The Sixth Amendment provides that "In all criminal prosecutions, the accused shall enjoy the right ... to be confronted with the witnesses against him." In Crawford v. Washington (2004), the Supreme Court held that this "bars 'admission of testimonial statements of a witness who did not appear at trial'" (with some exceptions): If I see you committing a crime, I'm supposed to testify in person about it, and be subject to cross-examination by you. The police can't just take my statement and then use it against you at trial when I'm not there to be cross-examined.

But what makes a statement "testimonial," and thus presumptively inadmissible, as opposed to "nontestimonial," and thus presumptively admissible? If someone overhears me saying, while you're attacking me, "Stop hitting me!," is that witness barred from reporting my statement at trial (assuming I'm not around to testify)? Crawford suggested that the answer is no, because the statement isn't really "testimonial," unlike (say) a formal statement given during a police interrogation at a courthouse, which is "testimonial." But Crawford didn't say where the line could be drawn.

In today's Davis v. Washington, the Supreme Court set forth this rule (emphasis added):

Without attempting to produce an exhaustive classification of all conceivable statements—or even all conceivable statements in response to police interrogation—as either testimonial or nontestimonial, it suffices to decide the present cases to hold as follows: Statements are nontestimonial when made in the course of police interrogation under circumstances objectively indicating that the primary purpose of the interrogation is to enable police assistance to meet an ongoing emergency. They are testimonial when the circumstances objectively indicate that there is no such ongoing emergency, and that the primary purpose of the interrogation is to establish or prove past events potentially relevant to later criminal prosecution.

In one of the cases that the Court decided with the Davis opinion, the Court found that a 911 call reporting on an in-progress attack by an ex-(?) boyfriend on an ex-girlfriend was basically nontestimonial. In the other, it found that statements gathered at the scene of a then-very-recent attack by a husband on his wife were testimonial:

There was no emergency in progress; the interrogating officer testified that he had heard no arguments or crashing and saw no one throw or break anything. When the officers first arrived, Amy told them that things were fine, and there was no immediate threat to her person. When the officer questioned Amy for the second time, and elicited the challenged statements, he was not seeking to determine (as in [the other case]) “what is happening,” but rather “what happened. ” Objectively viewed, the primary, if not indeed the sole, purpose of the interrogation was to investigate a possible crime— ....

As I read this, I wondered whether this distinction was quite sound. The trouble in domestic violence cases is that the victims often -- out of fear, misplaced loyalty, concern (whether sound or not) for their and their children's economic well-being, and a variety of other factors -- refuse to talk about the crime, or talk once and then stop talking. That's why the victims are often unavailable at trial; but it also suggests that when "Amy told [the officers] that things were fine, and there was no immediate threat," the officers might have reasonably had some doubts, especially when they saw some evidence of an altercation.

If you were an officer who was trying to figure out whether it was safe to leave a wife (and her daughter) at home with a husband who might have attacked them, would you simply accept the wife's word that all is fine, and that there's no immediate threat? Or, in trying to figure out what to do to immediately protect the possible victims (not just to or even not primarily to investigate the past crime with an eye towards prosecution), would you try to figure out if indeed things were fine and there was no immediate threat? There might be no "emergency" in the sense of an attack actually in progress that moment -- the police were there, after all, and many an attacker lies low when the police are present. But there might be an emergency in the sense of an attack that was merely suspended while the police were present, and that could resume shortly after they left.

In fact, Justice Thomas's partial dissenting opinion made this very point:

The Court draws a line between the two cases based on its explanation that [the husband-wife case] involves "“no emergency in progress,”" but instead, mere questioning as “part of an investigation into possibly criminal past conduct,” and its explanation that [the 911 case] involves questioning for the “primary purpose” of "“enabl[ing] police assistance to meet an ongoing emergency." But the fact that the officer in [the former case] was investigating [the husband's] past conduct does not foreclose the possibility that the primary purpose of his inquiry was to assess whether [the husband] constituted a continuing danger to his wife, requiring further police presence or action. It is hardly remarkable that [the husband] did not act abusively towards his wife in the presence of the officers, and his good judgment to refrain from criminal behavior in the presence of police sheds little, if any, light on whether his violence would have resumed had the police left without further questioning, transforming what the Court dismisses as "“past conduct”" back into an "“ongoing emergency." ...

Now I'm not sure whether the majority's opinion or Justice Thomas's dissenting opinion (which proposes a test that has its own flaws) is right. It may well be that on balance the majority's distinction, however mushy, is the best that can be created here. Still, it struck me that the majority's application of the distinction was less persuasive than the majority seemed to think. SCOTUSblog is posting what should be a very interesting discussion of the case.