Loading ...
Sorry, an error occurred while loading the content.

Re: [ANE-2] Re: Testis unus, testis nullus

Expand Messages
  • Robert Whiting
    I ve been waiting patiently for part II of this message that will explain what the connection is between the law and the historiography of the ancient world.
    Message 1 of 5 , May 22, 2006
    • 0 Attachment
      I've been waiting patiently for part II of this message that will explain
      what the connection is between the law and the historiography of the
      ancient world. Since it doesn't seem to be forthcoming, I will try to
      explain why this response (which is interesting if one is interested in
      the history of law or in comparative legal systems) does not address the
      point at issue (which is the historicity of persons and places attested in
      ancient sources).

      I have already made the point that using analogies between legal
      procedures and scholarship is, in most cases, not productive. This is
      because legal procedures are adversarial processes whereas scholarship is
      not (unless one has a Zoroastrian/Manichaean/Gnostic philosophy where all
      existence is an adversarial process between the forces of darkness and
      light). A legal procedure is a two-person, zero-sum game whereas
      scholarship is not. A legal procedure is a process that forces a decision
      to be made, for better or for worse, whereas scholarship is not.

      At best, in some places the rules of evidence of legal proceedings and the
      rules of evidence of scholarship overlap, but for the most part, they do
      not. So to disqualify "testis unus, testis nullus" as a historiographic
      principle it is not sufficient to point out that it "does not seem to be a
      very good maxim at law" and then demonstrate fairly haphazardly that
      Anglo-American has generally moved away from it. Since the concurrent
      claim that "testis unus, testis nullus" is "highly dubious as applied ...
      outside the law" was also made, it is necessary to demonstrate this
      outside of the efficacy of the maxim inside the law. This, in my opinion
      has not yet been done. See my other comments below.


      On Tue, 9 May 2006, E Bruce Brooks wrote:

      > I can no more resist comment than could Ariel Szczupak, despite equally
      > heroic attempts to let the matter pass. In a word, "Testis unus, testis
      > nullus" does not seem to be a very good maxim at law, and highly dubious as
      > applied (as it seems often to be applied) outside the law.

      Perhaps so, but you must keep in mind that it was God's idea:

      Deut 17:6: At the mouth of two witnesses, or three witnesses, shall he
      that is worthy of death be put to death; but at the mouth of one witness
      he shall not be put to death.

      Deut 19:15. One witness shall not rise up against a man for any
      iniquity, or for any sin, in any sin that he sinneth: at the mouth of
      two witnesses, or at the mouth of three witnesses, shall the matter be
      established.

      I suppose Anyone can have an off day and come up with what seems not to be
      a very good maxim at law, but before you commit yourself to how bad an
      idea it is, you ought to consider the source and see if your fire and
      brimstone policy is up to date and in full force. :)

      > LAW
      >
      > I have it on the authority of a learned colleague that, where Roman law
      > vaguely required plural "witnesses," Ulpian (c200) clarified by saying that,
      > unless an exact number is specified, two witnesses suffice. In a ruling of
      > 334, Constantine further proclaimed: "We have long required that witnesses,
      > before testifying, should be bound by an oath and that belief should
      > preferably be accorded to witnesses drawn from the better sort of people.
      > Likewise we have ruled that no judge in any case should readily allow
      > admission to a single (unsupported) witness. Now we rule explicitly, that
      > the utterance of a single witness should not be heard at all, even if he has
      > the prestige of membership in the honorable Senate." Both these
      > prescriptions are part of the Justinian Code (respectively Digest 22.5:12
      > and Code 4/20:9).
      >
      > Constantine's ruling simply affirms the negative of Ulpian's rule: if
      > "witnesses" are required, less than two will not suffice.
      >
      > Note that this applies to witnesses, not to all forms of evidence, and
      > than even in post-Roman legal tradition, it is presently restricted to
      > Continental (especially Church) law;

      Depends on what continent you are talking about. It is also part of
      Islamic law (since biblical law is one of the sources of Islamic law).
      But it is part of the law codes of many European countries.

      > the Anglo-American tradition (as has already been pointed out) never
      > admitted it.

      What never? -- Well, hardly ever:

      No person shall be convicted of treason unless on the testimony of two
      witnesses to the same overt act, or on confession in open court.
      Constitution of the United States, Article III, section 3

      In addition to the requirement for two witness to treason written into the
      constitution, statutes for perjury in the US also generally have a
      "two-witness rule" associated with them. While it is an anomaly created
      by historical happenstance, and some states have rejected it, it is still
      part of the law in most states. For a history of this rule in
      Anglo-American jurisprudence, see Leo Kearney O.Drudy, Jr., "The Offense
      of Perjury in the Military," Military Law Review 58 (1972), 1-70 [58 Mil.
      L. Rev. 1-70 (1972)]. I suppose that whether the "two-witness rule" is a
      "good maxim at law" or not depends on whether one is prosecuting or
      defending, but it has given perjury the reputation of being "the most
      difficult crime to prove."

      All this makes your statement an unwarranted generalization. Your
      quotation from Wigmore below should be sufficient to show this since he
      restricts his comments to "modern" Anglo-American law and specificly
      points out that relics remain (apparently he has read the Constitution of
      the United States). This clearly shows that it was at one point part of
      Anglo- (if not necessarily of American) law. Otherwise, there could not
      be relics remaining.

      > Wigmore on Evidence (1935ed, p296): "The modern Anglo-American law has
      > fortunately escaped all of these numerical rules of thumb; only an
      > occasional relic remains. / The fundamental principle, then, is that
      > credibility does not depend on numbers of witnesses, but on the quality
      > of their testimony when analyzed by cross-examination and otherwise;
      > that in general the testimony of a single witness may suffice on any
      > issue for submitting it to the jury."
      >
      > Even at law, and even in Continental law, "testis unus" will be seen as an
      > absurd rule if applied to material evidence. At a trial for murder, a gun is
      > produced, which by ballistic test was the murder weapon, and which also
      > contains the fingerprints of the defendant. Would the case be strengthened,
      > or weakened, if a second gun with these characteristics were introduced in
      > evidence?

      It would neither strengthen nor weaken the case. It would only show that
      the accused had a gun in each hand as he blazed away at the victim. Or
      perhaps that after emptying one weapon into the victim the accused picked
      up another weapon and continued. But you make the rule absurd by
      misapplying it. In terms of physical evidence, already you have more than
      one witness in the ballistics match of the weapon and the fingerprint on
      it, each of which is independent of the other.

      But in relation to the ancient world, let us assume that we have a seal
      bearing the name of the servant of a certain king who is otherwise
      unknown. Now let us assume that we find a second seal with the name of a
      differnt servant of this same king. Would the case for the historicity of
      the king in question be strengthened or weakened by the evidence of the
      second seal?

      > I think the verdict of legal practice, considering all types of evidence
      > together, and also the lesson of good humanistic practice generally, and I
      > here have particularly in mind the philological field,

      In defense of philologists, who do indeed use single witnesses to
      establish points of grammar and lexicon, it needs to be pointed out that
      philologists have a technical term, hapax legomenon, that indicates that a
      word or phrase occurs only once in the corpus. It is used as a flag to
      signal that interpretations based on this form should be viewed with
      suspicion, and philologists who neglect to point this out will quickly
      earn the ire of their colleagues.

      > is that evidence is best evaluated on its character, and not on its
      > numerosity. So if we have to take a Latin maxim into the field or into
      > the fray, I suggest, with Wigmore, Constantine (note his qualifications)
      > and, allegedly, the ancient Hungarian parliament, that a better one
      > would be:
      >
      > "Non numerantur, sed ponderantur."

      The trouble with aphorisms like this is that they are like potato chips --
      you can't have just one. While it is doubtless good that evidence should
      be weighed rather than counted, nothing here tells you how to weigh it
      (whereas practically everyone knows how to count). So you need other
      rules as well, like "falsus in uno, falsus in omnibus" (or perhaps its
      contrary "verus in uno, verus in omnibus"). Nothing can tell you how much
      a single piece of evidence has to weigh before it is acceptable.
      Therefore determination of historicity is subjective as long as there is
      only one witness.

      Evidendce should always be weighed, but witnesses must be tested. Note
      that Wigmore expects that it will be possible to cross-examine witnesses
      or to test their testimony by other means. Furthermore, in law, witnesses
      and physical evidence are different categories. Witnesses are expected to
      be under oath and subject to cross-examination. Physical evidence is
      usually introduced by an expert witness who guarantees its bona fides
      under oath. In such cases it is quite possible to weigh the quality of
      the evidence.

      In determining historicity, however, witnesses (textual or art historical
      testimony) and physical evidence (archaeological remains) are usually, if
      not the same, then part of the same thing.

      In unravelling the ancient world, both the message and the medium are
      witnesses. Clearly the potential for cross-examining long-dead scribes
      and artists is nil as is placing them on oath or detemining that they are
      "drawn from the better sort of people." How then are such witnesses to be
      tested?

      One way is by interrogating the medium carrying the message. Physical
      analysis of the object may prove its bona fides. If the medium on which
      the message is found turns out to be modern, then the witness is
      impeached. But even if the medium proves to be bona fide, this does not
      guarantee the authenticity of the message or image on it. Spurious modern
      texts and images have been written or drawn on ancient materials. Forged
      coins have been struck on ancient planchets.

      Another way of testing the witness is through its provenience. Excavated
      materials found with other similar materials and in contexts that indicate
      that they have lain where they were found since they came there in
      antiquity bear a cachet of authenticity. Then it only remains to
      determine that they were "drawn from the better sort of people," that is,
      that the ancient witness was in fact telling the truth rather than
      spinning a yarn for whatever purpose. But still, evidence has been known
      to have been planted into its ancient context (cf. Piltdown man) so even
      material that is found in situ needs to be verified, especially if it is
      the only witness to a significant historical fact and, again, the
      antiquity of the medium is not necessarily a guarantee of the antiquity of
      the message.

      Yet another way to test witnesses is against each other. If the witnesses
      agree then they are assumed to be correct; if they do not, then their
      testimony must be weighed to see which is more likely to be factual.

      But one continually comes back to the fact that, unlike in the law, it is
      very difficult to test a single ancient witness to a presumed historical
      fact. Indeed, the only reasonable method is to test the witnesses against
      each other (and still this requires that the witnesses be separate and not
      dependent on one another). What is implicit in "testis unus, testis
      nullus" is that there is nothing to test a single witness against. I have
      no difficulty with your claim that evidence should be weighed. My only
      question is: Against what?


      Bob Whiting
      whiting@...
    • E Bruce Brooks
      To: ANE-2 Cc: CGC In Response To: Robert Whiting On: Testis Unus From: Bruce Bob has made a detailed response to my earlier message about the principle of
      Message 2 of 5 , May 22, 2006
      • 0 Attachment
        To: ANE-2
        Cc: CGC
        In Response To: Robert Whiting
        On: Testis Unus
        From: Bruce

        Bob has made a detailed response to my earlier message about the principle
        of Testis Unus. It seems to me (in a word) that his response adds detail but
        does not unseat my earlier conclusion, which is that Testis Unus Testis
        Nullus is (a) not a legal maxim, but a maxim based on certain legal
        practice, and that whatever its relevance for the general philosophy or
        practice of law, it is (b) not validly applicable to questions of evidence
        in humanistic investigations. I think all will agree that the more evidence,
        the better. The question is whether one piece of evidence is of any value at
        all. I think that it is, and I think that good humanistic practice agrees
        with me. I will pursue that thought (and try to keep the discussion focused
        on that point) in what follows.

        BOB: I have already made the point that using analogies between legal
        procedures and scholarship is, in most cases, not productive. This is
        because legal procedures are adversarial processes whereas scholarship is
        not . . . A legal procedure is a two-person, zero-sum game whereas
        scholarship is not. A legal procedure is a process that forces a decision
        to be made, for better or for worse, whereas scholarship is not.

        BRUCE: One might choose different points of dissimilarity, or phrase these
        points otherwise, but let it stand as agreed that borrowing from law into
        scholarship is not automatically valid. This is my final point, and I
        consider it gained at the outset. The rest of this discussion concerns
        matters of detail and application.

        BOB: So to disqualify "testis unus, testis nullus" as a historiographic
        principle it is not sufficient to point out that it "does not seem to be a
        very good maxim at law" and then demonstrate fairly haphazardly that
        Anglo-American has generally moved away from it. Since the concurrent claim
        that "testis unus, testis nullus" is "highly dubious as applied . . .
        outside the law" was also made, it is necessary to demonstrate this
        outside of the efficacy of the maxim inside the law. This, in my opinion
        has not yet been done. See my other comments below.

        BRUCE: To start with (and I apologize in advance for the length of this
        note), it strikes me as relevant that the principle of Testis Unus is not of
        universal validity in systems of ancient law; that is, that those who borrow
        it from the law are relying on more cogency in the source culture (the
        culture of world law) than the maxim in fact possesses. I note also the
        previous area of agreement: law and humanistic scholarship are different
        things. But there is nothing wrongful in exploring how far the one may
        provide useful methodological precedents for the other, and so it is worth
        continuing. I continue herewith.

        BOB [to my comment that Testis Unus "does not seem to be a very good maxim
        at law"]: Perhaps so, but you must keep in mind that it was God's idea:

        Deut 17:6: At the mouth of two witnesses, or three witnesses, shall he
        that is worthy of death be put to death; but at the mouth of one witness
        he shall not be put to death.

        Deut 19:15. One witness shall not rise up against a man for any
        iniquity, or for any sin, in any sin that he sinneth: at the mouth of
        two witnesses, or at the mouth of three witnesses, shall the matter be
        established.

        BRUCE: I appreciate the citation of two pieces of evidence, but I wish to
        point out that they do not support each other very well. Deut 19:15
        eliminates one-witness judgements altogether; Deut 17:6 does so only in
        capital cases. What exactly is God's mind in the matter? Or to put it in
        philological terms, what is the stratification of the Book of Deuteronomy?
        There are also conflicts between both these maxims and large areas of (eg)
        Exodus, which make no specification at all as to numerosity of witnesses.
        What, then, is the stratification of the Pentateuch? That question has been
        worked on. If there is a generally agreed answer, it would be highly
        relevant to general legal history, and I would like to know what it is
        (ANE-2 members, who collectively are expert in this matter, are invited to
        respond; by copy of this note, I also wish to query my legal history
        colleagues on this point). Pending an answer, I suggest that, whatever the
        directionality in the Pentateuch may be, whether from no rule to a
        one-witness rule or the reverse, the rule does not seem to be of universal
        application in the legal tradition of Israel. This does not refute my
        perhaps too vague statement that "the rule does not seem to be a very good
        maxim at law;" on the contrary, it brings the law of early Israel into the
        group of instances defined by the Continental > Anglo-American legal
        evolution, in which the Testis Unus principle is not always articulated, and
        where articulated, it sometimes applies to all trials, and sometimes only to
        capital cases, or to certain especially heinous kinds of capital cases. That
        is, the supposed universal principle fails on examination to be universal in
        its home territory. That, in my view, makes it perilous to borrow into
        another territory.

        BOB [to my comment that in Roman law, the rule explicitly forbidding a
        single witness to be heard was articulated no earlier than by Constantine in
        334]: Depends on what continent you are talking about. It is also part of
        Islamic law (since biblical law is one of the sources of Islamic law). But
        it is part of the law codes of many European countries.

        BRUCE: Already noted, but I think it also bears repeating that in the Roman
        tradition which underlies a good many of those modern European codes, the
        specific Testis Unus principle was articulated only relatively late. That it
        was taken over in that late form by derivative traditions does not refute
        that history within Roman law.

        I note again my comment at this point, not addressed by Bob, that the
        principle in question applies to witnesses, but not to all forms of
        evidence. It is at this point that the legal vs humanistic rules of evidence
        might be compared with special relevance. We have no living memories of
        (say) Cleopatra, and so as far as ancient history is concerned, all
        witnesses are mute witnesses, and all evidence is to some extent material
        evidence.

        BOB [on the degree to which the Anglo-American tradition differs from the
        Continental in this respect]: What never? -- Well, hardly ever:

        No person shall be convicted of treason unless on the testimony of two
        witnesses to the same overt act, or on confession in open court.
        Constitution of the United States, Article III, section 3

        In addition to the requirement for two witness to treason written into the
        constitution, statutes for perjury in the US also generally have a
        "two-witness rule" associated with them. While it is an anomaly created by
        historical happenstance, and some states have rejected it, it is still part
        of the law in most states. For a history of this rule in Anglo-American
        jurisprudence, see Leo Kearney O.Drudy, Jr., "The Offense of Perjury in the
        Military," Military Law Review 58 (1972), 1-70 [58 Mil.L. Rev. 1-70
        (1972)]. I suppose that whether the "two-witness rule" is a "good maxim at
        law" or not depends on whether one is prosecuting or defending, but it has
        given perjury the reputation of being "the most difficult crime to prove."

        All this makes your statement an unwarranted generalization. Your quotation
        from Wigmore below should be sufficient to show this since he restricts his
        comments to "modern" Anglo-American law and specificly
        points out that relics remain (apparently he has read the Constitution of
        the United States). This clearly shows that it was at one point part of
        Anglo- (if not necessarily of American) law. Otherwise, there could not be
        relics remaining.

        BRUCE: If some legal traditions have the rule, and others do not (or have it
        only in sharply limited contexts), then the generalization that not all
        legal traditions have the rule is entirely warranted.

        I am not a lawyer, but my impression is that Wigmore, right or wrong (and
        his book has been continually updated, which is at least a sign of ongoing
        relevance), is widely regarded as authoritative for current Anglo-American
        practice. Adding to the list of "relics" (as Wigmore calls them) does not
        refute Wigmore; it just details Wigmore. And I note that here again, the
        crimes to which those relics of Testis Unus apply in Anglo-American practice
        are specially heinous crimes (eg, treason) or involve specially stringent
        contexts (eg, military law; I forbear to quote the jibe on "military music
        is to music as . . ."). The implication, and for that matter the direct
        statement of Wigmore, is that (a) the maxim is rare in practice, and (b) in
        his opinion, it is undesirable in practice. It is invoked only in cases
        where a heightened degree of stringency is thought appropriate. It does not
        apply to normal situations. That leads directly to my comment in the
        preceding paragraph.

        BOB [to my thought that two smoking guns rather than one might actually
        weaken the evidence in a murder case, meant to illustrate the
        inapplicability of the Testis Unus rule to material evidence]: It would
        neither strengthen nor weaken the case. It would only show that the accused
        had a gun in each hand as he blazed away at the victim. Or perhaps that
        after emptying one weapon into the victim the accused picked up another
        weapon and continued. But you make the rule absurd by misapplying it. In
        terms of physical evidence, already you have more than one witness in the
        ballistics match of the weapon and the fingerprint on it, each of which is
        independent of the other.

        BRUCE: As long as we are using imaginary instances, I appeal to the
        judgement of the Court, or anyway that of Conan Doyle, in one of whose
        Holmes stories a piece of corroborative fingerprint evidence, itself of
        seemingly unmistakable import, turns out to destroy the prosecution's case
        by being shown to rest on a physical impossibility. The print was not there
        when the person supposed to have made it was arrested; ergo, he could not
        have made it. Corroboration is not always valid per se, hence the principle
        that corroboration is cogent, or even (to some) necessary, has a weak
        foundation. It is the probity, not the numerosity, of the evidence that
        ultimately decides.

        BOB: But in relation to the ancient world, let us assume that we have a seal
        bearing the name of the servant of a certain king who is otherwise unknown.
        Now let us assume that we find a second seal with the name of a differnt
        servant of this same king. Would the case for the historicity of the king
        in question be strengthened or weakened by the evidence of the second seal?

        BRUCE: Strengthened. The more evidence (as I think I began by saying), the
        better. I can't imagine anyone disagreeing with this very general principle.
        But that is not the question. The question in this example is whether the
        first seal is *any evidence at all* for the historicity of that king, or
        whether (as was implied by an earlier invocation of "Testis Unus" on this
        list) a single piece of evidence is a priori inadmissible. I should say that
        it was admissible, and once admitted, that it goes to support the existence
        of that king. I may also note, in view of the long duration of "trials" in
        these humanistic matters, that if the first seal is disallowed as evidence
        on grounds of its singularity, then there is nothing for the evidence of the
        second seal (perhaps coming much later) to corroborate.

        A weak case is made stronger by a second piece of evidence. The history of
        the "Testis Unus" rule, so far as it is revealed in the present discussion,
        amounts to certain jurists saying, at certain times and places, "In
        especially dire crimes [or, for some, in all crimes]" a conviction in a weak
        case will not be allowed." It is agreed in all variants (as I suppose) that
        bad evidence is no evidence. So if the first seal were epigraphically or
        iconographically dubious, it would be properly disallowed even in humanistic
        adjudications. But what we are here discussing is the worth of one
        unsupported piece of good evidence. Does the fact that it is the only piece
        of evidence weigh against it? Undoubtedly. Does the fact that it is the only
        piece of evidence prevent its being introduced in evidence in the first
        place? That is the question here, and I would answer, No. It makes a weak
        case, if you like, but it still makes a case. Humanistic practice (in
        agreement with at least some areas of legal practice) allows that case to be
        heard, and to be decided. I believe it is right in so doing.

        BOB: In defense of philologists, who do indeed use single witnesses to
        establish points of grammar and lexicon, it needs to be pointed out that
        philologists have a technical term, hapax legomenon, that indicates that a
        word or phrase occurs only once in the corpus. It is used as a flag to
        signal that interpretations based on this form should be viewed with
        suspicion, and philologists who neglect to point this out will quickly
        earn the ire of their colleagues.

        BRUCE: I would like to distinguish between one text and the entire corpus of
        texts in a language. (1) A unique occurrence of a word in a particular text
        (say, one of the Pauline epistles) is sometimes urged against the
        genuineness of that text (say, as a true writing of Paul). But in any finite
        text, where will always be some words used only once, just as there are some
        words used frequently. That is, all texts have one or more unique words or
        usages. That fact then cannot be used against authenticity. If the unique
        word is common in texts of, say, anti-Pauline tendency, that strengthens the
        value of its rare occurrence (when there should be no occurrences, except as
        adversely quoted) in a given dubious Epistle. (2) In the larger case of the
        compiler of a dictionary of an entire language, based on the corpus of known
        texts in that language, it is also relevant that some words are common, and
        others rare (the curve for cumulative frequency has been wrongly called a
        hyperbola; the actual equation is a little more complicated, but it remains
        true of all languages that there are few common words, and many rare words,
        with words of single occurrence being the most numerous category of all. In
        lexicography, it does recommend itself as good practice to regard a single
        attested usage as dubious evidence for the existence of that form in the
        first place (the change of scribal error, or of misinterpretation, is too
        great). Thus, the words in the Oxyrhynchus papyri firmed up previous lexical
        doubts as to many words in NT and allied texts. But what does good
        lexicography do in the pre-Oxyrhynchus period, with a word attested only
        once? Given especially that general frequency predicts that many genuine
        words will be singly attested? As far as I know, and Bob's statement seems
        to confirm my suspicion, good practice notes the singularity of the evidence
        for the word, *but it still includes the word.* It does not fail to record
        the word. That is, a weak (uncorroborated) case is still a case, and the
        evidence for that case is still received. That it is flagged as weak is just
        good practice. The point is that it is there to be flagged. That is, a weak
        case can be allowed to stand in philological practice. The actual practice
        of careful philologists is thus not well described by Testis Unus Testis
        Nullus. It would be better described by Testis Unus Testis Unus.

        I thus find that in this lexicographic example, Bob is actually giving
        evidence on my side of the question. I should not omit to be grateful, and I
        here record my gratitude. A cautiously admitted word is still an admitted
        word. It is not an excluded word, as would be the case if the Testis Unus
        principle actually obtained in good lexicographical practice. No?

        BOB [on my summary suggestion that a more appropriate Latin maxim for
        humanistic evidence would be, "ponderantur, non numerantur"]: The trouble
        with aphorisms like this is that they are like potato chips --
        you can't have just one. While it is doubtless good that evidence should be
        weighed rather than counted, nothing here tells you how to weigh it (whereas
        practically everyone knows how to count). So you need other rules as well,
        like "falsus in uno, falsus in omnibus" (or perhaps its contrary "verus in
        uno, verus in omnibus"). Nothing can tell you how much a single piece of
        evidence has to weigh before it is acceptable.
        Therefore determination of historicity is subjective as long as there is
        only one witness.

        BRUCE: Granted that a treatise on methodology is better than a maxim on
        methodology, and that a rule is not a complete education, I would still be
        inclined to think that an adjuration to consider the weight (the probity) of
        evidence is better than an adjuration to reject all uncorroborated evidence
        out of hand.

        BOB: In determining historicity, however, witnesses (textual or art
        historical testimony) and physical evidence (archaeological remains) are
        usually, if not the same, then part of the same thing.

        BRUCE: Already noted above; ancient situations are different from
        contemporary trials; there are no witnesses properly speaking. The term is a
        metaphor. Our job is to see whether we can save it as a *useful* metaphor.

        BOB: [Describes various ways of lab-testing physical evidence for ancient
        matters].

        BRUCE: Already noted summarily above; these are some of the ways one
        assesses the genuineness of physical evidence. The root question remains,
        after passing that test (the analogue, as Bob notes, of cross-examining
        human witnesses in real time), can a single piece of physical evidence
        validly be relied on for a conclusion? Does the single seal mentioned in his
        example attest the existence of the king in question? Or do we need two
        seals, both found to be authentic by the relevant laboratories, before we
        begin to consider the case? I think the judgement of historians will be that
        the single authentic seal suffices to open the matter, and indeed to reach a
        conclusion in the matter. That conclusion is strengthened by a second such
        seal (equally found to be authentic), meaning that the two-seal case is
        stronger, meaning that the one-seal case is weaker, but a decision can still
        rest on the first seal alone (the weaker case alone). That case *is not
        inadmissible.*

        It helps that historical inferences are not capital crimes, where the
        outcome is that somebody gets executed, and where a later exculpation cannot
        bring the wrongly executed person back to life: that is, where any
        injustices are permanent. Humanistic judgements, so to speak, are open
        judgements, always subject to reversal on the presentation (perhaps a
        century or so later) of adverse evidence, or to confirmation by later
        discovered supporting evidence. Historical judgements are less dire. We
        still want them to be rigorous. We don't want to wind up with a cardboard
        Cleopatra. But the evidence of one object, properly scrutinized and found to
        be authentic and correctly interpreted, can still be admitted to a rigorous
        determination. The rigor consists in the scrutiny, which is to say, in the
        "weighing."

        As far as I know, scientific practice as to hypotheses or conclusions is
        identical: all conclusions are in principle forever open to new evidence, or
        to a better interpretation of old evidence. The humanists do not sacrifice
        rigor by agreeing with the physical scientists in this regard. And the human
        and physical sciences thus agree in departing from the Testis Unus maxim
        (and from such legal practice as is validly summarized by it, which is not
        all legal practice) in this regard.

        So much for the single piece of good evidence. Consider for a moment the
        case where bits of evidence of unequal authenticity are introduced on both
        sides of a question. Suppose the more numerous inauthentic bits point to
        conclusion A, and the less numerous authentic bits point to conclusion B;
        are we then to adopt conclusion A? I should hope not. Are we to adopt
        conclusion A if there is only one bit of authentic evidence in favor of it,
        and many inauthentic ones on the other side? Does the inclusion of a line of
        the Gospel of Mark in ten thousand motel-room Gideon Bibles weigh against
        its absence in Codex Vaticanus? Again, I should think not. And I go on to
        suggest that it is this sort of situation that is well summarized in
        conveniently memorable maxim form by "ponderantur, non numerantur." Counting
        witnesses gets one nowhere; the correct procedure is to assess the worth of
        each witness.

        Of course, one has to go on to learn the specifics of particular situations,
        but that maxim will be a good guide in all those situations. It is a
        high-order principle, telling us how to handle ourselves amid the blizzard
        (or the dearth) of evidence in particular cases. I venture to recommend it
        once again.

        BOB: But one continually comes back to the fact that, unlike in the law, it
        is very difficult to test a single ancient witness to a presumed historical
        fact. Indeed, the only reasonable method is to test the witnesses against
        each other (and still this requires that the witnesses be separate and not
        dependent on one another). What is implicit in "testis unus, testis nullus"
        is that there is nothing to test a single witness against. I have no
        difficulty with your claim that evidence should be weighed. My only
        question is: Against what?

        BRUCE: I find this objection strange, coming as it does after a whole
        catalogue of examples (not here repeated) of how to scrutinize physical
        evidence (including archaeological context, in which I may point out that an
        artifact and its archaeological context are not two pieces of evidence, but
        rather aspects of one piece of evidence).

        Against what is the single piece of evidence to be tested? To answer by the
        book, "It is to be tested against itself; examined for its plausibility as
        really being what it purports to be." There is more to evidence than the
        agreement of witnesses. There is also, and much more important, the
        credibility of witnesses. And the credibility of witnesses is not determined
        by counting them, but by scrutinizing them one by one. Were they present at
        the scene of the crime, are they of known probity, have they (or in the case
        of objects, have their fabricators) an interest in the outcome? And so on.

        In the end, as it seems to me, *all* witnesses are single witnesses, and
        legal practice demonstrates this by cross-examining them singly and
        separately. It does not suffice in court to impugn a witness on grounds that
        there is only one other witness who agrees. The only point validly at issue,
        at any given moment, is the character, the probity, of that one witness.
        This, I should think, remains true whether there are twenty other witnesses,
        or no other witnesses, or anything in between.

        Bruce

        E Bruce Brooks
        Warring States Project
        University of Massachusetts at Amherst
      Your message has been successfully submitted and would be delivered to recipients shortly.