Scott Jones v. Jeri Taylor

FOR PUBLICATION UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT SCOTT DOUGLAS JONES, No. 13-36202 Petitioner-Appellee, D.C. No. v. 3:10 cv-1474 JO JERI TAYLOR, Superintendent, Two Rivers Correctional Institution, OPINION Respondent-Appellant. Appeal from the United States District Court for the District of Oregon Robert E. Jones, Senior District Judge, Presiding Argued and Submitted May 12, 2014—Portland, Oregon Filed August 19, 2014 Before: Arthur L. Alarcón, A. Wallace Tashima, and Sandra S. Ikuta, Circuit Judges. Opinion by Judge Tashima 2 JONES V. TAYLOR SUMMARY* Habeas Corpus The panel reversed the district court’s judgment granting a habeas corpus petition based on a freestanding claim of actual innocence in a case in which Scott Jones was convicted of unlawful sexual penetration for the sexual abuse of his sister, S.J. Jones’ conviction was based primarily on S.J.’s testimony that Jones inserted his finger inside her vagina on multiple occasions, as well as testimony by S.J.’s and Jones’ father, Ken Jones, and sister, Jennifer Pond, that Jones admitted to penetrating S.J. Jones brought his freestanding claim of factual innocence based on the recantations of all three witnesses. The panel chose to review de novo the district court’s conclusion that Jones is actually innocent, based on the panel’s holistic assessment of the evidence adduced at the hearings before the district court and at trial and the likely effect all this evidence would have on reasonable jurors in order to clarify how district courts should evaluate actual innocence claims. The panel did not resolve whether a freestanding actual innocence claim is cognizable in a federal habeas corpus proceeding in the non-capital context because, even assuming * This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader. JONES V. TAYLOR 3 that such a claim is cognizable, the panel concluded that Jones has not made a sufficient showing to merit relief. The panel explained that as a general matter, recantation testimony is properly viewed with great suspicion. The panel did not rely on the district court’s findings that S.J. and Ken Jones were credible because the panel was uncertain about the basis for some of the district court’s conclusions and unpersuaded that every reasonable juror would credit the recantations as the district court did. The panel explained that even if it accepts that the three witnesses testified truthfully based on their memory at the time of the evidentiary hearing, their recantations are insufficient to demonstrate that Jones is factually innocent, where neither Ken Jones nor Jennifer Pond witnessed the abuse, and where the panel could not say that every juror would credit S.J.’s recantation testimony over her trial testimony and the descriptions of the abuse she gave in her 2000 and 2002 interviews. The panel explained that evidence that merely undercuts trial testimony or casts doubt on the petitioner’s guilt, but does not affirmatively prove innocence, is insufficient to merit relief on a freestanding claim of actual innocence. The panel concluded that Jones has not made the extraordinarily high and truly persuasive showing required for habeas relief on a freestanding claim of actual innocence. 4 JONES V. TAYLOR COUNSEL Ellen F. Rosenblum, Attorney General of Oregon, Anna M. Joyce, Solicitor General, Paul L. Smith (argued), Attorney-in- Charge, Criminal and Collateral Remedies Appeals, Oregon Department of Justice, Salem, Oregon, for Respondent- Appellant. Ellen C. Pitcher (argued), Assistant Federal Public Defender, Portland, Oregon, for Petitioner-Appellee. OPINION TASHIMA, Circuit Judge: In 2003, Petitioner Scott Jones was convicted of unlawful sexual penetration for the sexual abuse of his sister, S.J. Jones’ conviction was based primarily on S.J.’s testimony that Jones inserted his finger inside her vagina on multiple occasions, as well as testimony by S.J.’s and Jones’ father, Ken Jones, and sister, Jennifer Pond, that Jones admitted to penetrating S.J. All three witnesses now recant their testimonies. Based on these recantations, Jones brought a federal habeas petition seeking relief based on a freestanding claim of actual innocence. The district court assumed that a freestanding actual innocence claim existed in this context and granted JONES V. TAYLOR 5 relief. We have jurisdiction under 28 U.S.C. §§ 1291 and 2253 and, for the reasons discussed below, we reverse.1 I. A. Jones was convicted of three counts of unlawful sexual penetration in the first degree pursuant to Oregon Revised Statute § 163.411 for the sexual abuse of his sister, S.J.2 He was sentenced to three concurrent 100-month terms of imprisonment, plus a consecutive 75-month term for a related offense. His conviction and sentence were affirmed on direct appeal. 1 Because Jones did not raise his actual innocence claim in state court, it was not “adjudicated on the merits in State Court proceedings,” and, therefore, the limitations imposed on our habeas review by 28 U.S.C. § 2254(d) do not apply. While a habeas petitioner in federal court must ordinarily exhaust his claims in state court, see 28 U.S.C. § 2254(b)(1), we review Jones’ unexhausted actual innocence claim pursuant to 28 U.S.C. § 2254(b)(2) (“An application for a writ of habeas corpus may be denied on the merits, notwithstanding the failure of the applicant to exhaust the remedies available in the courts of the State.”). 2 The statute provides: [A] person commits the crime of unlawful sexual penetration in the first degree if the person penetrates the vagina, anus or penis of another with any object other than the penis or mouth of the actor and . . . [t]he victim is under 12 years of age . . . . OR. REV. STAT. § 163.411(1) (2014). 6 JONES V. TAYLOR At Jones’ 2003 trial, S.J. testified that Jones went inside her vagina with his finger on multiple occasions in late 1998 or early 1999, when she was approximately nine years old and Jones was approximately seventeen. She testified that Jones “wiggled” his finger and moved it “back and forth” inside her vagina, and that it hurt when he did this. Her testimony at trial was consistent with her description of the abuse in two videotaped interviews that were played for the jury, one with a mental health counselor in 2000 and another with a police detective in 2002. In these interviews, as in her trial testimony, S.J. stated that Jones touched inside her vagina with his finger, though her memory seemed to have faded somewhat by the 2002 interview, and she was less sure in that interview of what had happened than she had been in the 2000 interview. S.J. and Jones’ father, Ken Jones, and their sister, Jennifer Pond, also testified at trial. Ken Jones testified that Jones admitted that he penetrated S.J., recounting that Jones said: “I admit, I penetrated her.” Jennifer Pond similarly testified that Jones admitted to penetrating S.J., stating that he said “I admit the full thing,” in the context of a conversation about allegations that he abused S.J. Jennifer Pond also testified at trial that in late 1998 or early 1999, S.J. came to her complaining of vaginal pain and that she noticed that S.J.’s vaginal area looked “a little red.” The late 1998 or early 1999 time period coincided with the time period in which the sexual abuse occurred.3 3 The 1998 or 1999 time period is based on S.J.’s testimony that the sexual abuse occurred while the family was living in a house in Toledo, into which the family moved in late 1998. JONES V. TAYLOR 7 All three witnesses have since recanted. S.J. came forward in 2012 saying that her description of the abuse at trial and in the 2000 and 2002 interviews was inaccurate. S.J. explains that her previous testimony that Jones put his finger inside her vagina was inaccurate and that he did not, in fact, put his finger inside her vagina. She says she was mistaken in her trial testimony and in the 2000 and 2002 interviews because she did not at the time understand that her genitals had an inside area capable of penetration. She explains that she was raised in a very conservative household, where she never received sex education, and therefore did not understand her sexual anatomy. She maintains that now that she better understands her sexual anatomy, she knows that Jones did not penetrate her vagina with his finger. Ken Jones and Jennifer Pond also recanted their testimony at around the same time S.J. came forward. Ken Jones now claims that Jones did not say he penetrated her, but rather that he said “I admit the whole thing.” Ken Jones explains that he previously assumed Jones was referring to sexual penetration of S.J., but upon further reflection, now realizes that Jones was actually referring to a burglary when he said “I admit the whole thing.” Jennifer Pond now similarly claims that the admission was made in the context of a conversation about a burglary, not sexual abuse. She also now claims that it was sometime between 1993 and 1997 when S.J. complained to her of vaginal pain and she observed redness in S.J.’s vaginal area, rather than 1998 or 1999. She explains the change in the timeline only by saying, “I guess that’s how I remembered it at the time.” Based on this new evidence, Jones claims that he is factually innocent of the crime of sexual penetration and seeks release on that ground. Jones does not contend that he 8 JONES V. TAYLOR did not sexually abuse S.J. S.J. testified at the evidentiary hearing that Jones touched her genitals, and Jones admitted to doing so in a separate hearing before the district court. Jones’ only contention is that he is innocent of the crime of unlawful sexual penetration because, although he touched S.J.’s genitals with his hand, he did not penetrate her vagina when he touched her genitals. B. On December 2, 2010, Jones filed a pro se petition for a writ of habeas corpus under 28 U.S.C. § 2254 in federal district court. On July 5, 2012, Jones, through counsel, filed an amended petition. On August 9, 2013, Jones filed a motion for release pending resolution of the habeas proceedings, and on August 28, 2013, the district court held a hearing on that motion.4 Jones and Ken Jones testified at the August 28, 2013, hearing. During his testimony in that proceeding, Jones, who did not testify at trial, stated that he never penetrated S.J.’s vagina. It was also at the August 28, 2013, hearing that Ken Jones recanted his trial testimony, explaining that his statement that Jones admitted to sexually penetrating S.J. was incorrect. The district court continued the hearing, and heard additional testimony on October 23 and 24, 2013. During those hearings, S.J. and Jennifer Pond testified. S.J. testified to the facts described above, stating that Jones had never penetrated her, and Jennifer Pond recanted her trial testimony, as described above. 4 This hearing on the motion for release pending resolution of the habeas proceeding turned into an unscheduled and non-noticed three-day evidentiary hearing on the merits of the actual innocence claim. JONES V. TAYLOR 9 On December 19, 2013, the district court issued an opinion and order granting Jones’ habeas petition. Jones v. Franke, 2013 WL 6780605 (D.Or. 2013). It denied relief on four of Jones’ claims, which are not relevant to this appeal, but granted relief on the freestanding actual innocence claim. Id. at *8–*10. The district court found S.J.’s recantation credible in full, id. at *10, and credited the portion of Ken Jones’ recantation in which he stated that Jones had never admitted to penetrating S.J., id. at *9. The district court further concluded that Jennifer Pond’s recantation was irrelevant to Jones’ actual innocence claim. Id. at *9–*10. Based on the recantations by S.J. and Ken Jones, the district court found that, assuming a freestanding actual innocence claim was cognizable, Jones had made a sufficient showing of actual innocence on the unlawful sexual penetration charge to merit habeas relief. It therefore ordered the State to release Jones from custody and discharge him from all adverse consequences related to the unlawful penetration convictions. Id. at *10–*11. The State timely appealed. II. In general, we review de novo the district court’s decision to grant or deny a habeas petition, while factual findings and credibility determinations underlying the decision are reviewed for clear error. Lambert v. Blodgett, 393 F.3d 943, 964 (9th Cir. 2004). However, as we recently recognized in Stewart v. Cate, No. 10-55985, 2014 WL 1707033, at *7 (9th Cir. May 1, 2014) (as amended), the standard of review applicable to claims of actual innocence “is not entirely settled in this circuit.” Jones contends that the district court’s conclusion that he is actually innocent of the crime of unlawful sexual 10 JONES V. TAYLOR penetration is a factual finding that we should review for clear error. Jones, however, overstates the effect of the district court’s decision: While we review for clear error the district court’s credibility findings as to the witnesses who testified at the evidentiary hearing, the district court’s conclusion that Jones is actually innocent of the crime of unlawful sexual penetration is a question we review either de novo or for abuse of discretion. See id. (contrasting the abuse of discretion review applied in Schlup v. Delo, 513 U.S. 298, 333–34 (1995) (O’Connor, J., concurring), and Paradis v. Arave, 130 F.3d 385, 396–99 (9th Cir. 1997), with the de novo review applied in cases like House v. Bell, 547 U.S. 518, 539–40 (2006), and Larsen v. Soto, 742 F.3d 1083, 1092 n.6 (9th Cir. 2013). As the Supreme Court has explained, an actual innocence finding “requires a holistic judgment about ‘all the evidence’ and its likely effect on reasonable jurors applying the reasonable-doubt standard.” House, 547 U.S. at 539 (quoting Schlup v. Delo, 513 U.S. 298, 328 (1995)). “[T]he inquiry does not turn on discrete findings regarding disputed points of fact, and ‘[i]t is not the district court’s independent judgment as to whether reasonable doubt exists that the standard addresses.’” Id. at 539–40 (quoting Schlup, 513 U.S. at 329 (emendations in original)). As in Stewart, “[w]e need not determine which standard is correct in this case . . . because under either standard [Jones] has failed to establish” a freestanding claim of actual innocence.5 2014 WL 1707033, at *7. We therefore choose 5 In its opinion holding that Jones established his actual innocence of sexual penetration, the district court abused its discretion by applying the wrong legal standard. See Koon v. United States, 518 U.S. 81, 100 (1996) (“A district court by definition abuses its discretion when it makes an error of law.”). While the Supreme Court has indicated that a court must make JONES V. TAYLOR 11 to review de novo the district court’s conclusion that Jones is actually innocent, based on our holistic assessment of the evidence adduced at the hearings before the district court and at trial and the likely effect all this evidence would have on reasonable jurors in order to clarify how district courts should evaluate actual innocence claims. III. A. We have not resolved whether a freestanding actual innocence claim is cognizable in a federal habeas corpus proceeding in the non-capital context, although we have assumed that such a claim is viable. See Osborne v. Dist. Attorney’s Office for the Third Judicial Dist., 521 F.3d 1118, 1130 (9th Cir. 2008), rev’d on other grounds, 557 U.S. 52 (2009); see also McQuiggin v. Perkins, 133 S. Ct. 1924, 1931 (2013) (noting that it is, as yet, unresolved whether a freestanding actual innocence claim is cognizable in a federal habeas proceeding); Herrera v. Collins, 506 U.S. 390, 417 (1993) (acknowledging the possibility that a freestanding actual innocence claim would exist in the capital context). We need not resolve this difficult question today, however, because, even assuming that such a free standing claim of actual innocence in a non-capital case is cognizable, we “a holistic judgment about ‘all the evidence’ and its likely effect on reasonable jurors applying the reasonable-doubt standard,” House, 547 U.S. at 539 (quoting Schlup, 513 U.S. at 328), here, the district court evaluated only the new evidence, the witnesses’ recantations, without engaging in a holistic assessment of all the evidence or explaining the likely effect of the new evidence on reasonable jurors. See Jones, 2013 WL 67800605, at *9–*10. 12 JONES V. TAYLOR conclude that Jones has not made a sufficient showing to merit relief. The standard for establishing a freestanding claim of actual innocence is “‘extraordinarily high’ and . . . the showing [for a successful claim] would have to be ‘truly persuasive.’” Carriger v. Stewart, 132 F.3d 463, 476 (9th Cir. 1997) (quoting Herrera, 506 U.S. at 417). We have held that, at a minimum, the petitioner must “go beyond demonstrating doubt about his guilt, and must affirmatively prove that he is probably innocent.” Id. (citing Herrera, 506 U.S. at 442–44 (Blackmun, J., dissenting)). While we have not articulated the precise showing required, we have discussed the standard for a freestanding actual innocence claim by reference to the Schlup “gateway” showing, which permits a petitioner to proceed on a procedurally barred claim by showing actual innocence. See, e.g., House, 547 U.S. at 554–55; Carriger, 132 F.3d at 477. In order to pass through the Schlup actual innocence gateway, a petitioner must demonstrate that “in light of new evidence, ‘it is more likely than not that no reasonable juror would have found [the] petitioner guilty beyond a reasonable doubt.’” House, 547 U.S. at 537 (quoting Schlup, 513 U.S. at 327). This new evidence must be reliable, and the reviewing court “may consider how the timing of the submission and the likely credibility of the affiants bear on the probable reliability of that evidence.” Schlup, 513 U.S. at 332. The federal habeas court “must consider all the evidence, old and new, incriminating and exculpatory, without regard to whether it would necessarily be admitted under rules of admissibility that would govern at trial.” House, 547 U.S. at 538 (internal quotation marks and citation omitted). “Based on this total record, the court must make ‘a probabilistic determination JONES V. TAYLOR 13 about what reasonable, properly instructed jurors would do.’” Id. (quoting Schlup, 513 U.S. at 329). The Supreme Court most recently applied this framework in House. See id. at 554–55. There, new DNA evidence showed that semen found on the victim was not the petitioner’s, negating the prosecution’s theory of motive and undermining the petitioner’s link to the crime scene. Id. at 540–41. Additionally, new scientific evidence proved that blood found on the petitioner’s clothes could not have come from the victim while she was alive. Id. at 542–46. The evidence in that case also included new testimony from multiple, disinterested witnesses who testified to facts implicating a different suspect and whose testimony was supported by independent evidence. Id. at 548–53. The Court found this evidence insufficient to meet the high standard required to merit relief on a freestanding actual innocence claim because it was not a case of “conclusive exoneration,” and several pieces of evidence remained uncontested that pointed to the petitioner’s guilt. For example, there was blood on the petitioner’s pants and testimony that could reasonably be interpreted as connecting him to the crime scene. Id. at 553–55. In Jackson v. Calderon, 211 F.3d 1148 (9th Cir. 2000), we similarly applied this framework and rejected a freestanding actual innocence claim despite new scientific evidence. There, the petitioner presented new expert medical testimony that, to a 95 percent medical certainty, the petitioner could not have had the requisite premeditation and specific intent to kill because he was intoxicated with PCP at the time of the murder. Id. at 1165. We concluded that although that evidence “certainly cast doubt” on the petitioner’s guilt, it was insufficient to warrant relief on a 14 JONES V. TAYLOR freestanding habeas petition because another doctor testified that a person with petitioner’s level of intoxication would “not necessarily be unable to process thought, premeditate, deliberate and intend to kill.” Id. Finally, in Carriger, we rejected a freestanding actual innocence claim based on our conclusion that the petitioner had “presented no evidence, for example, demonstrating he was elsewhere at the time of the murder, nor [was] there any new and reliable physical evidence, such as DNA, that would preclude any possibility of [his] guilt.” 132 F.3d at 477. There, we rejected the petitioner’s claim even though another suspect reliably confessed to the murder, described details of the crime that only a participant would have known, and boasted that the petitioner had been set up, and all the other evidence pointed as directly to the new suspect as to the petitioner. Id. at 478–79. B. With these cases as guideposts, we cannot say that Jones has demonstrated that he is probably innocent. Jones asserts his innocence based on recantation testimony alone. His petition is therefore lacking the type of proof Carriger implied might be sufficient; indeed, the proof is even less reliable than the evidence rejected in House, Jackson, and Carriger, because it is all in the form of recantation testimony, uncorroborated by any other evidence. As a general matter, “[r]ecantation testimony is properly viewed with great suspicion.” Dobbert v. Wainwright, 468 U.S. 1231, 1233 (1984) (Brennan, J., dissenting from denial of certiorari); see also Allen v. Woodford, 395 F.3d 979, 994 (9th Cir. 2004). “Recanting testimony is easy to find but difficult to confirm or refute: witnesses forget, witnesses JONES V. TAYLOR 15 disappear, witnesses with personal motives change their stories many times, before and after trial.” Carriger, 143 F.3d at 483 (Kozinski, J., dissenting). “It upsets society’s interest in the finality of convictions, is very often unreliable and given for suspect motives . . . .” Dobbert, 468 U.S. at 1233–34. For these reasons, a witness’ “later recantation of his trial testimony does not render his earlier testimony false.” Allen, 395 F.3d at 994; see also Christian v. Frank, 595 F.3d 1076, 1084 n.11 (9th Cir. 2010). Rather, a witness’ recantation is considered in addition to his trial testimony and in the context in which he recanted when assessing the likely impact it would have on jurors. See Christian, 595 F.3d at 1084 n.11 (considering the timing of the witness’ recantation and the contents of his earlier testimony in assessing the weight of the recantation); Graves v. Cockrell, 351 F.3d 143, 153 (5th Cir. 2003) (noting that a recanting witness had given numerous contradictory statements in assessing the weight to give to his new testimony). Although the district court found S.J. credible and Ken Jones credible in part, we do not rely on these determinations because we are uncertain about the basis for some of the district court’s conclusions and unpersuaded that every reasonable juror would credit the recantations as the district court did. See House, 547 U.S. at 539–40 (noting that it would consider the likely effect of testimony on reasonable jurors and was not bound by the district court’s assessment of the witnesses, especially given that its explanation for its credibility finding was unclear). The district court stated that it credited Ken Jones’ testimony, but also noted that it was apparent from the record that his motivation for recanting was his naive belief that Jones could not have digitally penetrated S.J.’s vagina 16 JONES V. TAYLOR because a subsequent physical examination demonstrated that S.J.’s hymen was intact. Jones, 2013 WL 6780605, at *9. From the record before us, it appears that the district court did not resolve this conflict in crediting Ken Jones’ testimony. Similarly, the district court credited S.J.’s testimony, but also noted that it could not “assess to what degree, if any [S.J.’s] present recantation has been influenced by her family—K. Jones and Pond in particular.” Id. at *10. The district court’s failure to resolve whether S.J. was influenced to recant by her family is particularly troubling in light of its description of Ken Jones as “imposing and controlling” and someone who “dominated” “a highly dysfunctional home.” Id. In other words, we cannot assess why the district court found S.J.’s testimony credible despite the possibility, which it recognized, that her father pressured her into recanting. Although the district court noted that her testimony about her conservative upbringing and limited knowledge was corroborated by the record, this conclusion does not preclude the possibility that S.J. was influenced to testify by her family and does not prove the truth of her assertion that Jones did not penetrate her. We therefore conclude that the district court’s bases for crediting S.J.’s and Ken Jones’ testimony are unclear, “a consideration that weakens our reliance on its determinations.” See House, 547 U.S. at 539–40. Several features of the recantations here persuade us that they are insufficient to prove Jones’ innocence. The recantations are all from Jones’ family members, which reduces their weight and reliability. See House, 547 U.S. at 552 (noting that testimony by friends or relations of the accused might have less probative value than testimony from disinterested witnesses); McCray v. Vasbinder, 499 F.3d 568, 573 (6th Cir. 2007) (noting that family members might have JONES V. TAYLOR 17 a personal stake in a defendant’s exoneration). Moreover, all three witnesses came forward with changed stories at roughly the same time, years after trial, and only one of them provided a reason for the delay. The other two changed their stories long after trial with no more explanation than that their memories and understandings of the events had changed. The timing of the recantations casts some doubt on their veracity, especially as to Ken Jones and Jennifer Pond, who provide no explanation for their delay in coming forward. See Christian, 595 F.3d at 1084 n.11 (noting that a witness’ recantation was “especially unreliable given that it was made more than a decade after his original [testimony]”); McCray, 499 F.3d at 573 (discounting evidence from witnesses who did not provide a good explanation for why they delayed in coming forward). But even if we accept that the three witnesses testified truthfully based on their memory at the time of the evidentiary hearing, i.e., if we accept, as the district court did, that they did not change their story for ulterior motives, their recantations are insufficient to demonstrate that Jones is actually innocent under the standard applied in House, Jackson, and Carriger. Neither Ken Jones’ nor Jennifer Pond’s recantation constitutes compelling evidence of Jones’ innocence, even assuming the jury were to believe their recantations over their trial testimony. Because neither witnessed the abuse, their testimony is of little weight in the actual innocence analysis. Cf. Schlup, 513 U.S. at 324 (identifying “trustworthy eyewitness accounts” as evidence that might be sufficient to show actual innocence); Carriger, 132 F.3d at 483 (same); see also Cox v. Burger, 398 F.3d 1025, 1031 (8th Cir. 2005) (discounting testimony from a witness who was not present at the scene of the crime in assessing an actual innocence claim). A reasonable juror 18 JONES V. TAYLOR could either convict or acquit based on S.J.’s testimony, regardless of whether Jones admitted to his father and sister that he penetrated S.J. or whether S.J. complained to Jennifer Pond of vaginal area soreness around the time of the abuse. That leaves S.J.’s recantation as the most compelling evidence of Jones’ innocence. See Gandarela v. Johnson, 286 F.3d 1080, 1086 (9th Cir. 2001) (suggesting that victim recantation or eye witness testimony may be given more weight in assessing an actual innocence claim than other types of evidence). But see Doe v. Menefee, 391 F.3d 147, 173 (2d Cir. 2004) (rejecting a Schlup claim despite the victim’s recantation). We do not, however, find that even her recantation is sufficient to establish Jones’ probable innocence because we cannot say that every juror would credit her recantation testimony over her trial testimony and the descriptions of the abuse she gave in her 2000 and 2002 interviews, even if they believed that she testified truthfully to the best of her present recollection at the evidentiary hearing. There are several impediments to fully crediting S.J.’s recantation testimony over her trial testimony. Like much recantation testimony, S.J.’s recantation occurred years after the events she describes. Jones abused S.J. in late 1998 or early 1999, when S.J. was nine years old; this was approximately thirteen years before she came forward with her recantation. A reasonable juror could very well believe that S.J.’s memory of the abuse faded or changed in the more than thirteen years since the incident occurred and, for that reason, credit the testimony that was closer in time to the abuse. See Herrera, 506 U.S. at 403–04 (noting that the passage of time diminishes the reliability of criminal convictions, in part due to the erosion of memory that occurs JONES V. TAYLOR 19 over time); McCleskey v. Zant, 499 U.S. 467, 491 (1991) (recognizing that witnesses’ memories erode over time); see also Christian, 595 F.3d at 1084 n.11. There is also some specific indication in the record before us that S.J.’s memory of the abuse might have faded or changed in the thirteen years that have passed since the abuse occurred. S.J.’s memory of the abuse appeared to have already faded somewhat by 2002. In her 2002 interview, S.J. recounted many details of the abuse, but also stated repeatedly that she did not remember exactly what had happened. A reasonable juror could infer that if S.J. forgot many details of the abuse three years after it occurred, her memory of the abuse had further faded in the additional ten years since that time. A reasonable juror could thus conclude that S.J.’s earlier testimony was more reliable and, therefore, credit her earlier description that Jones penetrated her vagina with his finger over her present assertion that he did not. Further, many of the issues raised in S.J.’s recantation were presented to the jury at trial. S.J. explains in her recantation that she did not understand her anatomy at the time of trial or the preceding interviews, so she was incorrect when she said that Jones penetrated her vagina. Jones further contends in his petition that the terms S.J. used at trial – the term “vagina,” in particular – were provided to her and that she did not understand what they meant, in part due to her conservative upbringing. These facts are of little relevance in establishing Jones’ innocence on a habeas petition, however, because the jury was presented with these issues at trial and convicted Jones anyway. The jury was aware when it rendered its decision that S.J. had limited knowledge of her sexual anatomy and that she learned the terms she used to describe the abuse during interviews about the abuse. That 20 JONES V. TAYLOR the jury nevertheless voted to convict Jones of unlawful sexual penetration suggests that it did not rely on S.J.’s knowledge of her sexual anatomy in concluding that Jones penetrated S.J. with his finger. See Wood v. Hall, 130 F.3d 373, 379 (9th Cir. 1997) (“That the jury nevertheless voted to convict [despite its knowledge that the victim had an intact hymen] suggests that they did not believe an intact hymen disproved [the defendant’s] guilt.”). This circumstance therefore undermines the inference that no reasonable juror would have convicted Jones in light of the purportedly new evidence about S.J.’s limited anatomical knowledge at the time of trial. See id. (discounting the relevance of facts of which the jury was aware in establishing actual innocence). Finally, a reasonable juror would not be required to rely exclusively on S.J.’s assessment of whether Jones penetrated her vagina in determining his guilt. A reasonable juror could conclude that Jones penetrated S.J. based on S.J.’s description of Jones’ actions and her sensations of them, regardless of whether she describes the abuse as penetration. In her interviews and trial testimony, S.J. said that it hurt when Jones touched her genitals and that Jones “wiggled” his finger and moved it “back and forth.” A reasonable juror could conclude that this description is consistent with penetration, even if S.J. did not know at the time what it meant to be penetrated.6 The most that can be said of the new testimony is that it undercuts the evidence presented at trial. Evidence that 6 The statute, Or. Rev. Stat. § 16.411(1), does not define “penetration,” but in the related crime of rape, “sexual intercourse, is defined as having “its ordinary meaning and occurs upon any penetration, however slight.” Or. Rev. Stat. § 163.305(7). JONES V. TAYLOR 21 merely undercuts trial testimony or casts doubt on the petitioner’s guilt, but does not affirmatively prove innocence, is insufficient to merit relief on a freestanding claim of actual innocence. See House 547 U.S. at 555 (rejecting freestanding actual innocence claim even though the petitioner had “cast considerable doubt on his guilt”); Jackson, 211 F.3d at 1165 (rejecting a freestanding actual innocence claim even though the petitioner’s new evidence “certainly cast doubt on his conviction”); Carriger, 132 F.3d at 477 (rejecting a freestanding claim when the postconviction evidence “serve[d] only to undercut the evidence presented at trial, not affirmatively to prove [the petitioner’s] innocence”). There is no “new and reliable physical evidence, such as DNA, that would preclude any possibility of [Jones’s] guilt.” Carriger, 132 F.3d at 477. Nor is there scientific or testimonial evidence even as persuasive as the evidence in House and Jackson, which was found to be insufficient. The recantations here are not from disinterested eyewitnesses, and, although victim recantation might in some instances be evidence of innocence, see Gandarela, 286 F.3d at 1086, for the reasons discussed above, the recantation here is not sufficiently reliable that we can conclude that every juror would credit it. Further, as in House, there is other testimonial evidence supporting the verdict, which further persuades us that habeas relief is not warranted in this instance. See House, 547 U.S. at 553–54. We, therefore, cannot say that “in light of the new evidence, no juror, acting reasonably, would have voted to find [Jones] guilty beyond a reasonable doubt.” Schlup, 513 U.S. at 329. Accordingly, we hold that Jones has not made the “‘extraordinarily high’” and “‘truly persuasive’” showing required for habeas relief on a freestanding claim of actual innocence. See Carriger, 132 F.3d at 476 (quoting Herrera, 506 U.S. at 417). 22 JONES V. TAYLOR IV. For the foregoing reasons, the judgment of the district court is REVERSED.