Miranda Rights and the Fifth Amendment: When You Must Speak to Stay Silent

Last updated 1 day ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.

Police read a suspect his rights, then questioned him for nearly three hours while he stayed largely silent. Near the end, he answered yes to a question about praying for forgiveness for a shooting. His attempt to say little had not brought the questioning to a halt.

In its 2010 decision Berghuis v. Thompkins, the Supreme Court said he had not clearly claimed his right to end the questioning and had knowingly and voluntarily waived his right by answering. The Fifth Amendment protects against being forced to give incriminating testimony, but ending an interview in police custody generally requires clearly saying that the questioning should stop.

Why silence did not stop the questioning

The Supreme Court’s 2010 Thompkins decision permits an implied waiver when police establish that warnings were given and understood and the resulting statement was uncoerced. Before the interview, a detective read four warnings aloud, and the suspect read another warning aloud but declined to sign the form. Refusing a signature did not settle whether he understood the warning or later chose to answer.

The detective’s question about forgiveness came about two hours and 45 minutes into the interrogation. The suspect answered yes, looked away, and refused to make a written confession before questioning ended about 15 minutes later. The consequential answer was spoken; a signed confession was unnecessary to create the waiver issue.

The majority explained that requiring an unambiguous invocation gives officers an objective rule when a suspect’s intentions are unclear. Justice Sotomayor’s dissent, joined by three other justices, objected that a suspect had to speak precisely to preserve silence and that the waiver ruling weakened Miranda’s protections. The disagreement concerned how police should recognize the exercise of an existing right, not whether the Fifth Amendment contains that right.

The Fifth Amendment protects more than defendants

The Fifth Amendment’s Self-Incrimination Clause says no person shall be compelled in a criminal case to be a witness against himself. Self-incrimination means providing testimony that can help establish the speaker’s criminal liability. The protection targets compelled testimony, not every uncomfortable question or every kind of evidence.

The Supreme Court applies this protection to states through the Fourteenth Amendment and recognizes it in other official proceedings when answers could incriminate the speaker in a future criminal case. The constitutional text speaks of a person rather than making citizenship the test. In Ohio v. Reiner, the Supreme Court explained that even truthful answers from an innocent witness can supply incriminating evidence. Claiming the privilege is therefore not an admission of guilt.

In its Miranda opinion, the Supreme Court traced resistance to compelled self-accusation to John Lilburn’s refusal to take the Star Chamber oath in 1637. That history concerns forcing someone to supply an accusation against himself, rather than giving investigators every answer they seek.

For federal felony charges, the grand-jury protection generally requires an indictment before trial unless the accused waives that right. The grand jury evaluates whether evidence supports probable cause to believe the accused committed a crime, issuing the formal charge if it finds sufficient evidence. The constitutional clause contains an exception for specified military cases.

Double jeopardy protects against being put in jeopardy twice for the same offense. The amendment requires due process before government deprives a person of life, liberty or property, and just compensation when it takes private property for public use. For a criminal defendant, due process guarantees a fair and impartial trial.

When police must give Miranda warnings

The Supreme Court decided Miranda v. Arizona on June 13, 1966 after considering four cases involving isolated custodial questioning without full, effective warnings. Miranda requires safeguards before custodial interrogation because the Court saw the pressure of that setting as undermining a person’s freedom to choose whether to speak. Two conditions do the work: custody and interrogation. An arrest is not itself a promise that warnings must accompany every word an officer says.

The required warnings cover the right to remain silent, use of statements in court, a lawyer’s presence during questioning, and an appointed lawyer before questioning if the person cannot afford one and wants one. The familiar warning about anything said being used in court describes the possible consequence of choosing to answer. Hearing the warning does not automatically stop the interview; the next issue is whether the person invokes or waives the right.

Courts begin by asking whether a reasonable person could end the questioning and leave, considering the location, duration, statements and physical restraints. For Miranda custody, restricting freedom of movement is not enough by itself; the setting must also create the kind of coercive pressure addressed by Miranda. Ordinary temporary traffic stops generally do not count as Miranda custody under Berkemer v. McCarty. That distinction explains why being temporarily unable to drive away does not alone resolve whether warnings are required. A child’s age belongs in the custody analysis when officers know it or it would be objectively apparent to a reasonable officer.

Under Rhode Island v. Innis, interrogation includes direct questions and other police words or actions they should know are reasonably likely to elicit an incriminating response, apart from ordinary arrest and custody practices. An officer need not phrase something as a question for it to have the legal effect of interrogation. Miranda does not bar freely volunteered statements.

In New York v. Quarles, the safety question concerned locating a missing gun, and the officer gave warnings before continuing with investigative questions after securing it. Quarles recognizes a public-safety exception allowing certain unwarned answers to be admitted in evidence. That exception does not turn an actually coerced statement into a voluntary one.

How to invoke silence or ask for counsel

The Thompkins majority said the suspect would have invoked his right by stating that he wanted to remain silent or did not want to talk with police. State the decision directly if the aim is to stop a custodial interview. A statement that leaves officers guessing does not communicate that decision as clearly as saying the questioning should stop.

The Salinas v. Texas plurality explained that invoking the privilege requires no ritualistic formula. A long phrase beginning with advice of counsel is therefore not necessary to make the request clear. The operative message is the refusal to answer, not a recital of a Supreme Court ruling or a claim that the justices were unanimous.

In Davis v. United States, after a knowing and voluntary waiver, the suspect’s statement “Maybe I should talk to a lawyer” did not require officers to stop questioning. The word maybe left the statement short of a clear request. Under Edwards v. Arizona, a custodial request to deal with police through counsel bars police-initiated interrogation without counsel, unless the suspect initiates further communication. Asking for counsel thus changes the rules for police reapproaching the suspect.

Under Maryland v. Shatzer, the Edwards restriction following a request for counsel ends after a break in Miranda custody lasting 14 days, when a new valid waiver can be obtained. A break in custody is essential to that rule; merely waiting while the suspect remains in uninterrupted Miranda custody is different.

A clear invocation of silence requires officers to honor the right to cut off questioning. Michigan v. Mosley permitted later questioning where officers immediately stopped the first interview, waited a significant interval, gave fresh warnings and asked about a different crime. Invoking silence does not create an absolute ban on every later interview. It also does not give officers permission to ignore the initial refusal and keep pressing for an answer.

A valid waiver must be voluntary and made with awareness of the right being abandoned and the consequences of abandoning it. Miranda allows a person who initially answers questions to invoke silence later. An initial choice to speak need not be treated as a commitment to finish the interview.

The Miranda counsel safeguard comes from the Fifth Amendment, while the Sixth Amendment counsel right begins with formal criminal proceedings and applies to the charged offense. A lawyer’s role in an interrogation should therefore not be explained solely by whether charges have already been filed.

Outside custody, silence follows different rules

Salinas involved a voluntary police interview without custody or Miranda warnings, during which the suspect answered questions but fell silent when asked about a possible ballistics match to his shotgun. Prosecutors used that silence as evidence of guilt at his murder trial, and the Supreme Court affirmed the judgment on June 17, 2013. The evidence at issue was his silence, rather than a later answer like the one in Thompkins.

Justice Alito’s opinion, joined by Chief Justice Roberts and Justice Kennedy, rejected the Fifth Amendment claim because the suspect had not expressly invoked the privilege. Justices Thomas and Scalia concurred separately in the judgment, and four justices dissented. The result was an affirmed judgment, but Alito’s invocation reasoning had a plurality rather than a majority of the justices.

Alito’s opinion left undecided whether prosecutors could use a person’s expressly asserted privilege during a noncustodial interview. Clear invocation should not be described as a nationwide guarantee that every noncustodial silence will be excluded from evidence. The practical warning from Salinas is narrower: simply falling quiet during a voluntary interview may leave the privilege unclaimed.

The Salinas plurality recognized that a criminal defendant need not take the stand to assert the privilege at his own trial and that governmental coercion can excuse a failure to invoke. Those exceptions are why the instruction to speak clearly cannot be turned into a rule that everyone must announce the privilege in every setting.

Hiibel v. Sixth Judicial District Court upheld application of Nevada’s stop-and-identify law and noted that such laws vary among states. The Supreme Court found no reasonable danger that stating a name would incriminate the suspect in that case. The privilege is not a blanket exemption from identification requirements; the applicable law and a genuine risk of self-incrimination matter.

In federal matters, 18 U.S.C. § 1001 prohibits knowingly and willfully making materially false statements within federal government jurisdiction. A false statement is material if it has a natural tendency to influence, or is capable of influencing, the decision of the body receiving it. Declining to supply an incriminating answer and inventing an answer are legally different choices.

What a Miranda violation changes in court

Miranda ordinarily prevents the prosecution from using statements obtained through custodial interrogation without the required safeguards. Excluding a statement concerns evidence in the case, rather than automatically canceling an arrest or dismissing every charge. Whether the prosecution can proceed depends on what admissible evidence remains.

Harris v. New York allows a legally trustworthy statement excluded from the prosecution’s case in chief to be used to challenge the defendant’s trial testimony. The case in chief is the prosecution’s presentation establishing the charge; impeachment challenges a witness’s credibility. United States v. Patane does not require exclusion of physical evidence derived from an unwarned but voluntary statement. An omitted warning and actual coercion must therefore be kept separate when describing the remedy.

Doyle v. Ohio bars using a defendant’s post-arrest, post-Miranda silence to impeach a later explanation at trial. That protection follows warnings assuring a right to silence, a different setting from the unwarned voluntary interview in Salinas.

Vega v. Tekoh holds that violating Miranda rules does not itself support a federal civil-rights damages claim under section 1983. A rule excluding evidence and a right to recover damages are distinct forms of protection.

The Supreme Court decided McCarthy v. Hernandez on June 22, 2026. It held that no clearly established federal law required the Miranda-related jury instruction at issue and reversed the Second Circuit’s grant of federal habeas relief, returning the case for further proceedings. Federal habeas review is a challenge to a conviction in federal court; the ruling limited that remedy rather than removing the warning or invocation rules. The Supreme Court docket records issuance of judgment on July 24, 2026 following the reversal and remand.

Testimony, physical evidence and the limits of silence

Schmerber v. California distinguishes compelled testimonial or communicative evidence from the blood sample involved in that case. The right concerns being made to communicate incriminating information; it does not make every physical examination a Fifth Amendment violation.

Under Hoffman v. United States, the privilege reaches answers that could supply a link in the chain of prosecution evidence when the witness reasonably fears incrimination. The court decides whether the asserted danger justifies refusing to answer. A witness need not wait for a question asking for a complete confession, but merely declaring that an answer is incriminating does not settle the issue.

Kastigar v. United States allows compelled testimony with immunity against using that testimony or evidence derived from it in a later prosecution. The prosecution must prove that evidence it later uses came from a legitimate source wholly independent of the compelled testimony. This immunity removes the incriminating use of the compelled answers; it is not a promise that the witness can never be prosecuted.

Baxter v. Palmigiano permits adverse inferences in civil cases when a party refuses to testify in response to probative evidence, while a criminal jury cannot be invited to treat a defendant’s silence as substantive evidence of guilt. An adverse inference means drawing an unfavorable conclusion, so silence can carry a consequence outside the criminal trial even when an answer risks incrimination.

The interview that led to Thompkins turned on a difference the warning alone did not resolve: declining to say much was not the same as clearly asking for questioning to end. Protecting silence begins with identifying the setting and communicating the decision the law needs officers to recognize.