*769 OPINION

Appellant John Tuy Pham was convicted by a jury of the offense of murder, and was assessed a punishment of life imprisonment in the TDCJ Institutional Division. Appellant Chance Derrick Gonzales pled guilty to the offense of murder in the 208th District Court of Harris County, Texas. He was sentenced to 45 years’ confinement in the Texas Department of Criminal Justice (“TDCJ”) institutional division pursuant to a plea bargain.

We consolidated Appellants’ cases, as the same issues are raised in each of their appeals, and both cases are being appealed from the First Court of Appeals in Harris County. We will affirm the judgment of the court of appeals in both cases.

I. Facts

Appellant Pham was sixteen years old at the time he became a suspect in the drive-by shooting that resulted in the death of the victim, Dung Van Ha. At 2:35 p.m. on September 9, 1998, Houston police officers arrested Appellant at Clear Brook High School. At 3:35 p.m., a magistrate gave Appellant legal warnings as required by section 51.095 of the Texas Family Code. Subsequently, Appellant was taken to the police station and questioned by an investigator. Appellant admitted to his involvement in the shooting at approximately 4:38 p.m. He was then taken to a juvenile facility for processing by the police officers. An officer from this processing facility first notified Appellant’s family at about 8:15 p.m. when he spoke to Appellant’s sister. No one from the police department spoke with Appellant’s parents until 9:50 p.m. Appellant’s parents did not come to see him until the following day.

Appellant Gonzales was arrested in connection with a shooting death during a robbery at a convenience store when he was 15 years old. Appellant was identified by two witnesses who confirmed that he shot the victim while attempting to steal beer from the store for a gang party. Police arrested Appellant at a party, sometime between midnight and 1:30 a.m. Before police officers took Gonzales to a juvenile processing office at approximately 2:30 a.m., the officers made a stop at a sheriff’s station where they left him for 20-30 minutes so they could pick up a surveillance tape from a convenience store. This tape showed Appellant committing a similar type of robbery the same night. Appellant was given his Miranda warnings in the car on the way to the processing facility. The officer’s then took Appellant to a municipal judge at 3:35 a.m., where he was given the warnings required by Texas Family Code section 51.095. It was in the judge’s chambers with the police officers that Appellant Gonzales then gave his written statement. The arresting officer testified that he did not notify Appellant’s parents that their son had been arrested. Appellant Gonzales’ parents did not know that he had been arrested until he was processed into the juvenile arresting facility five to six hours after he was arrested, and after he gave his statement to police.

II. Procedural History: Pham

Appellant Pham originally appealed his conviction to the First Court of Appeals. The court of appeals in Pham v. State1 (Pham I), reversed the conviction based on the conclusion that the trial court erred in admitting Appellant’s statement. The court held the statement inadmissible due to the State’s violation of Texas Family Code § 52.02(b)2, and remanded the case *770to the trial court. The State subsequently-filed a petition for discretionary review with this court, which was granted. This court vacated Pham I in Pham v. State3 (Pham IT), and remanded the case to the court of appeals in light of the decision in Gonzales v. State (Gonzales II).4 Upon reconsideration, the court of appeals decided Pham v. State (Pham III)5, on November 26, 2003. In this decision, the court held that the trial court had properly admitted Appellant’s confession, and that Appellant was not entitled to a jury instruction on the admissibility of his confession. Appellant subsequently filed a petition for discretionary review to this Court.

III. Procedural History: Gonzales

Appellant Gonzales appealed to the First Court of Appeals from a plea of guilty. The court of appeals handed down its original decision (Gonzales I) on November 4, 1999, holding that Appellant’s confession was inadmissible because the State had not met its burden of proving that Texas Family Code § 52.02(b) had not been violated.6 We granted the State’s petition for discretionary review, and on February 13, 2002, vacated the decision of the court of appeals (Gonzales II). This Court held in Gonzales II that Appellant’s statement was not inadmissible merely because the State failed to follow the requirements of Texas Family Code § 52.02(b). We held that an exclusionary analysis under Texas Code of Criminal Procedure Article 38.23, including a causal connection analysis, had to be undertaken by the court, and had not been.7 Therefore the case was remanded. Upon reconsideration of the case, the court of appeals handed down an opinion on November 26, 2003 (Gonzales III),8 holding that, in an article 38.23 exclusionary analysis, the initial burden was on the appellant to demonstrate a violation of a statutory requirement as well as a causal connection between the violation and his ensuing confession.9 Finding that Appellant Gonzales presented no evidence to demonstrate this causal connection, the court of appeals affirmed the judgment of the trial court and held Appellant’s statement admissible.10 Appellant then filed a petition for discretionary review with this Court.

IV. First Court of Appeals’ Decisions and Grounds for Review

The court of appeals held in Pham III that the State did obtain Appellant Pham’s confession in violation of Texas Family Code section 52.02(b), however, in light of our opinion in Gonzales II, the court also held that the confession was not automatically inadmissible without first conducting an analysis under Texas Code of Criminal Procedure article 38.23 to determine *771whether there was a causal connection between the illegal conduct and the acquisition of the evidence.11 The court of appeals interpreted our remand in Gonzales II as implying that an attenuation-of-taint analysis was an insufficient analysis to satisfy the requirement that a causal connection be established. Therefore the court of appeals determined that the causal connection analysis required by this Court in Gonzales II must be separate from an attenuation-of-taint analysis.12 The court of appeals further relied on our opinion in Roquemore v. State, 60 S.W.3d 862 (Tex.Crim.App. 2001) for this conclusion, where this Court first conducted a causal connection analysis and found a causal connection between the recovery of stolen property and the illegality of police conduct, but failed to conduct an attenuation-of-taint analysis because the State did not raise the argument.13

The court of appeals next addressed the issue of which party has the burden of proving a causal connection, noting that no direct authority establishes who takes on this burden.14 The court pointed out that the issue is not who has the ultimate burden of persuasion as to the admissibility of a confession, as this clearly rests with the State in an attenuation-of-taint analysis, but rather who has the initial burden of production of proof of a causal connection between the violation of the statute and the ensuing confession. The court ultimately held that the defendant should shoulder this initial burden of proof, analogizing the situation to that of a violation of the statute requiring that a defendant promptly be taken before a magistrate. The court also based its conclusion upon more practical considerations, such as the fact that it is more “reasonable to place the burden on the defendant to produce evidence to which only the defendant has access.”15 The court concluded that the State has the burden to demonstrate attenuation of the taint once the defendant has satisfied the initial burden of production of some evidence that shows a causal connection between the police illegality and the recovery of the evidence.16 The court found that Appellant Pham produced no evidence of a causal connection, and therefore did not meet his burden, making it unnecessary to conduct an attenuation-of-taint analysis, and rendering his confession admissible.

In Appellant Gonzales’ case, the court of appeals, citing its decision in Pham III (discussed above), again held that when there is a violation of section 52.02(b) of the Texas Family Code, the initial burden is on the defendant to demonstrate a violation of the statutory requirement and a causal connection between that violation and the ensuing confession. Again, the court held that Appellant Gonzales produced no evidence that would demonstrate a causal connection between the police violation of the Family Code and his ensuing confession, therefore his statement was admissible.

Both Appellants argue that a causal-connection analysis is part and parcel of an attenuation-of-taint analysis, and that the burden of proof falls on the State to negate the causal connection between the violation and the confessions of the two Appellants. Appellants claim that the court of appeals ignored cases stating that when voluntari*772ness is an issue, the burden of proof for admissibility of confessions lies with the State, and argue that this Court did not create a separate causal connection analysis in Gonzales II.

The State argues in these cases that the court of appeals did not err in placing the burden of production on Appellants to offer evidence to prove that violations of the Family Code occurred, and that they were causally connected to the ensuing confessions. The State contends that Appellants failed to meet the burden of production, i.e. they did not produce any evidence that their statements were obtained in violation of any laws. Additionally, the State argues that causal connection and attenuation-of-taint are two separate analyses by which a court will determine whether evidence was obtained in violation of the law as set out in article 38.23.

We granted review in both of these cases to clarify the causal connection analysis which must be undertaken in article 38.23 suppression of evidence claims. In Appellant Pham’s case, we granted the following two grounds for review: 1) did the court of appeals err in holding that causal connection and attenuation-of-taint constitute separate analyses, and 2) did the court of appeals err by requiring Appellant to prove a causal connection between the violation of section 52.02(b) of the Texas Family Code and Appellant’s confession? In Appellant Gonzales’ case, we granted the following ground for review: did the court of appeals adopt the wrong standard by which a causal connection must be established under article 38.23 to justify suppression of evidence seized in violation of the Family Code?

IV. Law

Neither party argues with the settled law that the burden of proof is initially on the defendant to raise the exclusionary issue by producing evidence of a statutory violation, and that this burden then shifts to the State to prove compliance. The main issue we face here is whether the defendant has the burden of producing evidence that shows the violation is connected to the obtaining of the evidence sought to be suppressed.

Appellants argue that a causal connection analysis cannot be separated from an attenuation analysis, and thus, since it is well settled that the State bears the burden of proving attenuation of taint, the burden is also upon it to produce evidence of a causal connection. We disagree. If we follow this circular argument, anytime an appellant asserted a statutory violation of Family Code section 52.02(b), a court would immediately have to conduct an attenuation-of-taint analysis because it is part of the causal connection analysis. This would further the assumption that once Appellant shows a violation of the statute under 52.02(b), the evidence is automatically assumed inadmissible unless the State demonstrates attenuation-of-taint, an assumption which we expressly rejected in Gonzales II.

We have held that the State may make an attenuation-of-taint argument which is included under an Article 38.23 analysis.17 However, this argument is discussed by a court only if the State raises it.18 Analysis of causal connection and attenuation-of-taint are not the same. An attenuation-of-taint analysis is not always required and therefore need not always be conducted. We have expressly held that a causal connection between a violation of section *77352.02(b) and the obtaining of evidence must be shown before the evidence is rendered inadmissible.19 If there is no causal connection shown in the first place, there is no reason for the State to argue that the taint of the violation is so far removed that the causal connection is broken. Attenuation-of-taint breaks this connection. . It does not negate the existence of the causal connection.

If the defendant produces evidence that there is a causal connection, the State may either try to disprove this causal evidence, i.e. disproving that there is a causal connection in existence at all, or, the State may make an attenuation-of-taint argument. Attenuation-of-taint is evaluated under the four-step Brown v. Illinois20 analysis, in which the State may argue that although the defendant has demonstrated evidence of a causal connection, the taint of the violation was so far removed from the obtaining of the evidence that the causal chain the defendant demonstrated is in fact broken. In short, without first establishing that there is a causal connection between the violation and the obtaining of the evidence, there can be nothing for the State to assert has been broken through the attenuation-of-taint factors. Thus, we uphold the court of appeals’ conclusion that a causal connection analysis regarding Family Code section 52.02(b), as required by this Court in Gonzales II before evidence may be deemed inadmissible, is separate from an attenuation-of-taint analysis, which may be used by the State to rebut a defendant’s causal connection argument.

We also uphold the court of appeals’ distribution of the burdens of proof in both of these cases. We have long held that “the burden of persuasion is properly and permanently placed upon the shoulders of the moving party. When a criminal defendant claims the right to protection under an exclusionary rule of evidence, it is his task to prove his case.”21 In Russell v. State, we again cited this holding. Recognizing that this analysis was used for federal claims of illegal search and seizure under the Fourth Amendment, and because Texas statutory law is silent as to how the burden of proof is distributed on a motion to suppress, this Court adopted some of the rules followed by federal courts in distributing burdens of proof.22

Appellant argues that the burden should be on the State to show the causal connection, and analogizes this situation to the admissibility of confessions when issues of voluntariness are raised. Although the burden is on the State to prove that a defendant’s confession was voluntary once the issue has been raised, that situation may be distinguished from the statutory violation of the Texas Family Code we have here. All a defendant must do on a claim of involuntary confession is to demonstrate there is a cognizable violation, and the confession is immediately presumed inadmissible unless the State can prove by a preponderance of the evidence that it was made voluntarily.23 There is no *774requirement that a defendant establish any causal connection between the illegal conduct and the ensuing confession; the defendant simply must raise the voluntariness issue. In this case, it is not enough for the defendant to merely establish a violation. Under Texas case law, it is required that a causal connection be established, and we hold that the defendant, as the moving party wishing to exclude the evidence, is responsible for the burden of proving this connection.

Thus, the court of appeals correctly held that the burden is on the defendant, as the moving party in a motion to suppress evidence obtained in violation of the law under Art. 88.23, to produce evidence demonstrating the causal connection which this court required in Gonzales II. The burden then shifts to the State to either disprove the evidence the defendant has produced, or bring an attenuation-of-taint argument to demonstrate that the causal chain asserted by the defendant was in fact broken.

Y. Conclusion

In the case of Appellant Pham, we hold that the court of appeals conducted the appropriate analysis, and that the court did not err in requiring Appellant Pham to produce evidence to prove a causal connection between the violation of section 52.02(b) of the Texas Family Code and his ensuing confession. The decision of the court of appeals is affirmed.

In the case of Appellant Gonzales, we hold that the court of appeals adopted the correct standard by which a causal connection must be established under Art. 38.28 to justify the suppression of evidence seized in violation of the Family Code. The decision of the court of appeals is affirmed.

dissenting.

The Court concludes that an attenuation-of-taint analysis is different from a causal-connection analysis. Because the only support for this conclusion is a troublesome substantive footnote, and the conclusion is in conflict with Supreme Court jurisprudence, I dissent.

The United States Supreme Court introduced the attenuation concept over 65 years ago in Nardone v. United States.1 There the Court explained that “[sjophisti-cated argument may prove a causal connection between information obtained through illicit wire-tapping and the Government’s proof,” but “[a]s a matter of good sense ... such connection may have become so attenuated as to dissipate the taint.”2 In other words, an analysis of the “causal connection” between the violation of the law and the discovery of the evidence is part and parcel of the “attenuation of taint” analysis. They are one and the same. If the evidence is obtained as a direct result of the illegal action, then there is a causal connection between the two. If the discovery of evidence is a remote result of the illegal action, then there is an attenuation of the taint. The two events are either causally connected or they are attenuated. The doctrines are two sides of the same coin.

We recognized this in Bell v. State3 when we addressed the “taint attenuation to determine whether the causal chain between the illegal arresi/first confession and subsequent events was broken.” Later, in *775State v. Daugherty,4 we discussed “what Article 38.23(a) means by ‘evidence obtained in violation of the law.”5 We explained that the “ordinary meaning” of “obtained” accommodated the attenuation doctrine because “depending on how removed the actual attainment of the evidence is from the illegality, the ordinary person would not consider that evidence to have been ‘obtained’ by that illegality.”6 Even when evidence would not have been obtained but for an illegality, we said, if the eventual discovery of the evidence was far removed from the illegality then the evidence would not need to be excluded “because the ordinary meaning of ‘obtained’ does not extend to such a remote, or ‘attenuated,’ causal relationship.”7 So we, too, have recognized that the attenuation doctrine encompasses a causal connection analysis.

But a few years ago in Roquemore v. State,8 we dropped a mischievous footnote. After analyzing whether there was a causal connection between the illegality and the discovery of the evidence, we then said that “we need not do an attenuation of taint analysis because the State did not raise the argument.”9 With this footnote, we implied that an attenuation analysis is separate from a causal-connection analysis. This is not true, of course, and the footnote itself is dicta,10 but those facts have not stopped this footnote from having a negative impact on the law.

The footnote has certainly confused the First Court of Appeals. On original submission in Pham, the First Court analyzed whether there was a causal connection between the illegality and the discovery of evidence, and the court concluded that there was. The court said the following:

If the arresting officers had promptly notified appellant’s parents of his arrest approximately two hours before his confession, there would have been time for them to get to the juvenile processing office at 1200 Travis before the confession. As in Comer,11 we cannot say with any degree of confidence that if appellant had access to his parents or his attorney, he would still have chosen to confess to the crime.12

Despite the court’s analysis, we granted the State’s petition and remanded the case for the Court of Appeals to consider the issue again in light of Gonzales.13 The First Court was somewhat perplexed, saying:

In our original opinion, we conducted a taint-attenuation analysis, tracking Comer. This was apparently an insufficient analysis to avoid a remand for reconsideration in light of Gonzales,14

The court determined from our remand that, “at the outset,” it should “determine whether causal connection and attenuation *776of the taint constitute separate analy-ses.” 15 It then found that the case “most clearly demonstrating separate analyses for causal connection and attenuation of the taint, and the order in which they are to be undertaken, is Roquemore.” 16 The court concluded that, “[biased on Roque-more, and based on the fact that our attenuation-of-the-taint analysis in Pham I was found inadequate to satisfy a Gonzales causal-connection analysis, we conclude that there are separate analyses for causal connection and attenuation of the.taint.”17

But the footnote in Roquemore was misguided and misleading. Additionally, a remand from this Court in light of another case does not necessarily mean that the appellate court was wrong in its original analysis. It means only that we want the court to consider the issue again and determine if its analysis needs to be altered in light of the new opinion. In this case, there was no need for the Court of Appeals to alter its analysis in Pham, I. It had already done an attenuation/causal connection analysis, which is all that Gonzales required.

In other jurisdictions, it is a given that an attenuation-of-the-taint analysis and a causal-connection analysis are one and the same. The United States Court of Appeals for the Tenth Circuit says that “the government must prove, from the totality of the circumstances, a sufficient attenuation or break in the causal connection between the illegal detention and the consent.” 18 The Eleventh Circuit says that “challenged evidence will be admissible under the ‘attenuation’ doctrine if the causal connection between the constitutional violation and the discovery of the evidence has become so attenuated as to dissipate the taint.”19 The Fifth Circuit refers to the analysis as the “causal-taint attenuation” analysis.20 State courts have held similarly.21

*777Once we acknowledge that the attenuation analysis and the causal-connection analysis are one and the same, the question regarding burden of proof is answered. In Brown v. Illinois, the United States Supreme Court set forth some of the factors to be considered in a causal connection/attenuation analysis, then added that “the burden of showing admissibility rests, of course, on the prosecution.”22 We have also recognized that “the burden of showing attenuation and thus admissibility rests with the prosecution.”23

In both Pham and Gonzales, the appellate court erred in analyzing the causal connection separately from the attenuation of the taint. The court also erred in placing the burden of showing a causal connection on the appellants. I would reverse both judgments. Because the majority holds otherwise, I dissent.