dissenting.
The majority holds that the trial court erred in concluding that child was within the jurisdiction of the juvenile court. I disagree with the majority’s application of our standard of review and its reading of the evidence. Accordingly, I dissent.
As the majority notes, we review juvenile matters, as we do suits in equity, de novo. ORS 419A.200(5); ORS 19.125(3). After stating that standard, however, the majority describes our review as follows:
“We give deference to the trial court’s credibility findings. State ex rel Juv. Dept. v. Greenwood, 107 Or App 678, 680-81, 813 P2d 58 (1991). The degree of deference we give [‘]varies with the importance of the trial court’s opportunity to observe the [demeanor of the] witnesses!’] Brunswick v. Rundell, 126 Or App 582, 586, 869 P2d 886 (1994), and our ability to discern the trial court’s specific credibility determination.” 131 Or App at 319. (Emphasis supplied.)
*324My first concern is with the majority’s focus on “credibility findings.” Although we previously have discussed giving deference to “credibility findings,” see, e.g., State ex rel Juv. Dept. v. Greenwood, supra, or “credibility determinations,” see, e.g., State ex rel Juv. Dept. v. Beyea, 126 Or App 215, 217-18, 867 P2d 565 (1994), those phrases are somewhat imprecise statements of the nature of our review. The essence of de novo review is that we may disregard the factual findings of a jury, or a judge acting as a factfinder, and decide the facts anew upon the record. However, the reality is that when we are faced with conflicting testimony, we run the risk of drawing inferences from a cold transcript that we might not draw if we had the opportunity to see the witnesses ’ demeanor and to hear the tone of their voices during the course of the trial. Consequently, Oregon appellate courts have long recognized that, even though our review is de novo, when there is conflicting testimony on factual issues, we give considerable weight to the trial court’s findings of fact, which reflect the court’s judgment as to the witnesses’ credibility. See Rohrbacher v. Strain, 95 Or 1, 12, 186 P 583 (1920) (concluding that the trial court’s findings “have very strong advisory weight” because the court heard the witnesses, observed their manner and appearance, and had “a much better opportunity to judge of their truthfulness * * * than we can possibly have”); see also Krueger v. Ropp, 282 Or 473, 478, 579 P2d 847 (1978) (“Although we review equity cases de novo, we give substantial weight to the trial court’s findings where * * * those findings depend on the resolution of conflicting testimony and the credibility of the witnesses.”); Ollman v. Active Homes, 278 Or 113, 115, 562 P2d 1217 (1977) (“As we have frequently said, in reviewing evidence de novo considerable weight is to be given to the trial judge’s decision where it rests upon his appraisal of the credibility of the witnesses.”); Adamson v. Adamson, 273 Or 382, 389, 541 P2d 460 (1975) (“[W]here the testimony is conflicting and the evidence is in equipoise, the credibility of witnesses is decisive and we give considerable weight to the findings of the trial court even though we try the case de novo.”); Seitz v. Albina Human Resources Center, 100 Or App 665, 674, 788 P2d 1004 (1990) (same, in context of unlawful employment practice); Hampton v. Sabin, 49 Or App 1041, 1047, 621 P2d 1202 (1980), rev den 290 Or 519 (1981) (same, in context of *325rescission of contract); State v. Watkins, 35 Or App 87, 89, 581 P2d 90 (1978) (civil commitment); State ex rel Juv. Dept. v. Maves, 33 Or App 411, 416, 576 P2d 826 (1978) (termination of parental rights).
Our past comments about giving deference to “credibility findings,” then, are more properly understood as restatements of the above principles. In this case, however, the majority confuses what is already a less than precise statement, by requiring that a trial court make a “specific credibility determination” in order to merit deference. If the majority is suggesting that we defer to credibility determinations only when such determinations are explicit in the record, that proposition makes little sense because, as discussed above, our deference is more properly accorded to factual findings, not to credibility findings. If, on the other hand, the majority is suggesting that a trial court’s factual findings are entitled to deference only when they are based on explicit credibility determinations, that proposition represents a significant change in our review of cases like this, a change for which the majority offers no authority or rationale.1 In my view, although it is helpful to this court in reviewing a trial court’s decision, the court’s failure to outline specifically who it finds to be credible and why does not preclude us from giving deference to the trial court’s firsthand appraisal of the testimony adduced at trial and its consideration of that appraisal when resolving factual disputes that turn on the witnesses’ credibility.
In accordance with our established standard of review in these cases, I believe that we should give substantial weight to the factual findings of the trial court when the credibility of the witnesses is decisive in making those factual findings. That does not mean, however, that the resolution of every issue of disputed fact will depend on the trial court’s judgment regarding which of the witnesses to believe. Of course, when the record contains other persuasive evidence as to a factual issue, witness credibility is not decisive. Further, we may accord less weight to a trial court’s findings when the conflicting evidence comes from expert witnesses, *326whose manner and demeanor during trial maybe less crucial to evaluating credibility than would be an analysis of their reasoning. See McCoy and McCoy, 28 Or App 919, 924, 562 P2d 207, mod 29 Or App 287, 563 P2d 738 (1977). Finally, regardless of how well a witness may comport himself or herself while testifying, we may conclude that other factors, such as the inherent improbability of the testimony, the possible internal inconsistencies in the testimony or extrinsic evidence of the witness’s credibility, are so persuasive as to render the trial court’s findings untenable. In such cases, the weight we accord the trial court is appreciably diminished.
In this case, the majority rejects the trial court’s factual findings, apparently because it concludes that internal inconsistencies in C.’s testimony and extrinsic evidence of her lack of credibility are so overwhelming that those findings are insupportable. I disagree on both counts.
As to the internal inconsistency of C.’s testimony, the majority first identifies “major discrepancies” between what she told law enforcement officers and what she said at the hearing. The majority first notes that C. did not tell anyone about child’s knife. C. testified that during the sexual encounter, she was afraid that child was carrying, and might use, a knife that she had previously seen in his possession. However, she also testified that child never actually mentioned or displayed a knife and that she did not discuss the knife with the police. Given that C. was reporting what did happen to her, not what she feared might have happened, and that her fears about the knife might have been unfounded, I do not see her failure to mention the nonexistent knife to the police as an “inconsistency” in her testimony.
Although I agree with the majority that the statement C. made to the officers about how she ended up on the ground was inconsistent with her testimony at the hearing, the resolution of that factual issue is immaterial to the crime for which child was found to be within the jurisdiction of the juvenile court, sexual abuse in the second degree. Further, I do not believe that that inconsistency is such persuasive evidence of C.’s general lack of credibility that we should disregard the trial court’s factual findings on material issues, such as whether C. had consensual sex with child.
*327The majority also states that it is troubled by the fact that C. told three different people that she had consensual sex with child. At the hearing, C. did not deny that she told those people that she had willingly had sex with child. She testified that, shortly after the encounter, she told Christina, child’s girlfriend, and Melinda, a girl that child had been talking to while C. was on the phone with Christina. She also testified that, several hours later, she told her father. When asked why she told those people that she had wilhngly had sex with child, C. testified that she was in a state of shock and did not know what to make of what had happened to her and that she was embarrassed.
In rejecting the trial court’s finding that C. did not consent to having sex with child, the majority is also rejecting that court’s belief in C. ’s explanation for why she initially told people that the encounter was consensual. Unlike the majority, I do not believe that her explanation was so improbable, nor the above inconsistencies so compelling, that we should take it upon ourselves to deem her testimony noncredible, based solely on our reading of a cold transcrjpt. In my view, this is the very reason why, in appropriate cases and despite our de novo review, we give substantial weight to the trial court’s factual determinations.
In addition to what it views as internal inconsistencies in C.’s testimony, the majority apparently believes that the record also contains extrinsic evidence of C.’s lack of credibility. For instance, the majority states:
“Although C. testified that she was raped after beingpushed, falling hard and being held down virtually naked on the ground in a wooded area, the physical examination the next day revealed no scratches or bruises on her hack, legs or buttocks.” 131 Or App at 320.
The majority, however, cites only part of the evidence relevant to that issue. First, C. testified that, on the day of the incident, she wore leggings underneath her jeans, a turtleneck, a big sweater and a jacket. Contrary to the majority’s implication, she had on several layers of clothing when she hit the ground. Second, although C. testified that she could not move when child pinned her to the ground after removing her clothing, she also testified that she did not try to get up, that *328she was “basically paralyzed” and that she did not do anything physically to resist him.
The majority places little reliance on the testimony of Dr. Pendleton, who conducted the physical examination of C. It concludes that any corroboration of C.’s testimony by Pendleton’s description of her physical trauma was “equivocal.” 131 Or App at 320.1 strongly disagree with that characterization of Pendleton’s testimony. He testified that the exam closely supported C.’s statement of what had happened. On cross-examination, child’s counsel questioned the doctor further:
“Q. Well, given the fact that she reported to you that her clothing was entirely removed, with the exception of her bra, and that she was raped on the ground, it’s inconsistent, is it not, that she had no bruises or abrasions on her buttocks or her back?
“A. I can say that I have seen individuals in similar circumstances and have read about them and have heard about them in similar circumstances who make the same claims, who have the same findings that she does. So I can’t say that it’s inconsistent. No, sir.
“Q. But you would expect to see those types of...
“A. One might expect it. Certainly if you had seen it, it wouldn’t be inconsistent with... with her test... with what she says. But in all honesty, I don’t know that I can say that it’s entirely inconsistent. No.”
Reading the evidence as a whole, I do not believe that the absence of scratches or bruises is persuasive evidence of C.’s lack of credibility on the issue of consent.
I believe that the majority’s characterization of that evidence as equivocal is based on its failure to consider Pendleton’s entire testimony; instead, it relies on selective quotations of his testimony. On direct examination, Pen-dleton described in detail C.’s injuries. In response to the prosecutor’s question as to how he would characterize an encounter that would produce those kinds of injuries, the doctor responded:
“I think that the entry of the pene ... of a penetrating object into the vagina would have to be something that was forceful and persistent. And ... and then maybe I should clarify that a little hit, and .. . and I apologize for adding on. *329It’s . . . it’s possible to have in ... to have penetration of a vagina, even of one that is not sexually experienced, and not cause trauma like this, but it has to be veiy, very slow * * *. This is something that was done with... in my opinion, with force and not slow at all. It caused injury.”
On cross-examination, the following exchange took place:
“Q. Doctor, many of the things that you’ve noted, for instance the bruising [to the hymenal ring] in your diagram and so forth, are all consistent with consensual sex. Correct?
“A. No. Consensual sex — it’s something that can he seen with particular rough consensual or rough athletic sex, but it is no way consistent with routine consensual sex.
“Q. Okay. Consistent to a greater degree with consensual sex for the first time. Correct?
‘ ‘A. No. I don’t think that the... the amount of injury to the perineum is consistent with consensual sex in somebody who has a virginal introi... or virginal vagina. * * * That .. .that typically may involve minor irritation, but this was a significant amount. * * *
it* * * * *
“Q. Okay. So it seems... seems to me that what you’re saying is those injuries could be caused even by — or these types of injuries at least could be caused even by an exam if you’re not very careful about it?
“A. Again, I think that this kind of injury is out of bounds of the common examiner. I think that... that the exam, if this was done by a physician, would have to be extremely, extremely aggressive.
“A. * * * I think that... that in this situation certainly what she was saying and what my observations — and. .. and the nature of my obser... observations led me to believe that this was not consensual, that this was assaultive.
“Q. But give[n] your findings, you certainly can’t say that conclusively, can you?
“A. I think that this is one of those situations where perhaps it gets down to who says what. From my perspective, the findings, the... the history and physical were consistent with assault with force applied to the perineum. I think that the context that consensual sex occurs in does not cause this type or this extent of injury.” (Emphasis supplied.)
*330Pendleton’s testimony was not “equivocal.” His statement that trauma could be avoided if penetration was “very, very slow” does not detract from his opinion that C.’s injuries were “in no way consistent with routine consensual sex” and that “this was assaultive.” (Emphasis supplied.) In my view, reading Pendleton’s testimony as a whole, the physical evidence was strongly corroborative of C.’s testimony that she did not have consensual sex with child.
In summary, I believe that this is a classic case of “his word against hers,” in which the trial court is in the best position to resolve factual issues which turn on the witnesses’ credibility. There is more than adequate evidence in the record to support the court’s factual findings and, absent persuasive reasons to do so, we should not disregard those findings in the light of the trial court’s unique opportunity to observe the witnesses throughout the course of the hearing. Thus, I believe that the trial court did not err in finding that child committed acts which, if committed by an adult, would constitute sexual abuse in the second degree.
I respectfully dissent.
Interestingly, the majority never addresses whether it has discerned this trial court’s “specific credibility determination’ ’ and, if not, what impact the lack of such a determination has on our review.