dissenting:
Although I agree with the majority that the videotaped interview between Proctor and Heather was not admissible pursuant to the medical diagnosis exception, I dissent from the majority’s conclusion that Heather’s videotaped statements were not admissible pursuant to the residual exception because they lacked particularized guarantees of trustworthiness.
The Supreme Court has held that particularized guarantees of trustworthiness are to be determined from “the totality of circum*1394stances that surround the making of the statement and that render the declarant particularly worthy of belief.” Idaho v. Wright, 497 U.S. 805, 820, 110 S.Ct. 3139, 3149, 111 L.Ed.2d 638 (1990). The Court concluded there is no exclusive set of factors to be evaluated to determine the trustworthiness of a statement. Consequently, it advised that courts are to be given “considerable leeway in their consideration of appropriate factors.” Id. at 822, 110 S.Ct. at 3150.
In this case, the majority reviews the facts surrounding Heather’s statements and holds that the only possible conclusion to be drawn from Heather’s use of age-appropriate language and the spontaneity of her statements is that Heather’s statements were unreliable. I disagree and believe that the majority’s method of adopting a singular inference from factors which support multiple inferences is contrary to Wright and deprives courts of their broad discretion to consider the relevant factors in each particular case.
I disagree that Heather’s failure to use language “unexpected of a child of similar age” supports the conclusion that her statements were unreliable. The Court in Wright did note that the declarant’s use of terminology unexpected of a child of similar age is a factor indicating reliability. Id. at 821, 110 S.Ct. at 3149-50. However, relying on the converse, the majority states the fact that Heather’s language was not unexpected of a child of similar age as a basis for its conclusion that Heather’s statements lacked guarantees of trustworthiness.
It is true the anatomical terms used by Heather were entirely consistent with a child of similar age of two and one-half years. She referred to both male and female genitals as “things,” to the posterior area as “butt,” to female breasts as “boo-boos,” and specifically to Webb’s penis as “pee-pee.” In Swan v. Peterson, 6 F.3d 1373 (9th Cir.1993), cert. denied, — U.S. -, 115 S.Ct. 479, 130 L.Ed.2d 393 (1994), we approved of the trial court’s reliance on the child declarant’s description of “sexual acts in specific terms, using age-appropriate language” as a factor supporting reliability. Id. at 1380 (emphasis added). We noted consideration of this factor was consistent with the Supreme Court’s statement in Wright that courts have considerable leeway in their consideration of appropriate factors. Id. However, contrary to Swan, the majority establishes a bright-line rule that only language inappropriate for a child of similar age indicates reliability.
I do not agree that Heather’s use of age-appropriate terminology supports the conclusion that her statements were unreliable. A child molester cannot be expected to offer a course in correct anatomical terminology while molesting a child, nor could a child as young as Heather be expected to remember such terminology even if it were used. The crucial inquiry should not be whether the language is appropriate or inappropriate for a child of similar age, but whether the language employed by the child adequately describes a sexual encounter. Heather did describe explicit sexual encounters with specificity in age-appropriate language. An example of one such statement is Heather’s claim that Webb pee-peed in her mouth. Surely, this language is explicit enough to ensure the truthfulness of Heather’s claim that Webb had sexually abused her, regardless of the fact that Heather did not use sophisticated language.
I also disagree with the majority’s conclusion that no incriminating statement can be considered spontaneous if the child has been previously questioned about the abuse. The Supreme Court in Wright cites with approval a ease relying on spontaneity and consistent repetition as factors indicating reliability. See Wright, 497 U.S. at 821, 110 S.Ct. at 3149-50. However, according to the majority, the spontaneity of Heather’s statements to Proctor is a factor indicating unreliability. The majority reasons that because Heather had been questioned by her care-givers and investigators regarding her accusations prior to the videotaped interview with Proctor, Heather’s spontaneous accusation during the interview “must be stamped as insufficiently reliable and as the product of previous questioning.” Slip op. 1392. The majority’s reasoning stands for the proposition that if there has been any prior investigation, spontaneity can never be a factor supporting reliability.
Adherence to such a proposition would immunize sexual abusers from all but one accu*1395satory statement of those children they abuse, i.e., the first spontaneous statement of the abused child. After a child’s first spontaneous statement alleging abuse, an investigation of the child’s claim will necessarily ensue, and the focus of the investigation will be on the child. Pursuant to the majority’s reasoning, however, investigators may not question the child, otherwise, any future statements by the child are unreliable. Conversely, if the investigators do not question the child, they cannot determine if there is any substance to the child’s complaint. Thus, prosecution of a sexual abuser of a very young child such as Heather can never occur, either because the child cannot be questioned, or because the child has been questioned. The fact that the child is willing to talk without prompting at subsequent discussions regarding the abuse, as Heather did during her interview with Proctor, cannot be a factor indicating unreliability. If it were, every investigation would inevitably lead to the conclusion that the ease cannot be successfully prosecuted.
Finally, the majority is critical of Proctor’s description of Webb as bad. However, the majority fails to place Proctor’s comments in context. Proctor’s first comment about Webb, i.6., that Webb should be told not to pee-pee in Heather’s mouth, occurred twenty minutes into the interview after Heather had already said that 1) Webb pulled down his pants; 2) he put medicine1 in “there” (pointing to her vaginal area); 3) his penis “got little” afterwards; and 4) he pee-peed in her mouth. While Proctor’s comment was inappropriate, it should not distract us from the proper inquiry: Were Heather’s statements prior to Proctor’s first negative comment sufficiently shown to be reliable? I conclude that they were. Heather’s statements were in age-appropriate language; they were internally consistent; and they were spontaneous or in response to appropriate, non-leading questions. The majority’s statement that the Proctor’s repetitious questions would not have been permitted by a court is incorrect. Trial courts have great latitude in handling young or vulnerable witnesses. See United States v. Castro-Romero, 964 F.2d 942, 944 (9th Cir.1992) (holding district court did not abuse its discretion by allowing Government to ask leading questions of eight-year-old witness reluctant to testify regarding sexual abuse).
Under the totality of the circumstances surrounding the interview, up until Proctor began to denigrate Webb, I believe Heather’s videotaped statements met the test of Wright. Thus, the only remaining inquiry is whether Proctor’s comments rendered the admission of the tape an error of constitutional dimension. Proctor made three negative comments about Webb. After the first negative comment, Heather said that Webb did it more than one time; Cindy watched but did not say anything; Webb’s “thing” touched her vaginal area; and Webb took pictures while he was doing a bad thing. Heather later retracted the statement about the pictures.
Proctor’s gratuitous negative comments about Webb were clearly improper. They had the capacity to color Heather’s following statements. However, I do not believe the admission of this later portion of the videotape rose to the level of a constitutional error. The majority of Heather’s incriminating statements were elicited before Proctor began to counsel her. At most, admission of the last minutes of the videotaped interview was harmless error. As the Supreme Court in Wright noted, “the presence of corroborating evidence ... indicates that any error in admitting [a] statement might be harmless.” 497 U.S. at 823, 110 S.Ct. at 3150-51 (footnote omitted). In this ease, corroborating physical evidence existed to support Heather’s claim that she had been sexually abused: Heather’s examining physician testified that Heather’s physical condition was consistent with sexual abuse.
*1396After killing the tape and burying it, the majority digs it up and buries it again by discussing all the extrinsic evidence demonstrating its unreliability. I dissent from the several pages of dicta offered by the majority during this second burial. In this dicta, the majority concludes that extrinsic evidence bolsters its conclusion that Heather’s statements are unreliable reasoning that “Wright does not forbid recourse to other evidence that confirms the presumptive unreliability of the hearsay.” Slip op. 1392. In Wright, the Supreme Court held that extrinsic evidence could not be used to bolster the reliability determination. See 497 U.S. at 823, 110 S.Ct. at 3150-51. It did not hold, however, that extrinsic evidence could be used to bolster the presumptive unreliability of the hearsay. Having already held that the presumption of unreliability of Heather’s statements had not been overcome, there is simply no reason for the majority to decide the flip side of Wright in this case. Further, in doing so, the majority ignores our previous statement in Swan regarding the use of extrinsic evidence to establish unreliability:
Wright forbids using other corroborating evidence at trial to show that an initial hearsay statement is reliable. The issue here is subtly different: can arguable non-corroborating evidence be used to demonstrate the unreliability of a hearsay statement? Wright’s caution against reference to other evidence at trial suggests not.
Swan, 6 F.3d at 1381 (footnote omitted).
In sum, I cannot agree with the majority’s method of adopting singular inferences from factors to support its conclusion, thereby ignoring many other reasonable inferences to be drawn from those factors. Furthermore, I dissent from the extensive dicta offered by the majority. In the future, the question of whether noncorroborating evidence may be used to demonstrate unreliability should be addressed in a case where the answer is needed, unhampered by the decision in this case.
. The majority contends Heather’s statements were inconsistent because she used the term "medicine” when referring to hair gel and also when referring to what “could have been understood to mean ... semen.” Maj. op. 1392. The majority, without basis, presumes Heather was referring to semen when she indicated Webb put medicine in her vaginal area. However, what it fails to recognize is that perhaps what Heather meant by medicine was a clear liquid gel — such as a lubricant.