Concept
What Counts as Testimonial
Confrontation and presence
What Counts as Testimonial
The first question on any Confrontation Clause fact pattern is whether the out-of-court statement is testimonial, because the two conditions from the preceding parts, unavailability and a prior opportunity to cross-examine, apply only to statements of that kind. The text supplies the reason. The Sixth Amendment gives the accused the right "to be confronted with the witnesses against him."1 In Crawford v. Washington the Court read witnesses to mean those who bear testimony, and testimony to mean a solemn declaration made to establish or prove some fact.2 On that reading, an accuser who makes a formal statement to government officers bears testimony, while a person who makes a casual remark to an acquaintance does not.3 Declarants in the second group are therefore not witnesses against the accused in the constitutional sense.
Davis v. Washington turned that reading into a holding: the Confrontation Clause applies only to testimonial hearsay.4 Only testimonial statements cause the declarant to be a witness, and it is the testimonial character of a statement that separates it from other hearsay, which stays subject to the traditional limits of hearsay law but falls outside the Clause.5 The Court later confirmed in Whorton v. Bockting that nontestimonial hearsay cannot violate the Clause.6 The Clause thus works as a testimonial gate: a statement that clears it faces the unavailability and cross-examination conditions, and a statement that does not is governed by the rules of evidence alone.7 Students lose this issue when they treat any hearsay offered against the accused as a confrontation problem, since a statement can be damaging, unreliable, and inadmissible hearsay and still raise no constitutional question if it is nontestimonial.
The gate sits where it does because of the danger the Clause was written to meet. The historical record shows that the principal evil was the civil-law practice of examining witnesses privately and then using the written results against the accused, and government involvement in producing testimonial evidence recreates the same risk.8 The trial of Sir Walter Raleigh is the paradigm, because the prosecution there used affidavits from eyewitnesses that were prepared with its help.9 Two features of such practices carry the danger. Officers who take a statement choose the questions, decide what is recorded, and shape the account into a form useful for conviction, and a declarant who knows the account will be used against someone has a reason to color it. The defendant, absent from the room, cannot probe either problem. Statements made to a friend or a neighbor have neither feature, because no one is assembling evidence, so ordinary hearsay law can govern them without constitutional supervision.
Two applications show how the listener and the purpose move the answer. In Crawford, Sylvia's statement was recorded in custody after Miranda warnings, in response to structured questioning by detectives, and it is testimonial even under a narrow definition of the term.10 A standard illustration holds the declarant constant and changes only the listener. Ellen, who took part with Harvey in an embezzlement, describes the scheme to police and implicates him, and she is unavailable at his trial; the statement is testimonial, and its admission violates the Clause.11 If Ellen instead confides the same account to her best friend Ben, no police are involved and nothing suggests her primary motive was to have the statement used in a prosecution, so the statement is nontestimonial and Ben's testimony about it raises no confrontation problem.12 Whether Ben may repeat it then depends on the hearsay rules, where the exception for statements against penal interest is the natural candidate.13
The Court has never produced a complete definition of testimonial. Crawford identified several formulations that share a common nucleus without selecting one, and it said the term covers at a minimum prior testimony at a preliminary hearing, before a grand jury, or at a former trial, and police interrogations.14 The features that run through the formulations are that the statement is aimed at proving a fact for a criminal prosecution, has some structure or formality, and is directed at or prepared by government officials.15 Later decisions have moved the boundary somewhat as the Court addressed police emergencies, statements to listeners who are not police, and forensic material.16 Because the answer depends on the kind of statement, it is best sorted in an order that matches the categories taught below.
Each of the next four parts takes one category through that sequence: formal statements and prior testimony that qualify under any definition, police questioning and the ongoing-emergency line, statements to listeners who are not police, and laboratory reports offered to prove the matter asserted.
Core Testimonial Statements
The gate question has easy cases and hard ones, and the law has been built outward from the easy ones. Crawford v. Washington declined to give a comprehensive definition of testimonial and left that task for another day, but it fixed a floor: whatever else the term covers, it applies at a minimum to prior testimony at a preliminary hearing, before a grand jury, or at a former trial, and to police interrogations.22 Call this the minimum core. These practices were chosen because they are the modern ones closest in kind to the ex parte examinations the Clause was written to stop. In each of them, government officers receive an account and produce a record while the accused is absent, and the record is ready for use against him.23 Because these statements sit at the center, a court can classify them as testimonial without settling any dispute about where the outer edge of the definition lies.
Depositions appear in the second formulation, a detail that matters later in this chapter when Rule 15 depositions are taught. The third formulation, the objective-witness standard, asks what a reasonable person in the declarant's position would expect to happen to the statement, rather than what the declarant privately intended. That framing has a practical reason: a test keyed to private intent would require proof of the state of mind of an absent declarant, the one person the defense cannot question. Formality is therefore one route into the core and not a precondition. Affidavits are formal, while statements made during on-the-street police questioning can be informal and still fall within the category.25 Which of those street-level statements actually qualify depends on a further test, taken up in the next part.
The facts of Crawford itself show the core operating. Sylvia Crawford's recorded statement was given in police custody in response to structured questioning, and the prosecution offered it to rebut her husband's claim of self-defense. The Court treated it as testimonial under any conceivable definition, because it was knowingly given to officers who were building a case.26 Prior testimony is the other anchor, and grand jury testimony is its clearest instance. It is prepared for use in a criminal prosecution, it is even called testimony, and the accused has no right to cross-examine at a grand jury, so admitting it against him when the witness is unavailable at trial violates the Clause no matter how reliable it appears.27 Testimony at a former trial or preliminary hearing sits in the same core, but it differs on the second condition, because the defendant may have had the chance to question the witness there. Prior confronted testimony offered when the declarant is unavailable at the current trial satisfies the Clause, and the unavailability requirement for such statements is enforced rigorously.28
The lunch-table version marks the opposite end of the line. Crawford contrasted an accuser who makes a formal statement to government officers with a person who makes a casual remark to an acquaintance, and only the first bears testimony.29 Casual statements to a neighbor are likely nontestimonial and fall beyond the reach of the Clause.{{cite:c403039c-858c-48a0-bd87-9d9e1652c7d1)}
Classifying a statement as testimonial opens the Crawford analysis and does not end it. Students lose points when they treat a testimonial label as a verdict of exclusion, because the label only triggers the two conditions. A testimonial statement of an absent declarant comes in if the declarant is unavailable and the defendant had a prior opportunity to cross-examine, as with confronted former-trial testimony.30 Once a statement lands in the core, the remaining work is to ask whether both conditions are met. A statement belongs outside the core only when its listener and circumstances point away from use in a prosecution.
Police Questioning and Emergencies
The minimum core taught above includes statements taken by police in the course of interrogations, but interrogation takes many forms. An officer's questions at a doorstep or a dispatcher's questions over the telephone look very different from a recorded station-house session. A rule is needed to sort those settings, and Davis v. Washington supplied it in two consolidated cases that together define the primary-purpose test.32 Statements made in response to police questioning are nontestimonial when the circumstances objectively indicate that the primary purpose of the interrogation is to enable police assistance to meet an ongoing emergency, and they are testimonial when the circumstances objectively indicate that there is no such emergency and that the primary purpose is to establish or prove past events potentially relevant to later criminal prosecution.33 The paired facts of the two cases show how a single moved variable changes the outcome, and a short checklist at the end turns the test into steps.
Each half of the test follows from the danger the Clause was written to meet. A caller who reports an emergency in progress is asking for rescue and speaks about events as they unfold. The operator who answers needs to learn what is happening so that responders can act, which means the exchange lacks the formal, prosecutorial character of testimony.34 Once the emergency has passed and officers are asking what happened, the setting changes. The questioner is investigating, the declarant is recounting history, and the answers work as an accusation that can be carried to trial in the declarant's absence. The test also looks to the objective circumstances and not to the private motives of the participants, and that choice has a practical basis. A declarant who does not appear at trial cannot be asked what the declarant meant to accomplish, and motives in a tense encounter are often mixed, so the court asks what the setting and the questions show the exchange was for.
The word interrogation in this test carries its everyday meaning. Davis recalled that Crawford used the term in its colloquial sense and declined to give it a technical definition.35 That separates it from the custodial interrogation that triggers warnings under Miranda v. Arizona, which is questioning initiated by law enforcement after a person has been taken into custody or deprived of freedom in a significant way, and which protects the privilege against self-incrimination.36 The two inquiries ask different questions. Custody and warnings concern the compulsion of a suspect, while the primary-purpose test concerns what an exchange between a declarant and a questioner was for, so a complaining witness who is not a suspect can give a testimonial statement, as the second of the two cases shows.37
In Davis, the victim called 911 and reported that her former boyfriend was beating her. The operator asked about the nature of the emergency, the identity of the assailant, and his current location, and then dispatched police.38 The Court held the recording nontestimonial. The victim was facing an ongoing emergency and describing events as they were actually happening. The operator needed to know whether the assailant was still present, whether he was armed, and whether he posed a continuing threat, and those answers let police respond to the present danger.39 Because the primary purpose was to obtain police assistance and not to create evidence for a later trial, the Clause did not bar the recording, and any remaining objection to it belonged to the ordinary hearsay rules.40
Hammon v. Indiana reached the opposite result on facts that began in a similar way.41 Police responded to a reported domestic disturbance and found the victim, who had a fresh injury, together with her husband. The officers separated the two, kept the husband in another room, and questioned the victim while she completed a battery affidavit stating that her husband had assaulted her.42 The Court held these statements testimonial. The husband was elsewhere, the victim was not in immediate danger, and the officer was acting in an investigative capacity to establish what had happened so that police could decide whether to arrest the husband and what charges to file.43 The victim did not appear at trial and the defendant had no earlier chance to cross-examine her, so admitting the affidavit and the officer's account of her statements violated the Clause.44
The battery affidavit in Hammon is the visible product of the third variable. A written, signed account is what an investigation produces, and it is ready for use at trial.
The emergency category is wider than a caller in the middle of an attack. An emergency can include the need to ensure that a victim is safe or to find out where a violent person is located, and in Michigan v. Bryant a statement obtained by police was held nontestimonial because its primary purpose was to get help during an ongoing emergency.45 How long an emergency lasts depends on its nature, since some dangers persist for a long time, and a later part of this chapter returns to that point when the listener is a teacher and not an officer.46 The test also cuts in the other direction for 911 itself. A call to an emergency operator can be testimonial if the operator's questions are directed at establishing the facts of a past crime for use in a later prosecution.47
Classifying a statement as nontestimonial ends the constitutional inquiry without admitting the statement. Davis separates testimonial statements from other hearsay that remains subject to the traditional limits of hearsay law, and in the Indiana courts the victim's oral statement in Hammon was treated as an excited utterance for state-law purposes, which is a different question from the constitutional one.48 The usual error on an exam stem is to let a label decide the result, for instance by treating every 911 call as nontestimonial or every answer given to an officer as testimonial. The setting does not decide the question. What decides it is whether the circumstances objectively show an emergency to be met or a past event to be proved.
Statements to Non-Police Listeners
The primary-purpose test above was built for questioning by police and by operators who work with them, but many accusations never reach an officer at all. A child tells a teacher who caused a bruise, a witness confides in a friend, an owner files a form with a government office. Whether the Clause reaches those statements depends on the same inquiry into what the exchange was for. The Clause can reach them, though far less often than it reaches the station-house interview.
The Clause reaches them in principle because the trigger is the character of the statement and not the listener's title. A declarant becomes a witness against the accused only by making a testimonial statement, and that character is what separates such a statement from other hearsay that only the traditional hearsay limits govern.53 Nothing in that formulation requires a badge on the listener. In Ohio v. Clark the Court accordingly declined to adopt a categorical rule that places statements to people who are not law enforcement officers outside the Clause.54 The rule for non-police listeners is therefore probabilistic. The listener's role makes a testimonial purpose less likely, and it does not remove the purpose inquiry.
The odds shift because of what each kind of listener is ordinarily doing. The Clause guards against accusations assembled for use at trial, and a police officer's job is to uncover and prosecute crime, so the questions an officer asks tend to be shaped by what a case needs and the answers tend to function as proof. A teacher, a neighbor, or a relative is not in that business. Their questions usually serve another need, such as keeping a child safe or understanding a friend's trouble, and that gives less reason to think the conversation was a way of building a case. When the listener is not principally charged with uncovering and prosecuting crime, the inquiry therefore comes out nontestimonial more often. It remains an inquiry into the objective circumstances, and the listener's job is only evidence on that question.
The facts of Clark show a non-police listener at work. Preschool teachers noticed injuries on a three-year-old boy and asked him who had caused them, and he named the defendant. When the prosecution offered what the boy told them, the question was whether his statements were testimonial. The Court held they were not, because the teachers' questions were primarily aimed at identifying and ending a threat to the child.55 The teachers also had a legal obligation to report suspected abuse to the authorities, and that obligation did not change the result.56 The threat they faced was that someone in the child's life was hurting him and might do so again, a danger that lasts until an adult acts on it, so questions meant to learn whether abuse was occurring fit the emergency rationale of the earlier cases.57
Confidences to a friend sit at the other end of the range. In one illustration, Ellen tells her best friend Ben that she and Harvey carried out an embezzlement, and she describes how each of them took part. Ellen is unavailable at Harvey's trial, and the prosecutor calls Ben to repeat her account. Her statements are nontestimonial because they were not made to police and nothing suggests a primary motive to have them used in a prosecution, and a statement that is not testimonial cannot violate the Clause.58 What remains is the hearsay question, and the same absence of any wish to curry favor with law enforcement helps support admission as a statement against interest.59
The listener's title does not settle the matter in the other direction either. In the prosecution of Esparza for smuggling marijuana hidden in a car registered to Diana Hernandez, the government did not call Hernandez. It offered a form she had filed with the California Department of Motor Vehicles stating that she had sold the car to Esparza six days before his arrest, and it used the form to prove that he owned the car and so knew about the drugs.60 A motor vehicles office is not a police agency, and the form was a standard notice of sale. Yet Hernandez filed it only after law enforcement notified her that the car had been seized for smuggling, and that notice asked an owner who no longer held the car to identify its new owner.61 The court of appeals held the statement testimonial, so Esparza had a right to confront her, and it vacated his conviction because the error was not harmless.62
Clark and the motor-vehicles filing involve two listeners who are not police and two opposite results, and the difference lies in purpose and in the government's hand in producing the statement. The teachers asked their questions to protect a child from a threat, and no prosecution was driving the conversation. The filing was set in motion by a law-enforcement seizure notice and then used by the prosecution to prove ownership and knowledge, which is the work testimony does at trial. Students lose this issue when they treat a non-police listener as an automatic exemption, because the Court refused a categorical rule and a civil office can receive a testimonial statement. The opposite error is to treat a listener's duty to report as a switch that makes a statement testimonial, since the teachers in Clark had such a duty and the statements were still nontestimonial.63 In both directions the listener's status is a clue to purpose, and the purpose decides.
Forensic Reports
Laboratory results do not fit the categories examined so far, because no officer questions a declarant and no conversation takes place. An analyst tests a sample, records what the test showed, and signs the result, and the prosecution later offers that writing to prove an element of the charge. The Clause still reaches the writing, since the analyst is asserting a fact that bears on guilt, and the Crawford rule decides the outcome: a testimonial statement of a witness who is absent from trial comes in only if the declarant is unavailable and the defendant had a prior opportunity to cross-examine.66 Hearsay law offers no escape here, because testimonial hearsay is barred even when the rules of evidence would otherwise admit it, unless the defendant had that chance to cross-examine.67 Two questions remain for forensic evidence: when a laboratory result counts as a testimonial statement, and who must appear to answer for it.
In Melendez-Diaz v. Massachusetts the defendant was charged with distributing and trafficking in cocaine, and the prosecution introduced the bags of white powder found in his possession along with three certificates of analysis from the state forensic laboratory reporting that the substance contained cocaine.68 The certificates were sworn before a notary, and under state law their sole purpose was to serve as prima facie evidence of the composition, quality, and net weight of what had been analyzed.69 A certificate made for that purpose does the work a witness does on direct examination, because it announces a fact the prosecution needs in order to convict, and the Court held that admitting the certificates violated the Sixth Amendment unless their authors were subjected to cross-examination.70 The result follows from the general definition of a testimonial statement, which covers statements that would lead an objective witness to believe they would be available for use against the accused at a later trial, and affidavits are the standard example.71 Written reports of forensic analysis fall in the same class, which is why a drug-analysis report offered as evidence of an element of the crime triggers the Clause.
The second decision closes a tempting workaround. In Bullcoming v. New Mexico the defendant was convicted of driving while intoxicated on the strength of a forensic report certifying a blood alcohol concentration of 0.21 grams per hundred milliliters, well above the legal limit.72 The State did not call the analyst who signed the certification. It called another analyst who had neither performed nor observed the test and was familiar only with the laboratory's general testing procedures.73 The Court declined to accept that substitute even though he was a knowledgeable representative of the laboratory, stating that "The accused's right is to be confronted with the analyst who made the certification."74
The insistence on the analyst, and not on someone who can describe the laboratory's methods, follows from what cross-examination is for. The questions that matter about a forensic result concern what happened to this sample on this occasion: whether vials were switched, whether the sample was contaminated, whether the test was competently run, and whether the result was accurately recorded. Only the person who handled the sample can be questioned on those points, and a witness who knows only the general procedures can recite them without being able to say whether they were followed here. The dissent in Williams v. Illinois, a decision taken up below, made the point with an earlier incident in which an analyst, once confronted, discovered and disclosed her own error, a mistake that probably would not have surfaced had the prosecutor simply admitted the report or asked a third party to present its findings.75
This rule should not be confused with an older right-to-counsel point. In United States v. Wade the Court treated scientific analysis of fingerprints, blood samples, clothing, and hair as stages at which the accused has no right to have counsel present, reasoning that the accused can still confront the government's case at trial by cross-examining the government's experts and presenting experts of his own.76 That holding concerns counsel's presence at testing, which is a different right from the one at issue here. What carries forward is the premise that trial cross-examination is how a defendant tests scientific evidence, and the later cases enforce that premise by requiring that the person cross-examined be the one who did the work.
Students lose this issue when they accept a knowledgeable substitute from the same laboratory as satisfying the Clause. The stand-in's competence and familiarity with the report are beside the point, because the Clause gives the accused the right to question the person who made the accusation, and the substitute made none.77
The limit on this rule comes from Williams, and it must be learned with candor about its weakness. Illinois prosecuted Sandy Williams for rape based in part on a DNA profile created in the laboratory of an outside firm, Cellmark, and the State offered the results through an expert witness who had no idea how Cellmark generated them, in a trial before a judge.78 The Court allowed the evidence but could not agree on why. Four Justices reasoned that the expert's summary of the Cellmark report was not offered for its truth and that the report was not the kind of statement that triggers the Clause. Five Justices rejected every aspect of that reasoning, and the fifth vote to admit the report came from a Justice who found it nontestimonial on a different ground that no other Justice joined.79
Because no rationale won a majority, Williams is best learned for its result and its limiting idea, and the plurality's reasoning should not be presented as the holding. The limiting idea is that a result offered for some purpose other than its truth makes no testimonial accusation for the accused to confront, so no Confrontation issue arises. How far that idea reaches when an expert recounts a report to support an opinion remains disputed, since the Court is widely divided on whether the bases of expert testimony offered for something other than their truth should be analyzed under the Clause.80 The dissent's objection shows the difficulty: an expert's account of a report supports the expert's opinion only if the account is true, so the factfinder can use it only by assessing its truth, and that is so whether the factfinder is a judge or a jury.81 An exam answer should therefore state the Melendez-Diaz and Bullcoming rule without hedging and treat Williams as a recognized limit with an unsettled basis.