The Influence of the Legal Paradigm on the Development of Logic

by Robert H. Schmidt

I. Introduction

II. The Legal Paradigm in the Historical Development of the Logical Discipline
          A) Aristotelean Syllogistic Demonstration & Induction
          B) Bacon and the Logic of Experimental Science
          C) Viète and the Symbolic Formalization of Algebraic Reasoning

III. The Uniqueness of Legal Reasoning & the Special Problems It Poses
          A) The Pastness of the Events in Legal Cases
          B) The Particularity of the Events in Legal Cases
          C) The Contingent Nature of Events in Legal Cases

IV. Conclusion

References

I. Introduction

Why do we say that reasoning follows laws? Why do we also use the term 'judgment' in logic, which was originally and still is a legal term? Why do we mingle legal and logical concepts in this way?

Argumentation and debate were practiced in the Athenian law courts long before the reasoning process was itself subjected to thematic treatment by Plato, Aristotle, and the Stoics, resulting in the discipline that is now called "logic." This legal background surely provided these philosophers with numerous concrete examples of both valid and fallacious reasoning. (1) And even though in the Prior Analytics Aristotle developed his doctrine of the syllogism (and related logical matters) quite generally and in apparent independence from any special application, we see from his Rhetoric that side by side with its usefulness in philosophical and scientific inquiry, one of the main areas for the application of this new logical discipline remained reasoning in the law court.

Throughout the long history of the development of logic, legal reasoning continued to be a major area of application, even though the new concerns of later logicians were the inferential procedures appropriate to experimental science and the formalization and generalization of symbolic mathematical reasoning. At the same time, a number of thinkers whose work contributed to the profound transformation of the classical logical tradition, most notably Bacon and Leibniz, who subjected it to the most penetrating criticism in accordance with their own programs were either jurists or had significant legal training.

These preliminary observations naturally raise the question of how important the legal context was in the initial framing of logic and its subsequent reconceptualizations. In this paper I will suggest that legal reasoning provided the basic paradigm and point of reference for Aristotle's thinking about logical reasoning in general, and that the legal paradigm reasserted itself in equally fundamental ways at key moments in the later development of logic.

Whereas before (as I will show) the legal paradigm was often invoked in the rethinking of the logical discipline itself, in the present modern revival of interest in the nature of legal reasoning and its connection with logic, there is one important difference. Today the main concern has primarily been how to apply or adapt logic to the concerns of the courtroom. In the modern attempts to create a science of evidence or science of proof, at least three distinct starting points or areas of concentration have been suggested for an examination of the problem of courtroom evidence and the determination of its relevance to some allegation; these are all currently foci of lively debate, as may be seen from the other papers in this volume. The three starting points are: 1) the role of generalizations, explicit or implicit, insofar as concrete legal inferences are drawn from them; 2) the role of stories or narratives as the means by which jurors determine the overall plausibility of evidence; and 3) the suggested employment of probabilistic reasoning, whether mathematical or non-mathematical, in the analysis of facts.

These modern discussions have raised interesting and fundamental problems, although they for the most part take it for granted that logic in its present state is adequate to the task of devising a science of evidence applicable to the courtroom. This is supported by a popular opinion that logic has undergone a continuous and progressive development since Aristotle, and that it is an autonomous and perfectible discipline virtually free from ontological assumptions. These are highly questionable presuppositions, to say the least.

In fact, the history of logic is a chronicle of hotly debated and still unresolved controversies, which do not merely concern details such as the number of valid figures of inference or the Aristotelean modal syllogism, but issues just as basic as those faced by researchers in the problem of legal evidence itself. The first commentators on Aristotle were split on the very question of whether logic was an exact discipline (epistm or techn) in its own right, and if so what was its proper subject matter; or whether it was simply a tool (organon) that could be used in the service of any of the special sciences. Today, there is still disagreement as to whether logic has to do with words, thoughts, or objects. Furthermore, the various contemporary logical schools each favor assimilating logic to either linguistics, psychology, epistemology, or mathematics. (2)

Again, even a cursory survey of the history of logic would have to distinguish at least three moments, or three distinct conceptualizations of logic, which overlap, intertwine, and cross-reference one another in very confusing ways. First of all, there is the classical tradition represented by the Organon of Aristotle and the few surviving fragments of Stoic propositional logic. (3) Here it is important to note that the Aristotelean logical writings were effectively out of circulation for nearly three hundred years following the death of Aristotle. When they were finally published in the first century B.C.E., it was Stoic logic that held the field. Thus, the logical writings of Aristotle were given their first and guiding interpretation by commentators who were steeped in Stoic propositional logic. Stoic logic then effectively drops out of the picture. An intense development of the quasi-Aristotelean logic followed through Scholastic times well into the Renaissance, producing many different schools that disagreed on a surprising number of details of this syllogistic logic.

Next, there is the logic that developed side by side with modern experimental science. Beginning with Francis Bacon's critique of deductive syllogistic and his rethinking of the Aristotelean method of induction, and supplemented by a later reinterpretation of the hypothetical syllogism (as briefly mentioned by Aristotle) into the method of hypothesis or abduction, this logic becomes formalized into the "scientific method" so simplistically presented in elementary school textbooks.

Finally, there is modern symbolic logic, sometimes called "logistic." Although an initiative in this direction may already be seen in the work of Ramon Lull, the real sustained treatment begins with the dream of a mathesis universalis or scientia universalis in Descartes and later John Wallis (prompted by Proclus' remarks about a katholou pragmateia in the preface to his commentary on the first book of Euclid 's Elements). The paradigm for this was already provided by Viète's symbolic formalization of algebra. This dream was turned into a project or program by Leibniz with his universal characteristic, later to be executed by De Morgan, Boole, Jevons, Peano, Frege, Pierce, Russell & Whitehead, and a host of others.

Most modern textbooks of logic, whatever their orientation, in one way or another recognize or acknowledge these three moments in the history of logic, although they usually give the impression that these moments have been conceptually reconciled to each other in the subsequent "development" of logic as an exact discipline. However, it should be pointed out that the originators of the new movements in logic did not consider themselves to be mere continuators, perfecters, or revisionists of the logical tradition, but rather innovators striking out in entirely new directions. The reasons they themselves gave for rejecting the traditional Aristotelean logic are quite different from the reasons now adduced in its criticism. (4)

By opposing themselves to the tradition in various ways, these innovators successively reinterpreted Aristotelean syllogistic at the same time. Additional reinterpretations of syllogistic were later required in order to welcome it back into modern logic. Behind all of this lies Aristotle's own conception of logic, whatever that may have been. (5) Furthermore, modern textbook accounts usually mask profound changes in the nature of abstraction, symbolization, and intentionality (6) that make it highly questionable whether modern logic is anything more than a collage of conceptually incompatible pieces. Far from being one seamless whole, from this historical perspective modern logic seems to be riddled with conceptual fault lines.

Given this state of affairs, I believe that modern logic could itself profit not only from a rethinking of the role that the legal paradigm has played in its development, but also from a fresh confrontation with the unique problems posed by the attempt to apply logical reasoning to legal cases. I can think of no better way to expose some of the fault lines mentioned above.

II. The Legal Paradigm in the Historical Development of the Logical Discipline

In this section I would like to show how the legal paradigm played an instrumental role in the conceptualization of logic at each of the three key moments of its history articulated above.

A. Aristotelean Syllogistic Demonstration & Induction

In his Rhetoric, Aristotle addresses the problem of finding the proper means of persuasion in public speaking. He regards rhetoric as a branch of both dialectic and ethics, since the rhetorician must not only know how to reason correctly, which is something provided for in dialectic, but he must also understand human character and emotions, and interestingly enough Aristotle's most substantial treatise on the emotions is found in this book. Aristotle further subdivides rhetoric into three kinds, depending on the class of listeners to which it is directed: political when directed to assemblies or governing bodies that have to make decisions about future courses of action; forensic when directed at judges or juries that have to decide about past events; and ceremonial, concerned with praising or blaming men for their present behavior in front of any sort of observer.

The proper subject matter of rhetoric is further restricted to events that are not necessary, but which instead allow of being otherwise and happen for the most part. These are the contingent events about which we can deliberate or make inquiries, most notably the actions of human beings. Nevertheless, rhetoric does not deal with these events in their particularity, but insofar as they are common to men of a certain type. Aristotle adds, somewhat mysteriously, that whereas dialectic proper is concerned with the truth, rhetoric is concerned with what is "like the truth."

As far as forensic rhetoric is concerned, the materials with which the rhetorician has to deal are five: the law, witnesses, written documents such as contracts, evidence obtained under torture, and oaths. Strictly speaking, since these are givens and not invented by the rhetorician himself, they are called "non-technical" (atechna). This is to contrast them with the reasonings that he does produce, those which are "technical" (entechna) or properly within the province of his discipline, which are divided into two major classes: enthymemes, which are rhetorical syllogisms; and examples (or paradigms), which are rhetorical inductions. Enthymemes are distinguished according to whether their premises are likelihoods (eikota, often and misleadingly translated as 'probabilities') or signs (smeia), both usually expressed as commonly accepted opinions; the maxim or gnome (gnm) is a simple statement referring to practical conduct that Aristotle regards as a defective enthymeme, since one can construe it as either a premise or conclusion in a potential syllogism where the listener automatically supplies the missing parts. (7) Examples (paradeigmata) fall into two classes: actual historical parallels; and parallels invented by the speaker, which latter are either illustrations (parabolai) or fables (logoi).

But what is most immediately important for our purposes is that Aristotle claims to have largely pioneered the systematic treatment of rhetorical reasoning, whereas earlier writers on rhetoric had mainly concerned themselves with the subject of how to arouse emotions in their hearers.

I have gone into this extensive summary partly because Aristotle's treatise was to be the definitive text on rhetoric for centuries and thus had a profound effect on the development of legal reasoning, but also to show that since Aristotle understands the enthymeme to be a rhetorical syllogism, and the example a rhetorical induction, it might easily be thought that he is merely adapting his general logic as developed in the Prior Analytics to the concrete concerns of the political forum or the courtroom. However, when we examine the language that Aristotle uses in his Categories and On Interpretation, which are preliminary to the Prior Analytics itself, we find that Aristotle's fundamental conceptualization of syllogism is expressed in terms of language umistakeably drawn from the Athenian law courts.

In the Categories, each of the fundamental modes of predication, such as species, genus, quantity, quality, etc., is called a 'category' (katgoria), a Greek word that most concretely means an 'accusation' or 'allegation' in the legal sense. (8) A sentence such as 'The man is tall' predicates or accuses the man of being tall, according to the category of quantity. Implied in this term is the notion that simple predication holds the subject responsible in some way, treats it as a cause (aitia).

In On Interpretation, Aristotle distinguishes declarative speech (logos apophantikos, or simply apophansis) from other uses of language, such as questioning or requesting, on the grounds that only declarative speech is capable of being true or false. Such declarative speech is either affirmation (kataphasis) or denial (apophasis). Affirmation is basically saying "yes" to a question posed, denial saying "no" to a question posed. Because Plato and Aristotle regard thinking (dianoia) as holding a dialogue with one's self — answering one's own questions, as it were — it is possible for an affirmation or denial to be the simple assertion of one's opinion (doxa) as to the truth or falsity of some claim. In fact, the verb (apophanesthai), from which Aristotle's term for declarative speech derives, can mean simply that: the expression of one's opinion on some matter. (9)

Here it is important to note that according to Metaphysics, Book Delta, (10) which explains the different Greek meanings of the Greek verb 'is' (einai), the 'is' of accusation or predication is different from the 'is' that affirms or denies something to be true or false. Thus, in the sentence 'The man is tall', the 'is' can be either that of affirmation or of accusation, depending on the intention of the speaker.

Now, it is very clear that Aristotle regards a premise in a syllogism as just such a declarative affirmation or denial. (11) At the same time, he characterizes the two terms into which a premise may be resolved as 1) the accusation made and 2) that against which the accusation is made. (12) Thus, it seems that the affirmation or denial is the assertion, within the same sentence, of the truth or falsity of the accusation being made against something. But if someone expresses his opinion about some accusation, either through his own inner dialogue or as a response to a question put to him, this is effectively testimony (marturia). So it seems that for Aristotle even the simple sentence or logos is a collapsed dialectic of accusation and testimony, reflecting a deep connection between the 'is' of predication and the 'is' of affirmation.

Accusation and testimony are correlative terms, implicitly referencing each other. The accusation requires testimony, and the testimony is about, or for the sake of, the accusation. Each is incomplete by itself. It is mainly when the testimony for an accusation may be adduced through direct observation or reporting that these two functions may be legitimately coalesced in the manner mentioned above, as if one were really saying that "I affirm such and such to be the case, for I have seen it." But often the affirmation and testimony are distinguishable, as when one charges the sum of the internal angles of a triangle with being equal to two right angles, based on the affirmation that the sum of these angles is equal to a straight angle. In other words, in the syllogism the conclusion is the accusation and the premises the testimony. It is worth noting that an argument was commonly called a marturion in Greek (that is, a 'testimonial'), indicating that the premises were regarded as the evidence for the conclusion, even though Aristotle does not himself often use such language. Also, it is important to point out that the evidential character of the premises is conceptualized in terms of witness testimony rather than in terms of one of the other kinds of primary evidence enumerated by Aristotle in the Rhetoric.

The traditional syllogism is sometimes described as a certain kind of correlation of propositions. This description is ultimately based on Aristotle's own characterization of the syllogism as logos in which certain things having been posited, something different from them results necessarily from their being so posited. To speak of the syllogism in this way as a correlation of propositions is merely a formal characterization. It does not let us concretize or intuit the nature of this correlation. But Aristotle explicitly calls the syllogism a logos, a kind of higher order logos, to be sure, but a logos nonetheless. Now, I have tried to show that even a simple logos was regarded by him as an implicit correlation of testimony with an accusation; and again consistent with the legal paradigm, that the conclusion of a syllogism was regarded as an accusation while the premises were regarded as testimony. That is, the legal paradigm helps make clear and distinct the exact nature of this correlation. It is, in effect, the primary evidence for the correlation itself.

Aristotle's term for the inductive process, epagg, also had concrete legal applications. It was used for the act of adducing the testimony of witnesses, of quoting or citing an author in testimony to a claim, etc. The exact distinction between syllogistic demonstration and induction in Aristotle's own writings is sometimes obscure, and he occasionally even seems to use these words interchangeably. This may be because one of the primary meanings of apodeixis, the Greek word for 'demonstration', is likewise the production of witnesses. The modern understanding of induction as a process of passing from particulars to generalities is only one of the senses in which Aristotle uses this word. Suffice it to say here that this process must still ultimately be explicated as a correlation of testimony to allegation.

Attending to the concrete meaning of katgoria as 'accusation' also helps us get a clearer concrete understanding of the process of syllogistic inference Aristotle calls "demonstration" (apodeixis). In Greek, the verb apodeiknumi most concretely means 'to point away from other objects at some one object' , or else 'to exhibit or produce something from something'. An accusation is a demand to select one predicate rather than its contradictory, to hold the subject guilty or not guilty of the accusation, as when we must decide whether Theatetus is flying or is not flying. Then, when the accusation is decided upon through the demonstration process, it effectively selects one of these two contradictory predicates while excluding the other; it effectively says, "this, not that." At the same time, the conclusion or accusation is exhibited from (13) the premises or testimony. Given the generally multivalent character of technical terminology in all classical Greek writers, I would suspect that both of these meanings are intended simultaneously.

Plato also employs the term apodeixis in his dialogues Sophist and Statesman in the context of trying to come up with a definition of these two types of men. There he employs the method of dichotomous division, where he begins with a very general class, subdivides it into two differentia, subdivides these diffentia themselves, in a relentless search for the defining characteristic of the sophist or the statesman. (14) But when Plato runs down his total hierarchy of divisions, taking care to take the right fork at every crossroads of the division, he calls this procedure the "demonstration," presumably because he is selecting the correct path while neglecting the other in his exhibition of the entity being defined.

Logos the most common Greek term for 'speech' but also the root of the word 'logic', in its most fundamental sense refers to the picking out or selecting of something, "abstracting" in its most original sense. But selection can be done either by disregarding what one is not interested in, by saying "this, not that," or by choosing something according to its likeness to something else, by saying "this, like to that." Of course these two procedures also imply one another. But we can emphasize either the disregarding or the comparing. If we concentrate on selection through disregarding, we have apodeixis, or demonstration; if we favor the comparison, we have paradeigma, or example. I maintain that both these acts are originally understood by Aristotle as two different ways of correlating evidence (whose paradigm is legal testimony) with an accusation. Therefore, we can come to a provisional characterization of evidence as that which steers us toward an accusation by disregarding the alternative, or that which helps us select it by means of a comparison.

Finally, let me also mention that Aristotle understands the sense or meaning belonging to individual isolated words to be established "by convention" (kata sunthkn), but this conventionality is explained in terms of their character as symbols (symbola), where he is almost certainly playing on the original concrete meaning of the Greek word, which refers to each of the two broken halves of some object, one of which was kept by each of the parties of a contract or indenture. One way of interpreting Aristotle's choice of this word is that the spoken word as an expression of the speaker's meaning, and the meaning understood by the listener through this word, must be in agreement; it is as if the spoken word and the heard word are the two copies of a contract.

B. Bacon and the Logic of Experimental Science

Underlying the inductive logic of experimental science as originally proposed by Bacon is the conception of the experiment as a means of "proof" that goes beyond the mere observation or contemplation of nature. Nature will not give up her innermost secrets unless she is put into unnatural and artificial constraints. To quote four telling passages in which Bacon articulates this view:

"I mean it [his Natural History or Physics] to be a history not only of nature free and at large (when she is left to her course and does her work in her own way) ... but much more of nature under constraint and vexed; that is to say, when by art and the hand of man she is forced out of her natural state, and squeezed and molded. Therefore I set down at length all experiments of the mechanical arts, of the operative part of the liberal arts, of the many crafts which have not yet grown into arts properly so called, so far as I have been able to examine them and as they conduce to the end in view. Nay (to say the plain truth), I do in fact (low and vulgar as men may think it) count more upon this part both for helps and safeguards than upon the other, seeing that the nature of things betrays itself more readily under the vexations of art than in its natural freedom. (15)

Nature exists in three states, and is subject, as it were, to three kinds of regimen. Either she is free and develops herself in her own ordinary course, or she is forced out of her proper state by the perverseness and insubordination of matter and the violence of impediments, or she is constrained and molded by art and human ministry. The first state refers to the "species" of things; the second to "monsters;" the third to "things artificial." For in things artificial nature takes orders from man and works under his authority; without man, such things would never have been made. But by the help and ministry of man a new face of bodies, another universe or theater of things, comes into view. Natural history therefore is threefold. It treats of the "liberty" of nature, or the "errors" of nature, or the "bonds" of nature, so that we may fairly distribute it into the history of "generations," of "pretergenerations," and of "arts;" which last I also call "mechanical" or "experimental" history." (16)

Among the parts of history which I have mentioned, the history of the arts is of most use because it exhibits things in motion and leads more directly to practice. Moreover, it takes off the mask and veil from natural objects, which are commonly concealed and obscured under the variety of shapes and external appearance. Finally, the vexations of art are certainly as the bonds and handcuffs of Proteus, which betray the ultimate struggles and efforts of matter. For bodies will not be destroyed or annihilated, rather than that they will turn themselves into various forms. (17)

As soon, however, as I have leisure for it, I mean to draw up a set of questions on the several subjects, and to explain what points with regard to each of the histories are especially to be inquired and collected, as conducing to the end I have in view — like a kind of particular topics. In other words, I mean (according to the practice in civil causes) in this great plea or suit granted by the divine favor and providence (whereby the human race seeks to recover its right over nature), to examine nature herself and the arts upon interrogation. (18)

At its most extreme, the Baconian experiment can be regarded as means of torturing nature so that she will give her truest and most genuine testimony. In a milder form, it is understood as an interrogation of nature that compels her to answer questions of our own devising. In any case, the legal overtones are unmistakable here, and we must remember that Bacon was himself a renowned jurist. In order to obtain evidence about nature, the experimenter puts nature on trial.

In later centuries, this experimental testimony comes to be conceptualized as data on the one hand, and facts on the other.

As far as I have been able to determine, the scientific concept of a datum is a conflation of the logical concept of the conceded or granted, a characterization originally of the syllogistic premises of argumentation, and the mathematical concept of the given, referring to numbers or geometrical figures that can be found or constructed and thus are available for further calculation or construction. So conflated, the testimony provided by experimentation is understood to be material for the inductive or hypothetical processes used in science, and no longer for syllogistic inference.

Collaterally with the concept of a datum there emerges the scientific concept of a fact, which has an even more colorful history. In classical Latin, a factum was originally anything that was done or made, but even in early Latin, this meaning began to be transferred to any occurrence or event, as if it were the result of an impersonal agent. In accordance with its meaning of something done or made by man, it was thematically adopted by Vico as the proper object for his New Science of History. Relying on a Leibnizian distinction, Vico gave the characterization of concrete and certain (certum) to the facts of history. This is in opposition to the truths of mathematics, which were designated as abstract and common. Because the most certain and concrete matters are ones determined by direct observation, the fact concept was opposed to inference. So characterized, and harkening back to the general sense of any occurence, the fact concept was appropriated by the physical sciences. Today the terms 'fact' and 'datum' are used virtually interchangeably.

Originally, then, the act of proving belongs to the same class as acts of testing, trying, assaying, etc. (19) When the term 'proof' is appropriated by mathematics and logic, it begins to lose some of this sense and comes to be used almost interchangeably with the term 'demonstrate,' even though it originally derives from a radically different semantic field.

As forms of reasoning, Baconian induction, and the method of hypothesis (or abduction) developed later as a means of reasoning from facts or data already known by experience to facts not yet experienced, cannot be understood apart from the compelled testimony afforded by experimentation. The exact correlation between the data or facts that this kind of testimony produces and the inductions and hypotheses drawn from it must be explicated in at least as much detail as we attempted to provide in the case of the Aristotelean syllogism. This is a somewhat subtle problem that we will take up at another time.

C. Viète and the Symbolic Formalization of Algebraic Reasoning

Juridical notions may also be detected operating behind the scenes in the early stages of the development of symbolic logic. The defining characteristics of this kind of logic are the employment of ideographic or typographic symbols, and the formalization of the reasoning process in terms of axioms that permit the "derivation" of propositions symbolically expressed from certain definitions symbolically expressed. Both of these features are already present in a mathematical context in the Introduction to the Analytic Art of François Viète (1591). It is to Viète that we owe the systematic and self-conscious introduction of literal symbolism into algebra, and the first formal axiom system for the manipulation of this symbolism. For Descartes, Wallis, and Leibniz, this algebra became the paradigm for a more general formal and symbolic discipline that later develops into symbolic logic.

In the preface to this work, Viète, who was also a jurist, declares that it was necessary for him to think out an entirely new vocabulary for his new algebra. In particular, Viète called the proper object of his algebra as designated by letters species, rather than number or magnitude. Viète never gave a clear explanation of why he chose this term, except to say that "logistice speciosa is what is exhibited through species or the forms of things, as for example, through the letters of the alphabet;" (20) and there was considerable disagreement among later mathematicians over just what Viète meant by this term. However, John Wallis, who took up the dream of a mathesis universalis after Descartes, had this to say:

The name of Specious Arithmetick is given to it (I presume) with respect to a sense wherein the civilians use the word Species; for whereas it is usual with our Common Lawyers to put Cases in the name of John-an-Oaks and John-a-Stiles or John-a-Down, and the like (by which names they mean any person indefinitely who may be so concern'd) and of later times (for brevity sake) of J.O. and J.S. or J.D. (or yet more shortly) of A, B, C, etc. In like manner, the Civilians make use of the Names of Titus, Sempronius, Caius, and Mevius or the like, to represent indefinitely, any person in such circumstances. And cases so propounded they call Species. Now with respect hereunto, Viète (accustomed to the language of Civil Law) did give, I suppose, the Name of Species to the letters A, B, C, etc., made use by him to represent indefinitely any Number or Quantity, so circumstanced as the occasion required. And accordingly, the accommodation of Arithmetical Operations to Numbers, or other Quantities thus designed by Symbols or Species, was called Arithmetica Speciosa or Specious Arithmetick; the word Species signifying what we otherwise call Notes, Marks, Symbols, or Characters, made use of for the compendious expressing or designation of Numbers or other Quantities. (21)

Here we see that the symbolization intrinsic to symbolic logic is first conceptualized in a manner consistent with the legal practice of the time. According to Wallis' view, the literal symbols such as A, B, etc., in the law are understood to be little more than general verbal abbreviations standing for any case you please. This is in fact a bit of an anachronism, since one of the defining characteristics of the literal mathematical symbolism employed by Viète and later mathematics is that it is operational, meaning that it is directly manipulated in the derivation of equations. Operational symbolism has more the nature of the counters on an abacus or other calculating machine. The reinterpretation of the operational symbolism of mathematics back into verbal abbreviations has contributed much to the confusion over the nature of symbols in modern symbolic logic. Here we see the role that the legal paradigm played in this reinterpretation. The implications of this are too complex to be pursued in this paper.

Furthermore, when Viète imports from Euclid's Elements the common notions and the general theorems of ratio manipulation found in Book V into his new algebra as axiomatic, he collectively calls them symbola (perhaps best translated in this context as 'stipulations'), drawing on the most original and concrete legal sense of the Greek word as one of the two parts of a contract or "indenture" that will perfectly match the other (see above). This term is almost certainly introduced by Viète to convey a subtle sense of how the two sides of an algebraic equation are related to one another in algebraic analysis, and also how the equality in successive derivations is insured by the stipulated axioms. (22)

Therefore, we might consider referring the evidential character afforded by derivation in the reasoning of symbolic logic back to the primary evidence that a contract offers. Here again, we will leave the detailed exposition of this problem to some other occasion.

In this section I have suggested that the fundamental, and fundamentally distinct modes of reasoning described by syllogistic demonstration, proof, and derivation correlate conclusions with evidence in very different ways, and that all of them were originally influenced by the legal paradigm. Syllogistic demonstration is clarified by the legal notions of accusation and testimony; proof relates its inductions and hypotheses to testimony obtained by the physical analogues of interrogation and torture; derivation is originally conceptualized in terms of contracts and stipulations. These are in fact three of the primary kinds of evidence cited by Aristotle in the Rhetoric.

Even if my historical interpretation is incorrect, or even if the legal paradigm was merely invoked metaphorically to provide terminology for the new concepts of the developing logic, I think I have shown that we do not need to be simply satisfied with the application of existing logic to legal cases, but we can use the legal paradigm to initiate a fresh rethinking of logical reasoning itself. (23)

III. The Uniqueness of Legal Reasoning & the Special Problems It Poses

In this section I would like to discuss the somewhat paradoxical circumstance that even though the legal paradigm has been several times invoked at crucial moments in the development of logic, legal reasoning remains something of a second class citizen, or negative reference point, when it is measured against the normative formulations of logic. At the same time, this issue gives me a chance to discuss the qualifications of some other disciplines that may wish to claim the science of legal evidence as falling within their own jurisdiction.

From a metalogical point of view, what is unique about the problem of legal reasoning is that it deals with events that are past, particular, and contingent (possible/probable but not necessary); consequently, legal reasoning has been understood to deal with belief or conviction rather than true knowledge. Each of these characteristics poses a special problem for the application of normative logical reasoning, which has implicitly been based on what is atemporal or eternally present, general, and necessary.

A. The Pastness of the Events in Legal Cases

As stated above, Aristotle distinguished forensic from other kinds of rhetorical reasoning on the ground that it concerns events that are inherently past rather than present or future. Although this is perfectly obvious, it has some interesting, albeit subtle, implications. Mathematical reasoning may be characterized as inherently atemporal, in the sense that its truths are independent of time. Alternatively, they may be regarded as eternal in the sense that they are true at all times. In either case they are not inherently past. And the inferential process itself, insofar as the syllogism is defined in terms of necessary consequences, is also regarded in this way.

Now, it may also be argued that the modern physical sciences deal with events in their presentness. Not only do they seek to account for present effects in terms of past causes or to predict future effects on the basis of present causes, but they do this as if the past cause or future effect could have been observed by the scientist, as if the scientist could have been present with the past or future phenomena. This is implied by the statement that in order for a phenomenon to fall within the province of science, it must be replicable. Even if the characterization of modern science that I have presented here seems a bit contrived, it should serve to underscore the point that forensic events are irreducibly past — that the legal reasoner must assume that he cannot even in principle have been in the past to observe some event. (24)

This suggests that we might look to historical science as a means of founding a science of evidence relevant to legal cases, since this is the one discipline that seeks to understand the past in its pastness. After all, in its reliance on the testimony of earlier historians and chroniclers, other surviving written documents, and surviving material exhibits such as coins and statuary, for the purpose of reconstructing past events, it faces almost exactly the same problems as a potential science of courtroom evidence would, where the primary evidence is witness testimony, written documents, and material exhibits. (25)

However, most historical writings of the Greeks and Romans can hardly be said to be historical in the sense that they treat human existence as composed of events that have some intrinsic property that entitles them to be considered historical rather than scientific, economic, or social. A sign of this is that history had no place in the ancient classifications of the sciences — that is, it had no subject matter peculiar to itself that would entitle it to be an autonomous discipline. Instead, history had its own muse like the other imitative or "fine" arts. The ancient historical writings are mostly animated by a moral, inspirational, chauvinistic or other practical agenda, selecting their events accordingly.

Of historical writings that have any claim to being theoretical rather than merely descriptive or moralistic — that is, of historical writings that purport to understand the coherence, integrity, and lawfulness of the events in human history in terms that are inherently historical — there are those writings in the classical tradition wherein human history is understood to derive its meaning from a divine plan and take its direction from a spiritual order. This divine plan and perfect order is regarded as eternal or outside of ordinary time, whereas imperfect human existence extends this plan in and through time according to God's providential ordering of all human events. In other words, human events are referred to a perfect paradigm.

As outstanding examples of this kind of historical writing, we cite Augustine's City of God and Dante's Divine Comedy. For Augustine, the paradigm is the drama of creation, fall, redemption, and salvation, and human history plays this out in its own imperfect way. For Dante also, secular history duplicates the spiritual history of man: the history of Empire follows the history of the Church.

However, it is also important to note that for Augustine and Dante historical succession is not an accidental property of the lives of the men who make history, but belongs to their very essence as temporal beings. The coherence, integrity, and meaning of the events in human history do not result from the selection of certain events according to the subjective yardstick of a given historian; rather, it has an objective basis in the temporality of human existence. In other words, if we understood this human temporality properly, we would have access to the truth of human history. We would know which events were essential and which accidental. We would be able to discern the natural articulation moments in an apparently continuous historical process. We would know how different events in human history were interconnected. And all the other things that would make for a true science of history.

Vico may perhaps be regarded as the founder of scientific history in the modern sense. His New Science of History, published in the early 18th century, comes to grips with some of the conceptual problems involved in making history an autonomous discipline.

For Vico, history has its own proper subject matter in the facts of human life. These facts (facta) are understood to be the things that human beings have done or made. They are characterized as certa, which literally means 'settled' or `decided upon', because they are the result of deliberation and decision. Thus, when a historian of Vico's stamp looks at the events of human history, the historical succession itself, he does not regard these events as "happenings" or events in the same way that natural science would. He sees them as something made and decided upon rather than something that merely is; this is the essential property that entitles them to be called "historical."

Again, since the proper subject matter of history is these concrete and certain facts, the external history of human events is due to what men have thought, their customs and ideas. History derives its intelligibility from the circumstance that men deliberate upon, do what they do, and make what they make, for various purposes or ends — we might say because of their various "motives" — and history accordingly and appropriately employs final causes, or the doctrine of means and ends in order to make sense out of the succession of human events, whereas natural science for the most part confines itself to efficient causes.

Now, for Vico, the historical order is not outside of time as it was for Augustine; that is, the historical succession is not an imitation of some timeless order. Rather, it has its own laws that govern the succession of events in ordinary temporality. Instead of contrasting the historical order to a timeless or divine order, Vico contrasts the laws of the historical order to the laws that govern the natural world.

What insights might we draw from this discussion of historical science that might have relevance for a potential science of legal evidence?

First of all, historical science almost by definition has the task of examining past events in ways radically appropriate to their pastness. To the extent that it adopts the perspective of mathematics or natural science (as opposed to accepting the evidence they can provide in certain cases), it begins to capitulate on the possibility of becoming an autonomous discipline. If historical existence is essential to man's being, if he in some sense is what he has been, then might there not be some special ambiguity and uncertainty in the reconstruction of events that results simply from their pastness and not from the mere inconvenience that we were not there to observe them ourselves? Perhaps it is a consequence of the pastness of events that our interpretations of them seem contradictory; or perhaps they are inherently so, as the Hegelian historical dialectic might be interpreted to mean. Again, by what right are we entitled to understand past events in terms of the chains of cause and effect so dear to modern science, rather than in terms of some other more appropriate connective principle? Surely the past is no more accessible than the sub-atomic world where ordinary causality may break down. These are, to be sure, difficult metaphysical questions. But if we take the pastness of the events studied in legal cases to be an essential and not merely an accidental characteristic, then a proposed science of legal evidence can hardly avoid them.

Secondly, historical science has previously felt the need to invoke paradigms in order to give itself an "objective" character. Here it is appropriate to cite the recent observation that jurors frequently create stories or narratives to string evidence together in a plausible way and come to their decisions. Aristotle himself recognized the role that historical parallels, fables, and other sorts of analogical reasoning could play in the arguments of lawyers. The question then becomes what guides the selection and ordering of the evidence in the story-telling processes of lawyers and jurors. It has been pointed out that consciously or unconsciously they often invoke parables for this purpose. But what paradigms and parables could have sufficient generality while retaining the necessary specificity to give legal reasoning a convincing objectivity comparable to that postulated by Augustine and Dante with their single, perfect and eternal paradigms? Obviously there could not be one paradigm that would serve to clarify human behavior in all possible legal cases. We must then ask ourselves the question whether there is a finite number of patterns of human behavior, or at least a collection of common and basic behaviors that could give such legal reasoning the needed objectivity. Clearly, in the story-telling process mentioned above jurors draw on the patterns they have noticed in their own individual and subjective experience of human life. However, to make these patterns thematic in a way that would be appropriate to a science, we would almost have to go in the direction of modern archetypal psychology, and I am not sure many legal theorists would wish to take this step.

However, this may not be as far-fetched as it may seem at first. In the selection, ordering, and articulation of detail the story-telling process is quasi-numerical. The earliest meanings of the Greek word logos, from which our word 'logic' derives, have to do with recounting, giving an account, telling a tale, etc. Now, Jung late in his life attempted to understand the archetypes in terms of arithmological structures, as Plato had earlier tried to understand the structure of the ten fundamental Ideas in terms of eidetic numbers. So there is a kind of logic associated with story-telling insofar as it is guided by an archetypal paradigm having an intrinsic numerical structure. Finally, I would like to add that the words 'rite' and 'ritual' are cognate to both Greek and Old English words (arithmos and rim, respectively) that mean number. Given this network of interconnections, we might very cautiously infer that the succession or interconnection of human events that imitates or spreads out through time the numerical structure of these archetypes, might be conceptualized in terms of the moments of a ritual rather than causal chains.

In the ancient descriptions of ritual, the first moment of the ritual, such as the washing of hands or the shearing of the hair of the sacrificial victim, is called the katarch — the ceremonial event in accordance with which the beginning is made. The conclusion of the ritual is called the telesma, the concluding flourish. Other moments may be articulated in between this beginning and ending. It is interesting that these two words parallel the important Greek words arch and apotelesma. Now, each of the four causes of Aristotle is called an arch, the beginning that rules over what comes later; (26) apotelesma is a common Greek word for 'outcome' or 'effect'. But the ritualistic terminology does not point to a causal connection of events, but rather to another way of rendering their connection intelligible; for, the sequence of moments in a ritual is prescribed and unvarying. And if one objects that the effect must follow immediately upon the cause, then the ritualistic connection would be an articulation of that crucial moment. After all, it is supposed to spread within and throughout ordinary temporality an archetypal timeless pattern.

  Arch , usually translated as "principle" was often   rendered by Jacob Klein as "ruling beginning" or   "commanding origin" to bring out a sense which we   no longer notice in the Latin derived word.

This is not to say that cause and effect are not operative in the past. However, it may be that the proper way of regarding events in their pastness is in terms of something like this ritualistic connection. These two connective principles need not be mutually exclusive. It is possible, for example, that the ritualistic connection is intended to articulate or lay out the structure by which the cause is itself connected to the effect.

B. The Particularity of the Events in Legal Cases

Legal cases are also concerned with particular events rather than general truths such as the Pythagorean theorem. Again, this is perfectly obvious, but it poses a special problem since according to the classical logical tradition stemming from Aristotle, there can be no true knowledge of particulars in their particularity. (27) Aristotle even goes so far as to say that most arguments and inquiries are concerned with predicates that can be properly predicated of other things and have still other things predicated of them. (28) In other words, there is both an upper and a lower limit of predication, and neither the particulars in their particularity nor the highest genera in their generality are really the proper subjects of syllogistic argumentation. This would indicate that Aristotle's logical norm is designed to deal with the intermediate entities, and indeed Aristotle's examples throughout the Prior Analytics bear this out.

It is true that Aristotle makes provisions for reasoning syllogistically about the highest genera, such as species, genus, quantity, etc., which are in the province of philosophy or metaphysics, by means of the art of dialectic. The dialectical syllogism is distinguished from the demonstrative or scientific syllogism according to the principle that the dialectical syllogism is based on premises that are merely commonly accepted opinions, whereas the demonstrative syllogism is based on principles that are true and knowable in themselves. For example, if we premise that which is more permanent and more constant to be more real, this might be considered a commonly accepted opinion with a wide range of application. Now, it is understood that the commonly accepted opinions apply to these highest genera only for the most part, or only by analogy. So if we tried to conclude from this premise that time was less real than the species or forms of things because time was essentially impermanent, we may not be saying the truth, because the premise may not apply to time at all or perhaps only by analogy with the way in which it applies to other things.

Moreover, it is the hallmark of dialectical argumentation to draw up arguments both for and against a proposition. However, this should not be understood as mere debate. Dialectical argumentation is itself understood by the Greeks to be a higher order logos, just as the syllogism was. Consequently, it is still conceptualized as an interplay of evidence and accusation, except that in this case the evidential component is not the two premises of the syllogism, but rather the contradictory arguments themselves. And even though Aristotle does say that the truth should be more easily supported by such argumentation than what is false, it is by no means the case that the side with the most or better arguments routinely gets the prize. Dialectical argumentation above all seeks the truth. But the less supported side of such an argument may often be more important in this search because the attempt to understand why it is false draws attention to why the generally accepted premise does not apply in this case, and in this realization there is new knowledge. Thus, dialectical arguments in Plato and Aristotle are often deliberately orchestrated so as to come to an antinomy or impasse (aporia). As Aristotle says, an impasse in the argument points to an unrecognized knot in the matter under investigation. It may also be discovered that the logical opposition assumed in the contradictory propositions being argued was itself falsely drawn. I believe that this original intention of dialectical argumentation, to serve as a tool in the investigation of truth, was lost in subsequent centuries, to be replaced by the debased presupposition of modern debate that the best supported argument has proved the truth of its proposition.

It is very tempting to make a parallel here with Bacon's conception of a physical experiment. Whereas the scientist must put nature into an unnatural condition, impose stresses and strains on her, compel her to give up her deepest secrets, the dialectician must perform a similar experiment in his own soul. Through his objections to an argument, he must put himself into an unnatural condition. He must deliberately and artificially bring arguments into conflict. Only by means of such cross-examination and impasse will the key insight be obtained.

But there are other logical tools that can be employed when dealing with the highest genera, matters of definition, or other matters of first principle in philosophical inquiry, and these need not be mediated by general statements at all. First of all, there are hypothetical syllogisms, where the connection between two propositions is conceded on the basis of some deep-lying affinity recognized by the philosopher, but this connection is unmediated by any other premise. Here we can adduce Descartes' infamous "Cogito ergo sum," or "I think, therefore I am." This statement is formulated as an argument, although there is clearly no major premise. It would be a profound error to suppose that this was intended to be a full syllogism where the major premise "All men who think exist" was simply omitted. In fact, the context of this statement in Descartes' Meditations makes it clear that such a major premise is not even in principle to be supposed, since Descartes has resolved to doubt everything he can, from the evidence of his senses to the truths of mathematics, and if he brings into doubt any evidence drawn from his senses, he cannot suppose that other men exist, let alone think. In fact, I think the best way to interpret this statement is to paraphrase it as "I cannot be thinking unless I exist. I am thinking; therefore I exist." (29) This is a perfect example of a hypothetical syllogism, unmediated by any generalizations.(30)

In fact, it is so fundamental that Descartes considers the movement from premise to conclusion in this hypothetical argument to be the preeminent intuitive act of the mind; other propositions are to be accepted as true to the extent that they come up to this standard of clearness and distinctness. The philosophical literature is full of similar arguments. (31)

When it comes to the particulars, we must make a distinction. Truths about particulars in mathematics, and to some extent in the science of nature, may be adduced as instances of general truths, and such instancing can be put into syllogistic form: e.g. The sum of the interior angles of a triangle is equal to two right angles; the figure ABC is a triangle; therefore, the sum of its interior angles is equal to two right angles. Here the major premise is true and known from prior demonstration, while the minor premise simply relates this particular triangle to the universal, triangle. However, this is merely the most trivial kind of inference. (32)

However, here too Aristotle makes provisions for syllogistic argumentation concerning particular events such as those encountered in legal cases, by employing the device of commonly accepted opinions as premises. As examples of such premises, Aristotle offers: "The envious hate" and "The beloved show affection." (33) These general premises, which Aristotle agrees can be the result of inductive generalization, are clearly to be used in the examination of particular events. If we can establish that A was envious of B, we can with some conviction conclude that A hated B. Aristotle also allows a kind of hypothetical syllogism here, which he calls a 'sign'; for example, the fact that some woman is producing milk is a sign that she has recently given birth.

It seems that Aristotle is bent upon forcing all paths to knowledge and means of persuasion into the mold of syllogism or induction. (34) Commonly accepted opinions are invoked as a kind of makeshift to deal with the upper and lower limits of predication in a quasi-syllogistic manner. But the fact that hypothetical argumentation, which has so often been resorted to in philosophical inquiry, is not properly speaking syllogistic (35) and has, so to speak, a logic of its own, may be an indication that there could be developed a kind of inference scaled to the lower limit of predication, or the level of the particulars, that likewise is not mediated by inductive generalizations in the manner of syllogism — a kind of logic of the concrete and particular. However, I would be the first to admit that such a logic has not yet been developed. This should not cause us to abandon the quest for this kind of logic, however, which must make provisions both for some art of persuasion for the attorney (as an analogue to the syllogism), as well as an art of deliberation and decision (as the analogue of dialectic at the scale of particulars).

To a certain extent, the attempt to apply the methods of probability theory to legal cases might be regarded as an attempt to deal directly with the particulars in their particularity. If we can enumerate and give relative weight to all possible outcomes in a given situation simply by examining that situation (and this is of course a very big "if"), we can avoid the necessity of invoking commonly accepted opinions or inductive generalizations at all, unless we consider the commonsense reasoning (36) and theorems underlying the mathematical theory of probability to be generalizations applying to all situations where there is a variable outcome. But isn't that in fact just to say that it applies to a certain class of particulars insofar as they are particular?

C. The Contingent Nature of Events in Legal Cases

The particular past events subject to legal inquiries belong to the class of those that are possible rather than necessary. Aristotle frequently characterizes the possible as that which allows of being otherwise, while the necessary is that which does not allow of being otherwise; so the possible and the necessary in this sense are opposites.

But more exactly, legal reasoning wants to show that some allegation is likely, not merely possible. To do so, it must reason from premises that are likelihoods (eikota), those things that happen "for the most part." Here there is an occasion for confusion. In some translations, eikota is rendered as "probabilities," and a modern reader may very easily think that Aristotle would accept general statements about the behavior patterns of given individuals as probabilities that could be used for premises; for example, such a statement as "Whenever Albert goes out to eat, he manages to get drunk." However, Aristotle makes it very clear that the kind of likelihoods that he is referring to are those that apply to a certain class of men for the most part, not to a given individual for the most part, as with the examples quoted earlier: "The envious hate" and "The beloved show affection." (Remember that Aristotle states that there can be no systematic knowledge of particulars in their particularity.) This kind of likelihood (or verisimilitude) is very different from the probabilities of the modern mathematical theory, which are used to calculate the odds for a particular event.

Moreover, rhetorical induction is not, strictly speaking, a generalization from instances that properly belong to some class, but rather an argument from paradigms, which are only analogues or likenesses to the particular case under examination.

Aristotle goes so far as to say that rhetorical argumentation in general must content itself with arguing from what is merely like the truth, from the verisimilar, which is yet another reason why I have spoken of the influence of the legal paradigm on the logical discipline itself. This is not the same as sophistical reasoning, which is based on what only appears to be true, but it is an unsettling concession nonetheless, since Aristotle recognizes that there is a special kind of sophistical reasoning to which the lawyer is vulnerable—that is, reasoning that only appears to be verisimilar.

IV. Conclusion

I have tried to show that in its concept formation logic has borrowed repeatedly from the legal paradigm at the same time that it has defined itself in opposition to that very same paradigm. (37) Let me briefly recapitulate and summarize the results of the above investigation.

From Aristotle's original understanding of the correlation of premises and conclusion in demonstration as testimony and accusation, respectively, Bacon introduces the concept of proof to accompany experimentation in nature, where the primary evidence is obtained by careful and direct sensory examination of nature's behavior under compelled conditions likened to torture. The inductions (and later, abductions or hypotheses) of experimental science were conceptual modifications of their counterparts in Aristotle's logical writings (Aristotelean induction and hypothetical syllogism, respectively), designed to accommodate the new experimental evidence concept. It is significant that the characterization of the conclusion as an accusation is left behind at this point, when the classical syllogism is rejected on the grounds that it is incapable of providing new knowledge in the context of the experimental investigation of nature. Concurrently with Bacon's work, we have the "new algebra" of Viète, the prototype of the symbolic formalization of the reasoning process, where the original characterization of the subject-predicate relation in terms of accusation is reinterpreted as equation, itself conceptualized as contractual agreement. At the same time, the movement from premise to conclusion in derivation is itself insured by axioms that are likewise characterized as stipulations.

But even from the beginning, the legal paradigm came into conflict with the mathematical paradigm. Curiously enough, in Aristotle this did not affect the form of the syllogism as a correlation of evidence and accusation, but only the certainty with which the premises could be held. Since the truths of mathematics were eternal rather than past, general in some sense rather than particular, and necessary rather than contingent, the premises of the demonstrative or scientific syllogism could lead to real knowledge, and not merely result in persuasion. As a matter of fact, for many centuries thereafter logicians tried to explicate mathematical reasoning in terms of the Aristotelean syllogism. With Bacon, the legal paradigm comes into conflict with a new program for the investigation of nature through experiment. Here the events in nature are understood to be present and subject to direct observation by the senses, rather than past; particular, but instances of general truths discoverable through induction; and actual rather than contingent. At this point, the character of the new objects to which experimental inquiry is directed begins to affect the form of reasoning itself, with the consequent rejection of the syllogism in favor of the twin methods of induction and hypothesis. Finally, a fresh confrontation with the truths of mathematics results in a complete overhaul of the reasoning process itself in the development of symbolic logic.

The final act in this story involves the readmission of the traditional syllogism into formal symbolic logic, with the understanding that insofar as it was confined to the subject-predicate relation, and did not recognize many of the other kinds of relation found in mathematics, it can now only play a very restricted role in a far more general formal logic. However, by this point no trace survives of the original conception of the syllogism in terms of testimony and accusation. And legal reasoning, which specifically deals with the relation of testimony and evidence to an accusation, is relegated to one small corner of the traditional province of the syllogism.

So what can we say about the historical development of logic? Can we say that it has been itself a logical process? It developed in accordance with a paradigm, originally the legal paradigm, but later the paradigms of experimental inquiry and mathematics took over. But reasoning in accordance with a paradigm is one of the hallmarks of rhetorical reasoning. In the systematic selection of its norm or the yardstick against which all other reasoning methods were measured, it deliberately and decisively turned away from the field of objects studied in rhetoric itself, in preference for those found in science and mathematics. But to select something in this way, by turning away from certain things to point at something else — that is, to declare "this, not that" — is apodeixis or demonstration in the most original sense. And the most concrete instance of such a decision process has always been the judgments and decisions of the courtroom. So even in turning away from the legal paradigm as it did, logic was following the legal paradigm in a more profound way.

In the dialogue called Statesman, Plato calls paradigm and demonstration the warp and woof of argument. It seems to me that in this study of one episode in the history of thought, they have also been the warp and woof of the historical development itself.

But again, has this historical development of logic followed a logical path? Has there been a smooth transition from evidence to conclusion? Has history successively subjected the earlier tradition to a piercing and torturous cross-examination to bring out what the earlier logicians were really after? Have the later developments been derived from the former with a formal consistency that must compel agreement now that logic has reached its present condition? Or is the logical development full of paralogisms, misleading and unjust trials of the original evidence, and flawed derivations due to inapplicable axioms?

Or has the history of logic been rather the ritualistic extension, elaboration, and articulation through time of some eternal archetypal paradigm that has its most concrete expression in the reasoning that binds together evidence and accusation in the highly ritualized setting of the courtroom? For, it is there that the evidential and inferential processes, the application of some yet undiscovered science of evidence, are regulated and governed by the rules of evidence — the union of the laws of thought with the lawfulness of thought.

In the above account or historical narrative, I have examined some of the historical evidence for the influence of the legal paradigm on the logical tradition. I trust that my interpretations of this evidence have been sufficiently torturous. I sincerely hope that I have not made a fetish of foolish consistency. Finally, I here explicitly reject the prevailing modern presupposition—what we might call a kind of regulative historical generalization — that the modern sciences and disciplines contain some self-regulating or self-corrective measures that will insure their perfectibility, their own intrinsic ability to repair the fissures and fault-lines that continually emerge in their applications. I believe that this is particularly true of logic in its modern state. It cannot claim to be itself the science of evidence. A sign of this is that it cannot itself answer the question: What would be the evidence for a potential science of evidence? This question is too fundamental for logic itself to answer, no matter what degree of self-testimony it indulges in, no matter how much it tortures itself, no matter how self-consistent it is, and no matter what oaths it takes. After all, we can hardly expect logic to incriminate itself!

One way to bring this out is to ask here the more specific question: What would constitute evidence for the correctness or adequacy of a proposed theory of legal evidence itself? Let us postpone for the moment the problem of defining the term 'evidence' in an exact manner — in the classical tradition, a definition should be the outcome of an inquiry into first principles, not the starting point — and merely pay attention to its etymology. 'Evidence' comes from the Latin evidentia, which originally simply means 'clearness' and 'distinctness', but later becomes associated with those signs, tokens, testimonials, etc., that make something else clear and distinct.

With this provisional characterization in mind, we can explicate evidence in the legal context in a somewhat broader sense than has customarily been understood.

Suppose we follow the traditional division of legal evidence into testimony, written documents, and material exhibits. Let us call this the primary evidence, analogous to the primary evidence or data resulting from direct observation in the physical sciences. The primary legal evidence is evidence both of some fact and for some accusation, which can be a source of some confusion. These two meanings correspond to the credibility of the testimony, on the one hand, and to its relevance, on the other.

It would seem that we must distinguish this primary evidence from the inferences drawn from it by the attorneys themselves, which can also be called evidence insofar as they can apply to the accusation. In general, it is the inferential procedure that transforms the primary evidence of some fact into evidence for some accusation.

Next, there is the evidence resulting from the deliberations undertaken by the jurors, which resolves conflicting testimony of witnesses or conflicting inferences made by the attorneys; this could be some compelling fact or inference that is either discovered in the deliberation process or something already in the record that is seen anew in their attempt to come to a decision — just as an impasse reached in dialectical reasoning may point to the knot in the problem.

Finally, we must ask the question whether there is a theory that directs, guides, and validates the correlation of the reasoning methods mentioned above to the primary evidence. Of course, then we must ask ourselves what is the evidence for the correctness of this theory that could potentially add so much persuasiveness to legal reasoning about the primary evidence — as predictability and replication are regarded as evidence of the validity of some scientific theory.

In this exposition we have isolated at least four different senses of the term 'evidence' in the legal context alone. These four types of evidence do not appear to be merely coordinated differentia of some general evidence concept, as discrete and continuous quantity are primary differentia of the category quantity; nor does one type of evidence seem to have such priority over the others that they can all be reduced to a single basic concept of evidence. It seems more likely that these four types of evidence constitute equally important "moments" or "components" in some more complex conceptual evidence structure that has yet to be fully articulated.

Obviously, one of the difficulties with the very idea of a theory of evidence is that all theories require evidence to validate them. So there would be a certain introverted or self-referential character appertaining to a theory of evidence that would make it quite unique. Logic belongs at the second level of the hierarchy mentioned above since its proper concern is the correlation of evidence with accusations, or more generally, premises with conclusions. As I indicated earlier, modern logic has not yet even systematically treated of the concerns of the third level, which deals with the evaluation of a collection of inferences in their totality, and the deliberations and decisions based on them, this being also the level of ancient dialectic. But what we require is a metalogical discipline corresponding to the fourth level.

The most far-reaching attempt to devise a metalogical discipline in modern times, phenomenology, was apparently developed in complete independence of the legal paradigm; it was rather the direct result of Husserl's reflections on the foundations of logic in his Logical Investigations. As a matter of fact, phenomenology might be understood to have already attempted to give a foundation for a science of evidence in the most general sense, although I do not believe that it has yet been introduced into contemporary discussions on legal evidence. In phenomenology we find not only a definition of evidence, but the ever-deepening explication of this definition at levels that roughly correspond to the stratification of the legal evidence concept as I have laid it out above.

However, it is very surprising that the problem of legal evidence is not thematically addressed by the phenomenologists, since the central concept of phenomenology is intentionality, a term that in its original scholastic context referred to the willing that ensues upon deliberation, and thus was in the general framework of the deliberation and decision process that has its most concrete exemplification in the legal context. It seems to me that theorists searching for a science of legal evidence could gain some preliminary insight and direction in their inquiries by studying the results of phenomenology, but at the same time I believe that phenomenology itself could profit from a confrontation with the problem of legal evidence. This is a subject that I intend to pursue in a subsequent paper.

In other words, I am here making an argument for the fundamental character of the problem of evidence, particularly as it is posed in a legal context. By this I mean that the problem of evidence is one that can bring us into a confrontation with first principles in an original way. I hesitate to call such inquiry metaphysical, not just because of the suspicion with which such inquiry is generally viewed in modern times, but also because a fair examination of this problem may well take us beyond the framework of recognized metaphysical inquiry, ancient or modern. I would prefer to think of it simply as an inquiry into first principles.

References

Aristotle. Categories. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Aristotle. Metaphysics. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Aristotle. On Interpretation. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Aristotle. Prior Analytics. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Aristotle. Posterior Analytics. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Aristotle. Rhetoric. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Bacon, Francis. The New Organon and Related Writings. Ed. F. H. Anderson. Indianapolis : Bobbs-Merrill, 1960.

Cohen, Morris R. and Nagel, Ernst. An Introduction to Logic and Scientific Method. New York : Harcourt, Brace & Co., 1934.

Croce, Benedetto. The Philosophy of Giambattista Vico. Trans. by R. G. Collingwood. New York : Russell & Russell, 1964.

Husserl, Edmund. Logical Investigations. Trans. by J. N. Findlay. New York : Humanies, 1970.

Kant, Immanuel. Critique of Pure Reason. Trans. by Norman Kemp Smith. London : MacMillan & Co.,1964.

Klein, Jacob. Greek Mathematical Thought and the Origin of Algebra. Trans. by Eva Brann. Cambridge MA : The M.I.T. Press, 1968.

Klein, Jacob. Lectures and Essays. Annapolis , MD : St. John's College Press, 1985.

Pierce, Charles S. Essays in the Philosophy of Science. Indianapolis : Bobbs-Merrill, 1957.

Plato. Sophist. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Plato. Statesman. Oxford Classical Edition. Oxford : Oxford Univ. Press.

Spinoza, Benedictus De. The Principles of Descartes' Philosophy. Trans. by H. H. Britan. LaSalle , IL : Open Court , 1961.

Sprague, Rosamond Kent, ed. The Older Sophists. Columbia , SC : University of South Carolina Press, 1972.

Vico, Giambattista. The New Science of Giambattista Vico. Trans. by T. G. Bergin & M. H. Fisch. Garden City, NY: Doubleday & Co, 1961.

Viète, François. Introduction to the Analytic Art. Trans. by J. Winfree Smith. Klein p. 328.

Footnotes