Previously, I have argued that the learned treatise exception to the rule against hearsay poses significant and serious tensions with the requirements of Federal Rules of Evidence 702, 703, and 705. Permitting a “learned treatise” to be considered for its truth conflicts with the essential learning of the Daubert case; namely, qualifications do not suffice to establish reliability. The exception conflicts with science’s rejection of authority in favor of sound inferences from accurate data and facts. The learned treatise exception ignores that authors often have their own agendas and myriad biases, which cannot be adequately explored by counsel or fully seen by the trier of fact.
As troublesome as the learned treatise hearsay is on scientific issues, its use on issues of social science and history, is even more disturbing. The standards for what is published as “history” are distressingly poor. Perhaps the exception has a role for proving “almanac-type facts,” which would be the subject of judicial notice in any event.
The learned treatise exception can be invoked by a witness on direct examination to support his opinions, or on cross-examination by the examiner to challenge the witness’s opinions. If used on direct examination, the exception raises the question why the witness must rely upon the opinion of another expert, who is unavailable for confrontation. If the testifying witness is an expert, then what should matter is his opinion and how he arrived at it. If the testifying witness needs an opinion of a textbook author, then perhaps the testifying witness is not qualified after all. Of course, the testifying expert witness may wish to accredit his opinion by showing that “learned treatises” agree with him, but accrediting the opinion before it is attacked seems like a waste of time.
This last point has a further consequence for the form of expert witnesses’ opinions now allowed by trial courts. Expert witnesses sometimes carelessly state that they are relying upon the published article, but the typical scientific article is broken down into discrete sections: (1) a statement of the issue; (2) materials and methods; (3) results; and (4) discussion. What exactly do expert witnesses rely upon when they cite a published study? The materials and methods, along with the results, represent the essential constituents of a study. The statement of the issue, and the discussion section, are filled with extrajudicial opinions that need not, and should not, be admitted into evidence, or disclosed to the jury. There is no “necessity” for such a practice – although it is often countenanced – because the witness on the stand can provide his opinion of where the study fits into the available evidence, without becoming a conduit for the speculation and opinion that fills the introductions and conclusions of most published articles. Peer review, which is of little help in any event, is particularly useless in ensuring the quality of the speculative opinion in discussion sections. Authors are usually allowed much greater latitude in discussing the results than in reporting the data and data analyses.
The loose practice typically continues on cross-examination, when adverse counsel is permitted to ask whether the witness agrees or disagrees with the speculative opinions in “relied-upon” papers, on the theory that the papers appeared on the witnesses Rule 26 witness list. Strictly speaking, the witness has not relied upon, nor need he rely upon, the introductions and discussions of the studies that make up the basis for his opinions. It is time to tighten up the practice. Perhaps the Third Edition of the Federal Judicial Center’s Reference Manual on Scientific Evidence, which should come out next year, will help.
