(concurring in part; dissenting in part). I agree with the majority's analysis in part II. C., and the conclusion that the trial court correctly admitted Kidd's testimony that Johnson robbed him. With considerable reservations, I also accept the majority's analysis and conclusion in part II. B., regarding the newly-discovered evidence. Although I think it dubious to conclude that "it is not reasonably probable that there would be a different result on a new trial if both *495Davis' identification of Walker Johnson and Walker Johnson's recanted out-of-court statement were admitted[,]" majority op. at 489,1, like the majority, cannot say that "the trial court acted outside the bounds of its discretion," id., in reaching its conclusion. I do not, however, agree with the majority's analysis or conclusion in part II. A.
As the majority concedes, the trial court stated, "So there has to be some corroboration it seems to me which clearly indicates trustworthiness." Majority op. at 485. As the majority further concedes, these are "the forbidden words," id., because they invoke "the more stringent test of admissibility used by the federal courts," id. at 482-483, and, as the majority acknowledges, the more stringent test was rejected in State v. Anderson, 141 Wis. 2d 653, 416 N.W.2d 276 (1987). Nevertheless, the majority concludes that the trial court ultimately applied the correct standard and that the trial court's declaration of the incorrect standard "was merely a slip of the trial court's tongue." Majority op. at 486.
I give the trial court more credit. I accept that after discussing Anderson, the trial court clearly, and erroneously, reached its conclusion:
So there has to be some corroboration it seems to me which clearly indicates trustworthiness.
(Emphasis added.) Indeed, the trial court's analysis conformed to the erroneous standard it stated. The trial court addressed factors relevant to whether the statement "clearly indicates trustworthiness," specifically discussing Walker Johnson's possible motives to fabricate, his possible communication with the defendant while they were incarcerated, and his recantation of the statement. In addressing these factors, the trial *496court assumed the role against which Anderson warned:
By not requiring evidence which "clearly" indicates trustworthiness, our standard does not thrust the judge into the jury's role of assessing credibility and weighing evidence.
Anderson, 141 Wis. 2d at 666, 416 N.W.2d at 281.
Like the admirable character of the Seuss story, the trial judge "meant what he said [a]nd he said what he meant . . . ." Dr. Seuss, Horton Hatches the Egg (1940). The trial court explicitly invoked the wrong standard and then applied it. In concluding that the trial court's explicit error was an inconsequential "slip of the tongue," the majority itself slips.
Unfortunately, this is not the majority's only slip. Without explanation, the majority alters the Anderson standard. As accurately cited in the majority's opinion, majority op. at 482, Anderson held that in deciding the admissibility of a statement under sec. 908.045(4), Stats.:
The question before the judge is not whether the judge personally believes the statement is true, but rather whether there is sufficient corroboration for a reasonable person to conclude that it could be true.
Anderson, 141 Wis. 2d at 665-66, 416 N.W.2d at 281 (1987) (emphasis added). One paragraph later, however, the majority inaccurately paraphrases Anderson’.
Anderson recognized that a statement against penal interest offered to exculpate the accused must be excluded if the trial court... concludes that... no reasonable jury could find that the statement could be true.
*497Majority op. at 483 (emphasis added). Thus, altered by the majority, the standard (1) focuses on a unanimous jury rather than on any one of twelve reasonable persons who may comprise the jury and may, quite reasonably, view evidence in different ways; and (2) converts a positive standard that facilitates admission of evidence to a negative standard that virtually mandates exclusion of evidence.
Next, the majority slips again by adding a requirement to the statutory prerequisite for corroboration. The majority states that "[t]he trial court appropriately considered . . . the total lack of corroborating information that was independent . . . ." Majority op. at 486. Section 908.045(4), Stats., however, provides:
A statement tending to expose the declarant to criminal liability and offered to exculpate the accused is not admissible unless corroborated.
Now, the majority implies, the corroboration must be "independent" of what the declarant could have learned from the defendant. Hypertechnically, one could argue that, in some cases, this addition does not conflict with Anderson because it goes directly to the sufficiency of the corroboration rather than to the credibility of the statement. In this case, however, the details of the statement itself provide much of the corroboration for the statement. Thus, to reject the corroboration for lack of sufficiency because it is not "independent" is to reject the statement itself for lack of credibility.
Although the corroboration considered in Anderson was independent of the statement itself, Anderson did not address whether corroboration must be separate from the statement. The majority suggests that perhaps it must be. That suggestion, however, is dubious, particularly in light of the theme of Anderson that *498would enlarge rather than restrict the potential for admissibility of such evidence. In some cases, an "independent source" requirement would preclude a jury's consideration of highly probative evidence. Moreover, in some cases the declarant's details of the offense may be the only available corroboration.
Consider the following two statements under sec. 908.045(4), Stats.: (1) "He didn't do it, I did," accompanied by no corroborating evidence or additional statement; and (2) "He didn't do it, I did," accompanied by no corroborating evidence except the declarant's additional statements detailing the offense. Clearly, the first statement is inadmissible; there is no corroboration. The second statement, however, may be admissible unless we add an "independent source" requirement. Further, as a jury is entitled to conclude, the second statement may he reliable even if the details correspond to those the declarant could have learned from the defendant.
Under Anderson, although the trial court still reviews whether the corroboration is "sufficient to permit a reasonable person to conclude, in light of all the facts and circumstances, that the statement could be true," Anderson, 141 Wis. 2d at 656, 416 N.W.2d at 277, the trial court must not "thrust [itself] into the jury's role of assessing credibility and weighing evidence." Id., 141 Wis. 2d at 666, 416 N.W.2d at 281. In some cases, no doubt, the source of the corroboration will be among the "facts and circumstances" the jury considers.
Some statements offered to exculpate the accused are reliable. Others are not. Their reliability, however, does not necessarily depend on whether the corroborating source is independent. After all, although details within an exculpatory statement may be unreliable, *499corroboration provided by an independent source may be unreliable, too. And although details within the statement may have come from collusion, corroborating evidence from an independent source can come from collusion, too. Thus, in any given case, while the "sufficiency" of corroboration under Anderson may logically depend, in part, on the source of corroboration, the "sufficiency" of corroboration under Anderson must not legally depend only on whether the source is independent of the statement, or whether the declarant's information could have come from the defendant.
In this case, understandably, the trial court considered the corroborating details of the statement to be suspect because of the potential collusion between Walker Johnson and the defendant. The trial court's suspicion, however, relates directly to the credibility of the declarant and the weight of his testimony, not to its admissibility. Exclusion of the statement because of the trial court's suspicion conflicts with the Anderson admonishment that "[t]he question before the judge is not whether the judge personally believes the statement is true ... ."Anderson, 141 Wis. 2d at 665-66, 416 N.W.2d at 281.
Finally, the majority slips by relying on exactly what Anderson proscribes — the judge's personal belief. Basing its decision, in part, on the trial judge's "doubt[ ] [about the declarant] based on the trial court's extensive experience with the criminal justice system," the majority assumes facts beyond this record. Majority op. at 486-87. It further implies that the same declaration of the incorrect standard from a less experienced judge might not be similarly excused as "merely a slip of the trial court's tongue." That implication would complete the conversion of the Anderson standard to no standard at all or, perhaps, to the very *500standard emphatically supported by the dissenters in Anderson.
Under Anderson, the corroboration need not "clearly indicate trustworthiness" of the statement. The corroboration, to be "sufficient," only requires that it "permit a reasonable person to conclude, in light of all the facts and circumstances, that the statement could be true." Anderson, 141 Wis. 2d at 656, 416 N.W.2d at 277. Dissenting from that very distinction, Justice Ceci, in an opinion joined by Justice Steinmetz, maintained:
The focus should be not upon the reasonableness of the statement's content and the possibility of its having been made, but upon the reasonableness of the corroboration and whether the corroborative evidence indicates that the statement was probably made.
Id., 141 Wis. 2d at 672, 416 N.W.2d at 284 (Ceci, J., dissenting). The supreme court rejected that position. Instead, it adopted the standard that "facilitates rather than restricts the defendant's ability to present evidence." Id., 141 Wis. 2d at 665, 416 N.W.2d at 281.
The merits of the distinction defined in Anderson may be fairly debated and, perhaps, the dissenters may prevail some day. The distinction, however, was identified and carefully considered in Anderson and the supreme court chose a standard that differs significantly from the federal standard endorsed by the dissenters. In this case, the majority's opinion reasonably could be read to support the dissenting position in Anderson.
In this case, the trial court confronted a difficult evidentiary issue complicated by the fact that Johnson sought to support the admissibility of an exculpatory *501statement with corroboration supplied, in part, by the details of the statement. The trial court erroneously applied the federal standard rather than the less stringent Anderson standard. At the very least, the majority now clouds the issue and, analytically, undermines the Anderson standard. At most, the majority now adopts the position unsuccessfully advocated by the dissenters in Anderson.
Applying Anderson as we must, I would reverse and require a new trial at which both the statement of Walker Johnson and the newly-discovered evidence would be admissible. Accordingly, on this issue, I respectfully dissent.