dissenting:
I.
I must disagree with the majority’s conclusions regarding the admissibility of the evidence relating to the telephone calls to the various bookstores. I believe this evidence was not relevant to the crimes charged and was, therefore, inadmissible. Because of the highly prejudicial nature of the evidence, I believe its admission at trial was reversible error.
The State offered this evidence to prove that the kidnapping of Cathy Ford was sexually motivated. It must be remembered, however, that the State’s case was based entirely on circumstantial evidence. The defendant was not seen with Cathy Ford on or near the date of her disappearance, and her body had not been found at the time of trial. There was absolutely no evidence that Cathy Ford had been sexually assaulted or abused following her disappearance.
The only evidence introduced by the State in this regard was a series of telephone calls to bookstores, placed from the defendant’s telephone, in which a male caller asked the salesperson to read selected portions of a book describing anal sex practices. None of the salespersons who testified were able to identify the defendant as the caller.1 From these meager facts, the *136majority concludes that “[t]he record shows that Paul Ferrell made numerous phone calls to bookstores and libraries, and that he regularly received sexual gratification from interaction with persons who were not voluntary providers of sexual gratification.” 184 W.Va. at 130, 399 S.E.2d at 841.
As an initial matter, I do not believe the majority’s characterization of the facts is accurate.2 Having twisted the facts, it comes as no surprise to me that the majority would then twist the law.
The majority begins by setting up a false premise, i.e., that “the State did not offer the phone call evidence to show that defendant was more likely to commit the crime of kidnapping.... The calls simply prove elements essential to the kidnapping case, namely motive and intent, and also tend to show a common plan or scheme.” 184 W.Va. at 130, 399 S.E.2d at 841. This would render the evidence admissible under Rule 404(b) of the West Virginia Rules of Evidence.
I believe, however, that the evidence of the telephone calls to the bookstores does not have a sufficient connection to the victim of the alleged crime to render it admissible under Rule 404(b). There was no evidence, for example, that similar phone calls or, indeed, any type of obscene or harrassing phone calls were made to Cathy Ford or to any other women in the area. There is nothing in the record to indicate that other women in the area had been abducted for sexual purposes during the period in which the telephone calls to the bookstore were made or that the defendant had ever attempted to coerce anyone to engage in the sexual practices described in the phone calls.
My review of the law of other jurisdictions reveals no case in which evidence of this nature, which did not clearly demonstrate some nexus to the offense charged, was admitted under the precepts of W.Va. R.Evid. 404(b). In cases involving sexual offenses, for example, evidence that the defendant made obscene or suggestive telephone calls before or after the crime has been held to be admissible to show motive or intent or to identify the defendant as the perpetrator of the offense. E.g., Ballweg v. State, 158 Ga.App. 576, 281 S.E.2d 319 (1981); Blanton v. State, 150 Ga.App. 559, 258 S.E.2d 174 (1979); State v. Stephens, 708 S.W.2d 345 (Mo.App.1986). In each of the above-cited cases, however, the phone calls were directed at the victim of the sexual offense.
In sum, then, I do not believe that this evidence was admissible as an exception to the general rule prohibiting the introduction of evidence of other crimes or wrongs. At most, the evidence demonstrated to the jury that the defendant had a propensity to commit the crimes charges because he is a “bad” man.3 That is the very result the rule seeks to prevent. E.g., State v. Hanna, 180 W.Va. 598, 378 S.E.2d 640 (1989); State v. Harris, 166 W.Va. 72, 272 S.E.2d 471 (1980); State v. Thomas, 157 W.Va. 640, 203 S.E.2d 445 (1974). This is particularly true in view of the highly prejudicial nature of such evidence. See State v. Dolin, 176 W.Va. 688, 347 S.E.2d 208 (1986), overruled on other grounds, State v. Edward Charles L., 183 W.Va. 641, 398 S.E.2d 123 (1990). Consequently, I conclude that the trial court abused its discretion in allowing the jury to hear the evidence of the telephone calls to the bookstores.
Moreover, the real question, left unanswered by the majority, is whether this evidence, if admissible, establishes an in*137tent or motive for the kidnapping. We have recognized that proof of a motive or intent is an essential element of the crime of kidnapping. E.g., State v. Hanna, supra. See State v. Fortner, 182 W.Va. 345, 387 S.E.2d 812 (1989); State v. Miller, 175 W.Va. 616, 336 S.E.2d 910 (1985).
This case lacks even the circumstantial evidence that was present in State v. Hanna, supra, where the defendant’s former live-in girlfriend had attempted to break off her relationship with the defendant and had begun dating another man. The defendant became enraged, broke into the man’s house, and took the girl with him against her will. She was never seen again. The State charged the defendant, among other things, with abduction, an offense which requires proof of a sexual motivation. See also State v. Miller, 175 W.Va. at 620 n. 3, 336 S.E.2d at 914 n. 3. Despite the prior intimate relationship between the defendant and the victim, we held that there was “no evidence of any overt act or statement indicating that the defendant was sexually motivated to remove [the victim] from [the] home.” 180 W.Va. at 605, 378 S.E.2d at 647.
An analogous situation arose in State v. Walters, 99 Or.App. 570, 783 P.2d 531 (1989), review allowed, 309 Or. 521, 789 P.2d 1386 (1990), in which the defendant was tried on charges of attempted kidnapping and attempted sexual assault after approaching a thirteen-year-old girl and offering her money to enter his vehicle to help him search for a lost dog. The prosecution offered evidence that the defendant had previously been convicted of the rape and sodomy of another thirteen-year-old girl whom he had approached in the same manner. In reversing the conviction for attempted sexual assault, the court stated:
“It is true that, from defendant’s persistent efforts, in spite of the victim’s refusal, to entice her into his truck to look for a nonexistent dog, the jury could conclude that defendant intended to interfere substantially with her liberty and to take her from one place to another.... The jury could also conclude from defendant’s behavior and remarks to the victim’s mother and the police officer that he was sexually attracted to the victim. There is nothing in the record, however, to permit the jury to infer that defendant intended to injure the victim physically or to rape or sodomize her.” 99 Or.App. at 575, 783 P.2d at 533-34.
Finally, in Stevens v. Commonwealth, 8 Va.App. 117, 379 S.E.2d 469 (1989), the defendant was convicted of murder and abduction. There was no affirmative evidence of a sexual relationship between the defendant and the victim, and the medical examiner was unable to determine if she had been sexually assaulted due to the decomposition of the body. The court found sufficient evidence to support the convictions based on several sexually suggestive remarks the defendant had previously made about the victim to third persons.
The evidence presented by the State below lacks even this minimal factual basis to prove the element of intent essential to the kidnapping conviction. As I have already remarked, there was simply no evidence to connect the telephone calls to the bookstores with the kidnapping of Cathy Ford. Consequently, because there was a lack of evidence to show that Cathy Ford was abducted for the purpose of obtaining some sexual advantage or concession, the kidnapping charge cannot, in my opinion, be sustained.
II.
The majority agrees that it was error for the prosecution to elicit from F.B.I. Agent Curtis his opinion that the defendant’s body language demonstrated his guilt of the crimes charged. Typical of its well-crafted opinion, the majority offers no reason for this conclusion. Certainly, as a minimum beginning point, the majority writer might have cited his recent opinion in State v. Woodall, 182 W.Va. 15, 385 S.E.2d 253 (1989), where the threshold rule for the admissibility of expert testimony in such circumstances was outlined in Syllabus Point 1:
“Under W.Va.R.Evid., Rule 702, expert testimony concerning generally rec*138ognized tests is presumptively admissible and the burden of excluding such testimony is upon the side seeking exclusion. However, when a test is novel or not generally accepted, that circumstance alone meets the threshold requirement of rebutting any presumption of admissibility under Rule 702 and, therefore, with regard to tests that are not generally accepted the burden of proof that the test is reliable remains on the proponent.”
It is perhaps an understatement to say that body language interpretation is novel and not generally accepted as a means of proving guilt.4 I am aware of no appellate court which has sanctioned expert opinion based on body language tests. The trial record is completely barren of any effort on the part of the State to make the threshold demonstration that the interpretation of body language has become recognized in the scientific community. Without such an initial showing, our cases are clear that the test results should not be admitted. State v. Woodall, supra; State v. Barker, 179 W.Va. 194, 366 S.E.2d 642 (1988); State v. Clawson, 165 W.Va. 588, 270 S.E.2d 659 (1980).
The majority, however, was undoubtedly not concerned with this point because it intended to conclude that the error was harmless, a result which borders on the absurd. While the majority relies on the harmless error rule set out in Syllabus Point 2 of State v. Atkins, 163 W.Va. 502, 261 S.E.2d 55 (1979), cert. denied, 445 U.S. 904, 100 S.Ct. 1081, 63 L.Ed.2d 320 (1980), it undertakes no careful analysis of the rule. In Atkins, we discussed at length some of the practical factors to consider in determining whether the identified error should be deemed harmless, including the principle that where the case “is basically a circumstantial evidence case, ... there is an increased probability that the error will be deemed prejudicial.” 163 W.Va. at 515, 261 S.E.2d at 63.
It must be recalled that the State’s case here was entirely circumstantial. Even the majority begins its opinion by acknowledging that this is a “difficult circumstantial evidence case.” 184 W.Va. at 126, 399 S.E.2d at 837. Cathy Ford’s body was never found. No witness ever saw the defendant with her. Contrary to the majority’s assertion, none of the women who had received telephone calls similar to those received by Cathy Ford on the day she disappeared could positively identify the defendant as the caller. Despite numerous interrogations, the defendant made no confession. Thus, when Agent Curtis, cloaked with the expertise of an F.B.I. agent, offered his opinion that the defendant had impliedly acknowledged his guilt, his testimony undoubtedly carried substantial weight with the jury. The crimes charged were heinous, the evidence entirely circumstantial, and what better place for a jury to repose confidence than in an F.B.I. agent’s expertise. I cannot find this error to be harmless.
III.
The majority does not even address whether the trial court erred in admitting the testimony of the F.B.I. serologist as to the analysis of bloodstains found at the defendant’s residence, relegating this issue to a footnote.5 I believe resolution of this question demonstrates further the majori*139ty’s disregard of the law in order to reach a desired result.
The evidence at trial revealed that approximately twenty samples of what appeared to be bloodstains taken from the defendant’s home and from Cathy Ford’s vehicle were analyzed. F.B.I. Agent Audrey Lynch, a forensic serologist, testified that her analysis of these samples demonstrated that one contained human protein, but could not be confirmed as containing blood, five contained blood which could not be confirmed as human, and the remainder contained human blood.
These stains were extremely minute in quantity and only three were found to contain recognized genetic markers. However, because of the small quantity available for test purposes, only one exhibit (Q 134) was found to contain more than one genetic marker.6 The testing process also destroyed the samples, leaving the defendant with no opportunity to have the test results verified. Finally, and most importantly, the State’s DNA expert admitted that because of the smallness of the sample, there was no way to determine whether these bloodstains came from the same person.
In addition, a cigarette butt of the brand smoked by Cathy Ford found in a corner of the defendant’s living room was tested. Traces of saliva found on the filter demonstrated that the smoker had been a Type A secretor. A DNA testing expert had previously testified that his analysis of exhibit Q 134 had revealed that the blood had come from a female.
There was no known blood sample from Cathy Ford which could be used for comparison. Agent Lynch, therefore, was given samples of blood from Cathy Ford’s parents. After testing for the same genetic markers, Agent Lynch testified that the bloodstains analyzed in the course of the police investigation could have come from an offspring of the Fords.7
Thereafter, Agent Lynch was asked to compute the percentage of the population which would carry all of the genetic markers found in the samples analyzed in the criminal investigation. She ultimately concluded that they would occur in approximately .75 percent of the population.8
*140The defendant does not challenge the scientific basis of the bloodstain and saliva analyses or of the DNA testing. Instead, the defendant challenges the conclusions of Agent Lynch, asserting that it was error to allow her to testify, in effect, that the bloodstains found at the defendant’s trailer could have come from Cathy Ford based on her analysis of the victim’s parents’ blood and to offer statistical analysis as to the percentage of the population with the same genetic markers revealed by the analysis.
This Court has recognized that the results of properly conducted and generally recognized blood tests are admissible at trial if they are relevant to some issue. State v. Wyant, 174 W.Va. 567, 328 S.E.2d 174 (1985); State v. Kopa, 173 W.Va. 43, 311 S.E.2d 412 (1983). See State v. Clements, 175 W.Va. 463, 334 S.E.2d 600, cert. denied, 474 U.S. 857, 106 S.Ct. 165, 88 L.Ed.2d 137 (1985). In State v. Woodall, supra, we stated the general proposition that statistical probability evidence is admissible with respect to blood analysis evidence.9
In Woodall, however, we also recognized the dangers inherent in the use of statistical evidence based on blood tests where the premise on which such evidence rests is faulty:
“In the celebrated case People v. Collins, 68 Cal.2d 319, 66 Cal.Rptr. 497, 438 P.2d 33 (1968), the defendants were an inter-racial couple. A witness described the robbers as a white woman and a black man who left the scene in a yellow car. The prosecution conjectured that 1 in 10 cars is yellow, and that 1 in 1,000 couples inter-racial. Thus the jury was given the figure of 10,000 to 1 odds the defendants were guilty.
“In Collins, the underlying statistics were wholly conjectural.... Fearing the jury might have seized on the 1 in 10,000 figure to convict, the California Supreme Court reversed the trial court. See generally G. Lilly, An Introduction to the Law of Evidence, 496-99 (1987).” 182 W.Va. at 23, 385 S.E.2d at 261.
I believe a similar flaw underlies the testimony of Agent Lynch in this case. The most obvious problem is that the three blood samples which were used to obtain the genetic markers could not be identified as coming from the same person. Yet, this was the very premise on which the State built its case. The parents’ blood was analyzed for the same markers as if they had come from a common source. I know of no court which has sanctioned such an approach.
The same distortion occurred in the statistical probability questions which are set out in note 8, supra. When the questions combine the genetic markers to elicit smaller probability answers, the fundamental premise is that the markers came from a common blood source. As previously pointed out, this is not the fact.
I believe it was error to permit the State’s blood expert to testify with regard to the genetic markers as they relate to the parents’ blood and to the statistical probabilities when combined.
I am authorized to state that Justice McHUGH joins me in this dissent.
. The defendant did admit making several calls from a different phone.
. There was, for example, no evidence of any telephone calls made to libraries. Moreover, one can only assume that the majority’s reference to "[in]voluntary providers of sexual gratification” is intended to describe the salespersons who took the telephone calls at the bookstores. In each instance, the salesperson was asked to locate and to read certain passages over the telephone. It is apparent that at any point along the way the salesperson could have terminated these voluntary activities by hanging up or by advising the caller that it was not customary to read books over the telephone.
. Indeed, in closing argument, the prosecution stated that the telephone calls demonstrated that the defendant had a “perverse or somewhat abnormal attitude” and that he was "depraved of heart and mind.”
. It is difficult to even conceive serious scientific consideration be given to such a subjective endeavor. Even if I could conceive of it, I would reject such body language interpretation for the same reason that we rejected the lie detector as appropriate evidence in State v. Frazier, 162 W.Va. 602, 252 S.E.2d 39 (1979), and the horizontal gaze nystagmus test in State v. Barker, 179 W.Va. 194, 366 S.E.2d 642 (1988).
. Note 2 of the majority opinion states that this issue was not fairly raised. 184 W.Va. at 126, 399 S.E.2d at 837. However, the issue was clearly raised by motion in limine prior to trial. In Syllabus Point 1 of Wimer v. Hinkle, 180 W.Va. 660; 379 S.E.2d 383 (1989), we stated:
"An objection to an adverse ruling on a motion in limine to bar evidence at trial will preserve the point, even though no objection was made at the time the evidence was offered, unless there has been a significant change in the basis for admitting the evidence.”
See State v. Weaver, 181 W.Va. 274, 382 S.E.2d 327 (1989); Bennett v. 3 C Coal Co., 180 W.Va. 665, 379 S.E.2d 388 (1989).
. A bloodstain taken from a floor joist of the defendant’s bedroom showed the presence of phosphoglucomutase, a blood enzyme (PGM type 1 + 2 +). A bloodstain on a paper towel taken from a trash bin in the defendant’s living room or kitchen exhibited the serum protein haptoglobin (Hp type 1). Exhibit Q 134, a wooden brace which ran from the bottom of the defendant’s trailer to the ground, was found to contain the PGM type 1 + 2 + marker as well as an enzyme, erythrocyte acid phosphatase (EAP type BA).
. The tests demonstrated that Cathy Ford’s father had Type O blood and carried the following genetic markers: PGM type 1 + 2 +; EAP type BA; Hp type 1. Cathy Ford’s mother had Type A blood with the following genetic markers: PGM type 1 +; EAP type BA; Hp type 1. Agent Lynch testified that the offspring of these two people would have to carry the Hp type 1 marker and that the other test results were not inconsistent with the characteristics that might be found in the blood of such offspring.
. The testimony at trial was as follows:
"Q. What percentage of the population has a blood type A?
"A. Approximately forty percent of the population.
"Q. What percentage of the population has a blood type A and are secretors?
"A. Approximately thirty-two percent of the population.
"Q. What percent of the population has a genetic marker of PGM 1 + 2 +?
“A Approximately twenty-two percent of the population.
"Q What percent of the population has a Haptoglobin of 1?
"A Approximately 16.5 percent.
"Q What percent of the population has a combination of PGM 1 + 2 + and BA, or EAP of BA?
"A Approximately nine percent.
"Q Agent Lynch, what percent of the population has a PGM of 1 + 2 +, an EAP of BA, and an HP of 1?
"A Approximately 1.5 percent.
“Q So that would include both male and female?
"A Correct.”
The .75 percent figure was arrived at by dividing the final number in half to reflect the DNA expert’s determination that the bloodstain on exhibit Q 134 had come from a female.
. In Woodall, 182 W.Va. at 23-24, 385 S.E.2d at 261-62, we stated:
"Blood type and enzyme tests have general scientific acceptance, and the distribution of particular blood traits in the population is ascertainable. [G. Lilly, An Introduction to the Law of Evidence 496] at 499-504 [ (1987) ]. The party seeking to impeach blood test evidence is free to cross-examine the proponent’s experts and offer experts of his own to discredit the conduct of the tests and the underlying statistical probabilities.”