Schutz v. State

957 S.W.2d 52 (1997) Ben Wayne SCHUTZ, Appellant, v. The STATE of Texas, Appellee. No. 726-95. Court of Criminal Appeals of Texas, En Banc. December 3, 1997. *56 Frank Blazek, Huntsville, for appellant. David P. Weeks, District Attorney, Kay Douglas, Assistant District Attorney, Huntsville, Matthew Paul, State's Attorney, Austin, for State. Before the court en banc. OPINION ON APPELLANT'S PETITION FOR DISCRETIONARY REVIEW KELLER, Judge, delivered the opinion of the Court, in which McCORMICK, Presiding Judge, MANSFIELD, PRICE, HOLLAND, and WOMACK, Judges, joined. A jury convicted appellant of the aggravated sexual assault of his six-year-old daughter, A.S., and sentenced him to thirty years confinement. In his fourth and fifth points of error on appeal, appellant complained that the trial court erroneously admitted expert testimony concerning the credibility of the child complainant. The Court of Appeals affirmed the conviction, Schutz v. State, No. 01-94-00257-CR, 1995 WL 325769 (Tex. App—Houston, June 1, 1995)(unpublished) and appellant petitioned for discretionary review. We will reverse. I. BACKGROUND A. The Testimony During the State's case-in-chief, Patricia Burns, a social worker,[1] and Dr. David Poole, a psychologist, testified about interviews they conducted with the complainant. When asked about the characteristics of a child who is being manipulated, Burns testified as follows: There are several things that I try to observe, and that is I first of all pay attention to the event that a child is describing, the way that that child describes it. If it's done in an appropriate—age appropriate manner.... I feel like if the child is being manipulated and coerced into something that often times the language would be more sophisticated than a child that age would know....I look for emotions and behaviors that are consistent with what *57 she is describing or with what a child is describing. If a child is describing a situational sexual abuse then she many times will have a very distressed, painful expression of emotions, emotions where she would be exhibiting emotions that she does not realize that she is supposed to be exhibiting and describing in describing such events. So there is some parallel between the emotions she has and the event they are describing or she is describing....I look for the description. I look at the child and her age. I examine her language and her knowledge of certain things. In the case of sexual abuse, if the child has an advanced knowledge in this area then that is indicative of something that perhaps did definitively occur....I feel that if the child is being manipulated in a situation such as sexual abuse the behaviors that I would predict would be more of enthusiasm on the part of the child in telling the story where she would want to use the events and the happenings as a tool to get attention....Or, as I said before, if I saw a child with her emotions did not match up to the events that she was describing I would examine that, also. (Ellipses inserted). When asked about the possibility of fantasizing, Burns responded: If I feel like that there is a situation where the child is making up a story or fantasizing the story I will continue questioning the child's difficulties. I think my feeling is if a child is lying about something like this then that is what I need to get a handle on because that's just as big a problem as, you know, telling the truth about this. (Emphasis added). Finally, the State elicited Burns' opinion about whether A.S. exhibited the characteristics of manipulation or fantasy: Q. Do you have an opinion as to whether [A.S.] exhibited any of those traits or characteristics of manipulation that you have described to the jury? .... A. Yes I have. Q. What is that opinion? ....[intervening objections by defense counsel] A. My opinion is that [A.S.] has not exhibited—in my professional opinion she has not exhibited behaviors that point to having been manipulated. Q. And based upon your experience, training and expertise, your close relationship and therapy of [A.S.], do you have an opinion as to whether she has exhibited any of the traits of fantasizing? .... A. Yes. Q. And what is your opinion? ....[intervening objections by defense counsel] A. My opinion is she has not exhibited any evidence of fantasizing. (Emphasis added). Likewise, Dr. Poole testified about the possibility of manipulation: The point of the material that the kid generates in the testing has to do with what is on their mind. Usually if a kid is prepared to dispose of a relationship with one parent, trash it, be a party to getting them in trouble, they are usually coming under some significant pressure from some other source. The kinds of things that you will find the kid worrying about is the source of the pressure. If they are getting threats, if they are getting told they might get abandoned, nobody loved anymore, they will never see them again or pressure like that you will see the kid worried about losing a parent who would be putting the pressure on them. It's depending upon where their anguish and anxiety is coming from. Concerning the issue of fantasizing, Poole testified as follows: Well, one of the aspects you look at in the testing is seeing if the reactions that the child has, the interpretation, shapes of the ink blots, the identification of the content of the picture has some relevance to what is really there. If they are talking about material that is actually depicted, if their perception of things is common and it's recognizable to other people, that is one of the ways you judge whether the kid is in touch with reality and whether their perception *58 of what is going on has any similarity to what is really out there. It's one of the things we evaluate. You usually find kids distorting those kind of interpretations in pretty dramatic ways if they are really out of touch with reality. And it's usually evident clinically in a lot of other different ways, too, if a child is so convinced about the reality of the situation that they're distorting what is really going on. After this testimony and defense objections, the State elicited Poole's opinion concerning whether the complainant's statements were the product of manipulation or fantasy: Q. Dr. Poole, based upon your experience, your training, your education in your profession, my question to you is do you have an opinion based upon your examination of [A.S.] whether she has been the subject of manipulation. A. Yes. Q. And what is that opinion? [renewed objections] A. The evidence I have available to me made that the less likely explanation. Q. Do you have an opinion based upon your experience, training and expertise and examination of [A.S.] whether her allegations are the subject of fantasy? A. My opinion is that they were not the result of fantasy. Defense counsel: Object to unresponsive. Court: Sustained. Q. Do you have an opinion? A. Yes. Defense counsel: Same objection.[2] A. That they were not to [sic] result of fantasy. (Emphasis and bracketed material added). We note two important differences between Burn's and Poole's testimony. First, Burns testified as to the characteristics of children who fantasize or are manipulated while Poole gave opinions as to whether the complainant had in fact been manipulated and whether her allegations were the result of fantasy. Second, Burn's testimony essentially equated "fantasizing" with "lying" while Poole's testimony about "fantasy" involved the ability of a child to perceive reality accurately. The significance of these differences will be explored later in this opinion. B. The Court of Appeals' Opinion and the State's Arguments To the Court of Appeals, appellant argued that the expert testimony regarding "manipulation" and "fantasy" constituted impermissible comments on the child complainant's truthfulness or credibility. The Court of Appeals advanced two reasons for rejecting appellant's position: (1) that the experts' opinions were not direct comments on the child's truthfulness, and (2) that appellant opened the door to the testimony by impeaching the child's credibility. These reasons are likewise advanced by the State in opposition to appellant's petition for discretionary review. With regard to its first reason, the Court of Appeals held that the experts' opinions, taken in context, merely showed that the child did not exhibit any of the traits commonly associated with manipulation or fantasy, and hence, were not direct comments upon the child's truthfulness. The State argues additionally that these opinions were not an expression of the experts' belief in the child's testimony and that their testimony did not foreclose the possibility that the child could have made up the allegations. In support of its holding that appellant opened the door, the Court of Appeals noted several instances in which appellant attempted to impeach the complainant's credibility through cross-examination. Appellant cross-examined the complainant about whether her grandmother told her what to say, about babysitters, and about whether she was afraid of an 11-year-old male cousin. Further, appellant cross-examined the complainant's mother about whether the child had ever fabricated or fantasized, whether the child was a truthful person, whether the child was unwilling to talk about abuse to her *59 mother "because it might not be true," and whether the child had ever told her mother that babysitters had hurt her. Both the child and the mother categorically answered these questions in the negative. The court also found that the defense "implied, through cross-examination of the child's pediatrician, that the mother and the grandmother had caused the child's physical injuries when they performed a `clumsy and unprofessional' physical examination of the child." Finally, the court stated that: "Throughout the testimony of all the witnesses was the suggestion that the child's allegations were merely the product of her parents' bitter divorce." In support of its argument that appellant opened the door, the State relies upon the facts detailed by the Court of Appeals and sets out the following colloquy from the defense counsel's cross-examination of the complainant's mother: Q. Has [A.S.] always been a truthful person? A. Very, very truthful. Q. Has she ever fantasized or fabricated? A. Nothing. Q. Have you ever talked to her about her fantasies or ideas? A. Nothing. In fact, she will not talk to me about this at all. The outcries that she had made was very, very—it was like I said, just—it was very—it wasn't very in detail at all. And the outcry she had made to her counselor was— Q. And you know— A. She would not talk to me about it today. Q. She won't talk to you? A. About it at all, what happened or anything. Q. Is it possible she won't talk to you because it might not be true? A. No it's not possible.... (Bracketed material substituted for original). We now turn to the merits of appellant's grounds for review. II. THE RELEVANT AUTHORITIES A. Direct Comment on Truthfulness? To be admissible, expert testimony must "assist" the trier of fact. Tex.R.Crim. Ev. 702.[3]Duckett v. State, 797 S.W.2d 906, 914 (Tex.Crim.App.1990). Expert testimony assists the trier of fact when the jury is not qualified to "the best possible degree" to determine intelligently the particular issue without the help of the testimony. Duckett, 797 S.W.2d at 914. But, the expert testimony must aid—not supplant—the jury's decision. Id. Expert testimony does not assist the jury if it constitutes "a direct opinion on the truthfulness" of a child complainant's allegations. Yount v. State, 872 S.W.2d 706, 708 (Tex.Crim.App.1993). The question we must resolve here is whether testimony concerning "manipulation" or "fantasy" is equivalent to testimony about "truthfulness," or if different, whether such testimony nevertheless merits similar treatment under the rules of evidence. An important aspect of this question is whether the concept of "truthfulness" encompasses nonmoral as well as moral considerations. For instance, a child who "fantasizes" could be doing one of two things: (1) the child could be making up events that he or she knows never occurred—i.e. the child could be lying (moral considerations), or (2) the child could be relating events that he or she truly believes occurred but did not in fact occur—i.e. the child could be suffering delusions, perhaps as a result of a mental disorder (nonmoral considerations). In the present case, Burns' testimony relates to the first category while Poole's testimony relates to the second. Clearly, where "fantasizing" is used as a synonym for "lying" the two should be treated the same. The more difficult question is how to treat "fantasizing" that involves nonmoral considerations, such as delusional behavior. The same moral/nonmoral question can be raised with respect to manipulation: Does the child knowingly make false allegations *60 due to manipulative pressure from others or has the child been manipulated into believing that certain false allegations are actually true? However, manipulation also involves the additional element of actions taken by persons other than the child complainant. We must determine whether the additional element of behavior external to the child has an effect on whether to treat "manipulation" the same or differently from lying or fantasy. With these considerations in mind, we turn to the question at hand. At least one court has noted that there is a "fine but essential" line between helpful expert testimony and impermissible comments on credibility. State v. Myers, 382 N.W.2d 91, 98 (Iowa 1986). Because we have never addressed the manipulation/fantasy question under the Rules,[4] we look to other jurisdictions for guidance in drawing this "fine line." Such guidance is especially appropriate here because the rules of evidence in Texas and most other states are patterned after the Federal Rules of Evidence.[5] Most jurisdictions specifically addressing expert testimony concerning manipulation and/or fantasy have held the particular testimony involved to be an impermissible comment on a witness' credibility. A few jurisdictions have held such evidence to be different from evidence regarding truthfulness. Other jurisdictions distinguish between general behavioral/competency evidence and evidence relating to specific allegations. Although there is some overlap between the concepts of manipulation and fantasy, there are also differences, and we will endeavor to discuss the concepts separately to the extent that this is practicable. 1. Manipulation While we have not found the term "manipulation" in the published opinion of any court other than the present case, various aspects of manipulation—coercion, coaching, and suggestion—have been addressed by many decisions. Several courts have held such evidence inadmissible, whether offered by the defendant or the State. In Commonwealth v. Trowbridge, 419 Mass. 750, 647 N.E.2d *61 413, 419 (1995), the Massachusetts Supreme Court upheld the exclusion of a defense psychologist's testimony that "the mother's animosity could have influenced the child to make false accusations of sexual abuse." The court held that the opinion "would have impermissibly intruded on the function of the jury to assess the credibility of the child witness." Id. See also Commonwealth v. Ianello, 401 Mass. 197, 515 N.E.2d 1181, 1185 (1987)(Whether a young sexual abuse complainant has been rehearsed by her mother or others is a matter of truthfulness for the jury to decide). In State v. Erickson, 454 N.W.2d 624, 627-628 (Minn.App.1990), the Minnesota Court of Appeals upheld the exclusion of defense expert testimony relating to "learned memory." The defendant's psychologist would have testified that a child with "learned," as opposed to actual, memory uses more new words, ambiguous language, and less detail. Id. at 628. Further, the psychologist would have testified that these characteristics were exhibited in this particular case by the child sexual abuse complainant and that, in his opinion, the child "related the sexual abuse incidents from learned instead of actual memory." Id. The defense intended to use this testimony to bolster its theory that the child had been coerced and suggestively interviewed. Id. The court held that the jury was capable of evaluating the credibility of the child complainant in light of this defense theory without the help of expert testimony. Id. The court concluded that the testimony was properly excluded because of the potential for undue prejudice: Although testimony may not be properly excluded based solely on the fact that it tends to show that a witness is not telling the truth, the potential for unduly influencing the jury's credibility assessment in this case is too great for such arguably useful information. Id. (emphasis in original). In State v. Michaels, 136 N.J. 299, 642 A.2d 1372 (1994), the New Jersey Supreme Court ruled that the trial court must hold a pretrial hearing to determine whether prosecutorial use of coercive or highly suggestive interrogation techniques had rendered an alleged child sex abuse victim's statements unreliable. The court cautioned that, if the statements were found to retain sufficient reliability to warrant admission into evidence, the defense may offer expert testimony to explain the coercive or suggestive propensities of interviewing techniques employed but may not comment on the credibility of the child witness. Id. 642 A.2d at 1384. In State v. Baymon, 108 N.C.App. 476, 424 S.E.2d 141, 145 (1993), affirmed on different grounds, 336 N.C. 748, 446 S.E.2d 1 (1994), the prosecution's expert witness stated that she "had not picked up on anything" to suggest that someone had told the child complainant what to say, and that she had no concerns that the complainant had been "coached." The North Carolina Court of Appeals held that these statements "bear directly" upon the complainant's "credibility" and were thus inadmissible. Id. However, the North Carolina Supreme Court declined to address the issue, holding that the defendant had opened the door. See this opinion post. As in the present case, Ohio and Oregon have faced expert testimony regarding both manipulation and fantasy. In State v. Boston, 46 Ohio St.3d 108, 545 N.E.2d 1220, 1240 (1989), the prosecution's expert witness opined that the child complainant had not fantasized her abuse and that she had not been programmed to make accusations against her father. The court held such testimony to be an inadmissible comment on the truthfulness of the complainant's statements. Id. Relying upon Boston, the Ohio court later upheld the exclusion of proffered defense expert testimony concerning a child eyewitness' susceptibility to suggestion or influence on the ground that such testimony was an opinion as to the truth or accuracy of the child's statements. State v. Moreland, 50 Ohio St.3d 58, 552 N.E.2d 894 (1990)(murder prosecution). In State v. Keller, 315 Or. 273, 844 P.2d 195, 201 (1993), a medical doctor testified in the State's case-in-chief that "there was no evidence of leading or coaching or fantasizing" during the five-year-old child sexual abuse complainant's interview. The Oregon Supreme Court held the statements to be inadmissible, finding that each *62 "amount[ed] to testimony that the child was credible." Id. (bracketed material substituted for original). Some jurisdictions have held manipulation testimony to be admissible. In United States v. Cacy, 43 M.J. 214, 217 (CAAF 1995), the Court of Appeals for the Armed Forces acknowledged that an expert witness may not testify "regarding the credibility or believability" of the complaining witness. The court nevertheless held that an expert witness may testify as to his "ability to discern whether a victim appears rehearsed or coached, or is feigning." Id. at 218. See also United States v. Scheffer, 44 M.J. 442, 446 (CAAF 1996). In Doe By and Through G.S. v. Johnson, 52 F.3d 1448, 1463 (7th Cir.1995), the Seventh Circuit held that an expert witness did not invade the province of the jury where his testimony "merely dealt with his opinion that [the complainant's] allegations were probably fabricated and were the product of parental and therapeutic suggestion." In State v. Koepsell, 508 N.W.2d 591, 593-594 (S.D.1993), the Supreme Court of South Dakota upheld the admission of expert testimony regarding the characteristics of the child sexual abuse complainants in that case. Some of the testimony involved "signs of programming or coaching" or whether the children appeared to have been coached. Id. While noting this aspect of the testimony, the court did not specifically comment upon its validity. Instead, the court addressed the admissibility of other aspects of the witnesses' testimony. Id. Although Koepsell may arguably support the admission of testimony regarding manipulation, it is not clear authority for that proposition. One intermediate appellate court has specifically distinguished manipulation from the issue of truth-telling. People v. Jones, 851 P.2d 247 (Colo.App. Div. II 1993). The Colorado court held that testimony that a complaining witness was not "suggestible" did not constitute an improper comment on credibility because the testimony merely showed that, "in the witness' opinion, [the victim's] story, whether true or false, was likely to have emanated from within the victim, not from someone else." Id. at 250 (emphasis in original; bracketed material substituted for original). We note, however, that the expert in Jones did not testify that the witness' actual allegations were the result of suggestion. Other jurisdictions have permitted testimony regarding manipulation but have placed limits upon the form that such testimony may take. The Supreme Court of Kansas upheld the admission of a statement by a social worker that "it would be highly unlikely for that [level of detail, vocabulary, sentence construction, and consistency over time] to happen in a child who's coached." State v. Wilson, 247 Kan. 87, 795 P.2d 336, 343-344 (1990)(bracketed material inserted). The court stated that this testimony did not impermissibly comment on the complainant's credibility where the social worker did not testify that the complainant "was or was not coached." Id. The Minnesota Court of Appeals upheld a trial court's decision to allow a prosecution expert witness to testify regarding a child sexual abuse victim's pathologic or chronic lying because the testimony did not address the veracity of the particular allegations before the court. Potter v. State, 410 N.W.2d 364, 369 (Minn.App.1987). The court also upheld the exclusion of defense expert testimony on whether the child had been coerced or pressured into recanting the allegations before the court because such testimony would have been an improper comment on the truthfulness of the those allegations. Id. Vermont's high court addressed expert testimony discussing recent research on whether young children can remember distant events accurately and whether their memories can be altered by leading or suggestive questions. State v. Gomes, 162 Vt. 319, 648 A.2d 396 (1994). The court noted that the testimony related to children in general and that the doctor had not examined the child complainants and could not comment directly on their truthfulness. Id. In finding that the testimony was not "plain" (fundamental) error, the court stated that the experts did not testify that the children were telling the truth, that the children were resistant to suggestibility, or that their memories of abuse were accurate. Id. *63 The general characteristics/specific allegations dichotomy advanced by these cases appears consistent with the holdings in all of the "manipulation" cases cited, except for Cacy (Scheffer), Johnson, and possibly Koepsell. However, Cacy and Johnson appear to be dubious authority for distinguishing manipulation from lying. Despite Cacy's suggestion to the contrary, "feigning" appears to be a mere synonym for lying, and Johnson expressly approved testimony concerning fabrication. To the extent these cases support the admission of expert testimony regarding the phenomena of lying, they are clearly inconsistent with Texas precedent. See Yount, 872 S.W.2d at 709-712. Similar to these cases, at least one commentator who favors the admission of expert testimony concerning the truthfulness of a child complainant's allegations has treated lying, manipulation, and fantasy as aspects of the same credibility question: For example, in many child sexual abuse cases the following alternative hypotheses might all be considered: (1) The allegations are basically true. (2) The allegations are basically true but the child has substituted a different person for the perpetrator. (3) The fundamental allegation is true, but the child has made additional allegations that are false. (4) The child has been influenced or pressured to make a completely false allegation to serve the needs of someone else. (5) The child has made a false allegation for personal motives. (6) The child has fantasized the allegations, possibly because of psychological problems. Honts, Charles Robert, Assessing Children's Credibility: Scientific and Legal Issues in 1994, 70 N.D. L. REV. 879, 889 (1994)(numbering added). 2. Fantasy As in Boston and Keller, most jurisdictions addressing the issue of "fantasy" have refused to distinguish it from the issue of truthfulness. In State v. Brotherton, 384 N.W.2d 375, 378-379 (Iowa 1986), the Iowa Supreme Court held inadmissible an expert's opinion that a three-to-four-year-old child does not have the ability to fantasize a sexual experience. The court held that such testimony was an opinion concerning the truthfulness of the complainant. Id. See also Goodson v. State, 566 So.2d 1142, 1153 (Miss.1990)(citing Brotherton). The Supreme Court of Maine has upheld the exclusion of defense testimony by an expert in child psychology to the effect that "because the children's allegations involved unusual elements not found in a typical child abuse case, such allegations could be projections of childhood fantasies." State v. Mazerolle, 614 A.2d 68, 71 (Me.1992). The court stated that the jury was capable of drawing its own conclusions regarding the believability of the children's allegations. Id. Likewise, the Missouri Supreme Court found that a "psychiatrist's specific statement that the victim did not fantasize the rape was an express opinion about her credibility" and hence, inadmissible. State v. Taylor, 663 S.W.2d 235, 241 (Mo.1984)(adult victim), relied upon by, State v. Williams, 858 S.W.2d 796, 798-801 (Mo. App. E.D.1993)(child victim), Williams cited favorably by, State v. Silvey, 894 S.W.2d 662, 671 (Mo.1995)(child victim). In Commonwealth v. Davis, 518 Pa. 77, 541 A.2d 315, 316-317 (1988), the Pennsylvania Supreme Court did not distinguish between fantasizing and fabrication in holding a statement containing both terms to be an impermissible comment on truth-telling. As in the case of manipulation, some authorities expressly permit testimony about a general disposition to fantasize but exclude testimony about specific allegations. In People v. Stricklin, 162 Mich.App. 623, 413 N.W.2d 457, 464-465 (1987), the Michigan Court of Appeals held that a child psychologist could testify that a child who has been sexually molested before would be more likely to fantasize instances of sexual abuse "[a]ssuming that the psychologist's testimony would have been based on his knowledge of child development and not on an evaluation of the female child's credibility" (bracketed material substituted for original). The Minnesota Supreme Court has stated that it may be permissible for an expert witness to *64 express an opinion concerning a child's ability to distinguish reality from fantasy, but the expert could not express an opinion as to whether the complainant fantasized or fabricated the facts in issue. State v. Saldana, 324 N.W.2d 227, 231-232 (Minn.1982). Only one court has ever expressly distinguished "fantasizing" from "lying." In State v. Heath, 316 N.C. 337, 341 S.E.2d 565 (1986), the following colloquy was in issue: Q. Mrs. Broadwell [psychologist], do you have an opinion satisfactory to yourself as to whether or not [the complainant] was suffering from any type of mental condition in early June of 1983, or a mental condition which could or might have caused her to make up a story about the sexual assault? [Objection made and overruled]. Q. What is your opinion? A. There is nothing in the record or current behavior that indicates that she has a record of lying. Id. 341 S.E.2d at 567 (bracketed material inserted or substituted for original). The Supreme Court of North Carolina found improper both the question, addressing itself to "the sexual assault," and the answer, concerning a "record of lying." Id. at 568 (emphasis in original). The court criticized the testimony as referring to the truthfulness of the complainant's specific allegations. Id. The court also distinguished "fantasy" from "lying" as involving nonmoral versus moral considerations: It is one thing to ask an expert in psychology or psychiatry whether a victim fantasizes, but it is another thing altogether to ask whether a witness has made up a story or lied. One who fantasizes can honestly and subjectively believe in the reality of the fantasized-about occurrence, but "making up a story," or lying, denotes an affirmative or conscious intent to deceive, invent, or not tell the truth." Id. (emphasis in original). Thus far, intermediate North Carolina appellate courts addressing "fantasy" issues have not relied upon Heath's distinction between fantasy and lying. Instead, these courts have held that Heath prohibits testimony about specific allegations but not about a person's ability, in general, to distinguish truth from fantasy. State v. Jones, 99 N.C.App. 412, 393 S.E.2d 585, 588-589 (1990), dismissed, 328 N.C. 95, 402 S.E.2d 423 (1991); State v. Teeter, 85 N.C.App. 624, 355 S.E.2d 804, 807 (1987). 3. Capacity In essence, Heath characterizes fantasy as a comment on mental (nonmoral) capacity for truth-telling. But, as the facts of the present case illustrate, "fantasy" only sometimes relates to capacity. Even when it appears to do so, the question still remains: should we treat questions of mental capacity differently from questions concerning moral disposition? The drafters of the federal rules recognized the difficulty of distinguishing between mental capacity and character and opined that the rules of evidence covered both: Traditionally character has been regarded primarily in moral overtones of good and bad: chaste, peaceable, truthful, honest. Nevertheless, on occasion nonmoral considerations crop up, as in the case of the incompetent driver, and this is bound to happen increasingly. If character is defined as the kind of person one is, then account must be taken at varying ways of arriving at the estimate. These may range from the opinion of the employer who has found the man honest to the opinion of a psychiatrist based upon examination and testing. No effective dividing line exists between character and mental capacity, and the latter has traditionally been provable by opinion. Notes of Advisory Committee, Fed. R. Ev. 405 (emphasis added). In United States v. Shay, 57 F.3d 126, 131 (1st Cir.1995), the First Circuit related this commentary to Federal Rule 608. In accordance with this observation, the court held that Rule 608 governed expert testimony on whether a defendant suffered a mental disorder that would cause him to make false statements against his self-interest. Id. at 132. See also United States v. Butt, 955 F.2d 77, 82-84 & 84 n. 8 (1st Cir.1992)(severe mental disorder relevant to truthfulness under Rule 608); United States v. Hall, 93 F.3d 1337, *65 1344 (7th Cir.1996)(discussing Shay); United States v. Smith, 77 F.3d 511, 516 (D.C.Cir.1996)(agreeing with Butt that evidence of mental illness is relevant only when it may reasonably cast doubt on the ability or willingness of a witness to tell the truth). But one federal circuit court has opined that Rule 608 reflects only moral inducements for truth-telling, not mental capacity for truth-telling. United States v. Lindstrom, 698 F.2d 1154, 1162 n. 6 (11th Cir.1983). Most courts addressing the issue of capacity have not favored the admission of such evidence. Some courts have held that a witness' capacity to tell the truth is an issue only when determining a witness' competence to testify, to be determined by a judge outside the presence of the jury. According to these courts, absent a recognized mental disorder or deviant mental condition, an expert witness may not testify to the jury about another witness' mental capacity for truth-telling. State v. Oliver, 188 Ga.App. 47, 372 S.E.2d 256, 260 & 260 n. 2 (1988, cert.denied)(Expert may testify about ability of children of a certain age to distinguish truth from falsity but not about a particular child's ability to do so; possible exception for mental illness), cited by, Roberson v. State, 214 Ga.App. 208, 447 S.E.2d 640, 643 (1994)(Clinical social worker may not testify that child victims "were capable of distinguishing truth from fantasy"); State v. Lewisohn, 379 A.2d 1192, 1203 (Me.1977)(expert testimony about the ability of children age 10-14 to accurately observe and remember events inadmissible absent "medical" condition); State v. Friedrich, 135 Wis.2d 1, 398 N.W.2d 763, 770 (1987)("Where the basis for the challenge of the witness' credibility is not a physical or psychological disorder, however, the need for expert testimony is not recognized"); Gale v. State, 792 P.2d 570, 576 (Wyo.1990)(Expert witness may not comment on the effect an emotional disturbance might have on a complaining witness' ability to tell the truth; denying requests for psychological examinations); United States v. Wertis, 505 F.2d 683, 685 (5th Cir.1974), cert. denied, 422 U.S. 1045, 95 S.Ct. 2662, 45 L.Ed.2d 697 (1975)(Psychiatrist may not testify as to whether a principle prosecution witness "would have a tendency to be reliable in distinguishing truth from nontruth, realities from fantasies;" distinguishing cases concerning the existence of a "deviant mental condition" affecting a witness' capacity to see and hear), cited by, United States v. Jackson, 576 F.2d 46, 49 (5th Cir.1978) and First Nat'l Bank of Louisville v. Lustig, 96 F.3d 1554, 1575 (5th Cir.1996); United States v. Binder, 769 F.2d 595, 602 (9th Cir.1985 (State's expert witnesses improperly bolstered child complainants' testimony by testifying that the children could distinguish reality from fantasy and truth from falsehood); Hoult v. Hoult, 57 F.3d 1, 7 (1st Cir.1995)(Citing Binder); United States v. Cecil, 836 F.2d 1431, 1441 (4th Cir.1988)(Psychiatrist may not testify as to the credibility of a witness; citing Wertis); United States v. Pino, 827 F.2d 1429, 1430 (10th Cir.1987) (Psychiatric opinion generally not admissible as impeachment; citing Jackson); United States v. Beasley, 72 F.3d 1518, 1528 (11th Cir.), cert. denied, ___ U.S. ___, 116 S.Ct.2570, 135 L.Ed.2d 1086 (1996)(Rejecting proffered defense testimony of expert that witness was a psychopath who had no conception of the truth; "Absent unusual circumstances, expert medical testimony concerning truthfulness or credibility of a witness is inadmissible;" citing Wertis). Perhaps exemplifying these courts' concerns is the statement by the Fifth Circuit that an opinion regarding a person's ability to distinguish reality from fantasy is really a comment on truthfulness: "Peeled of its thin veneer of jargon, it amounts to no more than an inquiry whether the witness is to be believed by the jury or not." Wertis, 505 F.2d at 685. New Mexico's Supreme Court has cautioned that experts should testify about competency outside the presence of the jury to "insure that the jury does not confuse competency with credibility." State v. Fairweather, 116 N.M. 456, 863 P.2d 1077, 1081 (1993). See also State v. Casaus, 121 N.M. 481, 913 P.2d 669, 677 (1996)(Trial court did not err in permitting expert to testify that "most children have fairly good memories for any event that's novel," where the expert never commented specifically on the victim's memory). Other courts permit an expert to testify about a witness' or complainant's general *66 ability to distinguish reality from fantasy but not about that person's specific allegations. People v. Ainsworth, 45 Cal.3d 984, 248 Cal. Rptr. 568, 584-585, 755 P.2d 1017, 1033-35 (1988), cert. denied, 488 U.S. 1050, 109 S.Ct. 883, 102 L.Ed.2d 1006 (1989); People v. Castro, 30 Cal.App.4th 390, 35 Cal.Rptr.2d 839, 842 (1994)(citing Ainsworth), disapproved on other grounds, People v. Martinez, 11 Cal.4th 434, 45 Cal.Rptr.2d 905, 916, 903 P.2d 1037, 1048 (1995); Stewart v. State, 555 N.E.2d 121, 125 (Ind.1990)(Expert may testify that child is not prone to exaggerate or fantasize about sexual matters); State v. Colwell, 246 Kan. 382, 790 P.2d 430, 437 (1990)(Expert may testify that, as a result of extended observations and testing, he concluded that complainant was "capable of distinguishing the truth from a lie"); State v. Rimmasch, 775 P.2d 388, 389 (Utah 1989), quoting favorably, State v. Lairby, 699 P.2d 1187, 1199 (Utah 1984)(Defense expert could have testified about child complainant's psychological capacity for untruthfulness or delusional testimony, but "could not have offered any views on the testimony actually given")(emphasis in Rimmasch), Lairby overruled on other grounds by, State v. Ossana, 739 P.2d 628 (Utah 1987); Gomes, 648 A.2d at 404; United States v. Azure, 801 F.2d 336, 340 (8th Cir.1986)(Expert may: testify generally about a child's ability to separate truth from fantasy, summarize the medical evidence and express his opinion on whether it was consistent with the complainant's story, and perhaps discuss various patterns of consistency in the stories of child sexual abuse victims and compare those patterns with patterns in the complainants story); Westcott v. Crinklaw, 68 F.3d 1073, 1076 (8th Cir.1995)(Expert testimony that officer suffered from post-traumatic stress disorder, which often causes "incomplete, inaccurate or in some cases total memory lapses" was improper. Such testimony was directed at the reliability of the officer's statements and provided a psychological label or diagnosis as a way of excusing or justifying those statements); United States v. Scop, 846 F.2d 135, 142 (2nd Cir.), modified on rehearing in part, 856 F.2d 5 (2nd Cir.1988)(dicta: "even expert witnesses possessed of medical knowledge and skills that relate directly to credibility may not state an opinion as to whether another witness is credible, [citing Azure], although such witnesses may be permitted to testify to relevant physical and mental conditions" [bracketed material substituted for original). Minnesota distinguishes between adults and children. An expert may not testify about an adult's ability to distinguish reality from fantasy but may do so for children or mentally retarded persons. Saldana, 324 N.W.2d at 231. But even for children, an expert may not testify about the credibility of the particular allegations being tried. Id. Likewise, New York distinguishes "impaired witnesses." When a witness is impaired by some mental or physical condition (such as mental retardation), an expert may explain and describe the witness' condition. People v. Parks, 41 N.Y.2d 36, 390 N.Y.S.2d 848, 857-858, 359 N.E.2d 358, 366-68 (1976)(sixteen-year-old mentally retarded complainant). However, even when the witness is "impaired," an expert may not comment on the credibility of the witness' testimony: Thus it would be error for the court to permit a witness to express a view as to the credibility or believability of the impaired witness. Notably, this case does not involve a situation where the independent [expert] witness attempted to express a view, scientific or otherwise, on the degree of confidence, or lack thereof, to be placed upon the complainant's testimony. By our holding today, we do not suggest, or even intimate, that an expert may ever pass his own judgment on a witness' credibility and impart that judgment to the jury. Such testimony would be a serious usurpation of the jury's prerogative and responsibility. Id. 390 N.Y.S.2d at 858, 359 N.E.2d at 368 (bracketed material inserted). Although it is unclear whether young children would qualify as "impaired," some New York lower courts have treated them as such. In the Matter of Sanchez, 141 Misc.2d 1066, 535 N.Y.S.2d 937, 940 (1988), cited by, Matter of Kelly F, 206 A.D.2d 227, 621 N.Y.S.2d 698, 700 (1994), appeal withdrawn, 85 N.Y.2d 905, 627 N.Y.S.2d 327, 650 N.E.2d 1329 (1995). Sanchez held that a psychologist could testify about a child's ability to recall, report, or tell *67 the truth, "short of an opinion on whether or not she or he believes the witness." Id. Florida and Virginia hold ambiguous positions but would at least exclude expert capacity evidence concerning specific allegations. State v. Townsend, 635 So.2d 949, 958 (Fla.1994)(Trial court erred in admitting a psychologist's testimony that "the child's statements were truthful because, in her opinion, the child was capable of distinguishing between the truth and a lie and pretending and playing"); Fuster v. State, 664 So.2d 18, 20 (Fla.App. 3 Dist.1995)(Trial court properly excluded defense testimony regarding the effect of defendant's conditions of confinement and treatment by therapists on her ability to testify truthfully; relying upon Townsend); Coppola v. Commonwealth, 220 Va. 243, 257 S.E.2d 797, 803-804 (1979), cert. denied, 444 U.S. 1103, 100 S.Ct. 1069, 62 L.Ed.2d 788 (1980)(Defense psychiatrist could not testify that, due to personality disorder, complainant cannot determine the truth when she's testifying); Brown v. Commonwealth, 22 Va.App. 316, 469 S.E.2d 90, 93 (1996)(Expert can testify about effect of mental impairment on a witness' ability to relate matters at issue; distinguishing Coppola). While the authorities may differ concerning the outer boundaries of capacity evidence, they are largely in agreement that comments on whether a child's specific allegations are the result of mental incapacity should be treated the same as comments on whether a child's specific allegations are lies. Even courts that have not directly commented on the legitimacy of expert testimony relating to fantasy or capacity have nevertheless expressed, in varying degrees, a belief that comments on mental capacity for truthfulness should be treated the same as comments concerning moral disposition for truthfulness. "[E]xperts should not be allowed to give their opinion of the accuracy, reliability, or credibility of a particular witness in the case being tried." State v. Lindsey, 149 Ariz. 472, 720 P.2d 73, 76 (Ariz.1986)(emphasis added; bracketed material substituted for original). See also State v. Hyde, 186 Ariz. 252, 921 P.2d 655, 679 (1996), cert. denied, ___ U.S. ___, 117 S.Ct. 1091, 137 L.Ed.2d 224, (1997); State v. Walters, 120 Idaho 46, 813 P.2d 857, 866 (1990); Lawrence v. State, 796 P.2d 1176, 1177 (Okla.Crim.App. 1990); Caldwell v. State, 267 Ark. 1053, 594 S.W.2d 24, 28-29 (1980)(Expert opinion as to the reliability of a sexual abuse victim's perception and recall held to be inadmissible as invading the province of the jury), cited with approval by, Utley v. State, 308 Ark. 622, 826 S.W.2d 268, 270 (1992); Hall v. Commonwealth, 862 S.W.2d 321, 323 (Ky.1993)(Opinion that a child's testimony is most likely accurate is improper; relying upon Rimmasch); State v. Raymond, 540 N.W.2d 407, 409-410 (S.D.1995)("one witness may not testify as to another witness' credibility or truth-telling capacity because such testimony would invade the exclusive province of the jury to determine the credibility of the witnesses"). Some of these jurisdictions have also expressly stated that an expert witness may not, through opinion testimony, weigh the evidence in the case. State v. Rice, 261 Kan. 567, 932 P.2d 981, 997-98 (1997); Fowler v. Kootenai County, 128 Idaho 740, 918 P.2d 1185, 1191 (1996). In State v. Carlson, 80 Wash.App. 116, 906 P.2d 999, 1002 (1995), the Washington appellate court noted that the State could not have offered expert opinion to prove the child witness' credibility. In a footnote to this comment the court cited, among other cases, State v. Walden, 69 Wash.App. 183, 847 P.2d 956, 959 (1993). Carlson, 906 P.2d at 1002 n. 17. In Walden, the prosecutor, during cross-examination, asked a defense witness whether a prosecution witness was "mistaken" in his testimony. 847 P.2d at 958. Case law in Washington holds that it is improper to ask one witness whether another is lying because such a question is designed to elicit an opinion as to the credibility of the other witness. Id. The State attempted to distinguish the case by arguing that it merely asked if the other witness was "mistaken" not whether that other witness was "lying." The court responded that the distinction was irrelevant: Asking a witness whether another witness is lying is certainly more prejudicial than asking whether another witness is simply mistaken. In both situations however, the questioning is designed to elicit testimony in the form of one witness' opinion as to *68 the credibility or veracity of another witness, a determination which lies solely within the province of the jury. Id. at 959. As the Massachusetts Supreme Court explained: "The issue of credibility of a witness focuses on both the individual's ability and willingness to tell the truth." Ianello, 515 N.E.2d at 1184 (prohibiting expert comments on credibility). In a relatively recent treatise on the Federal Rules of Evidence, two commentators have observed that most courts treat capacity evidence as relating to truthfulness: Occasionally courts try to distinguish between expert testimony relating to capacity and expert testimony relating to disposition but usually they ignore the distinction. Apparently the reason is that trying to make the distinction raises unanswerable questions about human nature and does not coincide with lines drawn by experts. Deciding when expert testimony goes to capacity (credibility generally) and/or character (veracity in particular) promises difficulty similar to that involved in defining and applying concepts of mens rea and the defense of insanity. And the arguments for excluding expert testimony (it threatens the jury's role, is not scientifically sound, and may cause more confusion than it's worth) operate regardless which category applies. Mueller, Christopher B. & Kirkpatrick, Laird C., FEDERAL EVIDENCE, 2nd. ed., Vol. 3, § 312, p. 422-423 (1994)(footnotes omitted). Moreover, in another but similar context, the federal rules expressly exclude expert testimony regarding a person's mental condition: No expert witness testifying with respect to the mental state or condition of a defendant in a criminal case may state an opinion or inference as to whether the defendant did or did not have the mental state or condition constituting an element of the crime charged or of a defense thereto. Such ultimate issues are matters for the trier of fact alone. Fed. R. Ev. 704(b)(per 1984 amendment). While this addition to the federal rules does not address the specific question of credibility, it at least demonstrates hostility toward expert conclusions on a person's mental condition when that condition is an issue ordinarily considered to be a jury question. As the above discussion shows, the credibility of witness' or declarant's allegations has generally been considered an issue within the exclusive province of the jury. 4. Evaluation We are convinced that the concept of truthfulness properly includes mental capacity as well as moral disposition. The framers of the federal rules apparently anticipated this consequence, and our holding is consistent with the treatment of the issue by most jurisdictions. Moreover, we agree with the observation that attempting to distinguish moral and mental truthfulness raises unanswerable questions about human nature. Often the distinction between the two becomes blurred, and attempting to draw a bright-line distinction can be hazardous. Of course, in any given case, testimony may fall clearly into one category or the other, but this is not invariably so. The State contends that evidence regarding manipulation or fantasy is distinguishable from direct comments on truthfulness because such evidence does not foreclose the possibility that the child witness may be lying. We agree that ruling out manipulation or fantasy does not conclusively establish that the allegations are true. By the same token, however, ruling out a deliberate lie does not conclusively establish the truth of a witness' allegations because it is still possible for the witness' beliefs to be inaccurate—i.e. the result of manipulation or fantasy. See State v. Arrington, 251 Kan. 747, 840 P.2d 477, 482 (1992)(Although a close question, therapist's testimony did not cross the line when she testified that she did not think that her severely mentally retarded patient was capable of being purposefully deceitful); State v. Maggard, 1 Neb.App. 529, 502 N.W.2d 493, 498 (1993)(impermissible for expert to testify that child was capable of fantasizing but not capable of telling a deliberate lie). In order for a person's recitation of events to be true, the events depicted must be both honestly related and accurately perceived/remembered. *69 While neither honesty nor accuracy guarantees truthfulness by itself, each is nevertheless a direct comment on truthfulness. Both constitute "particularized testimony concerning the alleged victim's credibility." See Silvey, 894 S.W.2d at 671; State v. Moran, 151 Ariz. 378, 728 P.2d 248, 252 (1986). Moreover, our rejection of expert testimony on truthfulness is based in part on a belief that psychology is not an exact science but involves much uncertainty and is often subjective. This uncertainty manifests itself in the kinds of situations that arise in the present case, in which experts attempt to assess the accuracy of a child's statements. According to one recent article, there is significant disagreement among legal scholars and psychologists about what factors can be applied to guarantee the reliability of children's testimony. Questioning the Reliability of Children's Testimony: An Examination of the Problematic Elements, 19 LAW & PSYCH. REV. 203, 210 (1995). Even so, psychologists and others in the mental health profession have much expertise in the area of human behavior that can be of assistance to a fact-finder. For instance, we have recognized expert knowledge concerning the behavioral characteristics typically exhibited by sexual abuse victims. Cohn v. State, 849 S.W.2d 817 (Tex.Crim. App.1993). But, "[o]nce an `expert' imparts his scientific, technical or specialized knowledge to the jury concerning the area of his expertise, jurors are just as capable as the expert in drawing conclusions concerning the credibility of the parties in issue." Yount, 872 S.W.2d at 710 (bracketed material substituted for original). And credibility issues involving specialized expert knowledge will often also involve unspecialized knowledge within the purview of lay jurors. In determining whether a complainant's allegations have been fantasized, for instance, a jury may look not only to psychiatric and psychological information about mental illness and human behavior but also to the commonsense knowledge of laypersons about how people think and react. We should be cautious about permitting experts to draw conclusions that rest on both expert and lay knowledge. Likewise, we recognize that experts may have specialized knowledge concerning signs of coaching, coercion, and suggestion. Whether a child exhibits signs of manipulation seems similar to the question about whether a child exhibits signs of sexual abuse. Both involve behavioral manifestations of external influences or events acting upon the child. However, while expert testimony may be relevant, whether a child has in fact been manipulated also involves matters within the knowledge of laypersons. Once the expert has imparted his specialized knowledge to the jury, the jury can use that knowledge, along with its own lay knowledge of human nature, to arrive at its own conclusion. Moreover, while testimony about symptoms commonly exhibited by child victims only indirectly bolsters a child's credibility by circumstantially corroborating the child's story, testimony about manipulation comments directly on the accuracy of the story itself. See State v. Charboneau, 323 Or. 38, 913 P.2d 308, 313-314 (1996)(Citing previous Keller decision and distinguishing evidence that directly comments upon credibility from corroborating evidence that has the indirect effect of bolstering credibility). Hence, evidence of manipulation and fantasy, whether relating to mental capacity or moral disposition, should be analyzed under the same rules that govern evidence of truthful or untruthful character.[6] Under Rule 608(a), a party may attack a witness' or other declarant's[7]general capacity or disposition to tell the truth. The other party may respond to such an attack with evidence supporting that person's general capacity or disposition for truthfulness. An attack or response concerning a person's general capacity for truthfulness may be related to a particular time period. General capacity evidence includes whether a person *70 can distinguish between reality and fantasy and/or whether the person's physical or mental condition adversely affects a person's ability to accurately perceive and/or relate events. Where mental capacity and moral disposition can be clearly distinguished, a party may respond with evidence supporting mental capacity only when mental capacity has been attacked and likewise for moral disposition. Where the distinction is blurred or unclear, the trial court has discretion on whether to admit certain evidence supporting truthfulness and should be reversed only for an abuse of discretion. Likewise, a party may attack the credibility of a witness or other declarant by offering the following kinds of manipulation evidence: (a) evidence that the person is, in general, the kind of person who is easily manipulated, (b) common signs or symptoms of manipulation and evidence that the person displays some or all of these signs or symptoms, and (c) evidence of third person acts or words designed to manipulate. The other party may respond to such attacks with (a) evidence that the person is not, in general, the kind of person who is easily manipulated, (b) common signs or symptoms of manipulation and evidence that the person does not display these signs or symptoms, and (c) evidence rebutting the existence of third person acts or words of manipulation. A party may respond to any of the first three categories with evidence from any of the second three categories provided that the latter evidence is logically relevant to rebut the former. Of course, when evidence relating to capacity, fantasy, or manipulation is presented in the form of expert testimony, such evidence must also assist the trier of fact under Rule 702. To do so, the testimony must meet the proper tests for scientific reliability, and the testimony must reflect information outside the general knowledge of lay persons.[8] And, such testimony may be excluded under Rule 403 if the probative value of the evidence is substantially outweighed by the danger of unfair prejudice, etc.[9] Moreover, expert rebuttal testimony supporting truthfulness is ordinarily admissible even if the predicate attacks on truthfulness comes from lay rather than expert witnesses, so long as the above conditions are met. In addition, like New York's highest court in Parks, we recognize that some witnesses are viewed by society as "impaired": jurors may naturally tend to view with suspicion the competence of a class of people to testify because of a condition or disability embodied by the class. Some examples of this phenomenon would include young children and mentally retarded persons. When an "impaired" witness or declarant is expected to testify, expert testimony should be permitted in the offering party's case-in-chief concerning the ability of the class of persons suffering the "impairment" to distinguish reality from fantasy and to perceive, remember, and relate the kinds of events at issue in the case. Testimony admitted pursuant to this "impaired" witness theory should be limited to a discussion of the class rather than the individual witness and should focus on the ability of the class to accurately perceive, remember, etc. rather than any tendency to do so. See Yount, 872 S.W.2d at 711-712 (error for expert to express opinion that a particular class of persons is truthful). However, evidence that a person's allegations are the result of manipulation or fantasy is inadmissible. Such evidence never assists *71 the jury because the jury is just as capable as the expert of drawing the conclusions involved.[10] B. Opening the Door? Nevertheless, otherwise inadmissible evidence may be admitted if the party against whom the evidence is offered "opens the door." But, the party offering the evidence may not "stray beyond the scope of the invitation." Bush v. State, 773 S.W.2d 297, 301 (Tex.Crim.App.1989). "The wall, attacked at one point, may not be fortified at another and distinct point." Farris, 643 S.W.2d at 697. In accord with the above discussion, it is clear that "general" testimony asserting manipulation or fantasy, which is admissible, does not open the door to inadmissible testimony that specific allegations are not the result of manipulation or fantasy. See generally above discussion. See also People v. Gaffney, 769 P.2d 1081, 1088 (Colo.1989).[11] The question remains whether testimony about specific allegations necessarily opens the door to expert testimony. We have held that the introduction of defense expert testimony that a child was coached opened the door to testimony by the State's expert that the child was not coached. Kipp v. State, 876 S.W.2d 330, 335-336 (Tex.Crim.App.1994). Other courts have held that cross-examination of the State's expert may open the door to expert opinions on credibility that otherwise would have been inadmissible. Commonwealth v. O'Brien, 35 Mass.App.Ct. 827, 626 N.E.2d 892, 896 (1994), review denied, 417 Mass. 1102, 631 N.E.2d 58 (1994); State v. Baymon, 336 N.C. 748, 446 S.E.2d 1, 4 (1994); United States v. Rivera, 43 F.3d 1291, 1295 (9th Cir.1995); United States v. Arruza, 26 M.J. 234, 238 (CMA 1988). Minnesota has held that a defendant opened the door to expert testimony by eliciting testimony from the complainant's mother during cross-examination. State v. Myers, 359 N.W.2d 604, 611 (Minn.1984). The defense elicited testimony that the mother had disbelieved her child's allegations for many months. Id. The court characterized the child's mother as "the ultimate expert" with respect to the complainant, but it also noted that the evidence elicited was not merely reputation or opinion evidence but was evidence of the mother's disbelief of her child's specific allegations. Id. The lower Minnesota appellate court has stepped beyond Myers to permit expert testimony to rebut general opinion evidence elicited from the child's mother. State v. Coleman, 426 N.W.2d 889, 892 (Minn.App.1988). The court also noted, however, that the testimony was not prejudicial *72 because the statement was volunteered by the officer (not asked for by the State), was very brief, and was not argued to the jury. Id. Although Minnesota may view a child's mother as "the ultimate expert," we find more sound the reasoning of a Texas Court of Appeals. In Matter of G.M.P., 909 S.W.2d 198 (Tex.App.—Houston [14 Dist.] 1995), the State argued that the appellant in the case had waived his complaint about expert testimony by failing to object to similar testimony by the child's mother. In rejecting the State's contention, the Court of Appeals held that the effect of a mother's testimony was very different from that of an expert: A jury would expect a mother to testify that her son was truthful, and would likely view such testimony with natural skepticism. On the other hand, the testimony of a police officer qualified as an expert on the investigation of sexual assault cases would likely carry exceptional weight and an aura of reliability which could lead the jury to abdicate its role in determining [the complainant's] credibility. Id. at 206 (bracketed material substituted for original). As a general rule, we reject the idea that lay opinion testimony on the truthfulness of allegations can open the door to expert testimony, but we do not foreclose the possibility that in some instances this might be the case, such as where a lay opinion witness' testimony takes on "expert-like" qualities or perhaps where there is no available lay rehabilitation testimony. The Minnesota Court of Appeals has also addressed whether a defendant may open the door to expert testimony about the truth of allegations by cross-examining the complainant on her own credibility. The court held that Myers did not control the case and that cross-examination of the complainant did not open the door to expert testimony. State v. Miller, 377 N.W.2d 506, 509 (Minn.App.1985). On the other hand, a Texas Court of Appeals held that the defense opened the door to testimony as to the truth of allegations when it cross-examined the complainant and introduced several defense witnesses who testified that the complainant made up the allegations. Brown v. State, 756 S.W.2d 793, 796 (Tex.App.—Houston [14th Dist.] 1988). But Brown is not clear authority as it involved the significant additional element of several defense witnesses. In any event, we agree with the Minnesota Court of Appeals that cross-examination of the complainant—who is in essence testifying about her own, not another's, credibility—is even less likely to open the door than cross-examination of the complainant's mother. In determining whether the door is opened to the introduction of certain testimony, it important to distinguish proper rebuttal testimony under Rule 608(a) and comments on credibility that fall outside of that provision. Although some minimal relevance is required, general evidence supporting truthful character as outlined by Rule 608(a) may be liberally employed to respond to attacks on truthful character: there need only be a "loose fit" between the rebuttal evidence and the predicate attacks on character. Moreover, such predicate attacks on character may include not only the introduction of evidence but also the use of questions designed to call a witness' character for truthfulness into doubt.[12] But, evidence that falls outside of the recognized rebuttal category in Rule 608(a) is illegitimate and a much more specific showing of relevance to character attacks must be shown: there must be a "tight fit" between the rebuttal evidence and the predicate attacks on character. Moreover, merely asking questions designed to cast doubt upon a witness' character for truthfulness should not ordinarily be sufficient to open the door to evidence supporting the truthfulness of specific allegations; favorable answers to such questions must ordinarily be obtained in order to justify the introduction of such specific testimony. It is helpful to summarize our holding by classifying the different types of testimony that impact upon credibility and their admissibility under Rules 608 and 702; we have included *73 an appendix at the end of this opinion for that purpose. III. APPLICATION We now apply the foregoing rules to the case at hand. There are four segments of testimony to analyze: (1) Burns' testimony about manipulation, (2) Burns' testimony about fantasy, (3) Poole's testimony about manipulation, and (4) Poole's testimony about fantasy. Burns did not express an opinion about whether the child's allegations had been the subject of manipulation but stated merely that the child did not exhibit "behaviors that point to being manipulated." While some of the behaviors described by Burns would be commonly known by laypersons to be indications of manipulation, we find that at least some behaviors were not. We do not believe that laypersons would necessarily associate with manipulation "more enthusiasm" in telling the story or the use of more sophisticated language.[13] We hold that Burns' testimony that the complainant did not exhibit the traits of manipulation did not constitute a direct comment upon the truth of the complainant's allegations. Burns' testimony about fantasy is a different matter. Although Burns was asked whether the complainant displayed the "traits" of fantasizing, Burns had not described any "traits" of fantasizing in her testimony. Instead, she merely equated fantasizing with lying. Her testimony that the complainant had not exhibited any evidence of fantasizing was therefore a direct comment on the truthfulness of the complainant's allegations. Poole testified that whether the complainant had been the subject of manipulation was the "less likely explanation." He did not testify merely that the complainant did not exhibit characteristics associated with manipulation nor did he even testify that the complainant was not easily manipulated. Instead, he expressed an opinion on whether the complainant had in fact been manipulated. The State argues that Poole's conclusion is not a comment on truthfulness because it is logically possible for someone to be manipulated against making allegations. But that claim ignores the reality of the testimony at trial. The child made allegations of abuse, and the defense raised the issue of manipulation by cross-examining the child about whether her grandmother told her what to say. By saying that it was less likely that the child had been manipulated, Poole clearly conveyed to the jury that the child's allegations were not the result of manipulation. We hold that this testimony was a direct comment on the truthfulness of the complainant's allegations. Likewise, Poole's testimony that the allegations "were not the result of fantasy" constituted a direct comment on the truthfulness of the complainant's allegations. Having found three out of four statements to be inadmissible, we face the remaining question of whether appellant opened the door to direct comments on the truth of the allegations. We hold that he did not. Appellant's cross-examination of the complainant about her own credibility did not open the door to another person's testimony concerning the truth of specific allegations. Moreover, cross-examining the complainant about her grandmother's actions falls within the guidelines we outlined above concerning permissible manipulation testimony. While the State was permitted to respond, as it did, with evidence concerning whether the child exhibited traits of manipulation, testimony about the grandmother's actions did not open the door to direct comments on whether the child's allegations actually were the result of manipulation. Nor was the door opened by appellant's cross-examination of the complainant's mother. Appellant's question concerning *74 whether the complainant was a truthful person was a valid attempt to impeach the complainant's character through general opinion testimony. Obviously, the complainant's mother knew the complainant well enough to give a lay opinion as to character. The question concerning whether the child ever fantasized or fabricated, while perhaps an attempt to elicit a general opinion as to character, may also have been an attempt to elicit extraneous specific instances of conduct. If the latter, such testimony would be barred by Rule 608(b). But, we do not think such a general question justifies a response concerning the actual truthfulness of the allegations for the offense being tried. The question which asked "Is it possible she won't talk to you because it might not be true?" does appear to be an attempt to elicit from the mother a direct comment on the child's truthfulness. But the question did not call for an expert-like answer from the mother. Moreover, the question was open-ended, giving the complainant's mother the opportunity to support the truthfulness of the child's allegations. The open-ended nature of the question distinguishes this case from the Minnesota opinion in Myers because the defense counsel in Myers asked questions concerning the mother's prior disbelief in the child's allegations. Such questioning might have created the impression of untruthfulness without giving the mother the opportunity to respond to that impression. But in the present case, the mother had the opportunity to respond and in fact responded negatively to appellant's question. Further, the State could have remedied any misimpression created by the defense cross-examination by eliciting testimony that the allegations were in fact truthful in a redirect examination of the complainant's mother. And, of course, at that point the State would have been free to introduce general testimony as outlined above in Part III(A)(4) (Evaluation). The State did not need to respond with expert testimony about the truth of the allegations. To the extent that appellant's cross-examination of the complainant's mother may have opened the door to some type of testimony, the State exceeded the scope of the invitation by introducing expert testimony concerning the actual truth of the complainant's allegations. As for the remaining reasons given by the Court of Appeals, we find that the questions involved were at most indirect, circumstantial attacks on credibility. Whether the child was hurt by cousins or babysitters and whether the child's injuries were caused by "clumsy" physical examinations are merely circumstances rebutting an inference of guilt, either by identifying other possible perpetrators or by explaining the existence of physical injuries. This testimony does not directly comment on truthfulness and therefore cannot open the door to expert testimony that does so. Finally we are not persuaded by the Court of Appeals' statement that the testimony in general suggests the child's allegations were merely the product of a bitter divorce. The defense is permitted to explore motives for third persons to manipulate the child or motives for the child to fantasize or lie so long as the testimony does not directly comment on the truthfulness of the child's allegations. We conclude that appellant did not open the door to the expert testimony of which he complains. Having found error, we deem it appropriate that the Court of Appeals conduct a harmless error analysis in the first instance. The judgment of the Court of Appeals is reversed, and the case is remanded for proceedings consistent with this opinion. OVERSTREET, J., dissents. BAIRD, J., filed a concurring opinion in which MEYERS, J., joined. MANSFIELD, J., filed a concurring opinion in which PRICE, J., joined. WOMACK, J., filed a concurring opinion in which McCORMICK, P.J. joined. *75 Appendix Type of Testimony Examples When admissible 1. Subtantive evidence of testimony about symptoms Admissible during offering guilt which incidentally exhibited by child abuse victims party's case-in-chief (or in impacts on credibility and whether or not cross-examination of other child in question exhibits party's witnesses) those symptoms physical evidence and whether such evidence is consistent or inconsistent with child's allegations 2. General testimony relating testimony in general about Admissible during offering to impaired witnesses the ability of a class of persons party's case-in-chief (or in or declarants recognized by society as cross-examination of other being impaired, such as party's witnesses) if an impaired young children or the mentally person is expected to retarded, to distinguish be a witness or declarant reality from fantasy and to perceive, remember, and relate the kinds of events at issue in the case 3. General testimony that testimony that child has a Admissible during offering directly attacks credibility general character for making party's case-in-chief (or untruthful or dishonest cross-examination of other statements party's witnesses) to attack credibility of witness or out-of-court testimony that child has a declarant general character for fantasizing or is the kind of person that is susceptible to manipulation; testimony that child has difficulty distinguishing between fantasy and reality testimony about the common symptoms or traits of a child who is fantasizing or being manipulated coupled with testimony that the child in question exhibits some or all of those symptoms or traits testimony that a child suffers from a mental handicap, mental disorder, the influence of a chemical substance, or some other physical or mental impairment coupled with testimony about adverse effects that such impairment may have on perception and/or memory testimony that certain third parties committed acts designed to manipulate the child into making certain allegations *76 4. General testimony that testimony that child has a Admissible as rebuttal to attacks directly supports credibility general character for making on credibility so long truthful or honest statements as there is a loose fit between the rebuttal testimony testimony that child does not and the predicate attacks on have a general character for credibility; predicate attacks fantasizing or is not the kind may include an attorney's of person that is susceptible questions as well as testimony; to manipulation; testimony expert testimony is generally that child does not have difficulty permitted to rebut lay distinquishing between testimony as well as other fantasy and reality expert testimony testimony about the common symptoms or traits of a child who is fantasizing or being manipulated coupled with testimony that the child in question does not exhibit those symptoms or traits testimony that certain third parties have not committed acts designed to manipulate the child into making certain allegations testimony that a child does not suffer from a mental handicap, mental disorder, the influence of a chemical substance, or some other physical or mental impairment that would adversely affect perception and/or memory 5. Specific testimony attacking testimony about specific instances May be admitted only to rebut or supporting in which the child other specific testimony credibility has lied, fantasized or had attacking or supporting credibility been manipulated; specific and only if there is a instances in which the child tight fit between the rebuttal told the truth, accurately testimony and the previous perceived reality, or resisted specific testimony attacking manipulation (Rule 608(b)) or supporting credibility; expert testimony is generally testimony that the child's allegations not permitted to rebut lay relating to the offense testimony at trial were the result of manipulation or fantasy, or were lies; testimony that the child's allegations were not the result of manipulation or fantasy, or were not lies testimony that a child did not in fact accurately perceive or remember events due to a physical or mental impairment *77 BAIRD, Judge, concurring. Issues of manipulation and fantasy are ultimately issues of truthfulness. The questions of whether an expert can testify as to the truthfulness of a particular child or children as a class were answered by Yount v. State, 872 S.W.2d 706, 710 (Tex.Cr.App.1993). As to an individual child, we specifically held: "[E]xpert testimony that a particular witness is truthful is inadmissible under [Tex.R.Crim. Evid.] 702." Id., 872 S.W.2d at 711. As to children as a class we held: ... A number of courts faced with testimony as to the truthfulness of a class of persons to which the complainant belongs have concluded that such testimony is tantamount to direct testimony that the complainant would not lie ... We agree. An expert who testifies that a class of persons to which the victim belongs is truthful is essentially telling the jury that they can believe the victim in the instant case as well. This is not "expert" testimony of the kind which will assist the jury under Rule 702. Yount, 872 S.W.2d at 711.[1] So there would be absolutely no confusion, the Yount opinion concluded: "We hold that Rule 702 does not permit an expert to give an opinion that the complainant or a class of persons to which the complainant belongs is truthful." Id., 872 S.W.2d at 712. Therefore, Yount is the controlling authority and the only citation necessary to resolve this case. Nevertheless, the majority both relies upon Yount to resolve the case and rejects Yount by stating: ... [w]hen an "impaired" witness or declarant is expected to testify, expert testimony should be permitted in the offering party's case-in-chief concerning the ability of the class of persons suffering the "impairment" to distinguish reality from fantasy and to perceive, remember, and relate the kinds of events at issue in the case. Ante, at 70, (emphasis in original). The majority's discussion regarding the admission of expert testimony about a "class" of persons is contrary to Yount, which is settled and controlling authority on that issue. Moreover, the aforementioned passage from the majority opinion is dicta because this Court did not grant review on that issue.[2] For the majority to go beyond what this Court agreed to hear has no precedential value. For the reasons stated in Yount, I join only the judgment of the Court. MEYERS, J., joins this opinion. MANSFIELD, Judge, concurring. I have strong doubts as to whether a social worker or a psychologist has the expertise to give a reliable opinion as to whether any witness—much less a child witness—is fantasizing or being manipulated. Given that such opinion testimony is essentially educated guesswork rather than based on quantifiable true science, I believe that we need to place strict limitations on the use of such testimony at criminal trials in Texas. See Tex.R.Crim. Evid. 702, 703. I also believe that any testimony relating to so-called "repressed memory syndrome" is inherently suspect and should not be admissible in Texas courtrooms. With these comments, I join the opinion of the Court. PRICE, J., joins. WOMACK, Judge, concurring. Children are not small adults. The younger they are, the less like adults they are. If the courts treat children as though they are adults, the courts are treating them wrongly. *78 The younger the children are, the more they are being mistreated. Texans have made progress in recognizing that child witnesses must be treated differently from adult witnesses. The legislative and executive branches have enacted several laws in this area. The corroboration requirement in sexual assault was deleted for child victims,[1] and the period of limitation in such cases was extended.[2] Those statutes showed the legislature's recognition that children do not always quickly tell someone that they have been abused. Child victims of certain offenses have been allowed to give evidence on videotape,[3] by closed-circuit television,[4] or by statement to an adult,[5] because it was recognized that there are better ways to get the truth from a child than by marching the child to the witness stand in open court to be questioned as an adult. This Court has not been altogether inactive. We approved closed-circuit televising of a child witness's testimony in all kinds of cases, even though it was not authorized by statute, for the same reasons the legislature approved it for victims of certain offenses.[6] Today the Court takes more steps forward. It permits the introduction of expert testimony about the ability of young children, as a class, to testify. Ante at 70-71 & Appendix at No. 2. It recognizes that credibility can be put in issue by questions, even when the answers are not impeaching. Ante at 72 & Appendix at No. 4. And the Court approves the introduction of expert "testimony about the common symptoms or traits of a child who is fantasizing or being manipulated coupled with testimony that the child in question does not exhibit those symptoms or traits." Ante at 76, Appendix at No. 4 ; also see ante at 70. Yet the Court still finds error in allowing the expert to testify that the child in question was not fantasizing or being manipulated, and it remands the case to the court of appeals for harm analysis. The result of that analysis is a foregone conclusion. The Court's rule is, when a child witness has been impeached, it is not error to admit testimony of the major premise of a syllogism (children who are fantasizing or being manipulated behave in certain ways) and the minor premise (this child did not behave in those ways), but it is error to admit testimony of the inevitable conclusion (this child was not fantasizing or being manipulated). Could jurors, who have heard the expert's opinion on the major premise and the minor premise, not be aware that the expert must hold the inevitable opinion about the conclusion? The conclusion is ineluctable after the premises have been admitted. It would be more natural and straightforward to admit the expert's opinion on the conclusion. I would not be deterred from doing so by the reasons the Court gives today. The Court says it is justified in excluding the expert's opinion on the credibility of a child witness because "expert testimony must aid—not supplant—the jury's decision." Ante at 59 (citing Duckett v. State, 797 S.W.2d 906, 914 (Tex.Cr.App.1990)). I would not follow the dictum of Duckett, or the holding of its successor, Yount v. State, 872 S.W.2d 706 (Tex.Cr.App.1993), which exclude the expert's opinion on the credibility of the child who reports sexual abuse. First, the expert opinion could supplant the jury's decision only if it were accompanied by an instruction from the court that the jury must agree with the expert's assessment. State v. Geyman, 224 Mont. 194, 199, 729 P.2d 475, 478 (1986). Second, this dictum is merely a restatement of the common-law rule that expert opinion must not "invade the province of the jury" or "embrace an ultimate issue." Texas Rule of Criminal Evidence 704 expressly did away with that common-law rule as, ironically, this Court recognized in Duckett, supra, at 914. Third, the Court's view that an expert's opinion on credibility does not assist the jury is based on its notion that jurors are just as capable as experts in drawing conclusions *79 about credibility.[7] It is important to remember what this Court forgot in cases such as Yount v. State, supra: The issue is not the credibility of witnesses generally or even the credibility of child witnesses generally. It is whether a child who says she has been sexually abused is fantasizing or being manipulated. This is not within the everyday experience of lay persons. Even if it were within the experience of some lay persons, none of them would be on the jury; the parties will have removed from the jury every person who has personal experience with sexually abused children. It would be helpful for the jury to hear the ultimate opinion of someone who has been trained and experienced in dealing with such children, and who dealt with the particular child at a time closer to the alleged event and in a setting that is more conducive than a courtroom to getting accurate evidence from children. Fourth, Texas Rule of Criminal Evidence 403, which makes relevant evidence admissible unless its probative value is substantially outweighed by the danger of unfair prejudice, supports the admission of the expert's ultimate opinion. (Ironically again, the Court in Duckett v. State, supra, at 914, recognized that Rule 403 "significantly altered" the law for admitting evidence.) If the jury has heard what the Court has today held to be proper—the expert's opinion that manipulated or fantasizing children display certain characteristics which the victim in this case does not display—how could there be any substantial danger of further prejudice in the expert's opinion that this child has not been manipulated or fantasizing? As I have pointed out above, the conclusion is ineluctable after the premises have been admitted. The Court says that there is disagreement on the subject of improving the reliability of children's testimony, "[O]ur rejection of expert testimony on truthfulness is based in part on a belief that psychology is not an exact science but involves much uncertainty and is often subjective."[8] This is true of many fields of expertise, such as medicine. The standard for the admission of expert testimony is not its "scientificness"; it is whether the expert has specialized knowledge that will assist the trier of fact to understand the evidence or to determine a fact in issue. Texas Rule of Criminal Evidence 702. Finally I want to record my surprise at the Court's skepticism about the value of psychological opinion. This skepticism has nowhere been evident on the question of predicting future dangerousness in capital murder trials.[9] The American Psychiatric Association says such prediction is not within the expertise of psychiatrists,[10] and this Court has held that lay persons (or at least persons with no formal training) are also qualified to predict future dangerousness.[11] But this Court admits opinions on the subject from psychologists and psychiatrists. Chambers v. State, 568 S.W.2d 313, 324 (Tex.Cr.App. 1978) (rejecting argument, in "well-written brief citing many authorities on the subject of psychiatric diagnosis," that the discipline of psychiatry is not sufficiently advanced to permit predictions of future violent behavior), cert. denied, 440 U.S. 928, 99 S.Ct. 1264, 59 L.Ed.2d 484 (1979). Accord, e.g., Barefoot v. State, 596 S.W.2d 875, 887 (Tex.Cr.App. 1980) ("This Court is well aware that the ability of psychiatrists to predict future dangerousness is the subject of widespread debate. However, we are not inclined to alter our previously stated view"), cert. denied, 453 *80 U.S. 913, 101 S.Ct. 3146, 69 L.Ed.2d 996 (1981). It would be more consistent with these decisions to admit the expert's ultimate conclusion on the testimony of child witnesses. The Court says that Montana is in the minority in holding that children require special consideration.[12] Although it may be in the minority, the Montana Supreme Court's holding recognizes that child witnesses are different from adult witnesses, it is faithful to the rules of evidence, it is logical, it meets the needs of the jury, and it furthers the goals of the justice system. I would be willing to follow Montana's lead on this issue, and today I join the Court's opinion which moves in that direction. McCORMICK, P.J., joins this opinion. NOTES [1] Burns also characterized herself as a "psychotherapist." In addition to being a "certified social worker," she also had an "Advanced Clinical Practitioner License." [2] Outside the jury's presence, appellant had already made specific objections to the proffered opinion testimony and obtained an adverse ruling. The trial court agreed to permit appellant to refer to these prior objections as "same objection." [3] All references to rules are to the Texas Rules of Criminal Evidence unless the context indicates otherwise. [4] In a pre-rules case, we did hold it improper for an expert witness to testify, based upon hypotheticals exactly matching the facts of the case, that "children nine and under are totally incapable of fantasizing about" the acts alleged, in the absence of impeachment of the witnesses. Farris v. State, 643 S.W.2d 694, 697 (Tex.Crim.App.1982). We referred to this error as "improper bolstering." Id. [5] We note that we have previously considered the law of other jurisdictions on whether an expert may give a direct opinion concerning the truthfulness of a witness' statements. Duckett, 797 S.W.2d at 911-912; Yount, 872 S.W.2d at 710-712 & 711 n. 8. The following states discussed in the present opinion have adopted versions of Rules 608, 702, and 704 that are the same in all respects material to the present case as the versions found in the Texas criminal rules: Arizona, Arkansas, Colorado, Idaho, Iowa, Minnesota, Mississippi, Montana, Nebraska, New Jersey, New Mexico, North Carolina, Oklahoma, Oregon, South Dakota, Utah, Vermont, Washington, Wisconsin, and Wyoming. See Ariz. R. Ev. (eff. 1977). Ark. R. Ev. (eff.1976). Colo. R. Ev. (eff. 1980). Id. R. Ev. (eff.1985). Iowa R. Ev. (eff. 1983). 50 Minn.Stat. Ann. Ev. R. (eff.1977). Miss. R. Ev. (eff.1986). Mont. R. Ev. (eff.1977). Neb. R. Ev. (eff.1975). N.J. R. Ev. (eff.1993). N.M. R. Ev. (eff.1973). N.C. R. Ev. (eff.1977). 12 Okl.Rev.Stat. Ann. §§ 2608, 2702, and 2704 (eff.1978). Or.Rev.Stat. §§ 40.350, 40.410, and 40.420 (eff.1981). S.D.C.L. §§ 19-14-9, 19-15-2, and 19-15-4. Utah R. Ev. (eff.1983). Wash. R. Ev. (eff.1979). Wis. Stat. Ann. §§ 906.08, 907.02, and 907.04 (eff.1974). Wy R. Ev. (eff. 1978). The rules in Maine are materially the same except that their versions of Rule 608 do not permit opinion evidence on character for truthfulness. See Me. R. Ev. (eff.1976). Florida excludes opinion evidence under its version of Rule 608 but permits evidence "Showing a defect of capacity, ability, or opportunity in the witness to observe, remember, or recount the matters about which he testified." See Fla. Ev.Code §§ 90.608(4) and 90.609 (eff.1977). The other relevant Florida rules are materially the same as those in Texas. See §§ 90.702 and 703. Kentucky excludes opinion evidence under its version of Rule 608 and it has no version of Rule 704. See Ky. R. Ev. (eff.1992). Michigan's rules are the same except that its version of Rule 702 requires that the scientific, technical, or other specialized knowledge be "recognized." See Mich. R. Ev. (eff.1978). When the Ohio cases cited in this opinion were decided, Ohio's rules were materially the same as the Texas rules, but the wording of Ohio's rule 702 has subsequently been significantly altered, although the alteration does not appear to substantively affect the holdings of the cases cited. See and compare Ohio R. Ev. (eff.1980) to Ohio R. Ev. 702 (eff.1994). Indiana adopted its version of the rules of evidence in 1994. See and compare Ind. Stat. § 35-37-4-2 to Ind. R. Ev. (eff.1994). [6] This holding does not govern expert testimony offered solely to determine a witness' competency when such testimony occurs outside the presence of the jury. [7] Attacking or supporting the credibility of a non-testifying declarant is subject to the same rules as a testifying witness. Rule 806. [8] Just as any lay witness, an expert may testify to events that he has observed, such as the demeanor of a child victim during out-of-court statements. See Trowbridge, 647 N.E.2d at 419; Kelly F., 621 N.Y.S.2d at 700. An expert may, perhaps, be able to testify as a lay character witness if he has known the child victim long enough to have an opinion. But see State v. Sims, 158 Vt. 173, 608 A.2d 1149, 1155 (1991)(expert cannot give opinion as to complainant's truthful character where witness knows complainant only through the case at trial). If an expert chooses to give a lay opinion, however, he should clearly inform the jury that his opinion is not an expert opinion. [9] Rule 403 provides: Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence. [10] Judge Womack contends that our exclusion of expert conclusions concerning the truthfulness of allegations is illogical. He isolates three different types of statements that an expert might make: (1) children who are fantasizing or being manipulated behave in certain ways, (2) this child did not behave in those ways, and (3) this child was not fantasizing or being manipulated. He claims that (3) necessarily follows from (1) and (2), and, because we permit expert statements of the type (1) and (2) variety, we must necessarily permit conclusions of type (3). But, for his argument to work, Judge Womack must assume that the expert testifies that all children who are fantasizing or being manipulated behave in certain ways. If only some exhibit the behaviors in question then a conclusion that a particular child is not fantasizing or being manipulated because he does not exhibit the behaviors does not necessarily follow. But, given the imprecision of psychological science and the variability of human nature, no competent and honest expert could make the global statement necessary to satisfy Judge Womack's syllogism. At most, statements of type (1) and (2) would provide some inductive support for a conclusion of type (3). But an expert opinion of type (3) would also be supported by personal observations and lay knowledge of human behavior. The latter is clearly the exclusive province of the jury, and the former, along with statements of type (1) and (2) can be imparted to the jury by the expert. The jury can then make the inferences necessary to determine whether a type (3) conclusion is warranted without the expert commenting on that issue. [11] At least one jurisdiction diverges from the general trend and apparently authorizes specific testimony if four criteria are met: (1) the victim is a child, (2) the victim testifies, (3) the victim's credibility is attacked, and (4) the expert is qualified. State v. Scheffelman, 250 Mont. 334, 820 P.2d 1293, 1298 (1991). Montana acknowledges that its position is the "minority view." Id. Montana appears to hold this view because it believes child victims require special consideration. But, this position raises awkward line drawing problems, on exactly how young a child must be. See State v. Hensley, 250 Mont. 478, 821 P.2d 1029, 1031-1032 (1991)(sixteen-year-old not covered by the child victim exception). [12] Rule 608(a) permits general rebuttal testimony if character "for truthfulness has been attacked by opinion or reputation evidence or otherwise" (emphasis added). [13] Although appellant appears to challenge the scientific reliability of this testimony, we decline to address his contention because it was not addressed by the Court of Appeals and is therefore not properly before us. If necessary to a harm analysis, the Court of Appeals may conduct such an inquiry upon remand. We do not express any opinion on the merits of the issue nor do we express any opinion on whether appellant properly preserved or presented this claim in the courts below. [1] All emphasis is supplied unless otherwise indicated. [2] Review was granted on whether the Court of Appeals erroneously decided that the trial court may allow a social worker and a psychologist to testify that the complainant's testimony was not the result of manipulation or fantasy. [1] Tex.Code Crim. Pro. art. 38.07. [2] Tex.Code Crim. Pro. art. 12.01(2)(D). [3] Tex.Code Crim. Pro. art. 38.071. [4] Ibid. [5] Tex.Code Crim. Pro. art. 38.072. [6] Gonzales v. State, 818 S.W.2d 756 (Tex.Cr.App. 1991). [7] Ante at 69, citing Yount v. State, supra. [8] Ante at 69. [9] Tex.Code Crim. Pro. art. 37.071, § 2(b)(1). [10] "In the view of amicus, this kind of inquiry about long-term future violence essentially does not involve medical analysis, and is not within the realm of established psychiatric expertise.... The psychiatrist's medical training and experience do not qualify him to provide reliable testimony about the likelihood of long-term future harmful acts." Brief Amicus Curiae, for the American Psychiatric Association, Smith v. Estelle, 602 F.2d 694 (5th Cir. 1979), quoted in J. Robitscher, The Powers of Psychiatry 205 (1980). [11] Jackson v. State, 822 S.W.2d 18 (Tex.Cr.App. 1990) (classification deputy at county jail); Matson v. State, 819 S.W.2d 839 (Tex.Cr.App.1991) (operator of halfway house); Esquivel v. State, 595 S.W.2d 516 (Tex.Cr.App.1980) (former district attorney); Duffy v. State, 567 S.W.2d 197 (Tex.Cr.App.1978) (priest). [12] Ante at 79 n. 11.