concurring and dissenting.
I dissent to the majority’s disposition of the inevitable discoveiy issue. The issue in this ease is whether Article 38.23(a), VAC.C.P., accommodates the federal “inevitable discovery” doctrine set out in Nix v. Williams, 467 U.S. 431, 104 S.Ct. 2501, 81 L.Ed.2d 377 (1984). This doctrine allows the admission of evidence the police would have discovered by means “wholly independent” of any constitutional violation. See Nix, 467 U.S. at 447-49, 104 S.Ct. at 2511. In Garcia v. State, 829 S.W.2d 796, 798 (Tex.Cr.App.1992), a plurality of this Court decided the federal inevitable discovery doctrine does not apply to Article 38.23(a). We granted the State Prosecuting Attorney’s petition for discretionary review to reexamine this plurality decision.
As I understand the majority’s analysis, the “plain” language of Article 38.23(a) does not accommodate the federal inevitable discovery doctrine because that doctrine “assumes” a causal relationship between the illegality and the evidence. Relying on the “ordinary” meaning of the word “obtained,” the majority tells us the “plain language of Article 38.23 does not provide for an inquiry into the potential legal acquisition of [the] evidence once it has been established that it *275was actually ‘obtained in violation of law.” The majority claims the issue presented here is “purely a question of statutory construction,” and criticizes what they characterize as “uncritical adherence to decisions of the United States Supreme Court” to construe a Texas statute.
But, we then learn the majority opinion ultimately relies on federal constitutional decisions to conclude the federal inevitable discovery doctrine assumes the evidence was “obtained in violation of the law.” ' Despite the majority’s criticism about “uncritical adherence to decisions of the United States Supreme Court,” their opinion and mine ultimately rely on federal constitutional decisions for our resolution of the critical question of whether the federal inevitable discovery doctrine considers the evidence “obtained in violation of the law.” The resolution of whether the remedy in Article 38.23(a) applies turns purely on a question of federal constitutional law.
And, I say the federal inevitable discovery doctrine does not consider the evidence “obtained” in violation of the Federal Constitution; therefore, the “plain” language of Article 38.23(a) does not exclude the evidence. Moreover, under the federal inevitable discovery doctrine, the word “obtained” takes on an entirely different meaning than its “ordinary” meaning. So, the majority errs to rely on the “ordinary” meaning of the word “obtained” in determining whether Article 38.23(a) accommodates the federal inevitable discovery doctrine.
The majority also says the federal inevitable discovery doctrine considers the evidence “obtained” in violation of the Federal Constitution but the evidence is admitted as an “exception” to the federal exclusionary rule. Therefore, according to the majority, the “plain” language of Article 38.23(a) excludes the evidence for purposes of Texas jurisprudence since the judiciary does not have the power to legislate “exceptions” to the “plain” language of Article 38.23(a).1
It is important to remember that Article 38.23(a) grants no substantive rights.2 It is purely a remedial statute that provides for the exclusion of evidence that is “obtained in violation of the law.” Here, appellant asserts a Fourth Amendment violation. Therefore, whether the remedy of Article 38.23(a) applies depends upon resolution of the substantive federal question of whether the federal inevitable discovery doctrine considers the evidence “obtained” in violation of the Fourth Amendment when the State can prove by a preponderance of the evidence that the police would have discovered the evidence by means “wholly independent” of any constitutional violation. See Nix, 467 U.S. at 447-48, 104 S.Ct. at 2511.
Appellant argues, and the majority apparently agrees, the federal inevitable discovery doctrine considers the evidence “obtained” in violation of the Fourth Amendment. The State disagrees, and argues evidence is not “obtained” in violation of the Fourth Amendment in the absence of a sufficient causal nexus between the evidence and the illegality, and this rule applies whether the causal *276break comes in the form of “attenuation,”3 “independent source,” 4 or “inevitable discovery.” 5 The State also argues that since the attenuation and inevitable discovery doctrines are specific variations on a common theme and share essentially the same rationale,6 it would make no sense to hold Article 38.23(a) incorporates the attenuation doctrine, but not the inevitable discovery doctrine.
This argument is consistent with this Court’s recent decision in Johnson v. State, 871 S.W.2d 744, 749-51 (Tex.Cr.App.1994), which held the federal attenuation doctrine is applicable to Article 38.23(a). By a 7-1 vote, a clear majority of this Court reached this decision by determining that the federal attenuation doctrine does not consider the evidence “obtained” in violation of the law because that doctrine considers the evidence sufficiently attenuated from the taint of the police illegality. See Johnson, 871 S.W.2d at 751 (the attenuation doctrine is a method for determining whether the evidence was “ ‘obtained’ ” in violation of the law).
Similarly, under Nix, the federal inevitable discovery doctrine considers the evidence “sufficiently attenuated” from the taint of the police illegality, if the State can prove the evidence inevitably would have been discovered by means wholly independent of any constitutional violation. See Nix, 467 U.S. at 448, 104 S.Ct. at 2511 (when the challenged evidence inevitably would have been discovered “without reference to the police error or misconduct, there is no nexus sufficient to provide a taint ”) (emphasis supplied). Therefore, if the State meets its burden of proof by a preponderance of the evidence, the federal inevitable discovery doctrine, just like the federal attenuation doctrine, does not consider the evidence “obtained in violation of the law.” See Johnson, 871 S.W.2d at 750-51.
The issue in this case really involves a straight-forward application of Johnson. In Johnson, seven members of this Court held the federal attenuation doctrine is applicable to Article 38.23(a). Johnson, 871 S.W.2d at 751. Now, barely two years later, four of these seven members decline to follow Johnson.
However, the majority says it “makes perfect sense to say that Article 38.23 accommodates the attenuation doctrine, but not inevitable discovery.” They explain that nothing in Johnson “regarding the attenuation of taint doctrine alters the fact that the plain language of Article 38.23 does not provide for an inquiry into the potential legal acquisition of evidence once it has been established that it was actually ‘obtained in violation of law.” So, according to the majority, Johnson should be read for the proposition that Article 38.23(a) does not call for a determination of whether the evidence was “obtained in violation of the law” once it has been established the evidence was actually “obtained in violation of the law.” I do not read Johnson as supporting this kind of circular legal reasoning. The critical thing about Johnson was that in determining whether the remedy in Article 38.23(a) applied, Johnson relied on federal constitutional law to conclude the federal attenuation doctrine does not consider the evidence “obtained in violation of the law.” See Johnson, 871 S.W.2d at 751 (the attenuation doctrine is a method for determining whether the evidence was “ ‘obtained’ ” in violation of the law).
The majority also attempts to distinguish this case from Johnson by explaining that under the attenuation doctrine the causal connection between the illegality and the acquisition of the evidence is broken because “there is a point beyond which the ordinary understanding of ‘obtained’ just does not apply.” However, the attenuation and the inev*277itable discovery doctrines begin with the premise that the challenged evidence is “ ‘in some sense’ the product of illegal governmental activity.” See Nix, 467 U.S. at 444, 104 S.Ct. at 2509 (emphasis in original). But, the United States Supreme Court has declined to adopt a “but for” rule “that would render inadmissible any evidence that came to light through a chain of causation that began with an illegal arrest.” United States v. Leon, 468 U.S. 897, 919, 104 S.Ct. 8405, 3414, 82 L.Ed.2d 677 (1984); see also Brown, 422 U.S. at 601-03, 95 S.Ct. at 2261. Under the attenuation doctrine, the evidence is admissible if the State can prove it “obtained” the evidence by means that “purge the primary taint.” See Brown, 422 U.S. at 602, 95 S.Ct. at 2261. This, and not necessarily some actual break in the chain of causation, is what breaks the causal connection between the illegality and the acquisition of the evidence under the attenuation doctrine. See id.
Similarly, the State’s ability to prove it inevitably would have discovered the evidence by legal means is what breaks the causal connection between the illegality and the acquisition of the evidence under the federal inevitable discovery doctrine. See Nix, 467 U.S. at 447-49, 104 S.Ct. at 2511 (when the challenged evidence inevitably would have been discovered “without reference to the police error or misconduct, there is no nexus sufficient to provide a taint”) (emphasis supplied). Under this doctrine, whether the illegality is the eause-in-faet or a but-for cause of the acquisition of the evidence is simply irrelevant. See Nix, 467 U.S. at 447-49, 104 S.Ct. at 2511; Brown, 422 U.S. at 601-03, 95 S.Ct. at 2261. If the State meets its burden of proof by a preponderance of the evidence, then federal law treats the evidence, as if, the police, in fact, discovered the evidence “wholly independent of any constitutional violation.” See Nix, 467 U.S. at 443, 104 S.Ct. at 2508. Therefore, the attenuation and inevitable discovery doctrines call for the same treatment under Article 38.23(a), and Johnson should control the disposition of the question presented in this ease.
The majority also justifies its holding by creating some kind of distinction between what they call an “exception to the federal exclusionary rule” and an “exception to core prohibitions of the Fourth Amendment.” See generally Garda, 829 S.W.2d at 798. The majority says the federal attenuation doctrine falls under the latter while the federal inevitable discovery doctrine falls under the former. As I understand the majority’s analysis, if federal law says the evidence is admissible as an “exception to the federal exclusionary rule,” then the “plain” language of Article 38.23(a) excludes the evidence. But, if federal law says the evidence is admissible as an “exception to core prohibitions of the Fourth Amendment,” then the “plain” language of Article 38.23(a) does not exclude the evidence. This also is more or less what Garcia says. From all of this flows the majority’s holding in this case.
The majority relies on Leon in support of its “exception to federal exclusionary rule” vs. .“exception to core prohibitions of the Fourth Amendment” distinction.7 The majority claims Leon supports the proposition that “whether evidence has been illegally obtained is a different and separate question from whether [the federal] exclusionary rule should apply.” The majority then claims Leon supports the proposition that the federal inevitable discovery doctrine assumes the evidence was “obtained in violation of the law.” This is not exactly correct. What Leon really says is:
“Whether the exclusionary sanction is appropriately imposed in a particular ease, our decisions make clear, is ‘an issue separate from the question whether the Fourth Amendment rights of the party seeking to invoke the rule were violated by police conduct.’” Leon, 468 U.S. at 906, 104 S.Ct. at 3412 (emphasis supplied).
The federal attenuation and inevitable discovery doctrines assume only the existence of a constitutional violation. When the federal inevitable discovery doctrine is applied to allow the admission of evidence, no one is *278claiming a defendant’s Fourth Amendment rights have not been violated. But, the federal attenuation and inevitable discovery doctrines legally break the causal connection between the constitutional violation and the acquisition of the evidence.
Leon sets out various situations where evidence “obtained in violation of the Fourth Amendment” is admissible as an “exception” to the federal exclusionary rule. See Leon, 468 U.S. at 910, 104 S.Ct. at 3414-16. However, it is clear the Supreme Court in Leon considered the federal attenuation doctrine, like the federal inevitable discovery doctrine, to break the causal connection between any police illegality and the acquisition of the evidence. See id,.; see also Nix, 467 U.S. at 447-48, 104 S.Ct. at 2511 (when the challenged evidence inevitably would have been discovered “without reference to the police error or misconduct, there is no nexus sufficient to provide a taint”) (emphasis supplied). Therefore, it really makes no difference whether the federal inevitable discovery doctrine is characterized as an “exception to the federal exclusionary rule” or as an “exception to core prohibitions of the Fourth Amendment” because, no matter how it is characterized, the federal inevitable discovery doctrine does not consider the evidence “obtained in violation of the law.” See Leon, 468 U.S. at 910, 104 S.Ct. at 3414-15; Nix, 467 U.S. at 447-48, 104 S.Ct. at 2511; Brown, 422 U.S. at 602-03, 95 S.Ct. at 2261.
Moreover, the majority’s distinction between “exception to the federal exclusionary rule” vs. “exception to core prohibitions of the Fourth Amendment” is a distinction with no legal significance. The Supreme Court has made clear that its various constitutional decisions in this area have taken into account considerations relating to the exclusionary rule and the constitutional principles it is designed to protect. See Leon, 468 U.S. at 910-11, 104 S.Ct. at 3414 (the perception underlying various constitutional decisions— that the connection between the police misconduct and evidence of crime may be sufficiently attenuated to permit the use of the evidence at trial — is a product of considerations relating to the exclusionary rule and the constitutional principles it is designed to protect).
Also, my understanding of how the federal exclusionary rule operates is that it provides a remedy of excluding evidence that is “obtained” in violation of the Constitution. See Mapp v. Ohio, 367 U.S. 643, 672-76, 81 S.Ct. 1684, 1701-08, 6 L.Ed.2d 1081 (1961) (Harlan, J., dissenting). If the evidence is admissible under the federal inevitable discovery doctrine or any other “exception” to the federal exclusionary rule, then the evidence has not been “obtained” in violation of the Constitution though there may have been a constitutional violation. See Leon, 468 U.S. at 910-11, 104 S.Ct. at 3414.
A fair reading of Leon, Nix, and Broum is that the various “exceptions” the Supreme Court has created to the federal exclusionary rule, especially the attenuation and inevitable discovery “exceptions,” break the causal connection between any illegality and the acquisition of the evidence. This is done for policy reasons such as striking a proper balance between society’s interest in deterring police misconduct and the public interest in having juries receive all probative evidence about a crime. See Nix, 467 U.S. at 443^5, 104 S.Ct. at 2509. The Legislature also has invited this Court to consider these kinds of factors when construing a provision of the Code of Criminal Procedure. See Article 1.03, VJLC.C.P.; Article 1.26, V.A.C.C.P. Based on the foregoing, I would hold it is unnecessary to “judicially legislate” the federal inevitable discovery doctrine as an “exception” to Article 38.23(a), because the “plain” language of Article 38.23(a) accommodates this federal doctrine.
Finally, principles of stare decisis alone require this Court to hold Article 38.23(a) accommodates the inevitable discovery doctrine. The majority even acknowledges the inevitable discovery doctrine has been a part of this State’s jurisprudence since at least 1969 — before and after the Supreme Court decided Nix. See Bell v. State, 724 S.W.2d 780, 787, 793 (Tex.Cr.App.), cert. denied, 479 U.S. 1046, 107 S.Ct. 910, 93 L.Ed.2d 860 (1987); Dickey v. State, 716 S.W.2d 499, 505 (Tex.Cr.App.1986); Wicker v. State, 667 S.W.2d 137, 141-42 (Tex.Cr.App.), cert. de*279nied, 469 U.S. 892, 105 S.Ct. 268, 88 L.Ed.2d 204 (1984); Vanderbilt v. State, 629 S.W.2d 709, 722-28 (Tex.Cr.App.1981), cert. denied, 456 U.S. 910, 102 S.Ct. 1760, 72 L.Ed.2d 169 (1982); Pitts v. State, 614 S.W.2d 142 (Tex.Cr.App.1981); McMahon v. State, 582 S.W.2d 786 (Tex.Cr.App.1978); Nicholas v. State, 502 S.W.2d 169 (Tex.Cr.App.1973); Johnson v. State, 496 S.W.2d 72 (Tex.Cr.App.1973); Noble v. State, 478 S.W.2d 83 (Tex.Cr.App.1972); Santiago v. State, 444 S.W.2d 758 (Tex.Cr.App.1969); see also Garcia, 829 S.W.2d at 801-03 (Miller, J., concurring) (inevitable discovery doctrine was applicable in this State long before the Supreme Court decided Nix).
These are this Court’s majority opinions which have applied the inevitable discovery doctrine. If a majority of this Court agrees on a principle of law, then that constitutes a “holding” of this Court. See 40 G. Dix & R. Dawson, Criminal Practice and Procedure at 14-15 (Texas Practice 1995). Principles of stare decisis require this Court to follow its earlier majority holdings. This is so basic that it requires no citation to authority.
With the exception of Bell, the majority declares that this Court’s prior majority holdings “can hardly be considered stare de-cisis” because they “appear to have been decided on the basis of the federal exclusionary rule.” My response to this is, “so what.” The point is a majority of this Court has held on numerous occasions that the inevitable discovery doctrine is part of the jurisprudence of this State. This Court should continue to follow these holdings and leave any changes to Article 38.23 to the legislature. See Garcia, 829 S.W.2d at 802 fh2 (Miller, J., concurring) (once a court has interpreted a statute, subsequent legislative inaction amounts to approval of that court’s interpretation).
In addition, the majority’s distinction between “exception to the federal exclusionary rule” vs. “exception to core prohibitions of the Fourth Amendment” is a meaningless distinction. So, it really makes no difference whether these cases might have been decided on the basis of the federal exclusionary rule.
This leaves Bell. Even the majority has to acknowledge that a majority of this Court in Bell held the attenuation doctrine was applicable to Article 38.23(a) and “may have similarly held as to the inevitable discovery doctrine.” The majority needs to read Bell again. In Bell, a majority of this Court was addressing Article 38.23 when it held the evidence was admissible under the inevitable discovery doctrine. See Bell, 724 S.W.2d at 787, 793. However, the majority refuses to follow Bell because it contains no “analysis whether the language of Article 38.23 ‘accommodates’ ” the inevitable discovery doctrine.
Instead of following this Court’s prior majority holdings, the majority decides to adopt a four-judge plurality decision in Garcia because it is the “only opinion to conduct” an analysis of whether Article 38.23(a) accommodates the inevitable discovery doctrine. The following is the substantive “analysis” from that plurality opinion in Garcia:
“[Article 38.23(a) ], on its face, absolutely requires the exclusion of all evidence seized in violation of the Fourth Amendment, and because the federal inevitable discovery doctrine is not an exception to core prohibitions of the Fourth Amendment, but only an exception to the federal exclusionary rule, the mandate of article 38.23 to exclude any evidence obtained in violation of the United States Constitution does not, even by necessary implication, require a decision about whether questioned evidence would inevitably have been discovered anyway.” Garcia, 829 S.W.2d at 798.
This is more or less the extent of Garcia’s “analysis.” Cases such as Bell and Vanderbilt contain at least as much “analysis” as this. More importantly, this Court decided Johnson after Garcia and Garcia is completely inconsistent with this Court’s more recent majority opinion in Johnson.
In addition, that this Court’s earlier majority opinions do not contain enough “analysis” to suit the majority in this case should be irrelevant. The point is that the inevitable discovery doctrine has been a part of this State’s jurisprudence since at least 1969. And, people have come to rely on this. It should require more than a poorly-reasoned
*280plurality decision to depart from this Court’s well-settled jurisprudence.
This brings us to Judge Baird’s concurring opinion. Judge Baird’s position is “we must presume the Legislature approved of Garcia’s interpretation of art. 38.23.” However, a plurality opinion decides only the particular dispute before the Court and is of no prece-dential value. See Vargas v. State, 838 S.W.2d 552, 554 (Tex.Cr.App.1992).
In Garcia, a four-judge plurality decided the federal inevitable discovery doctrine does not apply to Article 38.23(a). However, apparently another four-judge plurality in Garcia — Judges White, Miller, Campbell and Baird — would have decided the federal inevitable discovery doctrine is applicable to Article 38.23(a). So, the question becomes which plurality position in Garcia are we to presume the Legislature approved? Legislative inaction after Garcia has ho bearing on the question presented by this case, because a majority of this Court in Garcia did not agree that the federal inevitable discovery doctrine was inapplicable to Article 38.23(a). Moreover, given this Court’s numerous majority holdings over the years and the Legislature’s failure to overturn them, Judge Baird’s current position that this Court should now presume legislative approval of a single plurality opinion in Garcia is untenable.
Citing Grunsfeld v. State, 843 S.W.2d 521 (Tex.Cr.App.1992), Judge Baird claims the Legislature “does respond to plurality opinions.” However, a majority of this Court in Grunsfeld did agree that the 1989 legislative amendments to Article 37.07, V.A.C.C.P.,8 prohibited the admission of un'adjudicated extraneous offenses as part of a defendant’s “prior criminal record.” See Grunsfeld, 843 S.W.2d at 522-48. And, as Judge Baird correctly points out, the Legislature acted to amend Article 37.07 to reverse this Court’s majority holding in Grunsfeld.
In any event, Judge Baird’s point is we should presume legislative approval of a plurality opinion in Garcia because the Legislature has failed to act on it. Obviously, the Legislature did act in the Grunsfeld situation. Judge Baird’s Grunsfeld analogy to this case is inapposite. The point is that if this Court is to presume anything from the Legislature’s failure to act, as Judge Baird claims we should, then it makes more sense to presume legislative approval of this Court’s numerous majority holdings dating back to at least 1969 instead of a single plurality opinion. With respect, Judge Baird has yet to advance a principled reason for deviating from the position he took in Garcia.
The majority also holds “a direct appellate court must conduct a Rule 81(b)(2) harm analysis sua sponte, without the necessity of prompting by the State” in cases like this. While I agree with the majority’s ultimate disposition of this issue, I must disagree with their “analysis,” such as it is.
The majority cites Weatherford v. State, 828 S.W.2d 12, 13 (Tex.Cr.App.1992), to support their holding. However, Weatherford has nothing to do with the issue presented here. “Much more analysis is necessary before the Court may confidently say” that an intermediate appellate court must sua sponte conduct a Rule 81(b)(2) harm analysis in cases like this, especially since the majority in disposing of the “inevitable discovery” issue has just finished trashing twenty-five years of this Court’s jurisprudence because it does not contain enough “analysis.” See Autran v. State, 887 S.W.2d 31, 42 (Clinton, J., concurring).
Finally, appellant’s brief advances two arguments, supported by citation to various opinions of this Court, claiming the Court of Appeals did not err to decline to sua sponte conduct a Rule 81(b)(2) harm analysis. First, appellant argues that since this Court has held Rule 81(b)(2) places the burden on the State to prove an error did not contribute to the verdict or punishment, then the State has the burden to raise the issue in a timely manner. Second, appellant argues the State is trying to argue an issue it waived in the Court of Appeals “under the guise of a *281‘harmless error analysis.’” The majority opinion wholly fails to “analyze” these contentions. The majority’s lack of “analysis” on the Rule 81(b)(2) issue creates a risk that twenty-five years from now this Court could decline to give stare decisis effect to that part of the majority opinion addressing the Rule 81(b)(2) issue because it does not contain enough “analysis.”
With these comments, I join only that part of the Court’s judgment remanding the case to the Court of Appeals to conduct a Rule 81(b)(2) harm analysis. I dissent to everything else.
WHITE and KELLER, JJ„ join this concurring and dissenting opinion.. The majority also relies on the sole “exception” to Article 38.23(a) in support of its position that "according to the rules of statutory construction, where a statute contains an express exception, its terms must apply in all cases not excepted.” See Article 38.23(b), V.A.C.C.P., ("good faith exception”). The reader might find it ironic that this legislative "exception" to Article 38.23 came about as a result of another erroneous interpretation of Article 38.23 by this Court. See Garcia, 829 S.W.2d at 801-02 (Miller, J., concurring); Green v. State, 615 S.W.2d 700, 711-13 (Tex.Cr.App.1980) (McCormick, J., dissenting).
. For example, Article 38.23(a), in relevant part, provides that no evidence "obtained by an officer or other person" in violation of the law shall be admitted. Some argue the phrase "or other person” is intended to put the same restrictions upon private persons that the law places upon state actors. This is not the case. Sometimes private persons can obtain evidence in violation of the law such as when they are acting at the behest of law enforcement personnel to get a confession. See, e.g., Alvarado v. State, 853 S.W.2d 17 (Tex.Cr.App.1993); Cates v. State, 776 S.W.2d 170 (Tex.Cr.App.1989); Paez v. State, 681 S.W.2d 34 (Tex.Cr.App.1984). Therefore, a defendant presents nothing for review when he complains that a private person obtained evidence in violation of Article 38.23(a), or that certain evidence was obtained in violation of Article 38.23(a). However, this is the species of statutory construction the majority seems to do in this case.
. Brown v. Illinois, 422 U.S. 590, 601-02, 95 S.Ct. 2254, 2261, 45 L.Ed.2d 416 (1975) ("attenuation” doctrine).
. See Nix, 467 U.S. at 441-43, 104 S.Ct. at 2508 ("independent source" doctrine).
. Nix, 467 U.S. at 447-49, 104 S.Ct. at 2511 ("inevitable discovery” doctrine).
.The rationale of these doctrines is that admitting the evidence strikes a proper balance between society’s interest in deterring police misconduct and the public interest in having juries receive all probative evidence of a crime. See Nix, 467 U.S. at 443-45, 104 S.Ct. at 2509; Brown, 422 U.S. at 590, 95 S.Ct. at 2254.
. Leon addressed the "good faith exception” to the federal exclusionary rule, which, of course, is not involved in this case.
. These amendments came about as a result of a prior erroneous interpretation of Article 37.07 by this Court. See Grunsfeld, 843 S.W.2d at 523; Murphy v. State, 777 S.W.2d 44 (Tex.Cr.App.1989) (op. on reh’g).