concurring.
I write separately to explain my reasons for concurring in the majority decision, both with respect to the physical evidence that police obtained when defendant, before being given Miranda warnings, consented to the search of his *487backpack and with respect to the statements that defendant made after being given Miranda warnings.
In urging that the marijuana seized from defendant’s backpack should not be suppressed, the state has advanced an ambitious argument seeking a per se rule that physical evidence is never subject to exclusion under Article I, section 12, reasoning that the protection against self-incrimination is directed to testimonial in-court evidence only. I concur in the majority’s rejection of that per se rule. As this court has described it, our state-law based Miranda rule is “a judicial means” to secure the guarantee against compelled self-incrimination, one that this court has devised because it is appropriate for this court “to specify the procedure by which [Article I, section 12’s guarantee against compelled testimony] is to be effectuated.” State v. Mains, 295 Or 640, 645, 669 P2d 1112 (1983).1 Thus, the rule is constitutionally grounded, even if Miranda warnings and waiver of them are procedures that the constitution does not itself mandate. When those procedural requirements are violated, we properly ask whether any subsequently obtained evidence, physical or testimonial, is sufficiently a product of that violation to require suppression along with any statements made in direct response to unwarned custodial interrogation. And because the Miranda doctrine is a judicially devised procedural protection, one this court has adopted as a matter of state law, it falls to this court to determine whether and *488under what circumstances to exclude evidence to serve the objectives of that procedural rule.
The difficulty here is that the state’s argument on review stops with its invitation to adopt a per se rule of no exclusion. The state does not explore what test this court should adopt to determine whether and when physical evidence obtained after a Miranda violation should be subject to exclusion. Because of that omission, I concur in the result that the majority reaches. I write separately, however, to identify the limits of the majority’s reasoning and to point out the work that remains for future cases.
The majority announces a “derives from” test to determine under what circumstances consent to search, when that consent follows a Miranda violation, must be suppressed. 348 Or at 475-76. The starting point of such a test should be the underlying Miranda violation, because the nature and extent of that violation necessarily bear on whether and to what extent other evidence that police obtain is connected to that violation.
Here, the investigating officer asked defendant a total of three questions without first giving Miranda warnings. The first question was whether defendant owned the backpack. At trial, and continuing through the layers of appeal, the state has conceded that the officer was required to give defendant Miranda warnings and obtain a waiver before asking that question. For present purposes, I will assume that concession is correct.2
*489The officer’s second question was whether there was marijuana in the backpack. That question was, quintessentially, interrogation. Any reasonable officer asking if a container has marijuana in it would know that the question could elicit an incriminating response — i.e., an inculpatory or an exculpatory response that the prosecution might seek to use at trial.3 To lawfully ask defendant — who was under arrest — that question, the officer was required to advise defendant of his Miranda rights. The officer did not do so.
The officer’s third question asked if defendant would consent to a search of the backpack. That was not impermissible interrogation. Asking for consent to search is a mere request for permission. The answer either gives permission or it does not; the response is neither inculpatory or exculpatory (although, to be sure, the results of the search can be). With apparent unanimity, courts throughout the country that have considered the question have held that asking for consent to search is not interrogation within the meaning of the Miranda doctrine. See, e.g., U.S. v. Smith, 3 F3d 1088, 1098 (7th Cir 1993), cert den, 510 US 1061 (1994) (so observing; citing representative cases). Defendant has never contended to the contrary in this case.
Thus, at best for defendant, the Miranda violation in this case consisted of two short questions, each of which called for a one word (yes or no) response. In analyzing whether the physical evidence seized pursuant to defendant’s consent was “derived” from that violation, the majority concludes — with only brief discussion — that it was. 348 Or at 475-77. It is unclear what test the majority’s “derived from” test entails. The majority does not explain whether the test turns on causation, or exploitation, or some other way in which an initial illegality may be said to “taint” evidence that police gather after that illegality. Knowing the nature of the “derives from” test that applies is important. The majority appears to conclude that there is, in fact, a “derived from” *490connection between the illegal questioning and defendant’s consent to search in this case, and therefore places the burden on the state to disprove that connection. 348 Or at 476-77. At the least, if the state is to have that burden, it must know what it must disprove. As important, at some point, both litigants and lower courts are entitled to meaningful guidance as to the analysis that applies. Compare Wong Sun v. United States, 371 US 471, 487-88, 83 S Ct 407, 9 L Ed 2d 441 (1963) (“but for” causation not enough to establish that consent to search is the product of a prior illegality), with State v. Hall, 339 Or 7, 34-35, 115 P3d 908 (2005) (defendant need only show a “minimal causal nexus” between consent to search and prior illegality).
The state’s argument, as advanced in this court, examines none of those issues, however.4 Instead, as I have already noted, the state argues only that, as a per se matter, physical evidence is never subject to suppression under Article I, section 12, no matter how directly and immediately derivative of a Miranda violation. In this particular case, given the fact that the state has made no contrary argument, I am prepared to conclude that, under the particular circumstances of this case, the officer’s request to search was part and parcel of the impermissible unwarned questioning, at least enough to place the burden on the state to point to the circumstances that either legally or factually break that connection. Because the state has not done so, I concur. But it remains for future cases to explore, in a way that we do not in this case, when evidence obtained after a Miranda violation properly can be said to “derive from” that violation.
With respect to defendant’s subsequent, post-Miranda warning statements, the majority essentially *491adopts the test articulated by the plurality decision in Missouri v. Seibert, 542 US 600, 124 S Ct 2601, 159 L Ed 2d 643 (2004). 348 Or at 480-81. That test asks whether, viewed from the perspective of a reasonable person, the Miranda warnings that police give after an initial Miranda violation were effective for purposes of informing a suspect of his rights and obtaining a knowing and voluntary waiver of those rights.
I have no objection to that test. I see, ultimately, little or no difference in the “totality of the circumstances” analysis used to analyze that issue and the totality of the circumstances test that has long been in place to analyze the voluntariness of a confession following prior illegal conduct by police. See, e.g., State v. Wolfe, 295 Or 567, 572, 669 P2d 320 (1983) (drawing from totality of circumstances test in Brown v. Illinois, 422 US 590, 603-04, 95 S Ct 2254, 45 L Ed 2d 416 (1975)).
What is important to emphasize, however, is the distinctive context in which the Seibert plurality fashioned that test. The Seibert plurality identified that context in the very first sentence of the opinion: the case involved what had become, at least at the time, an increasingly popular “police protocol for custodial interrogation that calls for giving no warnings of the rights to silence and counsel until interrogation has produced a confession,” after which warnings are given and police then “lead[ ] the suspect to cover the same ground a second time.” 542 US at 604. As the plurality noted, “the reason that question-first [technique] is catching on is as obvious as its manifest purpose, which is to get a confession the suspect would not make if he understood his rights at the outset[.]” Id. at 613. The plurality therefore, throughout its opinion, expressly tied its “effective waiver” test to the question-first technique of interrogation — e.g., a confession “so obtained” {id. at 611); one obtained through the use of a “question first and warn later” protocol {id.); and one in which Miranda warnings are “inserted in the midst of coordinated and continuing interrogation” {id. at 613). That distinctive context also was what led the plurality to observe that belated Miranda warnings might be unlikely to provide a defendant with the knowledge necessary for a valid waiver, reasoning that it “would ordinarily be unrealistic to treat two spates of integrated and proximately conducted questioning *492as independent interrogations subject to independent evaluation simply because Miranda warnings formally punctuate them in the middle.” Id. at 614.
The plurality in Seibert also expressly contrasted the circumstances before it with those that had been involved in Oregon v. Elstad, 470 US 298, 105 S Ct 1285, 84 L Ed 2d 222 (1985), noting that the Court in Elstad had taken care to characterize the officer’s initial failure to grve Miranda warnings in that case as an “oversight.” Seibert, 542 US at 614. The plurality considered the facts in Seibert to be “[a]t the opposite extreme” and by “any objective measure” to reveal a police strategy adapted to undermine the effectiveness of the Miranda warnings. Id. at 616. The plurality observed the “unwarned interrogation was conducted in the station house, and the questioning was systematic, exhaustive, and managed with psychological skill. When the police were finished there was little, if anything, of incriminating potential left unsaid.” Id. In addition, the circumstances of the questioning did not change in any meaningful way between the initial round of questioning and the later round. And when police resumed their questioning and asked the defendant to cover the same ground a second time, they “fostered” the impression that the later round was a mere continuation of the first by reminding the defendant of the confession that she had already given them. Id.
As other courts have emphasized in adopting the equivalent of the Seibert plurality’s “effective waiver” test, an examination of the totality of the circumstances for an objective person’s perception of those circumstances permits no “bright-line rule.” State v. O’Neill, 193 NJ 148, 181, 936 A2d 438, 457 (2007). But because the test focuses on the circumstances as they would be objectively perceived, we should acknowledge that the danger of rendering Miranda warnings a nullity is greatest at the Seibert end of the factual spectrum. Likewise, that danger is at its lowest ebb at the Elstad end of the factual spectrum, where no question-first technique or protocol objectively appears to have been at work, where the initial questioning is minimal and the unwarned statements are more limited and less detailed than later ones, and where, given what went before the warnings, there *493is some substantial break in the circumstances.5 The facts of this case readily bring it within the class of cases in which, objectively, the circumstances of the initial Miranda violation are not such that they rendered the belated Miranda warnings or defendant’s waiver of his Miranda rights ineffective. The majority reaches the correct conclusion in that regard.
For those reasons, I respectfully concur.
Balmer and Kistler, JJ., join in this concurrence.I accept that this court has adopted Miranda warnings as an independent requirement under Article I, section 12. The pedigree of that doctrine is, however, uncertain. In the one case in which the court attempted to decide, consistently with our announced methodology for interpreting original provisions of our constitution, whether a Miranda-type rule could be divined from Article I, section 12, three members (of a six-member court) concluded that the answer was no, while three others believed the answer was yes, but only two believed that the rule could extend to anything other than formal custody or arrest. See State v. Smith, 301 Or 681, 725 P2d 894 (1986). Since then, without exploring the interpretative basis for an independent state Miranda rule, this court has, for the most part, assumed the existence of that rule. See, e.g., State v. Magee, 304 Or 261, 744 P2d 250 (1987) (per curiam decision, with three members concurring). But see State v. Isom, 306 Or 587, 592, 761 P2d 524 (1988) (“majority of this court has not agreed whether Miranda-type warnings are required under the Oregon Constitution”); State v. Kell, 303 Or 89, 734 P2d 334 (1987). It may well be that, at some point, in deciding novel questions about the scope or content of our state Miranda rule, this court will have to engage in that interpretative exercise. For present purposes, however, it is unnecessary to do so.
It is worth noting, however, that the issue is potentially a close one. In many circumstances, when police conducting a field investigation need to ascertain the owner of property in order to ask for consent to search that property, their inquiry may not qualify as “interrogation” that must be preceded by Miranda warnings. Here, the officer needed to determine who owned the backpack to ask for consent to search it, among other reasons. The backpack could have belonged to the driver, even though she denied that it did. It could have belonged to defendant. It could also have belonged to the driver’s parents, who owned the car, or to some other third party. In such a circumstance, attempting to determine the identity of the owner may potentially qualify as the kind of routine questioning normally attendant to investigatory activities, or even to custody and arrest, that does not trigger the requirement of Miranda warnings. See generally Wayne R. LaFave, 2 Criminal Procedure § 6.7(b) (3d ed 2007) (general investigatory questions not ordinarily considered interrogation; citing representatives cases). In an appropriate case, that issue may merit closer attention than it has received in this case.
“Interrogation,” both for federal and state law purposes, is express questioning, as well as words or actions on the part of police (other than those normally attendant to arrest and custody), that the police should know are reasonably likely to produce an incriminating response, whether inculpatory or exculpatory. State v. Scott, 343 Or 195, 202, 166 P3d 528 (2007) (adopting test from Rhode Island v. Innis, 446 US 291, 301, 301 n 5, 100 S Ct 1682, 64 L Ed 297 (1980)).
The state also has not argued on review that the officers had probable cause to believe that the backpack contained marijuana, based solely on the odor of marijuana coming from it, and that they could therefore seize and search it under the automobile exception. See State v. Meharry, 342 Or 173, 149 P3d 1155 (2006) (discussing that exception to the warrant requirement). The prosecution relied on that theory in the trial court, and the trial court made an explicit finding that the automobile was mobile at the time of the stop. Even though the trial court may have relied on that exception as an independent and alternative ground for its ruling, the state did not raise the automobile exception in the Court of Appeals. The state therefore cannot (and, as noted, does not) rely on that exception in this court. See Tarwater v. Cupp, 304 Or 639, 644, 748 P2d 125 (1988) (on review in this court, party may not argue alternative ground for affirmance of trial court if that argument was not presented to Court of Appeals).
In a case that falls on the Elstad end of the spectrum, Miranda warnings themselves will often provide the needed substantial “break” in the circumstances. In a case that falls on the Seibert end of the spectrum, Miranda warnings alone may rarely suffice. In such a case, among other possible attenuating circumstances, advice that the earlier unwarned statements probably are not admissible against the defendant is a significant intervening factor that will weigh heavily on the side of rendering the subsequent warned statements admissible.