OPINION

Defendant Raymond J. Vigil appeals his conviction for burglary, a second degree felony, claiming the traffic stop and subsequent impound of the vehicle in which he was riding was a pretext and not based on reasonable suspicion. Defendant further claims that consent to the post-inventory search of the vehicle was not voluntary and knowing. We remand for the entry of adequate findings.

FACTS

Defendant was a passenger in a car stopped by a Davis County Sheriffs deputy on the east side of Bountiful on January 14, 1989. The deputy observed defendant and his companions, all young Hispanic males, driving in an older vehicle in an affluent neighborhood. The deputy testified that he noted the youths were dressed as “cholos.”1 Based on this observation, the deputy followed the car for approximately twenty-three blocks before observing a speeding violation, for which he then *1297stopped the car. He was not aware of any reports of criminal activity in the area at the time of the stop.

As the deputy spoke with the driver, he observed open containers of alcohol in the vehicle, a violation of Utah law, and arrested all three occupants. The deputy impounded the car and searched the interior, excepting the locked glove compartment. The car was then towed by a private company to the company’s secured impound facility.

Later that evening the owner of the car, Sally Salazar, the mother of a juvenile arrested with defendant, telephoned police to inquire about the return of her car. She was referred to the arresting deputy, who conducted an extensive interrogation before addressing how she could retrieve her car. The deputy told her he had already inventoried the contents of the car, but he would “need to look through [her] car before we could let it go,” and then asked if he could again search the car. Salazar asked if her presence was required for the second search. The deputy replied that it was not and that nothing would be taken, but that he “need[ed] to look through it,” and again asked her consent. She then replied: “Yeah, I guess so.” The car was searched and the deputy discovered evidence linking defendant to a residential burglary which had occurred earlier that day.2

The trial court determined that the stop was not pretextual and that the ensuing arrest and impound were lawful. The court also “found,” without additional elaboration, that Salazar’s consent to the search was voluntary. On appeal, defendant claims the stop was pretextual, thereby invalidating the arrest, impound, and search, and that, in any event, consent was not voluntarily given and did not purge the taint of the unlawful arrest. The state, conceding the pretextual nature of the stop for purposes of the present appeal,3 responds that we need only consider the vol-untariness of Salazar’s consent and whether her consent validated an otherwise improper search.

STANDARD OF REVIEW

The state must present proof that consent to search was knowingly and voluntarily granted, State v. Arroyo, 796 P.2d 684, 687 (Utah 1990), applying a totality of the circumstances approach.4 State v. Ia*1298cono, 725 P.2d 1375, 1377 (Utah 1986). The state’s burden is a substantial one when, as here, the state relies on consent as purging the taint of unlawful police conduct. Arroyo, 796 P.2d at 687-88.

The issue remains whether volun-tariness of consent is a legal or factual matter. Contrary to the suggestion made by Judge Bench in his separate opinion, this issue is properly before us and must be resolved. If voluntariness of consent is a question of fact, we would affirm since the trial court “found” voluntary consent and that “finding” is supported by evidence in the record. If it is a conclusion of law, we would remand since no subsidiary findings were made by the court which support that conclusion.

Utah appellate court decisions have not been entirely consistent in treating this key question. See State v. Carter, 812 P.2d 460, 468 n. 8 (Utah Ct.App. 1991) (Utah Supreme Court “seems” to consider voluntariness of consent as a factual issue and Utah Court of Appeals has sent “mixed signals” on the question). Recent panels of this court have treated — explicitly or implicitly — voluntariness of consent as a purely factual matter and strictly applied clearly erroneous review, eschewing analysis of how any ultimate conclusion in a case could properly be considered a matter of pure fact. See, e.g., State v. Sterger, 808 P.2d 122 (Utah Ct.App. 1991); State v. Grovier, 808 P.2d 133 (Utah Ct.App. 1991). Other panels have reviewed the trial court’s factual findings underlying its conclusion as to voluntariness for clear error, evaluating the ultimate conclusion that such facts establish voluntary consent, in a legal sense, under a correction of error standard. See, e.g., State v. Hargraves, 806 P.2d 228, 231 (Utah Ct.App. 1991); State v. Bobo, 803 P.2d 1268, 1272 (Utah Ct.App. 1990). See also State v. Palmer, 803 P.2d 1249, 1251 (Utah Ct.App. 1990); State v. Talbot, 792 P.2d 489, 493 (Utah Ct.App. 1990); State v. Johnson, 771 P.2d 326, 327 (Utah Ct.App. 1989), rev’d on other grounds, 805 P.2d 761 (Utah 1991).

The state proposes in this case that evaluation of voluntariness of consent is a question of fact, to which the clearly erroneous standard — nothing more and nothing less — should be applied.5 The state relies primarily on United States v. Mendenhall, 446 U.S. 544, 557, 100 S.Ct. 1870, 1878-79, 64 L.Ed.2d 497 (1980) and Schneckloth v. Bustamonte, 412 U.S. 218, 248-49, 93 S.Ct. 2041, 2058-59, 36 L.Ed.2d 854 (1973). Each of these decisions of the United States Supreme Court contains language, albeit somewhat cursory, to that effect. We remain persuaded that both reasoned analysis and prior precedent sustain our view that the determination of voluntariness of consent is a mixed question of law and fact and is appropriately resolved under a bifurcated examination of, first, the predicate historical facts found by the trial court, weighed against a clearly erroneous standard, and, second, of the emerging legal conclusion, evaluated for correctness. Bobo, 803 P.2d at 1272. The soundness of *1299this view is demonstrated by considering the core functions of trial and appellate courts, the need for some semblance of consistency in the law, the reason for insisting on detailed findings, and the analytical deficiency inherent in treating ultimate issues as matters of fact.

1. Trial and Appellate Functions

Examination of voluntariness of consent mandates a fact-intensive inquiry, see State v. Lovegren, 798 P.2d 767, 770 (Utah Ct.App. 1990), although it is essentially a mixed question of law and fact. Regrettably, fact-sensitive issues may spawn confusion surrounding the appellate standard of review. See State v. Ramirez, 817 P.2d 774, 781 n. 3 (Utah 1991) (Supreme Court “might refer casually” to standard of review of voluntariness of a confession as an “ ‘abuse of discretion’ standard. In fact, it is not. It is a correctness standard, which incorporates a clearly erroneous standard for the review of subsidiary factual determinations.”). A mixed question of law and fact is one in which “the historical facts are admitted or established, the rule of law is undisputed, and the issue is whether ... the rule of law as applied to the established facts is or is not violated.” Pullman-Standard v. Swint, 456 U.S. 273, 289 n. 19, 102 S.Ct. 1781, 1790 n. 19, 72 L.Ed.2d 66 (1982). The Court has “noted the vexing nature of the distinction between questions of fact and questions of law.” Pullman-Standard, 456 U.S. at 288, 102 S.Ct. at 1790. There is no “rule or principle that will unerringly distinguish a factual finding from a legal conclusion.” Id.

Uniquely within the province of the trial court is the establishment of the predicate facts which emerge after a consideration of all testimony and the weighing of credibility. See Lovegren, 798 P.2d at 771. Because “the truth is rarely pure and never simple,”6 the trial judge is in the best position to sift witness credibility and the accuracy of conflicting evidence. Johnson, 771 P.2d at 327. Thus, we will not intrude upon the trial court’s distillation of the facts from contradictory evidence unless such findings are clearly erroneous. Id.

Appellate courts, on the other hand, take the facts as found by the trial court and square them with the law, sometimes developing the contours of the law in the process. But that development of the law is necessarily more consistent when undertaken by collegial appellate courts than individual trial judges acting alone, with no input from colleagues, no staff support to speak of, and little time for reflection.

2. Consistency

The judiciary best serves society when the law is interpreted uniformly and in predictable fashion. See United States v. Maragh, 894 F.2d 415, 418 (D.C.Cir.), cert. denied, — U.S. -, 111 S.Ct. 214, 112 L.Ed.2d 174 (1990). This is particularly true when fundamental civil liberties are at stake. Id. (“soundest of jurisprudential considerations compel appellate courts not to shirk their responsibility independently to apply important constitutional standards” and such “responsibility cannot be delegated to the trier of fact”).

Appellate review of legal conclusions and “ultimate facts” under a correction of error standard promotes the consistent application of legal principles. Appellate courts do not possess some inherent higher wisdom nor are trial courts bereft of the ability to resolve difficult legal questions. Notwithstanding, the conventional wisdom of two heads being better than one is actualized when three, five, or more judges combine their collective experience and legal knowledge to steer a constant and steady course. Moreover, appellate courts are largely insulated from the pressing demands of managing heavy calendars and the need to make snap decisions, and are afforded the luxury of full deliberation of issues presented to the court through elaborate briefing and focused oral argument, and with the able assistance of personal law clerks.

Concern for consistent application of legal principles is manifest in the doctrine of *1300stare decisis. Permutations of legal rules become controlling precedent through written appellate opinions. Such precedent in the criminal law realm controls future actions by law enforcement and prosecutorial officials. Therefore, while the trial court is primarily concerned with the proper resolution of factual issues under the controlling law, the appellate court addresses itself to the clarity and correctness of the developing law in order to provide unambiguous direction to those whose future rights and responsibilities are affected.

Attainment of consistency in deciding mixed questions of law and fact necessarily implicates a bifurcated decisional approach. First, the underlying historical facts must be revealed through production of testimonial and documentary evidence, a function reserved exclusively for the trial court. Second, the facts must be measured under the appropriate rule of law. Responsibility for proper execution of this latter step is shared by both trial and appellate courts. The determinant factor in selecting the proper standard of appellate review for mixed questions of law and fact is whether the first step — elucidation of historical facts — or the second step — correct application of the proper rule of law — is challenged on appeal. While we normally do not intrude on the trial court’s resolution in the first step,7 we independently assess the legal conclusions generated in the second step for correctness. This precept applies across the broad range of legal actions. See, e.g., Hurley v. Board of Review, 767 P.2d 524, 526-27 (Utah 1988) (describing standard of review for administrative determination of mixed questions of law and fact); Morris v. Mountain States Tel. & Tel. Co., 658 P.2d 1199, 1201 (Utah 1988) (questions of meaning of terms and intent are questions of fact, but determination of whether contract is ambiguous is question of law); State v. Crestani, 771 P.2d 1085, 1089 (Utah Ct.App. 1989) (ineffectiveness of counsel a mixed question of law and fact and ultimate conclusion reviewed independently); B.R. Woodward Mktg. v. Collins Food, 754 P.2d 99, 101 n. 1 (Utah Ct.App. 1988) (facts leading to determination of waiver reviewed under clearly erroneous standard; final conclusion of waiver reviewable as question of law).

This approach remains appropriate even where those conclusions are characterized as “ultimate facts.” See Baumgartner v. United States, 322 U.S. 665, 670-71, 64 S.Ct. 1240, 1243-44, 88 L.Ed. 1525 (1944) (ultimate fact which involves appraisal of strength of entire body of evidence is independently reviewable under correction of error standard); Pullman-Standard, 456 U.S. at 288, 102 S.Ct. at 1790 (no “rule or principle ... unerringly distinguishes] a factual finding from a legal conclusion”); Parker v. State, 66 Md.App. 1, 502 A.2d 510, 515 (1986) (“we give great weight to the findings of the hearing judge as to specific, first-level facts ... we must make our own independent judgment as to what to make of those facts; we must, in making that independent judgment, resolve for ourselves the ultimate, second-level facts”) (quoting Walker v. State, 12 Md.App. 684, 694-95, 280 A.2d 260 (1971)).

3. The Point of Requiring Findings

Both this court and the Utah Supreme Court consistently stress the importance of adequate findings of fact. Ramirez, 817 P.2d at 783-789; Lovegren, 798 P.2d at 770. We have previously endorsed a requirement of specific factual findings in ruling on motions to suppress. Marshall, 791 P.2d at 882 & n. 1. Cf. O’Connor, 546 N.E.2d at 339 (requiring “a high level of specificity and detail in a judge’s findings and analysis of the facts”). Specific, detailed findings not only ease the burden of appellate review by communicating the steps by which the ultimate legal conclusions are reached, Marshall, 791 P.2d at 882 n. 1, they also enable appellate counsel to properly frame issues *1301on appeal, see Lovegren, 798 P.2d at 771 n. 11, and to comply with our rigid requirement of marshaling evidence in support of subsidiary facts when challenging a trial court’s findings. See State v. Christofferson, 793 P.2d 944, 947 (Utah Ct.App. 1990); Mountain States Broadcasting v. Neale, 783 P.2d 551, 553 (Utah Ct.App. 1989). If, in any event, we were required to defer to an ultimate conclusion cloaked as a finding of fact, it would be pointless to require detailed factual findings to support the ultimate conclusion since that conclusion would be accorded the same deference as the underlying factual findings.

4. Analytical Deficiency of Viewing “Voluntary Consent” as a Finding

The analytical deficiency in reviewing the voluntariness of consent as a factual issue is demonstrated by the state’s concession at oral argument that it simply would not do for the trial court to deny a suppression motion on the strength of a single finding: “Defendant voluntarily consented to the search.” Yet if the question were really one of fact, such a tack would be unimpeachable. The burden would then fall to the appellate court to canvass all evidence of record and see whether that “fact” was clearly erroneous. If “voluntary consent” is itself a fact, the making of detailed “findings” pertaining to voluntary consent would be superfluous. Indeed, such “findings” would not really be findings at all but just a survey of evidence supporting the one, true finding. Conversely, a cat-alogue of “findings” which suggests coercion on the part of police officers followed by a factual finding of voluntary consent would be anomalous, leaving the appellate court to somehow weigh or prioritize the inconsistent “findings.”

We reaffirm the approach described in State v. Bobo, 803 P.2d at 1272, and now proceed to review the instant case under the bifurcated approach outlined in that opinion. We defer to the trial court’s underlying factual findings; once those facts are ascertained, we review as a question of law whether those facts establish the vol-untariness of consent in a constitutional sense.

VOLUNTARINESS OF CONSENT

One of the well-established exceptions to the warrant requirement of the Fourth Amendment is a search premised on voluntary consent. Schneckloth v. Bustamante, 412 U.S. 218, 219, 93 S.Ct. 2041, 2043-44, 36 L.Ed.2d 854 (1973). Voluntariness of consent must be decided after consideration of a totality of the circumstances. State v. Whittenback, 621 P.2d 103, 106 (Utah 1980). Accord, Bobo, 803 P.2d at 1272. When consent is claimed as the basis for a search following police misconduct, the consent must not only be voluntary in a factual sense; it must also not have been obtained by police exploitation of the prior illegality.8 Arroyo, 796 P.2d at 688.

Remand for entry of findings is proper where adequate findings do not address these issues. See Lovegren, 798 P.2d at 770-71. No actual findings address the question of consent; the court merely “found” consent without explanation. This lack of meaningful findings necessitates remand unless the undisputed facts of record compel but one conclusion. See, e.g., State v. Lovegren, 798 P.2d 767, 770-71 & n. 10 (Utah Ct.App. 1990). See also State v. Ramirez, 817 P.2d at 787 n. 6 (1991) (even though it can be appropriate to infer findings of fact not actually made, it is inappropriate to do so where not reasonable to assume court actually made such findings, as where testimony is irreconcilably contradictory). As the evidence in this case is in dispute, remand is necessary.

*1302The conversation between Salazar and the deputy, as related at the suppression hearing, demonstrates Salazar’s initial perception that the police had full control over her vehicle. The state claims that Salazar gave her consent freely, and that her only hesitation was whether she would need to travel from Salt Lake City to Bountiful to be present at the search. However, the record is equally susceptible to the view that Salazar reasonably believed that she would not be able to retrieve her car until she consented to the search, as well as that she believed, based on what the officer told her, that the car would be re-inventoried before release, with or without her consent.9 We are precluded from meaningful review of whether the state met its burden to show voluntary consent as well as whether the initial stop was actually pre-textual,10 by the lack of findings, oral or written, on each issue.11 See Lovegren, 798 P.2d at 770-71. Accordingly, we remand for a detailed and complete statement of the facts the court found from the conflicting evidence before it and for the entry of an order consistent with those findings.

GREENWOOD, J., concurs.

(concurring in the result):

Because of the way the issues were framed in this case, I join the majority in remanding in order to obtain adequate factual findings that will enable us to determine whether the search was constitutional by reason of Salazar’s consent.1 The additional findings are necessary essentially for the reasons noted in the majority opinion and in State v. Lovegren, 798 P.2d 767 (Utah Ct.App. 1990). Findings must encompass all material issues and “show that the court’s judgment or decree 'follows logically from, and is supported by, the evidence.’ ” Acton v. Deliran, 737 P.2d 996, 999 (Utah 1987) (quoting Smith v. Smith, 726 P.2d 423, 426 (Utah 1986)); Linam v. King, 804 P.2d 1235, 1238-39 (Utah Ct.App. 1991). The findings should also “be sufficiently detailed and include enough subsidiary facts to disclose the steps by which the ultimate conclusion on each factual issue was reached.” Acton, 737 P.2d at 999; State v. Marshall, 791 P.2d 880, 882 n. 1 (Utah Ct.App. 1990). In fact-sensitive situations such as search and seizure, the evidence is ordinarily not clear enough to fill gaps in the underlying findings to a degree that this court can perform its reviewing function. See State v. Ramirez, 817 P.2d 774, 789 (1991). In this case, the findings concerning consent are so conclu-sory that there is nothing that this court can meaningfully review — regardless of the standard of review.

*1303As the majority holds, this case must be remanded for additional findings. The majority’s statements concerning how we should review those findings are therefore dicta. Since the standard of review is not an issue in this case, it need not be mentioned at all. If any discussion is necessary, I would not depart from the longstanding practice of Utah courts to treat the voluntariness of a consent to search as a question of fact reviewable under the clearly-erroneous standard. See, e.g., State v. Arroyo, 796 P.2d 684, 687 (Utah 1990); State v. Sterger, 808 P.2d 122, 126-27 n. 5 (Utah Ct.App. 1991); State v. Grovier, 808 P.2d 133, 137 (Utah Ct.App. 1991); Marshall, 791 P.2d 880, 882 (Utah Ct.App. 1990).

Because I do not agree with dicta in the main opinion concerning the standard of review, I concur only in the result.