70 B.U.L. Rev. 447 (1990)

The cornerstone of modern evidence law is relevance. Its rationalist tones may not engender the passion of notions like prejudice, due process, or even privilege. Yet its theoretical and practical importance cannot be denied. For example, the heart of the Federal Rules of Evidence, Rule 402, codifies the two fundamental principles of the common law of admissibility: the presumptive admissibility of relevant evidence and the inadmissibility of irrelevant evidence. n1 Relevance is the threshold admissibility issue, applicable to every piece of evidence offered. Moreover, identification of the mode of relevance is usually the prelude to proper application of other exclusionary rules. Much turns, then, upon the question of relevance.

As relevance gained ascendancy in the law of evidence, so did the notion that the relevance of evidence can be "conditional," in the sense of depending upon a favorable finding of fact. (1) This view, explained by reference to common-sensical examples, has provided an important source of exclusionary authority. Conditional relevance is said to underlie a wide variety of common-law decisions. Accordingly, it also appears in several places in modern codifications.

However, the notion of conditional relevance has recently been subjected to powerful theoretical criticism. (2) n3 It is argued that the idea rests upon a conceptual mistake, at least in terms of the examples by which it is regularly illustrated. Describing relevance as conditional in these examples simply misconceives the requirement of relevance. So convincing is this argument that conditional relevance now seems unable to bear the weight as an explanatory device that it had confidently assumed. The upshot of the criticism is that we should abandon the notion entirely.

If this is correct, then either the judicial decisions thought to be based upon conditional relevance must be rejected as mistakes, or alternative explanatory vehicles must be found. This article reflects a mixed evaluation. It will be argued that many judicial decisions exemplifying conditional relevance, including decisions that have come to be explained in such terms even though they were not decided in such terms, have been following, however inartfully, distinguishable principles and policies. Moreover, there is an identifiable core of good sense in these principles and policies, a residual force to the conditional relevance idea. A reinterpretation of the doctrine will be offered to account for this residue.

This is an important project. Several long-accepted and practically important rules of evidence -- notably, the requirements that documents be authenticated, that out-of-court speakers be identified, and that testimony reflect personal knowledge -- are now said to be based upon the notion of conditional relevance. Again, if that notion is not serviceable, viable substitutes are needed, or a major revision of evidence law may be in order. Here, too, this article will take a mixed course, suggesting a modified interpretive stance based upon the residual force of conditional relevance.

A major thesis of this article is that the "best evidence principle" -- the principle that parties should present to the tribunal the best reasonably available evidence on a disputed factual issue n4 -- is the vehicle that explains the residual force of conditional relevance. n5 It does not explain every application of that idea found in cases, statutes, and secondary literature. In the end, we must be prepared to acknowledge some mistakes in this area, especially since the apparent plausibility of conditional relevance has led to its general acceptance. What can be said, however, is that the best evidence principle accounts for the residue of authority that may justifiably entail the exclusion of evidence.

This thesis has wide-ranging implications, since conditional relevance problems, as conventionally understood, are ubiquitous. In a recent federal criminal case, for example, the United States Supreme Court resolved a conflict among the circuits concerning the admissibility of evidence of alleged prior similar acts by the defendant. (3) The decision came after years of dispute over an issue of considerable importance to the criminal justice system, one generally thought to be far afield from the problem of relevance. However, the Court's opinion relies heavily upon framing the admissibility issue as one of conditional relevance. If the theory presented here is correct, it has significant ramifications for the propriety of a decision that portends substantially increased practical importance for that concept.

Part I of the article will present the basic idea of conditional relevance and lay out in simple yet improved terms the criticism that has cast severe doubt upon it. Part II will reexamine the principal authorities relied upon to support the doctrine, in order to discover what they are really about. Part III will test the suggested reconstruction by looking at those well-established doctrines that have come to be assimilated into the concept of conditional relevance and by examining certain problems of statutory interpretation that are raised by the theory advocated here. This examination will lead to consideration of the Supreme Court decision mentioned above. The Conclusion will recapitulate the practical and theoretical significance of a "best evidence" interpretation of conditional relevance.


Evidence is said to be conditionally relevant if its relevance depends upon the satisfaction of a factual condition:

[I]t often happens that upon an issue as to the existence of fact C, a combination of facts A and B will be highly relevant, but either without the other will have no significance. (4) In such situations, the basic requirement of relevance means that the admissibility of evidence of either fact A or fact B must be conditioned upon some form of showing of the other. Unavoidably, evidence as to one must be introduced first, so administration of the trial requires the flexibility of allowing the proponent to "connect up" later with evidence of the other. Rule 104(b) of the Federal Rules of Evidence thus expresses the requirement in these terms:

When the relevancy of evidence depends upon the fulfillment of a condition of fact, the court shall admit it upon, or subject to, the introduction of evidence sufficient to support a finding of the fulfillment of the condition.

Significantly, Rule 104(b) does not specify when relevance is conditional, but only what to do with a conditional relevance issue once it arises. For clues to the identification of such problems, one must look to the examples provided in the notes of the advisory committee that drafted the rules. One standard example (hereinafter the "notice hypothetical") is presented as follows:

[W]hen a spoken statement is relied upon to prove notice to X, it is without probative value unless X heard it.

* * *

The judge makes a preliminary determination whether the foundation evidence is sufficient to support a finding of fulfillment of the condition. If so, the item is admitted. If after all the evidence on the issue is in, pro and con, the jury could reasonably conclude that fulfillment of the condition is not [sic] established, the issue is for them. If the evidence is not such as to allow a finding, the judge withdraws the matter from their consideration.

The clear, if poorly articulated, intention is that a "finding" by the appropriate standard of proof (such as a preponderance of the evidence) of "fulfillment of the condition" means, in the context of the example, a finding that X heard the spoken statement. If so, this seemingly sensible doctrine is mistaken.

The reason, stated in terms of the example, is that the trier of fact need not make, and ought not be required to make, a "finding" on the question of whether the utterance was heard unless evidence of the utterance is the only evidence of notice presented. Of course, one must evaluate the likelihood that it was heard, but even if, from all the evidence presented, there is only a small probability that X heard the spoken statement, it is still some evidence of notice that, together with the other evidence, may warrant a finding of notice. This follows directly from the liberally inclusive concept of relevance now employed. Specifically, Federal Rule 401 provides:

"Relevant evidence" means evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.

Suppose, for example, there were a hundred such utterances, independently made. The trier of fact might reasonably conclude that it is unlikely that X heard any particular utterance and yet also reasonably conclude that it is very likely that X heard at least one of them. More realistically, and more generally, there could be evidence of a variety of events, different in nature and weakly probative taken severally, that taken in combination make the existence of notice very likely but do not make it likely that the particular utterance was heard. n14

- - - - - - - - - - - - - - - - - -Footnotes- - - - - - - - - - - - - - - - - -

Indeed, even a "finding" of notice may be unnecessary, depending upon the applicable substantive law. If notice is merely an evidentiary proposition, rather than an ultimate proposition in the case, then the ultimate proposition toward which evidence of notice is directed may be found to be true notwithstanding insufficient evidence to warrant a finding of notice. In such a case, the evidence of notice must be considered together with other evidence that pertains to the ultimate proposition. The point is simply that the trier must make a finding, by the appropriate standard of proof, only as to the ultimate propositions in the case, not as to intermediate evidentiary propositions contained within inferential chains.

If, from all the evidence, one were certain that X did not hear the utterance, then the evidence thereof, taken in context, would have no probative value as to notice. n16 One might then rightly speak of "conditional irrelevance." But such certainty is rarely to be found in a litigated case, and in any event it would license a ruling of inadmissibility (whether in advance or after the fact on a motion to strike) in many fewer cases than the prevailing doctrine of conditional relevance theoretically prescribes. Under the conventional doctrine, any proposition that, if known with certainty to be false, would render the proffered evidence irrelevant (such as the proposition that X heard the utterance), is a proposition that must be proven true in order for the evidence properly to be considered by the trier of fact. Any item of evidence is vulnerable to such hypothesized conditions. Indeed, the only limit on the number of conditioning facts pertaining to each proffer is one's imagination.

This is not to say that hypothesized conditions are always useless constructs. Quite the contrary, it is often helpful to ask whether evidence is relevant given certain assumed factual propositions, propositions about which the trier of fact will have to make judgments in the course of its decision-making. As long as the assumption favors relevance, it generates a necessary (but not sufficient) condition for admissibility. For example, one might ask whether the evidence of the utterance made to X in the notice hypothetical would be relevant on the favorable assumption that X heard it. That is, did the content of the utterance have anything to do with the matter about which X is claimed to have had notice? Here, the point is that the evidence of the utterance may be irrelevant, and therefore inadmissible, even if the utterance was heard. As suggested above, one can also ask whether the utterance would be relevant on the assumption that its content appropriately concerns the underlying matter being litigated. This assumption, implicit in the notice hypothetical as given, points one to the question of whether the utterance was heard and understood. Either way, this analytical device must not be confused with the conventional notion of conditional relevance. The probabilistic character of the tests of relevance generated thereby remains fundamentally at odds with the conventional theory.

One is moved to conclude that the concept of conditional relevance either simply confuses the standards for sufficiency with those for admissibility or else serves some function distinct from spelling out the logical implications of the basic requirement of relevance. This conclusion is confirmed by another standard hypothetical (hereinafter the "agency hypothetical"), presented by the drafters of the California Evidence Code:

[ I]f P sues D upon an alleged agreement, evidence of negotiations with A is inadmissible because irrelevant unless A is shown to be D's agent; but the evidence of negotiations with A is admissible if there is evidence sufficient to sustain a finding of the agency. (5)

This supposed application is even worse than the previous hypothetical, or rather what makes it erroneous is more transparent. Unlike the notice hypothetical, here the conditioning fact is one upon which a finding must be made by the trier of fact. The ultimate factual propositions that P must prove include A's agency for D as well as the agreement between P and A. To be relevant, an item of evidence need only tend to prove, or disprove, one of these propositions, or some other ultimate proposition in the case. The adequacy of proof as to the others is of no consequence in making that determination. Conditioning the admissibility of evidence of one of these propositions upon prior or subsequent presentation of evidence sufficient to support a finding of the other makes no more sense than conditioning the presentation of evidence of one element of a negligence cause of action (such as negligent conduct) upon the prior or subsequent introduction of sufficient evidence to support findings on the other elements (such as cause in fact and compensable injury).

The conflation of admissibility and sufficiency is blatant in this and similar hypotheticals. Of course, if a proponent cannot present sufficient evidence to support a finding on one of the essential elements of the proponent's cause of action or affirmative defense, then presentation of evidence on the other elements becomes pointless. It is perhaps understandable that evidence pointless in this sense would be excluded as "irrelevant." This usage finds some surficial support in the language by which relevance is now defined, since those other essential elements are no longer "of consequence to the determination of the action" as required by Rule 401. However, properly speaking, the "facts of consequence," known at common law as the "material" facts, are determined by the substantive law invoked by the parties' claims, as modified by any stipulations or other judicial admissions, not by the adequacy of proof that is adduced on those material facts. Thus, the only apparent justification for such a rule of conditional admissibility is to save the tribunal from the consideration of a frivolous claim by insisting that counsel assure that sufficient evidence on the other ultimate issues will be presented in due course.

The foregoing criticism of the standard examples of conditional relevance is fully generalizable. Before concluding, however, that conditional relevance is simply a mistaken idea, one ought to reexamine how it came to be. In fact, the doctrine in its modern form, attributable largely to the work of Edmund Morgan, is more liberal in the admission of evidence than the doctrine sporadically invoked prior to its existence. In older cases raising the kinds of problems posed by the previous hypotheticals, some courts would employ a standard that required the judge to determine the existence of the conditioning fact. In our notice hypothetical, for example, the trial judge would have had to determine whether or not X heard the spoken statement. Professor Morgan offered the insight that employment of such a standard unwisely extended the procedure properly employed for determinations of preliminary facts affecting the application of other exclusionary rules, like the hearsay prohibition and its exceptions. Recognizing the loss of valuable evidence that such a standard entails, without the concomitant advantages that attend the judicial administration of the more technical "competency" rules, he argued successfully for a liberal standard like that of the Federal Rules. Thus, for example, to admit evidence of the statement, the court need only determine that there is evidence "sufficient to support a finding" that X heard the statement.

Morgan's efforts were salutary, but he did not go far enough, at least if the modern critique is right. Of course, if we ignore the prescribed results in the hypotheticals used to exemplify conditional relevance, then the standard expressed in provisions like Federal Rule 104(b) can be rendered formally compatible with the liberal modern definition of relevance. This is accomplished by accepting that the only "condition of fact" upon which "the relevancy of evidence depends" is the existence of a non-zero probability for each "conditioning" proposition of fact, that is, each proposition that affects the probative value of the original evidence in such a way that the evidence is logically relevant only if the probability of that fact being true is non-zero. In terms of the notice hypothetical, this would mean that the evidence should be excluded if the judge does not believe that a reasonable jury could find there to be a non-zero probability that X heard the statement. This construction would convert the doctrine into what was described above as one of "conditional irrelevance."

The test has been stated in a seemingly double negative form for a reason. One should not ignore the case in which it cannot reasonably be said either that there exists a non-zero probability of a conditioning proposition or that there exists a zero probability of that proposition. This is just a special case of the general class of situations in which the probative value of the original evidence is unknown. See infra note 28.

Such a reading of Rule 104(b) would reduce the frequency of exclusions under the conditional relevance idea. Indeed, it would render the doctrine virtually a dead letter. (6) Before we embrace this result, it is wise to reconsider the judicial decisions that gave rise to the hypotheticals discussed above and other cases in which conditional relevance has been used as a rationale, or post hoc explanation, of a ruling on the admissibility of evidence. This is the purpose of the next Part.

div1.gif (1531 bytes)
Home | Contents | Topical Index | Syllabi | Search | Contact Us | Professors' Pages
Cases | Problems | Rules | Statutes | Articles | Commentary