The Theoretical Foundations and Implications of Evidence

Ronald J.Allen

“Evidence”and“evidence law”are two quite distinct concepts. “Evidence”generally refers to those inputs to decision making that influence its outcome in what, to introduce a third concept, is normally referred to as a rational manner.In the United States, “evidence”also has a technical legal meaning to refer to the testimony and exhibits introduced at trial, but this is problematic.In the United States, fact finders may take into account their observations of witnesses（“demeanor”）, which obviously is“evidence”in any useful sense of the term, and more deeply no observation may be processed and deliberated upon without the use of a vast storehouse of preexisting concepts, observations, and decision makings tools（such as logic, abduction, utilities, and so on）.A useful concept of evidence must thus expand considerably far beyond the mere“trial inputs”—the observations of witness testimony and exhibits.What“rational”means here is putting all of the inputs and cognitive capabilities to the use of discovering as best can be done the way the world was at some prior time, and then to let rights and obligations be determined consistently with the preexisting state of affairs.

“Evidence law”, by contrast, refers to the manner in which the evidentiary process is organized, but obviously the organization of the evidentiary process is contingent on both“evidence”and the nature of“rationality.”The domain of evidence law, then, extends to the traces of the past that we colloquially refer to as“evidence, ”the manner in which such traces of the past are processed and relied upon in human decision making, and the regulation by law of the formal evidentiary process.Evidence law is thus contingent upon, and must accommodate, at least three things:Universal truths of the human condition, contingent aspects of the nature of government and its legal system, and highly specific policies to be pursued in addition or opposition to the pursuit of truth.I will say a word about each of these in turn.

Universal Truths:Although much of human culture is socially determined, cognitive capacities are not.How capacities are developed and employed may differ, but the underlying epistemological capacities to perceive, process，remember, and relate what was observed are part of the human condition.Obviously, they differ over individuals within societies, but they are universally present in all competent adults.Many of the tools that humans employ to assist in understanding their environment likewise are universal.Mathematics and logic do not vary from place to place, nor do decision tools such as utility functions and cost curves.Together, I will refer to these epistemological capacities and formal tools as the“tools of rationality.”These tools of rationality are what permit humans to understand and control their environment.They include such things as simple deductive reasoning, the capacity to generalize, abductive reasoning（the search for the explanation of a series of data points）, an understanding of cause and effect and of necessary and sufficient conditions, and many other things as well.These issues comprise the study of epistemology—the study of knowledge—and the law of evidence is in fact the law's epistemology.I should note that in some discussions of the foundations of the law of evidence a distinction is made between probability theory and epistemology.That may be a useful distinction for some purposes, but in my opinion probability theory is just one of the tools of rationality that facilitate pursuing epistemic tasks.

There surely are cultural and social influences operating on the basic tools of rationality, and at all levels.Two individuals from different cultures may experience the same perceptual event but understand it completely differently based on their respective familiarity with the type of event in question and their background knowledge.Similarly, the assumptions that begin logical processes may differ, and both the sets of costs and benefits and their relative weights may vary, as well.

I am less sure that there is a universal human nature beyond the epistemological capacities, frankly, although many think that there is.There is much lose talk about a universal sense of justice and universal human rights, but the twentieth century is a reproach to any who would see“human nature”as beneficent or concerned about the welfare of strangers.Perhaps, then, the economists are correct in a sense that people pursue, or should be conceived of as pursuing, their own self-interest, however those interests are conceived.Obviously, people do pursue their own financial and medical self-interest and those of their families.There certainly is a widespread desire for the conditions of a peaceful life with the possibility of human flourishing, but it is hard to see this as resting upon universals of the human condition.Not only is the twentieth century a reproach to such a view, so, too, is most of recorded history.And the twenty-first century is off to a start that suggests not much has changed from the millennium.

As I will elaborate below, one of the primary tasks of the law of evidence is to process and digest this elaborate set of considerations and create in their light a system of dispute resolution that serves the interests of the community.

Contingencies of Government and its Legal System:Although there is much that is common to humanity, the ways in which humans organize themselves varies almost infinitely.Legal systems are critical components of government, and they reflect the resolution of issues of deep political theory.One need look no further than China and the United States to see this clearly.Because of the political history of the United States, our founders concluded that political power should be diffused over the three branches of government, with each needing one or both of the others in order to be effective.This was designed to counteract what western observers almost universally believe is the centripetal force of all power centers and their tendency to aggrandizement.In brief, this is why we have a tradition of independent courts that we conceive of as being a potential brake on other branches of government.

And of course there are innumerable additional ways in which governments can be constructed.Whatever form of government is chosen, and more importantly whatever assumptions form its foundation, will obviously impact the nature of the legal system, which in turn will impact the way in which disputes are resolved and evidence is administered.

Having said all that, there is one universal aspect of dispute resolution, and it is not what one might think.There is a misconception in the West that the fundamental political insight of the Enlightenment, and the strongest plank supporting modern western governments, has something to do with rights and obligations.Citations to Hobbes, Locke, and Rousseau are found in abundance in legal scholarship and underscore this point.While rights and obligations are important, the more fundamental insight of the Enlightenment was the epistemological revolution that there is a world external to our mind that may be known obj ectively through evidence; however, citations to the epistemological work of Locke, Berkeley, Hume, or even Kant for this proposition, are few and far between.This reverses the actual relationship of facts and rights/obligations.Facts are prior to and determinative of rights and obligations.Without accurate fact finding, rights and obligations are meaningless.Consider the simple case of ownership of the clothes you are wearing.Your ownership of those clothes allows you the“right”to possess, consume, and dispose of those assets, but suppose I demand that you return“my”clothes.That is, I insist that the clothes that you are wearing actually belong to me.What will you do? You will search for a decision-maker to whom you will present evidence that you bought, made, found, or were given the clothes in question, and, if successful in this effort, the decision-maker will indeed grant you those rights and impose upon me reciprocal obligations.The critical point is that those rights and obligations are dependent upon what facts are found and are derivative of them.The significance of this point cannot be overstated.Tying the rule of law to true states of the real world anchors rights and obligations in things that can be known and are independent of whim and caprice.This is why the ideas of relevance and materiality are so fundamentally important to the construction of a legal system.They tie the legal system to the bedrock of factual accuracy.

This point is truly universa.l Neither rights or obligations, on the one hand, or policy choices on the other, can be pursued in the absence of knowledge of the actual, relevant states of affairs.Thus, even within the contingencies of ways of governing, we find a universal aspect of the law of evidence.Of course, how one might think that facts are most accurately or efficiently found, and what policies may offset the significance of factual accuracy, are matters of reasonable disagreement.

The Significance of Policy Issues:An enormous number of policy choices face the designer of a legal system.Some are consistent with the pursuit of factual accuracy, but many are in opposition to it.Note that I use the phrase“policy issues”to accompany all interests that society may pursue.It thus encompasses what some might call“value theory.”However, not all the policies governments pursue are moral; many are quite practical and utilitarian.Indeed，maybe most policies governments pursue are practical and utilitarian.It is surely acceptable to make the distinction between moral and utilitarian policies，but they are parts of the larger category of interests governments pursue and can effectively be lumped together when thinking about the law of evidence. Another distinction that could be made, but that I do not make, is between the sources of policy issues.The source of some are just the standard questions that all governments face everywhere, that involve the ordinary exercise of what we call , misleadingly, in the United States, the police power—the power of the State to regulate issues affecting health, safety, and welfare.By contrast, the source of others are explicit constitutional provisions, whatever the form a constitution may take in any particular country.Some commentators sort out constitutional questions from other kinds of policy questions, which again is coherent.However, the distinction is not helpful to understanding the law of evidence, and thus I do not bother with making it.Evidence law does some things because of constitutional commitments, but at the highest level of generality that is no different than fashioning evidence law to pursue an interest that is not embedded in a constitutional document.

I now turn to many of the policy issues that must be accommodated by the law of evidence.

Pursuit of Factual Accuracy.One might reasonably suppose that natural reasoning processes based on innate epistemological capacities work reasonably well, and thus typically should be deferred to in the pursuit of factual accuracy.However, there may be recurring situations that lead people to error.In such a case, rules of evidence may attempt to correct for that systematic error.This explains FRE 403's authorization to exclude evidence when it may be misleading or unfairly prejudicial.It also underlies other rules, such as limitations on character and propensity evidence, and the requirement that witnesses testify from firsthand knowledge.The circumstances under which individuals systematically make errors probably is heavily dependent on culture.

The Value of Accuracy.Factual accuracy is surely the most significant desideratum, but it is by no means the only one.It has a cost, and the cost can sometimes be too high.A legal system overly preoccupied with factual accuracy may undermine the very social conditions that the legal system is trying to foster.A dispute worth only a dollar that would take a thousand dollars to litigate to a factually accurate conclusion perhaps should not be litigated.Such litigation may very well reduce overall social welfare and discourage private settlement of disputes.Where the limit is reached is difficult to say, of course, and surely depends on local views.I will say much more about this in my second lecture.

The Value of Incentives.Factual accuracy competes not just with cost but with other policies that a government reasonably may pursue.The list of such policies is long, and again culturally contingent.The law of privileges may foster and protect numerous relationships（spouses, legal, medical, spiritual，governmental, etc.）.Litigation of an accident should not discourage reduction of risk（the subsequent repair rule）.Perhaps settlement of disputes is preferred to their litigation, which leads to the exclusion of statements made during settlement talks.The encouragement of settlement is also a reason not to price litigation too low.The more the public subsidizes litigation, presumably the more of it there will be, and the less of private negotiation.There are still other policies that can be pursued.In the United States, we rest a vast body of exclusionary rules on the perceived need to regulate police investigative activities.Rules of evidence also can encourage or discourage certain kinds of law suits from being brought.Again in the United States, we went through a period in which we thought rape victims were being overly discouraged from reporting crimes against them, and one response was to create rules of evidence that reduced the abuse at trial that such individuals may have been exposed to.

General Considerations of Fairness may also influence the law of evi dence, although the precise effect of this variable is often hard to sort out from more overtly utilitarian motivations.Some think that the limit on unfairly prejudicial evidence reflects not just the concern about accuracy but the concern about humiliation, as is also the case with rape relevancy rules.The limits on prior behavior and propensity evidence reflect in part a belief that an individual should not be trapped in the past.The hearsay rule to some extent reflects the values of the right to confront witnesses against you.

The Risk of Error.A mistake free legal system is not possible.It is critically important to recognize that two types of errors can be made—a wrongful verdict for a plaintiff（including a conviction of an innocent person）, which we call a Type I or false positive error, and a wrongful verdict for a defendant（including an acquittal of a guilty person）, which we call a Type Ⅱ or false negative error—and resource allocation and other decisions will affect the relationship between these two types of errors.

Reasonable people can disagree as to the significance of these two types of errors, but both must be taken into account in the construction of the legal system.In the United States, we structure civil litigation to attempt to both equalize the errors made on behalf of plaintiffs and defendants and to reduce the total number of errors.The criminal justice process, by contrast, is designed to reduce the possibility of wrongful conviction at the admitted expense of making more mistakes of wrongful acquittals.Although the matter is complicated, these perspectives explain in large measure the preponderance standard in civil cases and the standard of proof beyond reasonable doubt in criminal cases.In civil cases, an error either way results in identical misallocation of resources.If the plaintiff wrongly wins a $500 verdict, a citizen（the defendant）wrongly must part with $500.If the defendant wrongly wins a verdict that he or she does not owe $500, a citizen（the plaintiff）wrongly will be deprived of $500 that rightfully he or she should possess.These two cases are identical analytically.In criminal cases, by contrast, in the United States we view a wrongful conviction as a more serious harm than a wrongful acquittal, and thus make convictions hard to obtain by requiring proof beyond reasonable doubt.We do so even though it is possible（but by no means certain）that a side effect will be increased numbers of false acquittals and an overall increase in the total number of errors.Again, I will say much more about this tomorrow, and I will cast some doubt upon how well these simple ideas work out in practice.

Miscellaneous Policy Questions.There are many other contingent questions that must be answered by the architect of a legal system.Most importantly are those allocating responsibility over the various actors in the legal drama.These involve such questions as whether trials should be episodic events as is somewhat more prevalent in Europe or single shot events as in the United States, how much discretion should the trial judge have and how much should the parties control the process, what is the relationship between trial judges and appellate judges.Should there be trial de novo in the appellate court or is it limited to review of legal errors? Are small civil cases different from large commercial cases in ways that justify different treatment? What about criminal cases？

The matters discussed above indicate the breath of the foundations and implications of the law of evidence, and I now wish to make four analytical points, three of which are critical to understanding the foundations and implications of any body of law, and the fourth of which is critical to thinking clearly about the law of evidence.They involve：

1.The distinction between the law on the books and the law in action；

2.The relationship between procedural and evidentiary law, on the one hand, and substantive on the other law, and in particular how procedural and evidentiary law are in fact quite interrelated with rather than distinct from substantive law；

3.Economics, or as we say in the United States, there is no free lunch.If you use a dollar（or yuan）here for one purpose you cannot use it there for a different purpose.

4.Whether trials the ideal or instead are perverse.Is the legal system designed to encourage trials or settlement? What should it be designed for？

I will discuss in turn each of these variables and their significance.

1.The law on the books; the law in action.Constitutions are enacted，legislation is passed, executives issue orders and directives, courts decide，and one would think that the rest of us more or less obey.Unfortunately（or perhaps fortunately）, life is not so simple.When constitutions or laws are adopted in any multi-party decision making process, there will be multiple understandings of what the legal language connotes.Some legislators may vote for the passage of a law even though they do not believe it goes far enough in its coverage（or even though it goes too far）; others may vote against it for just the same reasons.There also may be serious disagreements as to precisely what a particular provision is supposed to mean or do.One person may think the legal language has one implication, and someone else may think it has a different implication.Statutory language in the abstract often will not resolve the meaning of those phrases.Compounding the difficulty even further, legal language is often deliberately left vague because of the inability to come to agreement as to precisely what it should say or because of the omnipresent inability to anticipate all possible scenarios in which a particular problem might arise.

In the United States, there is the added complexity of separation of powers.It is the legislature's job to enact law, including in most states the law of evidence, but it is the courts'job to put that law into effect.The judges may have different understandings of the implication of the language adopted by the legislature, and their institutional concerns will differ as well.Thus, the application of the law by the courts may differ from the idealized meaning of the law intended by a legislature or an individual legislator.

The law of evidence has one potentially unique structural aspect that exacerbates the problem of indefiniteness.Aspects of the law of evidence are rulelike in the sense of providing necessary and sufficient conditions for the operation of a rule, but important parts of the law of evidence simply allocate responsibility and discretion precisely because the relevant issue is too complicated for rule-like treatment.Perhaps the single most important aspect of the law of evidence—relevancy—has precisely this attribute.It is impossible to state a priori the necessary and sufficient conditions for the relevance of most evidence presented at any particular trial.Those determinations will necessarily be contingent on the unique characteristics of each trial, and it is literally impossible to articulate them in advance（how could we identify when a presently unknown witness will lie about a presently unknown topic?）.Thus, the law of evidence vests responsibility in someone—party or judge—to determine what evidence to offer, and does so under quite general guidelines.In the United States, relevant evidence is defined as evidence that may increase or decrease the probability of some material fact being true, but virtually no effort is made to specify when the condition may be met.

One last factor that may result in the law on the books being different from the law in action is that some areas of evidence law must try to accommodate quite opposed principles or impulses.This can result in part of the law making a promise and another part subverting that promise.Two important examples of this from American evidence law are the hearsay rule and the rule against character and propensity evidence.The hearsay rule promises to exclude hearsay, but there has been a unidirectional growth of the exceptions to the hearsay rule for centuries.In civil cases, the promise of the exclusion of hearsay is rarely redeemed, and even in criminal cases hearsay is routinely admitted.Similarly, the law of evidence promises the exclusion of character and propensity evidence but then creates broad avenues of admission.

2.The Relationship Between Substantive Law and Procedural Law.Substantive law is sometimes conceived of as quite distinct from evidentiary（and procedural）law, but this is misleading, for the two are in a complex and interactive relationship.This has become particularly clear, and is the subj ect of interesting legal research, in the United States due to the significance of the point for the protection of constitutional rights, but the point applies to general evidentiary matters as well.The decisions of the United States Supreme Court extending and enforcing individual rights have been viewed as imposing considerable constraints on the police and prosecutors, yet the legal system has not been greatly disturbed by these rulings.These systems are dynamic and infinitely adaptable and thus can and do respond to changes in unpredictable and astonishingly varied ways.Thus, “reform”to a dynamic process often cannot be imposed unproblematically through discrete measures that will have only the desired and no unintended consequences.One important aspect of this dynamic phenomenon is that legitimate substantive changes can blunt virtually any procedural innovation that emerges from courts or law reformers.

An example of this point in the United States involves the fourth amendment limit on unreasonable searches and seizures.Suppose the police want to stop cars to do cursory inspections for criminality, but courts rule that the fourth amendment requires that the police have probable cause that a crime has been committed before a car can be stopped.All the legislature need do to make this judicial command a practical nullity is to expand the criminal law to include more rigorous driving requirements.The legislature can essentially make it next to impossible to drive without violating a criminal statute（such as crossing the center line, driving too closely to the car ahead of you, not putting your turning light indicator on early enough or too early, etc.）.If the legislature passes such laws, the police will be able to stop virtually any car by following it until the driver violates one of statutes regulating driving.The stop will be on“probable cause”but the legislation will have expanded dramatically the potential sources of probable cause, thus subj ecting everyone to being stopped by the police whenever the police decide to do so, notwithstanding the attempt by the courts to forbid just that process.Similarly, if the government cannot seize certain information without probable cause, it can often instead require that individuals keep records of the information it wants and divulge those records to the government.

This point generalizes across evidentiary and procedural law.The most obvious example is materiality, which is directly determined by the substantive law, but the point goes deeper than that.By changing the elements of causes of actions, legislatures can make recovery under those causes of actions easier or more difficult.Whether oral testimony concerning the meaning of contractual provisions is allowed—what we call the“parol evidence rule”in the United States—obviously impacts the evidentiary regime.Equally obviously, the statute of frauds that requires certain contracts to be in writing dominates normally evidentiary principles, as does res ipsa loquitur in tort law.

Just as substantive law can affect the evidentiary process, evidence law can affect substantive law.The examples are legion.Rules of exclusion typically increase and rules of admission typically decrease the costs of litigation.As privileges expand, the cost of litigating and thus enforcing rights goes up in most instances.The ready admission of hearsay makes proof easier（although at the same time perhaps less reliable）, and so on.Discovery rules can dramatically affect parties'incentives to create and search for evidence.Individual rules like the rape relevancy rules can affect the ease with which cases may be prov en.Allocation of burdens of proof can encourage or discourage the bringing of certain causes of action, and so on.Again, we will spend most of tomorrow talking in greater depth about such matters.

There is one other interaction between substantive and evidence law that should be noted.In the United States, but perhaps not in China, evidence underlies everything the lawyer does, since in the United States everything can collapse into litigation.Wills, criminal matters（sentencing based in part on what's in record）, anti-trust, commercial work, everything.Evidence bears upon every other legal field, and the worst case scenario of every legal transaction is the collapse into litigation.In litigation, a crucial variable will be what can be proven.Thus every attorney, no matter how remote from the courtroom, must take the courtroom into account, which means taking the rules of evidence into account prior to litigation so that if litigation ensues the necessary facts can be proven.Good records must be kept and be in an admissible format, for example.

3.Economics.We have a saying in the United States that“There is no such thing as a free lunch, ”which means that, if someone“invites”you to lunch, he probably wants to talk to you about something or may expect a favor in return.Whatever one thinks about this in general, it is literally true for government that“there is no such thing as a free lunch.”Governments are constrained by their economies, and economies are finite.Of the many valuable things that, in theory, governments could do, they are able to choose to do only some of them.If resources are used for one purpose, there are simply fewer resources left to do other things, and there are in total too many different“things”to do to fund them all.Similarly, private resources are finite, whereas the ways in which they can be consumed are almost infinite.

In thinking about the litigation system, the finite limitation of resources is critically important.Difficult choices need to be made about the allocation of resources across the whole range of governmental interests, including the operation of the legal system.Again, let me give a dramatic example.Investments in the criminal justice process obviously compete with investments in other social goods.If government provides more judges, police, or funds counsel for poor people, less resources will be available for economic development or medical research, or whatever.The police face an aspect of this problem daily.Faced with a surplus of crime, they must constantly decide how to allocate their limited resources.Should the police patrol this part of the city or that part?Should they concentrate on economic crimes, crimes of violence, or fraud? Investments within the criminal justice process likewise compete with other investments in different parts of that same process.If government provides more judges, perhaps it can provide fewer police.

The implications of economics extend further and are dramatic.Consider for example an issue of importance to you in China right now—the right to counsel.If the active involvement of defense counsel increases the total time a trial takes, obviously there can be fewer trials, as the total time available to try cases is finite.In general, as the average cost of a case increases, the total number of cases that can be tried decreases.That in turn means that effective counsel will either reduce the number of convictions that occur or will cause the government to substitute other means of enforcement.

Analogous but subtly different issues exist with respect to civil cases.The optimal amount of civil litigation may differ from the optimal amount of criminal litigation, although the matter is complicated.If the costs of litigating civil cases go down, one would predict that the amount of litigation would increase.There is a sense in which this is a public good; indeed, both in your country and mine the inaccessibility, cost, and delay of the litigation process are often identified as problems to be rectified.Yet, they have positive effects as well, in particular through their encouragement of private resolution of disputes.Indeed what to some seems shocking, in the United States even most criminal cases are resolved through what is essentially private negotiation between an accused and the government that results in a plea agreement.

Which leads me to my fourth point.

4.Trials as the Ideal v.Trials as the Embodiment of Social Breakdown.As I have tried to indicate, the implications of the law of evidence go far beyond trial itself.Nonetheless, there is a close association between the law of evidence and trials both in fact and in the public mind.And trials in the west are often idealized and glorified.In part this is because of the role of juries in both our political theory and our self-conception; in part it is because of the glorifi cation（well deserved, let me add）of the rule of law, with trials embodying its public vindication.To what extent this is true in China, I do not know.It is obvious, though, that evidence law is written with an eye toward its use at trial, and consequently how one conceives of a trial may affect what one thinks are the optimal rules of evidence.

The relationship between one's conception of trials and the law of evidence holds no matter what one's conception of trials happens to be.If one accepts the idealized, media glossed view of trials in the United States, one would want to fashion rules of evidence that further the purposes of the public vindication of the rule of law and human rights.Great latitude should then be given to criminal defendants and to the“underdog”in civil litigation.Rights of allocution should be protected, and so on.

But it is not so clear that trials should be glorified in this way, even if（as I believe is true）the rule of law should be.For all their drama, trials reflect a break down in the rule of law.They occur because there have been accusations of wrong doing—civil or criminal—and the parties have not been able to come to an agreement to resolve their dispute.Trials thus are in a sense pathological.They may consume resources that could be better spent in more productive ways（although again getting the facts right one way or the other is critically important to the conditions of flourishing）.At a still higher conceptual level，there can be reasonable disagreements about the ontology—the nature—of disputes themselves.Are disputes primarily between individuals? Or do they have a social aspect？

I raise these questions here not to answer them but to further illuminate the depth and profundity of the questions that lurk under the deceptively straight forward label“the theoretical foundations and implications of evidence.”

I will now try to sum up the various considerations alluded to above.In my opinion, “the theoretical foundations and implications of evidence”pose the following five sets of issues and questions：

1.In its most straight forward manifestation evidence is the study of dispute resolution, with the focus being the interaction and regulation of the various actors in the drama:trial judge, jurors/lay assessors, attorneys, par ties, and witnesses（both lay and expert）.The rules of evidence structure the relationships between these individuals, but the rules are themselves relatively rough distillations of a complex set of factors, and that set of factors contains the real“foundations and implications”of evidence.The rules of evidence are reflective of those underlying ideas.

2.How one constructs trials, and thus the rules of evidence one fashions to facilitate trials, is a function of one's beliefs concerning one of the fundamental questions of human thought and philosophy—what does it mean to know something? A trial is an epistemological event at which claims of knowledge are advanced, considered, rej ected, accepted.Philosophy began with systemic thought on what it means to think and to know, and it remains the dominant question today.Modern relativist philosophies are driven by a fundamental skepticism, for example, as is the desire for revelation reflected in those philosophers with a religious orientation.But this is not the only tradition.There is the spectacularly successful example of modern science; and in order to think clearly about the legal process, one must locate one's thoughts somewhere in or near this spectrum.Whether you think truth of historical facts is obtainable will determine in large measure your view of particular evidence rules.Is there a difference between an“historical”reconstruction of facts, and a“social”one? Indeed, is it even coherent to talk about“knowing”the past?If not, what role should trials play, what values pursued? To what extent should fact finding be located in randomly selected fact finders, and to what extent should the“expertise”of the judges be relied upon? What are the risks of either approach.To what extent should rules of evidence favor one class of litigants over another? And so on.

3.There is another fundamental question:what the purpose of trials is.The typical response has something to do with fact finding, and as I have tried to make clear that typical respond has enormous significance. But, are trials like science? How do they differ? The analogy to science is somewhat tenuous.Legal decision making cannot defer judgment until more information is collected.Also, the judgment to be made is what actually happened rather than what the underlying universal laws might be.Most tellingly, perhaps, there is no organized body of knowledge that is applicable to the typical case, as there is in science.To the contrary, the fact finder has to import the necessary background knowledge for decision.If on reflection trials do not seem a lot like science（at least some types of science）, are they like history? The focus of history is on facts, but as a means, generally, of greater understanding.At trials, understanding is largely irrelevant（except as a matter of persuasion）.

4.But what about all the other purposes that trials may serve? What about the economic aspects of trials? Not only the parties but the lawyers, the judges, the court reporters, and all the court personnel have vested economic interests in trials. What about the symbolic purposes of trials? The political purposes? And so on.There is an extraordinarily complex set of issues that inform the nature of trials.In light of all this, what might be the purposes of the rules of evidence? And again, how do the answers to these questions affect how trials should be constructed? To what extent should trials look like free markets; to what extent should they be regulated by the government? Which model is likely to foster efficient truth-seeking, and why? What other values than efficiency and truth seeking are or should be at stake？

5.Evidence law does not just structure fact finding; it also create incentives of various kinds.How does the significance of accurate fact finding compete with that of other social values？

Evidence and evidence law are not normally thought of as among the many factors that enter into the pursuit of a just and fair society, but as I have tried to show they are critical to such a pursuit.In the West, this neglect of the significance of evidence rests, somewhat ironically, on just how important evidence is to the bedrock foundation of our institutions.It is so important that it is just taken for granted.You in China who are experimenting with different forms of civil organization cannot be so blase about the matter.Without accurate fact finding, and thus without a rational law of evidence, all other pursuits are made exceedingly difficult if not impossible.