Today we take on our final topic for discussion in the Confrontation Clause community thread.  The Court’s recent Confrontation Clause cases, beginning with its decision in Crawford v. Washington, are often viewed as a triumph for proponents of originalism, as Crawford’s basic rule is based on a historical view of the meaning of the Confrontation Clause at the Founding.  Yet as the late Chief Justice Rehnquist and Justice Kennedy have argued, in reaching this result Crawford implies that U.S. courts (including luminaries such as John Marshall) misinterpreted the Confrontation Clause for the first 218 years of its existence.  The response is that Crawford actually reinstates the proper understanding of the Clause, from which the Court departed in Ohio v. Roberts.  What does this debate tell us about originalism as a mode of constitutional interpretation?  Is it better to have a correct view of the Constitution late than never, or are originalists wrong to claim to a “correct” view of the Constitution in the first place?

A selection of our favorite comments from yesterday’s lively discussion of the Williams case follows the jump.

Richard Friedman –

I haven’t been able to keep up with this whole thread, but I would like to make a few quick points.

1. The most arresting part of the argument in Williams, in my view, was clearly comments by Justice Kennedy suggesting that he was willing to abide by Melendez-Diaz and Bullcoming. In dissenting in those cases, he predicted dire consequences. But one would expect that if he accepting them as given he would want to work towards applying them in a reasonable, practical, and principled way. And, though one can’t infer too much from comments at argument, his statements – especially the one comparing “supporting actors” to “the Hamlet in the play” — suggested that this is just what he is doing. I have posted more extensive thoughts on the argument on my blog, at

2. Paul’s circumstantial evidence theory, no matter the frequency and the conspicuous ability with which he and Michael Dreeben assert it, just doesn’t hold up. Let’s bear in mind: What Cellmark did was not just run stuff through a machine that spat out a chart. Rather, they compiled a profile, with the exercise of human judgment, and stated what that profile was. Circumstantial evidence cannot demonstrate what that profile was. Lambatos took the profile and compared it to that of the defendant, and a computer program did a match as well. If the profile generated by Cellmark had been different, it would not have supported Lambatos’s opinion. I don’t think anyone has yet responded to my point (forgive me if I’ve missed something) that the situation is exactly as if an eyewitness gave Lambatos a description of the assailant and Lambatos compared that to a description of Williams. I say “exactly as if” because in fact that is what Cellmark (though not an eyewitness) did – gave a genetic description of the assailant. Calling this circumstantial evidence simply distorts the actual situation.

3. Victory for Williams has no impractical consequences. The whole specter of a parade of DNA witnesses in every case fails to take into account the facts that (1) most of those who work on a DNA test do not make testimonial statements that must be conveyed to the trier of fact to present the results (but somebody at the lab does, which is a point that I think Albert Locher overlooks), and (2) states that have always followed the procedures for which Williams ocntends have managed just fine.

4. A principal virtue of Crawford is that it recognizes that the confrontation right, just like the right to counsel and the right to jury, is a basic procedural right that we adhere to as a uniform matter; we do not decide case by case whether it is worthwhile in the particular circumstances. I really do hope that the Court doesn’t edge back towards Ohio v. Roberts, as my senior colleague Rick Lempert suggests. We wold lose sight of the fundamental principle underlying the confrontation right, and (in part because most judges don’t have anywhere near the intellectual subtlety of Rick) over time the right would once again be severely impaired.

Richard Lempert 


Upon further thought there are some complexities which determine the practical consequences of a decision for Wlliams. Consider two kinds of DNA data bases; one with individually identified DNA and the other with crime scene DNA. The first one is used when DNA is recovered at a crime scene and the effort is to identify who left it. The question asked is, “Does the crime scene DNA match any of the DNA samples in the data base?” This is the most common kind of data base use. A decision for Williams along the lines you propose has little consequence beyond increasing system costs somewhat – a common price of liberty. The possibility of data base error does not loom large because once a defendant has been arrested because a DNA collected at a crime scene matches DNA associated with him in the data base, he will (should) following arrest be retested to confirm that the crime scene DNA does in fact match his. The second kind of DNA data base is a data base of crime scene DNA that has not matched any individual DNA in an extant/searched data base. As new individuals are arrested and their DNA typed, the crime scene data base can be searched to see if the newly acquired DNA matches DNA left at a crime scene. This is the less common use of DNA data bases but is the use more like that of the Cellmark analysis in Williams. Williams was, however, a low consequence case if the witness must be called since the two DNA analyses were in close temporal proximity and requested by the same police department. Moreover, not only the company but the analyst who who supervised the DNA analysis was known and, one assumes, available to testify if called. This will not always be the case. A crime scene analysis may have been done decades before matching DNA has been acquired from someone arrested/convicted for another crime. The lab that did the analysis may be located in a distant state or even overseas, and the analyst who supervised the analysis may be long gone. It may even be the case that the identity of the lab that did the crime scene analysis and forwarded it to the data base is not known. In these circumstances there can be serious consequences both financial and to justice if courts hold that the Confrontation Clause makes inadmissible the testimony of someone who analyzed the suspect’s DNA and found it matched a crime scene profile unless the producer of the the crime scene profile is called. Moreover, as I point out in my original comment the chance that the crime scene DNA did not match the suspect’s DNA but that errors in collection or analysis meant that the reported results matched precisely is so minuscule that it can be safely ignored. About the only error that is even remotely possible is that samples of crime scene DNA were confused and while the defendant left DNA at a crime, it was not the scene which the analyzed DNA was associated with. Not only is this quite unlikely, it is not the kind of error that an analyst who presents the DNA analysis for Confrontation Clause purposes would be aware of.

It is because I agree with your characterization of the testifying analyst’s use of the Cellmark DNA profile as hearsay reporting testimonial assertions that I suggest a special “not error because harmless” excuse from the requirement for confrontation is the best outcome in Williams. (I might add that since no one has suggested this resolution I do not expect the Court to arrive at it.) Otherwise I fear that Williams could lead the Court to erode in a potentially unfortunate way the meaning of testimonial and allow gaming the Confrontation Clause by clever ways of presenting evidence that relies on and implicitly conveys what is in fact a testimonial assertion by an uncalled witness.

More generally and speaking now more as a sociologist than a lawyer, a major reason for incoherence and inconsistency in the rules of evidence is that courts like to do justice – not a bad thing unless one teaches evidence.[:-)] Because only criminals can appeal convictions, most cases heard by appellate courts and,indeed, trial courts as well involve guilty parties. Hence, judges often produce tortured interpretations of rules of evidence to facilitate what they see as substantively just outcomes at the trial level and to avoid overturning what they perceive as factually justified convictions on appeal. (We can argue about whether they should do this, but that is a legal/philosophical and not a sociological question.) I think it is this tendency that largely gutted Roberts, both by watering down the reliability standard (which was low from the start) and abandoning the requirement to bring available witnesses to court before their hearsay could be admitted. This tendency also leads courts, including the Supreme Court, to characterize errors in evidence law or even instructions as harmless, although any reasonable person can easily imagine that but for the error a verdict might have been different. Nevertheless, I often find myself applauding (albeit with only one hand) findings of harmless error because at least the court is not mucking up the rules of evidence with a tortured interpretation that may secure justice in the case appealed but will do so at the cost of making the rules of evidence less coherent, fair, workable, protective, etc. For similar reasons, I think to preserve a realistic view of what is testimonial under Crawford, the Court would be wise to carve out the kind of “not error because harmless” standard I suggest. I do not call this “harmless error” because in in the kind of situation I describe the chance of any harm to the defendant is so small that not calling the person/lab rep who did the pre-identification crime scene DNA profile should not be called error. If my standard were adopted Illinois would have been able to admit the Cellmark lab report and would not have had to go through an elaborate charade to pretend that the Cellmark report was not important evidence crucial to the testifying analyst’s certain identification of Williams as the man who raped L.J.

Albert Locher –  

Professor Friedman, it is not correct, or at least fails to acknowledge the full and clear record, to say that Cellmark “compiled a profile, with the exercise of human judgment, and stated what that profile was.” Cellmark did do that, certainly; but the chart that their machine “spat out,” the electropherogram, was sent from Cellmark to ISP, where Lambatos personally reviewed it, evaluated it exercising her own human judgment, and based on her own review, also “compiled a profile.” She reached the same conclusion that Cellmark did as to the profile, because the genetic evidence, based on the electrophergram, was equally clear to both analysts (the one at Cellmark, and Lambatos); both could look at 2+2 and conclude the answer was 4. If the machine-produced data (the electropherogram) is not hearsay (as everyone seems to agree), then what is the testimonial statement that Cellmark made, that Lambatos relied on? It can only be two things: (1) the chain of custody matters (i.e. that this electropherogram chart indeed came from the L.J. rape sample); and/or (2) that the procedures undertaken to get the genetic material from the swab to the genetic analyzer machine that produced the electropherogram were properly done. If either or both of these are testimonial, then how can the issue be resolved under this line of analysis without calling all analysts (a result you have said will not be required in this case under Crawford)?

The chain of custody evidence may be weaker here than if Illinois had called someone from Cellmark to testify, but Melendez-Diaz said it is up to the DA to decide how to prove up chain of custody, and there is circumstantial evidence in this record as to chain of custody here. As the solicitor general pointed out, the fact that evidence may be weak is not a confrontation issue.

As to the propriety of the lab procedures at Cellmark, there is also circumstantial evidence supporting that they were OK. First, Cellmark is an accredited lab. Second, how could poor lab procedures have produced a sample that not only matched Williams to a probability of 1 in 8.7 quadrillion [remember, there was no other Sandy Williams DNA at Cellmark to serve as a possible source of contamination]. And also, third, that the person identified by that match would be the person L.J. would then pick out of a lineup? Again, this may not be the strongest circumstantial case that the lab procedures were OK, but neither is it the weakest, and as Paul has pointed out, if you are going to rely on an analysis that says that weak circumstantial evidence supporting this point is insufficient under the confrontation clause, then haven’t you reverted to Roberts, with its “reliability” standard?

That is why for me, one of the most telling points in oral argument was when Justice Kagan questioned Mr. Carroll, asking (page 16 of the transcript), “So why is that a Confrontation Clause issue? Why isn’t it just that the prosecutor has failed to prove an element of his case?”

Paul Vinegrad –  

First, I want to thank scotusblog for providing an opportunity for non-academics such as myself to engage brilliant professors like Richard Friedman & Richard Lempert in dialogue on constitutional issues. It is quite a privilege.

In response to Prof. Lempert’s implied assertion theory:

In Williams, in effect, what Lambatos said on the stand was the following: “I rely on the electropherogram (not any statement or assertion from a Cellmark employee regarding the accuracy of it or how it was generated) in forming my “match” opinion, because I believe, based upon my past experience with Cellmark as an accredited lab, that it was produced through a reliable scientific procedure.”

Lambatos did not directly relate any statement or assertion of any Cellmark employee. And her reliance upon the electropherogram (in forming her “match” opinion) was not intended by her as an assertion of anything that any Cellmark employee stated.

At most, Lambatos was the declarant — the only declarant — and any implied testimony/statement/assertion that the electropherogram was produced through a reliable scientific procedure was subject to cross-examination. Indeed, defense counsel confronted her (regarding any implied assertion by her regarding the reliability/accuracy of the electropherogram) by pointing out that (1) she had no personal knowledge how that graph was produced; (2) she wasn’t there when the graph was produced; and (3) she had no idea what procedures Cellmark actually used in generating this particular graph. Having confronted the “witness” “against” him with these facts, and thereby undermining any implied assertion by her regarding the reliability/accuracy of the electropherogram, the defendant could (and did) argue that her ultimate “match” opinion — an opinion that relied upon the electropherogram being accurate and reliable — was worthless. The defendant’s confrontation right was fully safeguarded through this adversarial process.

Addressing Prof. Friedman’s comments:

Point 2. Contrary to Prof. Friedman’s claim, circumstantial evidence (and the reasonable inferences that can be drawn therefrom) can (and did) establish that the Cellmark DNA profile from the victim’s vaginal swab — a graph not containing any express or implied statements or assertions by any human being — that Lambatos compared to the defendant’s DNA profile “matched.”

Prof. Friedman’s argument that the situation in Williams is “exactly” the same as an “eyewitness” giving Lambatos a physical description of the assailant, and her testifying that the description by the out-of-court declarant “matched” the defendant, completely misses the mark.

In Prof. Friedman’s hypo an out-of-court declarant (a human being) made a statement by asserting that the assailant was a particular height and weight, etc. In Williams, there is no statement, no assertion, by a human being, set forth in the graph that Lambatos testified matched the defendant’s DNA profile.

A more apt example of the scenario presented in Williams is the following: An ATM camera took a picture of the rapist committing the crime and Lambatos (or someone familiar with the defendant’s physical appearance) testified that the picture “matched” the defendant on trial.

The picture, like the electropherogram in Williams, could not have been produced by the ATM machine unless some human being properly built the camera, correctly set it up in the ATM, made sure that the lighting was such to ensure an accurate snapshot, and regularly maintained the camera to make sure it was in proper working order. Yet, despite the conduct of humans in ensuring an accurate picture, no one would seriously contend that the picture produced by the camera was a statement or an assertion within the ambit of the hearsay rule, would they?

Point 4. Prof. Friedman’s hope that the Court doesn’t edge back towards Roberts is a day late and a dollar short. The Court already did that in Bryant by its emphasis throughout the opinion on reliability-based hearsay exceptions. If there is to be more backsliding from Crawford’s “wodden” testimonial approach (Justice Kennedy’s description in Melendez-Diaz, not mine), I believe it will be members of the Court adopting Justice Thomas’s “formalized testimonial materials” theory — a narrower formulation than Justice Scalia’s and Prof. Friedman’s. And an estimation, a proxy, that is certainly more closely aligned with whatever relatively clear 18th century historical record exists regarding the meaning of the phrase “witnesses against.”

Charles Nesson –  

questions for richie friedman:

The moral force of the Raleigh story lies in Raleigh’s call to confront his accuser:

1. Is Cellmark Williams’ accuser?

2. Does it make any difference to you whether Cellmark is or is not his accuser?

3. Why should the admissibility of testimonial statements which are not accusatory be determined by judges at a constitutional level? Why not leave the issue of their admissibility to the lower realm of the hearsay and expert basis rules?

Sole reliance on such evidence to convict may raise a due process problem, but that is an issue of sufficiency of the evidence to convict, not admissibility of the evidence for the jury to consider.

Posted in Everything Else

Recommended Citation: Aaron Tang, Today in the Community: December 9, 2011, SCOTUSblog (Dec. 9, 2011, 9:07 AM),