The Smith argument and the meaning of "testimonial"

     I believe that the Supreme Court will decide Smith v. Arizona on the assumption, made in the Question Presented in the petition for certiorari, that Rast’s statements were testimonial.  (And I believe they clearly were testimonial.) The state did not challenge that proposition in the Arizona courts, at least not more than very inferentially, and it was not argued there.  I believe the oral argument revealed no disposition on the part of the Court to use this case, at least at this point, to do anything substantive with respect to the question of what statements are testimonial.  But it wouldn’t be surprising if a remand invited the Arizona courts to take another look; the Court sure was interested in the question at argument.  So I’ll offer some comments here. 

     Justice Kavanaugh repeatedly raised the question of whether the Court should adopt Justice Thomas’s view of what is testimonial, dependent on formality or solemnity.  Eric Feigin for the SG and Alex Samuels for the state were hospitable to the idea; Hari Santhanam said, accurately in my view, that the test is too narrow and would not apply to many statements that fit within the core of what is testimonial.  At least that is true, I believe, given how Thomas applied the test in Williams v. Illinois.  But I think that if the Court were to adopt a test based on solemnity and if – very big if – it would apply such a test appropriately, then it might get to a better place.

      I’ve often written on this subject, and I'll do so again, rather briefly, now.  Let’s focus on formality.  Formality is not what makes a statement testimonial; it’s what makes testimony acceptable.  The oath, presence of the accused, and cross-examinations may all be considered incidents of formality, and they are all essential for prosecution testimony to satisfy the standards of our system.  But that does not mean that a statement given informally is not testimonial.  Suppose a police officer says to someone who has witnessed a crime, “I’d like you to tell me what happened.  You can tell me very informally.  You don’t have to take an oath.  I won’t record it, and I won’t take notes.  That way, you won’t have to come to court.  I’ll remember what you say, and I’ll tell the jury.”  That has to be testimonial; if it isn’t, then we have a system in which a witness can knowingly create evidence for use in prosecution by having such an informal conversation and not take an oath, confront the accused, or be subject to cross-examination.  That’s pretty much unthinkable in our system.  Now, I suppose you could say well, that’s really sufficiently formal to be considered testimonial because the witness expected prosecutorial use of the statement – it was “made in contemplation of legal proceedings,” as Justice Thomas put it in his concurring opinion in White v. Illinois (1992).  That would be OK by me, and I think it would get close to a good definition of testimonial, but it seems clear to me that the word formality here is misplaced and is not really doing the work; it’s the expectation of evidentiary use that is.

     If we take a narrower view of formality, in which some set of hoops must be jumped through for the statement to be deemed formal, then we will have virtually nullified the Confrontation Clause, because any witness wanting to create evidence for use at trial, or any investigator seeking to generate such evidence, can simply avoid the hoops.  Is an oath or certification a prerequisite for a statement to be deemed formal, as Justice Thomas suggested in Williams v. Illinois?  Then just avoid oath and certification.  Frankly, I am still mystified by Justice Thomas’s characterization of the lab report in Williams as not testimonial because insufficiently formal.  You can find the report here.  Notice that it’s on letterhead stationery, identified as a report of a laboratory examination, addressed to a forensic science center, and signed by two reviewers, and signed by two reviewers, one of whom is identified as the director of the forensics laboratory and the other as director of the identity laboratory.  In addition, it bears a case number and refers to the materials tested as exhibits and as evidence.  If the Court were to adopt the view that this is not enough to render a lab report testimonial, then it would be a simple matter for labs always to avoid the rule of Melendez-Diaz v. Massachusetts (2009):  Just avoid saying sworn or certified, and you never have to come to court.

     Now let’s talk about solemnity.  It’s not entirely clear what Justice Thomas means when he uses this word.  If he means it as a synonym for formality, then the same concerns apply.  But another, more useful, view of solemnity is possible.  If one takes solemnity to mean appreciation of the gravity of the consequences of making the statement, then it seems to me that gets at the essence of what testimonial means.  That is, if the speaker understands that her statement will likely be used in prosecution, and that it might help convict a person and lead to punishment, then it ought to be deemed to be testimonial.  More on this below, but it seems far preferable to an inherently ambiguous “primary purpose” test.

     Apart from the question of whether the Court should move in the direction of adopting a test based on formality or solemnity, the Court was interested in the question of whether the statements at issue in Smith itself were testimonial.  Justice Barrett particularly pressed the point:  What if a police officer jots down notes only for her own use to help think about the case, without intending them to be incorporated into a report?  (Samuels argued that in fact the in-court expert, Longoni, relied only on the notes of the absent analyst, Rast, rather than on her report; Hari Santhanam, for Smith,  without conceding that point, argued that the two were closely interconnected.)  Justice Kagan said it must depend on the facts of the case; some notes would be testimonial and some not.

     I would take a more categorical approach.  Yes, I think that if a lab tech, understanding that a sample is being tested for possible prosecution, jots down notes before writing a report, or if a police officer jots down notes about a crime scene to help her think about the case, it should be deemed testimonial.  No, such notes might not be formal, in the narrow sense of the word, but that shouldn’t matter.  But they are written in contemplation of being used in prosecution.  And if they are held admissible in favor of a prosecutor, and admitted, then we have a system in which a lab tech or police officer can create evidence for use in court by writing down such notes, and everybody would understand that this is what can be done.  In any realistic sense, the lab tech or officer is testifying by marking down the notes.  Eric Feigin, arguing for the United States as amicus, appeared to acknowledge this point when he said that there is a “chicken-and-egg problem,” because expert opinions will be aware of the Court’s decisions.  In other words, if the Court says that such notes are not testimonial, and that therefore the Confrontation Clause does not restrain their admissibility, then presumably they will be admitted as a regular matter, and they will be created with the expectation – and indeed the purpose – of being used as evidence at trial.

    This points to a complication.  If we are trying to determine the expectation, or the purpose, with which a statement is made by one involved in the criminal justice process, then whether or not the statement would be admissible at trial is a critical factor.   But notice the apparent paradox: If the maker knows the statement would be deemed testimonial, and so not admissible, then it could not be made with the anticipation, or purpose, of being used as evidence at trial.  So wouldn’t that mean that it is not testimonial?  And if the maker knows that the statement would not be deemed testimonial, and so would be admissible, then it likely would be made with the anticipation of purpose of being used as evidence at trial.  So wouldn’t that mean that it is testimonial?  I think there are two plausible approaches that would avoid this paradox.

    One is to ask not about potential use as evidence at trial but only about potential use in the prosecutorial process.  Even if the statement is testimonial, the maker could well anticipate that it would be used as part of that process.  I think that should be enough to render the statement testimonial. If a person makes a statement knowing that it will be used in the criminal justice process, then I believe it is enough to say that the person is acting as a witness as part of the case, by contributing to its resolution, whether or not the statement is formally introduced at a formal trial.

      The other is to assume hypothetically that evidentiary or procedural rules would not prevent the statement from being admitted at trial and ask whether a person in the position of the maker of the statement would reasonably anticipate use at trial. If the answer is affirmative, the statement should be deemed testimonial; the combination of an affirmative answer and admissibility would mean that in our adjudicative system a person could make a statement knowing that it would be used as prosecution evidence at trial and never confront the accused or be subjected to cross-examination.

      I have spoken about purpose as well as anticipation because the Supreme Court has phrased its test in terms of "primary purpose."  But I think that is an unfortunate choice, and if the Court ultimately takes a hard look at the question it should  do away with that test.  Trying to pick one purpose as primary is, as Justice Thomas suggested in Davis v. Washington (2006), a fool's errand.   But more fundamentally,  I believe any test should be expressed in terms of anticipation rather than of purpose.  I gave reasons for this long ago in an article titled Grappling with the Meaning of "Testimonial", 71 Brooklyn L. Rev. 241, 251-53 (2005), which you can see here, and I won't add more now.  

The argument in Smith v. Arizona: The Question Presented

 

     I put up a short post on the argument in Smith v. Arizona  right after the case was decided, and intended to present a fuller account soon after that.  Time has gotten away from me, but better late than never.  (And by the way, there were some interesting comments to that short post from a forensic scientist; I just noticed those while preparing this post.  I’ll respond to those on the prior thread.)

     Perhaps the most notable aspect of the argument was how interested the Court was in the question of whether the statements at issue were testimonial, rather than in the question actually presented, whether the statements satisfy the Confrontation Clause because they were introduced in support of the testifying expert’s opinion.  In the end, I doubt that the Court will use this case to revise the standard for what is testimonial, or even that it will determine that the statements at issue here were not testimonial.  And I think it is highly likely that Smith will get a reversal, with the Court holding that the statements should be deemed for Confrontation Clause purposes to have been presented for the truth of what they asserted.  But it is possible that a remand will leave it open for the Arizona courts to consider whether the statements were in fact testimonial.  I’ll comment in this post on the issue actually before the Court.  I’ll then present another post on the question of whether the statements were testimonial and the standards that should be used in determining that.

     The essence of the case is that one lab analyst, Rast, performed tests on certain items and wrote notes and a report stating her findings, but a second witness, Longoni, who had nothing to do with the testing or report, testified to an opinion, on the basis of the facts reported by Rast, that the items contained controlled substances.  The state courts held that the references to Rast’s findings should be deemed to be presented in support of Longoni’s “independent opinion,” and not for the truth of what they asserted, and that therefore there was not a Confrontation Clause problem.

     The problem, though, is that in this case Rast’s statements supported Longoni’s opinion only if they were true.  When a statement is offered to prove the truth of what it asserts, that proposition is most often an intermediate point on the way to some other conclusion, and this is no different; here, the other conclusion (itself an intermediate point on the way to the conclusion of guilt) is Longoni’s opinion.  I think it was quite clear that most of the Court understood this to be so; Justice Jackson was particularly vigorous in pushing the point.  I’m not sure that I could point to a single justice who thought otherwise.

        Justice Alito did say that Hari Santhanam, Smith’s counsel, was going out of his way to “trash” Rule of Evidence 703.  (Federal and Arizona Rule 703 are substantively the same; I’ll just refer generically to Rule 703.) And Alex Samuels, arguing for the state, suggested that a holding in favor of Smith would undermine Rule 703 beyond criminal cases.  These concerns seem to me entirely unfounded.  Rule 703 is very useful in its sphere.  It allows an expert to form an opinion on the basis of otherwise inadmissible information, if it is of a type on which experts in the field reasonably rely, and sometimes it allows the expert to testify to that information.  That’s good.  A holding for Smith would not have any impact on the operation of the rule in civil cases, or in favor of an accused – nor even in favor of a prosecutor so long as it did not result in a testimonial statement being used for the truth of what it asserts.  (For example, an expert could form an opinion as to public attitudes on a given issue on the basis of statements by people in a survey, whether or not those statements would otherwise be admissible.)  But Rule 703, a relatively recent creation, has to be subsidiary to the Confrontation Clause:  It can’t allow a testimonial statement to be presented against an accused for the truth of what it asserts.  And if the statement supports the opinion only if the statement is true, that’s just what is happening.

        Samuels repeatedly emphasized the possibility of giving a limiting instruction, telling the jury not to use the statement for its truth but only in support of the opinion, and the fact that the defense in this case never asked for one.  But a limiting instruction in this context is meaningless:  If the jury is told to use the statement in support of the opinion but not for the truth of what it asserts, then it is being told it may do something that is in fact logically impossible.  Santhanam addressed this point on rebuttal when he said that a limiting instruction is appropriate only when there is a valid use of the evidence, and here there was none.

        A related matter: Samuels said that Arizona law is clear in this situation that Rast’s statements are not admissible for the truth of what they assert.  But remember Lincoln’s line about how many legs a dog has if you call a tail a leg.  Four; calling a tail a leg doesn’t make it a leg.  The state saying that the statement is not being presented for its truth doesn’t make that fact so if that’s the only use the jury could have made of it.  A court protecting the Confrontation Clause cannot be controlled by state pronouncements of this sort.  Santhanam gave a similar, and effective response to a question posed by Justice Alito as to what the result would be if a judge said, “I’m not going to use this for the truth, only in support of the expert’s opinion.”  The judge saying that’s so can’t make it so. 

        Justice Alito also seemed to suggest at one point that Smith’s argument would invalidate hypothetical questions posed to prosecution expert witnesses.  It’s possible that I misunderstood him, but in any event there is no cause for concern with that.  An expert can give testimony of the form, “If factual predicate X is true, then my opinion is Y.”  It is then up to the prosecution to present valid evidence of predicate X.

        As for the state’s repeated characterization of Longoni’s opinion as independent:  It is only independent in the sense that Longoni presumably drew his own conclusions from the facts reported by Rast.  But those facts were essential for the opinion.  I believe that Justice Jackson was driving at much the same point when she asked a long hypothetical – she is a former Breyer clerk, and his successor in more ways than one – that seemed aimed at presenting a situation in which the opinion of the expert testifying live was clearly dependent  on facts provided by the absent expert, but not on the absent expert’s opinion.  Samuels said the result would depend on whether the live expert was adding anything.  But I think that’s plainly wrong.  The question is whether a testimonial statement of the absent expert is being used for the truth of what it asserts, whether as a predicate for the live expert’s opinion or for any other reason.  Also, as I pointed out in my amicus brief, in this case given the facts asserted by Rast the conclusion to which Longoni testified followed pretty much immediately.  I think all the talk of an independent opinion is really a sham.  (I do not mean to be critical at all of Samuels, who I thought did a very good, professional job with a weak case.) 

        Eric Feigin of the Solicitor General’s Office, arguing for the United States as amicus, sought to strike a middle position.  (The fact that the United States was coming in on neither side, rather than in support of the state, as it usually does, was itself noteworthy, as Justice Kagan, a former SG pointedly remarked.)  He strongly intimated, without quite saying, that Smith was entitled to a reversal because Longoni explicitly endorsed statements by Rast.  But he argued that Longoni could have testified to the usual practice, and to the fact that he had no reason to believe anything different was done in this case.  He would then be exposed to cross-examination to emphasize that he knew nothing about the particular case from personal knowledge.  That seems right to me  – the jury can infer that something was done on a particular occasion from the fact that this is how the organization usually does it – but a couple of points.  First, the Chief Justice asked why that type of cross wasn’t sufficient in this case.  I think the answer is that in this case the state presented more – it presented the testimonial statements of an out-of-court witness and effectively asked the jury to believe the truth of those statements. So the fact that Longoni could be cross-examined on what he did or didn’t know is neither here nor there.  Second, if we take the SG’s suggestion, it still does not get the prosecution all the way home.  If the prosecution wants to present machine-generated data and graphs, which presumably are not testimonial in themselves, it would still have to satisfy a chain-of-custody problem.  And this usual-practice approach would not help getting Rast’s statements in.