle; that Acosta had to be threatened with a call to social services or the police before she would take the baby in for a long-lasting fever because she was worried the doctor would report her for the head bruise and cigarette burns. It also stated, erroneously, that there was no proof of a car accident, though we note that some of the proof suggested there had not been one. Regardless, the court stated, there was proof that Cecilia’s injuries could not have been caused by falls from a bed or the supposed car accident. From this proof, the court concluded that a jury could reasonably have believed .that Acosta had allowed Rankin to abuse Cecilia.
As to the first theory, intentional abuse by Acosta, the court acknowledged that the proof was less strong. Nevertheless, it noted, the evidence showed that the injuries were caused intentionally during a period in which Cecilia was in the exclusive custody of Acosta or Rankin. The court concluded: “While it seems more likely that Rankin inflicted the majority of the injuries, a jury could reasonably infer that [Acosta] committed at least one of the acts of abuse.”
[816]*816This Court granted discretionary review to determine whether circumstantial proof such as this, which establishes intentional abuse by someone and only an opportunity for the defendant to have committed that abuse, is sufficient to defeat a motion for a directed verdict.
II. Analysis
Acosta raises only that single issue: whether the trial court should have granted her a directed verdict of acquittal.2 We ultimately conclude that the jury should not have been instructed as to a direct-abuse theory but that a directed verdict should not have been granted because the proof was sufficient under the alternative theory. This still requires reversal of Acosta’s conviction, though it allows for retrial on the alternative theory.
When presented with a motion for a directed verdict, a court must consider the evidence as a whole, presume the Commonwealth’s proof is true, draw all reasonable inferences in favor of the Commonwealth, and leave questions of weight and credibility to the jury. Commonwealth v. Benham, 816 S.W.2d 186, 187-88 (Ky.1991). The trial court is authorized to grant a directed verdict if the Commonwealth has produced no more than a mere scintilla of evidence; if the evidence is more than a scintilla and it would be reasonable for the jury to return a verdict of guilty based on it, then the motion should be denied. Id. On appellate review, the standard is slightly more deferential; the trial court should be reversed only if “it would be dearly unreasonable for a jury to find guilt.” Id. (emphasis added).
A directed-verdict motion is reviewed in light of the proof at trial and the statutory elements of the alleged offense. Lawton v. Commonwealth, 354 S.W.3d 565, 575 (Ky.2011). The directed-verdict question is not controlled by the law as described in the jury instructions, but by the statutes creating the offense. Id. Thus, a directed verdict may be inappropriate even though the jury instructions were flawed. Id.
Under this approach, this Court is required to examine the evidence introduced at trial concerning whether Acosta committed first-degree criminal abuse against Cecilia and to compare that proof to the statutory elements of the offense. The first-degree criminal abuse statute3 sets out three essential elements: (1) the defendant intentionally abuses4 or permits a person in custody to be abused; (2) the [817]*817abuse either causes serious physical injury, creates a danger of serious physical injury, or causes torture, cruel confinement, or cruel punishment; and (3) the victim is less than 12 years of age or is helpless.
The first element is listed in the alternative, meaning there are two different ways to satisfy it: either by committing the abuse directly or allowing the abuse to occur. This case turns on whether the first element was satisfied.
Most of Acosta’s argument is directed toward the proof — or lack of proof — of direct abuse by her. Even though both alternatives — direct abuse and allowing abuse — were presented in the jury instructions, the proof need only satisfy one of the alternatives to survive a directed-verdict motion:
A motion for a directed verdict of acquittal should only be made (or granted) when the defendant is entitled to a complete acquittal!,] i.e., when, looking at the evidence as a whole, it would be clearly unreasonable for a jury to find the defendant guilty, under any possible theory, of any of the crimes charged in the indictment or of any lesser included offenses.
Campbell v. Commonwealth, 564 S.W.2d 528, 530 (Ky.1978).
As the Court of Appeals noted, the proof was stronger as to the theory that Acosta allowed the abuse to occur. And Acosta has not challenged whether she could have been convicted under that theory. Even if she is correct that the direct-abuse theory was not supported by the evidence, if the evidence would support the second, permitting-abuse theory, then she was not entitled to a directed verdict.
The proof, while extremely circumstantial, was sufficient as to the permitting-abuse theory to deny Acosta’s motion for a directed verdict. The Commonwealth’s expert witnesses testified that the child’s broken bones should have been noticeable and were the result of abuse, rather than disease or accident. Moreover, members of Rankin’s family testified that something was clearly wrong with the child, who acted in pain all the time. Acosta and Rankin were the child’s caregivers while she lived in Kentucky.
Other evidence showed Acosta’s consciousness of guilt. Several witnesses testified that Acosta was hesitant to take the child in for medical visits in Kentucky— despite a history of consistent medical visits before Rankin came back into her life. Witnesses also testified that Acosta was specifically concerned about medical personnel calling the police or social services upon seeing the child’s injuries.
Of course, Acosta did eventually take the child to a medical appointment after the injuries occurred. And while the medical professional was concerned about some of the child’s superficial injuries, she did not notice any broken bones. While such proof certainly could be the basis of a defense — if a medical professional did not notice the injuries, then how could the mother as a lay person? — it does not entitle Acosta to a directed verdict. As noted above, when considering the appropriateness of granting a directed verdict, the court is required to view the evidence in the light most favorable to the Commonwealth and questions of credibility and weight are to be left for the jury. Under this standard, Nurse Hays’s testimony could be ignored.
Acosta’s own brief states that the “evidence — if believed by the jury — might have supported a conviction for knowingly allowing Roy to abuse Cecilia.” This Court concludes that a jury could reasonably infer from the proof that Acosta knew that Rankin, who was the child’s caregiver while she worked, was abusing the child [818]*818and that she permitted it to occur, and therefore that she committed first-degree criminal abuse by permitting abuse to occur. Therefore, the trial court did not err in refusing to direct a verdict of acquittal based on the theory that Acosta permitted the abuse to occur.
That, however, does not mean that there was no error in this case. As this Court has stated in a similar case: “The fact that [the defendant] was not entitled to a directed verdict, however, does not necessarily mean that the trial court acted properly in instructing the jury on all the alternate methods of committing criminal abuse in the first degree.” Mason v. Commonwealth, 331 S.W.3d 610, 618 (Ky.2011).
The type of error alleged here is actually that the instruction under which the jury convicted Acosta was improper and unsupported by the proof, which is different from a directed-verdict complaint. Specifically, Campbell stated that when a directed verdict is inappropriate because the evidence was sufficient “under any possible theory ... of any of the crimes charged in the indictment or of any lesser included offenses ... [t]he proper method for obtaining relief ... would have been to object to the instruction upon which the jury’s finding ... was based.” Campbell, 564 S.W.2d at 530. This has consistently been the rule in Kentucky. See, e.g., Kimbrough v. Commonwealth, 550 S.W.2d 525, 529 (Ky.1977) (“When the evidence is insufficient to sustain the burden of proof on one or more, but less than all, of the issues presented by the case, the correct procedure is to object to the giving of instructions on those particular issues.... The appropriate procedure here would thus have been for appellant, at the close of the evidence and before the instructions were given, to apprise the trial court that he objected to the giving of an instruction or instructions ... for the reason that they had not been sufficiently proven.”).
In other words, here there was proof sufficient to present the charge of abuse to the jury based on permitting the abuse so that Acosta was not entitled to a directed verdict of acquittal on the charge. But should there have been an instruction on the theory that she directly and intentionally committed the abuse on the child under the state of the proof? Acosta thus was required to object to the first instruction on direct abuse if she believed that the evidence did not support such a theory.
This raises the question of whether Acosta has properly preserved this potential instructional error for ordinary appellate review. It is fairly clear that on appeal, Acosta has not distinguished between whether she could be convicted of the charge on any theory and whether there was proof sufficient to support an instruction on the theory of direct abuse (directed verdict versus an instructional theory not supported by the proof). That Acosta has confused the two issues is evinced by the discussion in her reply brief stating that the appellate court must focus on what the jury did and whether its decision was reasonable. But a directed-verdict motion is directed to the trial court, before the jury gets the case, and the decision on review then is the trial court’s decision, not the jury’s. This is why the directed-verdict standard focuses on whether the trial court was correct in concluding that a rational trier of fact could find guilt under any theory, even if the theory under which the jury actually found guilt was not supported by the evidence. When Acosta complains that this jury acted unreasonably under a specific jury instruction and that the evidence did not support giving that instruction, she is making a different type of claim. She is instead complaining about the appropriateness of the jury instruction, [819]*819and any error goes to the jury’s verdict as found under that instruction.
This Court’s review of the record does not show that Acosta objected to the giving of the jury instructions that covered both the direct-abuse and permitting-abuse theories of the crime. If a conference with the judge on the instructions was held, it either was not recorded or was not certified as part of the appellate record. And the proposed jury instructions include both theories, though it is unclear whether those proposed instructions were tendered by the Commonwealth or Acosta.
Practically speaking, there is not much difference between the directed-verdict question and the improper instruction question, since both focus on the sufficiency of the evidence. (Of course, there is one substantial difference: when a defendant is entitled to a directed verdict, it is a directed verdict of acquittal, whereas an improper instruction would require at most reversal of the conviction and allow retrial.) Nevertheless, our cases hold that a motion for a directed verdict does not preserve the type of error presented by this case.
Because the error was not adequately preserved for appellate review, reversal is allowed only if this Court finds the error to be palpable. See RCr 10.26. A palpable error occurs when the substantial rights of a defendant are violated and a manifest injustice results. RCr 10.26. Manifest injustice requires “showing ... [a] probability of a different result or error so fundamental as to threaten a defendant’s entitlement to due process of law.” Martin v. Commonwealth, 207 S.W.3d 1, 3 (Ky.2006). Elsewhere in that decision, we stated that the rule required deciding “whether the defect in the proceeding was shocking or jurisprudentially intolerable.” Id. at 4.
If Acosta is correct that the jury instruction under which the jury found her guilty was unsupported by the evidence, then the trial court committed palpable error.5 That the proof in a criminal prosecution must be sufficient to allow a reasonable jury to find guilt beyond a reasonable doubt is one of the bedrocks of the American justice system and is one of the core protections of due.process. That alone would make the type of error presented here jurisprudentially intolerable. It is not enough to speculate that the jury could have convicted her of permitting abuse. Here, the jury did not consider that charge, since it convicted her under the first instruction on direct abuse. Certainly, had the court instructed only on direct abuse, and had there not been evidence sufficient to submit the permitting abuse theory to the jury, Acosta would have been entitled to a directed verdict on the direct abuse under the proof in this case.
The evidence at trial showed only that Acosta, as the child’s mother and caregiver, had the opportunity to abuse her child.
The Commonwealth claims the question of sufficiency of the evidence in this case is like that in Mason v. Commonwealth, 331 S.W.3d 610 (Ky.2011). The Commonwealth goes so far as to claim that the present case offered more evidence to connect Acosta to the direct abuse than there was in Mason. Despite this claim, howev[820]*820er, the proof in Mason was substantially more compelling than in this case and included more than showing a mere opportunity by the defendant to have caused the injuries. Indeed, there was really no question in Mason whether the defendant caused the injuries or that they may have been caused by someone else. That case involved a single injury that the defendant claimed occurred when he tripped and fell on the child. In other words, the defendant admitted to causing the injury; he simply disagreed with whether the injury was caused by intentional abuse.
The Court of Appeals relied on Carpenter v. Commonwealth, 771 S.W.2d 822 (Ky. 1989), for its conclusion that a reasonable jury could have found guilt based solely on Acosta’s status as the child’s caregiver (and resulting opportunity to commit the crime). But Carpenter presented far more evidence than a mere opportunity to commit the crime. In that case, the defendant, along with another person, had exclusive control over the abused child, but the proof also showed that the defendant had abused the child (by shaking her and throwing her on a bed so that she bounced off and hit her head) shortly before the abusive act for which he was convicted. It was the combination of this additional evidence with the opportunity that was sufficient to avoid a directed verdict. Id. at 824.
But this case presents only an opportunity for Acosta to have committed the abuse herself. Mere opportunity, like motive or presence at the scene, is insufficient by itself to support a finding of guilt. See, e.g., Marcum v. Commonwealth, 496 S.W.2d 346, 349 (Ky.1973) (“neither motive alone nor motive plus opportunity (or presence at the scene) is enough to justify a conviction”); Hodges v. Commonwealth, 473 S.W.2d 811 (Ky.1971); Brison v. Commonwealth, 519 S.W.2d 833, 838 (Ky.1975). Indeed, the direct-abuse case against Acosta was even weaker than many of these “opportunity only” cases because in those, the defendants were at least present when the crimes occurred. Here, there is no proof that Acosta was present when the abuse occurred. At most, there is a question as to whether she should have taken her child to a doctor sooner, though she ultimately did do so. There is certainly no evidence that anyone saw or heard Acosta abusing her child. A finding that Acosta committed the abuse would require the jury to move beyond fair and reasonable inferences from the evidence to rank speculation. A verdict cannot be founded on nothing more than conjecture.
Because the Commonwealth’s proof at most showed that Acosta had the opportunity to commit the abuse, it was error to instruct the jury on intentional abuse by Acosta herself. Because such an error rises to the level of palpable error, as described above, Acosta’s conviction, having been found only under the direct-abuse instruction, must be reversed.
III. Conclusion
Because the Commonwealth’s proof was sufficient under at least one theory of first-degree criminal abuse, Acosta was not entitled to a directed verdict of acquittal on the charge of criminal abuse. However, the trial court erred in instructing the jury as to direct abuse by Acosta, and the resulting jury verdict under that instruction was erroneous. Acosta may nevertheless be retried under the alternative theory of permitting the abuse because the jury never reached that question, and she was not entitled to a directed verdict of acquittal. Therefore, the judgment of the Court of Appeals is reversed, and Acosta’s conviction for first-degree criminal abuse is reversed.
[821]*821MINTON, C.J.; ABRAMSON, CUNNINGHAM, SCOTT and VENTERS, JJ., sitting. All concur, except SCOTT, J., concurs in result only.