Criminal Justice Section  


Criminal Justice Magazine
Summer 2005
Volume 20 Issue 2

"Testimonial" and the Formalistic Definition - The Case for an "Accusatorial" Fix

By Robert Mosteller

Robert P. Mosteller is the Harry R. Chadwick, Sr., Professor of Law at Duke Law School. He is a coauthor of McCormick on Evidence. He writes and teaches about evidence and constitutional criminal procedure and has written extensively about hearsay and the Confrontation Clause, particularly as it relates to evidence offered in cases involving children.

In Crawford v. Washington, 541 U.S. 36 (2004), the United States Supreme Court radically changed confrontation law as it applies to hearsay statements. For the preceding two decades, Ohio v. Roberts, 448 U.S. 56 (1980), had provided the basic framework for analysis. Roberts made reliability the linchpin of the analysis, and admitted hearsay in the absence of confrontation if that hearsay was determined to be reliable.
In its summary passage, Roberts stated that “[r]eliability can be inferred without more in a case where the evidence falls within a firmly rooted hearsay exception. In other cases, the evidence must be excluded, at least absent a showing of particularized guarantees of trustworthiness.” (Id. at 66.) The system Roberts established closely linked hearsay and confrontation analysis. It applied the Confrontation Clause to all hearsay statements, but was extremely generous in finding the requirements of the clause satisfied. Indeed, it resulted in the automatic admission of hearsay that was admissible under long-established hearsay exceptions, generally without any showing of unavailability of the person who made the statement. (See White v. Illinois, 502 U.S. 346 (1992) (concluding that unavailability need not be shown for statements that under hearsay theory were likely to be as reliable as those made in court, such as excited utterances or statements for medical treatment).) Roberts’s system as it evolved even admitted much of the hearsay offered even under problematic hearsay exceptions, such as statements against penal interest offered against codefendants. (See Roger W. Kirst, Appellate Court Answers to the Confrontation Question in Lilly v. Virginia, 53 SYRACUSE L. REV. 87, 105 (2003); Crawford, 541 U.S. at 60 (criticizing the frequent failure of the reliability standard “to protect against paradigmatic confrontation violations”).)
Crawford gives the Confrontation Clause a dramatically different focus. It concluded that the “principal evil at which the Confrontation Clause was directed was the civil-law mode of criminal procedure, and particularly its use of ex parte examinations as evidence against the accused.” (Crawford, 541 U.S. at 50.) Within this area of primary concern, the Court indicated that the Confrontation Clause was to be applied with real rigor. For hearsay statements of this type, it rejected Roberts’s reliability analysis. Except in a limited set of circumstances, such as “forfeiture by wrongdoing” (id. at 62), the Court insisted upon actual confrontation—“testing in the crucible of cross-examination” (id. at 61), or it demanded exclusion of the statement. Figuratively, Crawford erects a prominent “stop sign” in the way of admitting such statements.
For this narrower, more rigorously applied Confrontation Clause analysis, the Court made “testimonial” statements the defining characteristic. Unlike the Roberts mode of analysis that covered all hearsay statements, not all hearsay was subject to the new rigor of Crawford. Instead, only those statements that are “testimonial” were covered.
The particular type of statement involved in the Crawford case was a tape-recorded statement of Sylvia Crawford, the wife of the defendant. The statement was made while she was in police custody as a potential codefendant of her husband and was being interrogated by the police using what the Court termed “structured . . . questioning.” (Id. at 53 n.4.) In summarizing its holding, the Court stated that “[w]hatever else the [‘testimonial’] 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.” (Id. at 68.)
The details of what constitutes a testimonial statement is at present extraordinarily complex and quite unsettled as the conflicting decisions in the lower courts demonstrate. I have offered a more expansive treatment of this subject in “Crawford v. Washington: Encouraging and Ensuring the Confrontation of Witnesses” (39 RICH. L. REV. 511 (2005). I continue the examination of this extremely important case here in the hope of providing a simplifying concept that should be helpful to both practitioners and courts and that provides an accurate and effective tool in defining the scope of this enhanced Confrontation Clause protection. The suggestion is to focus on accusatory statements to evaluate and supplement the term “testimonial.”

“Testimonial”: Decisive term without a definition
Remarkably, given the importance of the testimonial concept to its decision, the Supreme Court in Crawford did not give a definition. Instead of providing a single definition, the Court offered three possible definitions: one that is very narrow and on its face quite formalistic, and two others that are progressively broader. The broader definitions present some difficulties in terms of their complexity and the clarity of possible applications. My sense is that whether the complexity problems can be solved satisfactorily with those definitions may well determine the viability of any but the narrowest definition. Exacerbating the uncertainty, the Court chose to decide no confrontation case during the 2004-05 Term that could have provided further clarification while lower courts across the country are struggling in hundreds of criminal cases to apply an uncertain concept.
Although listed second by the Court, I start with the clear outlier in the group: “extrajudicial statements . . . contained in formalized testimonial materials, such as affidavits, depositions, prior testimony, or confessions.” (Id. at 51-52.) This definition is taken from Justice Thomas’s concurring opinion in White v. Illinois, 502 U.S. 346, 365 (1992). And while certainly at one extreme in the group of three, the fact that Justice Scalia, who wrote Crawford, adopted this definition a decade earlier by joining Justice Thomas’s opinion in White has to be significant. That point has not and should not be lost on anyone.
The petitioner suggested as a definition: “ex parte in-court testimony or its functional equivalent—that is, material such as affidavits, custodial examinations, prior testimony that the defendant was unable to cross-examine, or similar pretrial statements that declarant would reasonably expect to be used prosecutorially.” (Id. at 51.) The amicus brief, authored by Professor Richard D. Friedman of the University of Michigan School of Law, would define testimonial statements as “statements that were made under circumstances which would lead an objective witness reasonably to believe that the statement would be available for use at a later trial.” (Id. at 52.)
These last two definitions may be seen as reasonably similar in scope, although the amicus definition, perhaps somewhat modified, appears to be the primary competitor to the Thomas definition. While the petitioner’s definition is tethered to the formalism of the Thomas definition in its first part, it allows for some expansion by including “similar pretrial statements” and suggests the possibility of even more expansive application by covering such statements that the declarant would reasonably expect to use “prosecutorially.” Prosecutorial use would seem to be far wider and much more expansive than testimonial use, and they certainly do not appear to be the same.
The amicus definition generally appears to be the broadest in not beginning by limiting the statements to those that are formal in their tangible form, but it is narrowed by restricting its scope to statements that are fundamentally testimonial. The statement must be made under circumstances that would lead an objective witness to reasonably believe it “would be available for use at trial”—thus actual evidence that is admissible in the case and so very much serves the function of testimony. This limitation, in addition to calling for an evaluation of circumstances by an objective witness, directs that the statement must have an apparent use at trial—an anticipated evidentiary/testimonial use for the statement.
Thus, there are clear differences in these definitions, none of which was denominated as preferred by the Court. A notable difference in terms of ease of application—particularly since most of the declarants will be unavailable—is that the latter two focus on the expectation or intent of the “declarant” or “objective witness.” By contrast, Thomas’s definition focuses on an inherent characteristic of the statement—its form and formality. For that reason, it would prove much easier to apply consistently.
Before considering the question of how to define testimonial statements, we should consider the words of the Constitution. The Sixth Amendment nowhere mentions the term “testimonial.” Instead, the Sixth Amendment—the critical text—refers to “the right . . . to be confronted with the witnesses against him [the defendant].” How do we get from “to be confronted with the witnesses against him” to “testimonial”? How do we get to formalism? How do we get to the expectation of the declarant or the intent of an objective witness; or for use as evidence or prosecutorially? To my mind, not particularly clearly or easily.

A better approach: Accusatorial terminology
The Crawford opinion mentions but does not adopt the use of what I believe is a far more historically grounded term—accusers or accusatory statements. The opinion does, however, quote James F. Stephen’s History of the Criminal Law of England for Sir Walter Raleigh’s demand to have “his ‘accusers,’ i.e., the witnesses against him, brought before him face to face.” (Crawford, 541 U.S. at 43; Stephen at p. 326.) This quotation suggests the strong connection and some commonality of meaning between witnesses and accusers familiar to those concerned about the evils of the inquisitorial system.
Assuredly, the testimonial concept overlaps with these same core concepts of the Confrontation Clause. However, while alternative formulations of the Sixth Amendment constitutional principle dealt with witnesses and accusers (the Virginia Declaration of Rights included “accusers and witnesses”) explicitly, none of those alternative formulations call for exclusion of out-of-court statements based on their testimonial quality. Thus, while the concept behind the testimonial label has a substantial historical basis, the particular terminology has little or none. It is even less clear that it should be the entirety of the clause, particularly if that means that accusatory statements introduced against the defendant at trial are not included because they lack formality.
In Remaking Confrontation Clause and Hearsay Doctrine Under the Challenger of Child Sexual Abuse Prosecutions (1993 U. ILL. L. REV. 691, 747-49), I suggested that accusatory statements should form the core of a newly formulated Confrontation Clause doctrine. What I suggest is that accusatory statements are statements that are accusations, viewed at the time they were made, of conduct that is criminal. (Id.; see also Michael H. Graham, The Confrontation Clause, The Hearsay Rule, and Child Sexual Abuse Prosecutions: The Nature of the Relationship, 72 MINN. L. REV. 523, 593 (1988); Toni M. Massaro, The Dignity Value of Face-to-Face Confrontations, 40 U. FLA. L. REV. 863, 870 (1988).)
In Crawford, Justice Scalia observes that the three potential definitions for “testimonial” share a “common nucleus.” (Crawford, 541 U.S. at 43.) I contend that the better common nucleus across these three definitions and the history and purpose of the Confrontation Clause is the accusatorial statement concept. The testimonial element is not required when the statement is made, rather an accusatorial content, and the clause is implicated if such a statement is later offered in evidence, which satisfies whatever any element of testimony that might be required.
Most importantly, I want to argue here that the accusatorial concept is extremely useful even if the testimonial label is retained to denote statements covered by the Confrontation Clause. It at least helps supplement the testimonial concept and aids in developing explanations of why statements should or should not be considered within the core of the Confrontation Clause. Moreover, the accusatorial concept may be helpful in bridging a gap between the definition provided by Justice Scalia and that of the amicus definition and Richard Friedman. My personal view is that the appropriate definition is much closer to the functional definition offered in the amicus brief than the narrow formalistic definition offered by Justice Thomas, and that a focus on accusatory statements facilitates such a functional approach.
My immediate goal is through the accusatorial statement concept to encourage scholars and courts to achieve some modicum of convergence as to the scope of the Confrontation Clause in frequently encountered cases where the correct result is intuitively relatively obvious. As discussed below, my apprehension is that disarray will tend to lead to adoption of a narrow definition to avoid a massive release of prisoners, and in direct reaction to the disorder to bring about convergence.
Let me state my case for the accusatory theory in rough terms. No short definition can provide all the answers to the critical questions. A definition—a few words—whether simple or sophisticated, cannot achieve such a purpose. Moreover, words are too easily manipulated. Rigid restraint and consistent application through a simple definition will be particularly difficult if the task is to reconcile the intent of the Framers with the reality of a world more than 200 years distant and culturally eons away. In such circumstances, outside influences will constantly threaten to overwhelm language and will at least occasionally cause limited breaches of order.
However, I submit that the basic terminology used to define coverage of the Confrontation Clause should suggest answers to the obvious questions that are correct. It should also comport with basic intuition. And it should not frequently point in the wrong direction.
I believe that the term “testimonial” fails such a simple test. By contrast, I believe that a definition based in accusatory terminology suggests correct answers to most basic questions. In addition, a definition based on the concept of accusatorial statements should prove far, far simpler to flesh out. Most questions are answered immediately. Difficult ones will need some secondary development and a coherent explanation of why statements should be included or excluded from coverage. However, the additional explanation will typically not be complicated and certainly not as complicated as a testimonially based definition—that is, unless the testimonial concept is confined to an extremely narrow scope based on a formalistic conception.
Moreover, such a definition has a chance of being applied with some consistency by real courts in the rough and tumble of thousands of courtrooms each day. If Justice Scalia’s vision of testimonial remains, as it appears, to rest close to Justice Thomas’s definition in White, it can easily achieve a common application. However, it does so at the cost of frequently getting the answer wrong. I suspect that the amicus definition (in modified form supported by Richard Friedman), while reaching results I consider correct in most cases, appears to me too complex and far too unintuitive to be applied consistently as theoretically intended.

A case where “testimonial” permits the “wrong” result
I present a fact pattern of a case that I submit should be treated as “testimonial”; or as falling within Crawford; or as within the core of the Confrontation Clause; or as envisioned by the Framers as covered by the clause, whichever the formulation one wishes to use. In my judgment, the case I am about to describe should be absolutely a slam, bang winner as a violation of the Confrontation Clause after its reformulation in Crawford. But I must acknowledge that the Court, if it gives the clause a reading that I believe is inconsistent with its purpose but is instead trapped in a specific formalism, could do.
The fact pattern is not like Crawford in the sense there is no rigorous station-house interrogation. Thus, it may well fail a formalistic definition. Its facts are not like the Marian Statutes or Sir Walter Raleigh’s case. It raises the question of what fits within the core of the Confrontation Clause, and what doesn’t. Nevertheless, my argument is that this case should clearly be covered by the clause. If it isn’t, there is a failure in basic definition or the Confrontation Clause is not being defined as I believe it reasonably should be and had the situation arose at the time of the Framing, “its application would have been.” (Crawford, 541 U.S. at 43 n.3 (referring to Justice Scalia’s confidence that the Framers would have treated unsworn statements as falling within the confrontation right if, as the Chief Justice argued, they were inadmissible at the time and therefore not contemplated by the Framers).)
The fact pattern that I will discuss is from State v. Willie Forrest III, 596 S.E.2d 22 (N.C. Ct. App. 2004)), which was affirmed without opinion by the North Carolina Supreme Court (2005 WL1038811(May 5, 2005).) (The initial statement of the victim to a police officer in the hospital emergency room in People v. Cage, 15 Cal. Rptr. 3d 846 (Cal. Ct. App. 2004), review granted, 99 P.3d (2004), provides a similar misapplication with which others may be familiar.) I am particularly familiar with the facts of Forrest, having filed an amicus brief on Forrest’s behalf. The victim in the case is Forrest’s aunt, Cynthia Moore, who had been served with a subpoena but did not appear and did not testify. Forrest was convicted at trial of kidnapping and assault with a deadly weapon on Moore based on her hearsay statements.
For some reason—perhaps an earlier 911 call alerting the police to his presence, to reference a relevant, but quite distinct, body of case law—a police SWAT team surrounded a house where Forrest was located. They observed that house for about an hour. During that period, Forrest escorted his aunt outside the house on two occasions with escalating violence suggested. Inside the house, the two were apparently ingesting crack cocaine and Willie becoming “paranoid.” I say “apparently” because we know what happened inside the house only through the statements of the victim/aunt who never testified, but whose statements were introduced through the police detective with whom she talked after the incident was concluded.
When Forrest and his aunt left the house the third time, it was dark. They left the porch and started down a nearby sidewalk. At that point, a “take down” order was issued by the officer in charge of the SWAT team. Police officers surrounded Forrest and illuminated him with the lights attached to their “long guns.” Two officers put submachine guns to Forrest’s forehead. They separated him from Moore, who had small lacerations on her neck and an inch-and-a-half laceration on her arm.
When I speak publicly about this case, I suggest ironically that at this point that Willie Forrest was “probably” under arrest and “probably” his aunt recognized that the police were interested in prosecuting him. I believe that these are at least fair characterization of the facts. Moreover, I think that those facts should matter and, as a result, that this is a clear Confrontation Clause case.
Waiting nearby was Detective Melanie Blalock, who, according to her own testimony, was there to interview Moore. Her testimony on this point was perhaps more candid than it would be today after the Crawford decision suggested some of the possible boundaries for testimonial statements. At the time she testified, however, the Confrontation Clause posed little difficulty if the statement was admitted as an excited utterance, as police intention was largely if not entirely irrelevant to the statement’s admissibility. Blalock thus could not reasonably have anticipated that such candor might matter.
When Blalock moved from her nearby location to the crime scene, it had been secured and Forrest had been taken away. Moore was standing in the street with another officer, who brought her to Blalock. The victim was crying, and she had a cut on her arm that was bleeding. Blalock informed Moore that she was calling emergency medical services. The medics arrived and treated the wounds, but we do not know when that occurred relative to Moore’s statement. Later, Moore declined further medical attention when Blalock attempted to persuade her to go to the hospital.
Blalock stated that Moore “was nervous, she was shaking, she was crying and she was anxious to tell me that she had been held in the house.” (Transcript at 94.) She repeated that Moore “appeared anxious to tell me what happened. And by that I didn’t have to ask her what happened to you.” (Transcript at 95.) Blalock stated that Moore “just immediately abruptly started talking and telling me.” (Id.)
Moore’s statement, according to Blalock, lasted about a minute, during which Moore related that Forrest came to her house (at least an hour before the statement); he smoked crack cocaine (a later conversation revealed that Moore smoked crack with Forrest); he became paranoid and refused to let her leave; he took her from room to room at knife point; she attempted to run, but the door was locked; and Forrest cut her. (Transcript at 95-97.) Blalock indicated that she took notes of the conversation and referred to them as “exactly what she said.” (Transcript at 95-96.)
The North Carolina Court of Appeals opinion found the statement nontestimonial under Crawford because the statements were initiated by Moore and the conversation should not be deemed a police interrogation under that case. (596 S.E.2d at 27.) It asserted that under these circumstances, Moore “was not providing a formal statement, deposition, or affidavit, was not aware that she was bearing witness, and was not aware that her utterances might impact further legal proceedings.” (Id.)
As noted above, this case should be treated as testimonial or whatever terminology is needed to bring it within the Confrontation Clause. However, the use of the testimonial approach made the above error possible. That is an awful result in terms of what the Confrontation Clause should cover.
By contrast, accusatory analysis makes this case simple. The statement is accusatory. It is made to a person clearly identified as a police officer. Case closed. If more is needed, the conversation occurred after the defendant had been arrested, a fact clearly known to both the detective and to the declarant. Case really closed.
The Supreme Court in Crawford seemed to invite such possible results. It continued its unwillingness to be precise beyond giving multiple definitions of the testimonial concept. Even as to the type of statements involved in the case—statements to the police—it refused to provide a clear category of what was to be covered. It described the statements as the result of police interrogation and also used the term “structured . . . questioning.” Possibly the language is mere surplusage, but perhaps it is to substitute for a particular tangible form of the statement when that statement was obtained by the police from a witness but is neither written nor mechanically recorded. And, the statement in Forrest was not the result of structured questioning, if that is a requirement of it being testimonial.

Advantages of accusatorial terminology
In some ways all that I have suggested above is some degree of indeterminacy in the Crawford opinion. One appropriate response is “what do you expect with a new concept?” I accept that response in part. Any radically new approach will not and, indeed, should not attempt to answer all future questions. I think, however, it is clear that the problem with a nonformalistic testimonial approach is deeper. Not only is the answer unclear, but there is no clear basis on which to determine the answer. Once one leaves Scalia’s formalism, it is hard to find a background story that provides a source of definition.
An excellent scholar like Richard Friedman can develop a coherent system to address these, but it lacks a compelling central formulation that guides all the complex answers needed. Why is it that one looks only to the intention of the speaker and not to the intention of the receiver of the information as well? The Constitution protects the defendant, not the witness. Whether the statement takes the form of testimony because it was intended by the speaker or manipulated by the receiver does not matter to the party to be protected. If the declarant/witness believes the statement is informal and inadmissible, but the officer who is receiving it knows the opposite because she is tape recording it and knows evidence exceptions, why should the statement not be testimonial? Perhaps the factors I am mentioning—whether it is recorded and whether it is admissible—do not matter as to whether the statement is testimonial, but why in grounded theory are they irrelevant? Can we be sure that all of a complex system goes in the correct direction to make second order results come out the right way if they depend upon a set of as yet undecided first order determinations? The results under a complex testimonial approach cannot be predicted.
Contrast that situation with an approach based on accusatory statements made to parties at arm’s length being treated as covered by the new, stronger Confrontation Clause. Let’s look at some clear advantages.
• Probably the chief advantage is that whether the statement is covered, like Scalia’s formalistic approach, is determined relatively easily. Like the testimonial concept, whether the statement is accusatory is determined primarily by examining the statement itself, which in most situations makes for ease of application. But, while the “testimonial” concept may in Justice Thomas’s formulation point toward form and formalism, the “accusatory” concept points toward content and function.
• Under such a definition, there is no need to determine for most statements whether the intent of the speaker or the hearer or both matter, which can become a complex and difficult issue when determining anticipated testimonial use of an out-of-court statement. The form of the statement decides the issue. Given that the speakers are often unavailable, any system that requires determining the intent of the speaker is very problematic.
• The coverage of the Confrontation Clause is automatically modernized to fit the expanded admissibility of accusatory statements under contemporary evidence rules. Perhaps, at the time of the framing of the Constitution, the only statements that would be covered were those that were highly formalistic because those were largely the only accusatorial statements that were admissible. Even Justice Scalia, however, rightly indicates that such a limitation on the clause is inappropriate. He accepts that unsworn statements to the police, such as was admitted in Crawford itself, might have been inadmissible for evidentiary reasons, but that we can be confident the Framers would have wanted them covered by the Confrontation Clause if they had been admissible as evidence. (Crawford, 541 U.S. at 52 n.3.) A general use of the accusatorial concept allows the clause to move automatically to cover statements of the type the Framers would have known to cover in their day as the rules of evidence expanded over time to admit other similar statements. The term covers statements of the same character, impact, and need for adversarial confrontation without the formality required in the earlier day for admission.
• Under some formulations of testimonial, the speaker might be required to anticipate that the statement could be used in court as evidence and therefore some knowledge or expectation of admissibility might be required. The accusatory statement alternative eliminates this perverse issue as even potentially important.
• If the form of the statement is important to its testimonial nature, such as whether it was tape-recorded or written or made in response to structure questions, then governmental officials who receive the statement can consciously manipulate those circumstances to exclude the statement from coverage. Manipulation of this sort is made much more difficult when accusatory statements are used as a defining characteristic because it is the content, not the form, of the statement that is decisive.
• Statements by citizens to the police that provide incriminating information are answered clearly and I believe correctly regardless of whether they are made in responses to interrogation, questioning of an unstructured type, or are volunteered.
• When the testimonial concept is used, arguably irrelevant issues that may prove difficult to resolve arise with statements by children concerning whether the child is mature enough to understand that his or her statement will be used testimonially. Other issues must be resolved as to how much the child understands about the use that will be made of the statement and whether one views the statement and its circumstances from the perspective of a reasonable person or a reasonable child with limited ability. This inquiry may also require the court to determine whether the intent or knowledge of the adult receiving the statement as to the statement’s use may be considered and whether that intent or knowledge is irrelevant, relevant, or decisive (see State v. Snowden, 867 A.2d 314, 328-29 (Md. 2005) (concluding for several reasons that immaturity will not avoid testimonial treatment).) Such issues are either not involved when the focus is on statements that constitute accusations of crime or they are answered more easily.
• Coconspirator statements should be resolved accurately, and with potential nuanced precision. Most coconspirator statements would not be treated as testimonial, which is the appropriate result. This is because statements that are made in furtherance of a conspiracy are typically not accusatory and, therefore, would not be covered. However, those rare statements that are accusatory would be excluded from evidence, as they should be. Some formulations of this inquiry may require courts to examine the intention of the speaker. For example, must the statement be a focused accusation of crime or is the fact that it implicates the person spoken about in a crime sufficient even without an intent to accuse? However, if this dimension is examined, doing so to determine whether the statement is accusatory will lead to more intuitively satisfying results than will an inquiry as to whether the statement is testimonial.
• Statements to doctors might be resolved with precision as well. Those that accuse a particular perpetrator would be covered by the Confrontation Clause. Those made for another purpose, such as to receive treatment, might be handled differently. (See In re T.T., 815 N.E.2d 789, 815 (Ill. Ct. App. 2004) (treating statement to a doctor accusing a particular person of sexual abuse as testimonial because accusatory, but not other statements to the doctor concerning medical condition).)
• For those who think that all 911 calls should be admissible, this approach does not necessarily help to render them all outside the Confrontation Clause. But for those who think such calls when made for help should not be covered, the accusatory label provides a substantive explanation. Secondary development in this area will be needed, but I suggest that starting with an accusatorial rather than a testimonial focus will make resolution easier and can be used more easily to develop sensible dividing lines.
• Since the expectation or intent of the speaker that the statement has a testimonial use is not the defining element, in many cases there will be no need to determine whether that intent or expectation should be evaluated on a subjective or objective basis. The accusatory concept cannot be expected to eliminate all of these issues if it is to be applied in a nuanced fashion. Similarly, all issues concerning the actions or intention of the government officials receiving the information will not disappear as to difficult cases. The degree to which an accusation must be focused and intentional will be at issue in some situations, in addition to the objective fact that it implicates a person in criminal activity. Similarly, whether the conduct of the police in soliciting an incriminating statement that is not purposefully accusatory will need to be faced. The accusatory concept is not a “magic bullet” eliminating all difficult questions. However, it does better focus the inquiry on the substantive issues that are central to the concerns of the Confrontation Clause rather than the form and formalism of the testimonial concept. As noted earlier, the accusatory concept is at least a helpful supplement to the language of the inquiry regarding which statements fall within the core area of concern of the clause, which is the central question posed by Crawford.

The pressure to preserve existing convictions
I mentioned earlier that getting the results to converge relatively soon was important to the possibility of a broad definition of the scope of the Confrontation Clause. An alternative formulation of this proposition is that large numbers of cases decided using a broadly variant definition of testimonial will probably result ultimately in a narrow scope for the Confrontation Clause. This is an argument based on the practical cost of a decision, not theory. Let me sketch it out.
In the period when the Warren Court was at its most creative in developing and expanding new criminal procedure rights, its policy on the retroactivity of doctrine was extraordinary. One might have thought that an activist Court would want its precedents to apply as broadly as possible to the past to correct old wrongs. That was not the situation. Instead, in Linkletter v. Walker, 381 U.S. 618 (1965), the most extreme form of its narrow applicability, the Court applied the new precedent only to the case at hand as to prior conduct and to new police conduct (searches) occurring after the date of the decision. The Court, thus, let only one convicted defendant out of jail (assuming the error in that case was not harmless), but was able to change police conduct radically in the future. The Court could act like a legislature without the costs of general jail delivery. Police and prosecutors could adapt their conduct or seek other evidence and win error-free convictions.
Conservative jurists and constitutional theoreticians alike decried that result for a number of reasons, one of which was the license it gave the Court to be activist. And, beginning with United States v. Johnson, 457 U.S. 537 (1982), the Court changed its basic rule on retroactive application. All new decisions apply at least to all cases still on direct review at the time the decision is rendered. That is the regime under which we operate today, and Crawford applies at least that broadly.
What this meant is that, when Crawford established its extraordinary sea change in confrontation law, the new rule was applied to a huge number of cases that had been tried in courtrooms where everyone “knew” that the law was what Ohio v. Roberts (448 U.S. at 66) had said it was. Accordingly, many instances of testimonial statements had been admitted in full compliance with the then-existing law. If, however, the defendant had made a Confrontation Clause objection at trial, and the case was still working its way through the appellate system when Crawford was decided, the defendant stood to get a windfall. Indeed, the prosecutor may have had abundant other admissible evidence that he or she did not introduce because the confrontation issue was so clearly resolved under Roberts against the defendant. Thus, the prosecution may have done nothing wrong at the time, and even worse, it had been lulled into an unnecessary error that could now free a dangerous defendant.
Release is not required under Crawford, however, if admission was not error because the statement was not testimonial. The pressure to avoid release, which is particularly poignant when the police and prosecutors seem to have been abiding by the established law, thus puts pressure on courts to apply a narrow interpretation of the testimonial concept.

The pressure of discord on the Supreme Court
The pressure described above existed at the moment Crawford was decided, given the enormity of its change in Confrontation Clause law. What I now perceive is a new sort of pressure that is similar but distinct, created by delay and disorder.
This situation is unlike that with the sentencing uncertainty created by Blakely v. Washington, 451 U.S. 36 (2004), where, within a matter of months, the Supreme Court took and decided cases resolving much of the uncertainty. (See United States v. Booker, 125 S. Ct. 738 (2005).) Here, by contrast, the Supreme Court chose not to take and thus not to decide another Confrontation Clause case during the 2004-2005 Term. Thus, by the time it renders its first post-Crawford decision, at least 18 months will have passed—probably closer to two years. Hundreds of convictions will have been secured under trial court and lower court rulings without additional Supreme Court guidance.
A number of courts have reached decisions that are very formalistic in approach. Many have construed Crawford very narrowly. Some categorically treat excited utterances as exceptions to the testimonial concept. Some treat police questioning that is not “structured” as outside the scope of Crawford. (See generally Robert P. Mosteller, Crawford v. Washington: Encouraging and Ensuring the Confrontation of Witnesses, 39 U. RICH. L. REV. 511 (2005).) All of these and scores of other cases are inconsistent with a broad reading of the definition of testimonial such as that set out by the amicus in Crawford.
The more decisions resolved on what I contend are erroneously narrow, formalistic bases, the more costly will be a ruling by the United States Supreme Court that is more expansive in scope. By the time the next Confrontation Clause case is decided, the Supreme Court will be able to count with some accuracy the scope of jail delivery. It will likely be that some cases are reversed, but if the definition of testimonial is broad, whether the numbers are in the tens or the hundreds will be easier for the Court to calibrate. Certainly, a broad ruling could be incredibly costly in terms of jail delivery of dangerous convicts.
Because some of this cost could have been avoided by clarity at the beginning or a quickly rendered second ruling, one has to wonder whether the delay is purposeful. Whatever the merits of that specific analysis, I believe it is clear that lack of early convergence of decisions around a broad construction of the testimonial concept is likely to lead ultimately to a narrow interpretation of the Confrontation Clause.
The costs are particularly great as to cases that were tried after Crawford and are being tried each day as we move forward in this period of disarray and divergent opinions. Cases will be tried by prosecutors using narrow definitions of testimonial statements and trial courts will erroneously find accusatorial statements admissible because they use the same cramped definitions and analyses. Many of these cases would have resulted in conviction even if only those statements admissible under a broader definition of testimonial had been applied. But once a statement that should be excluded as testimonial under an appropriately broad definition is erroneously admitted and a conviction is secured based on such evidence, these cases provide a needless cost to rectifying the earlier error that will make it more likely that the Supreme Court will ultimately choose a narrow rather than a broader definition. These new cases are gratuitous injuries, and those interested in a broader scope for the Confrontation Clause need quick movement toward a more sensible definition that can be applied more uniformly. I suggest that a definition built around the accusatorial statement concept has some potential to achieve this important result.

The definition that the Supreme Court ultimately gives to the concept of testimonial statements will obviously be of critical importance in determining whether the new Confrontation Clause analysis adopted by Crawford affects only a few core statements or applies to a broader group of accusatorial statements knowingly made to government officials and perhaps private individuals at arm’s length from the speaker. I contend that the broader definition is more consistent with the anti-inquisitorial roots of the Confrontation Clause when that provision is applied in the modern world. If my sense of the proper scope of the clause is roughly correct, then the testimonial statement concept must be reoriented from its potentially formalistic definition to one that includes such accusatorial statements. Employing accusatorial statement language as part of the inquiry is one obvious and important step in this transformation.
I argue that a movement in the direction of accusatorial terminology and coverage needs to begin as soon as possible so that lower courts can demonstrate that convergence is feasible and to reduce the costs of general jail release. I contend that the accusatory concept is consistent with the core concerns of the Confrontation Clause and will help advocates and courts to reach sensible results that are consistent with history, the language of the clause, and its function in a modern and complicated world.

Return to Table of Contents - Summer 2005

Return to Criminal Justice magazine home page