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

Re: [Synoptic-L] balance of plausibility

Expand Messages
  • Stephen C. Carlson
    ... Thanks, I appreciate your interest. ... That s what I meant by instead of ; would something like ... but the Mark-Q replaces the third hypothesis with
    Message 1 of 4 , Oct 17, 2004
      At 12:03 AM 10/18/2004 +1000, Tim Lewis wrote:
      >I’d like to comment on one point of Stephen Carlson’s review of
      >Kloppenborg’s review [NTS 49 (2003): 210-236] of Mark Goodacre’s Case
      >Against Q.

      Thanks, I appreciate your interest.

      >In trying to level out the burden of plausibility between the Two-source and
      >Farrer theories, you point out in your weblog (Tuesday September 7th 2004
      >[http://www.hypotyposeis.org/weblog/2004_09_01_arch.html%5d) that the
      >Two-source theory (Mk-Q theory) actually requires two more hypotheses than
      >does the Farrer theory (Mk-Mt theory):
      >|The first two of these [hypotheses] are shared with the Mark-Q theory, but
      >|instead of the third [Lukan dependence on Mt], the Mark-Q theory adds three
      >|more: that Q existed, that Matthew used Q, and that Luke used Q.
      >But this is an overstatement. Working backwards:
      >1) The Two-source theory requires "that Lk used Q" instead of Mt so is not
      >quite an aditional hypothesis.

      That's what I meant by "instead of"; would something like "... but
      the Mark-Q replaces the third hypothesis with three others ..."
      be clearer?

      >2) On the Farrer theory Mt’s sources for the non-Markan material are simply
      >unknown. But this is not identical to supposing that Mt did not use any
      >sayings sources at all (Goulder is virtually alone here). The Farrer theory
      >remains silent about (yet implies?) Mt’s collection of sayings material from
      >somewhere (oral?).

      That's right. For all practical purposes (Streeter notwithstanding),
      the Mark-Q theory remains silent about the Matt and Luke's use of
      sources other than Mark and Q, and that's OK too. As far as the
      synoptic problem is concerned, an source to one of the synoptics that
      is not shared with any of the others is supererogatory: though they
      might explain the presense of some particular feature or Sondergut
      in one particular gospel, they are not strictly necessary to explain
      the literary interrelations between two or more of the synoptics.

      >3) Stephen you seem to be confusing Kloppenborg’s Q theory (where Q is a
      >unified document) with the standard supposition of the theory (whereby "Q"
      >is merely a symbol designating the body of common Mt-Lk material and need
      >not imply anything about Q as a supposed document). Thus "that Q existed" is
      >therefore not an "added" hypothesis for those agnostic about the character
      >of Q—it is the simple converse of not supposing Lukan dependence on Mt for
      >the double tradition.

      I do believe that Q as a single, presumably unified, document is the
      standard supposition of the Mark-Q theory, as least as far as the
      published literature is concerned. I am aware that many teachers
      are less committed to this aspect and view Q more broadly as merely
      a cover symbol for the Double Tradition. This may reflect a certain
      amount of unease in the standard Q hypothesis, but, given the highly
      extensive verbatim agreements in some of the Q material as well as
      large-scale agreements in sequencing of the sayings material, some
      recourse to a document that substantially accounts for these literary
      features must be made. Both the Farrer theory and the Q hypothesis proper
      propose such a document (Matthew and Q, respectively). An oral body
      of Q material cannot account for either the verbatim or the sequential
      agreements, while a fragmentary solution involving several documents
      or notes might handle the literary micro-agreements (at the word level)
      but not the macro-agreements (i.e. the larger, compositional level).

      That being established, the next issue is whether the hypotheses of
      "that Q existed" + "Matt used Q" + "Luke used Q" can be compared with
      the one of "Luke used Matt." Unfortunately, only a "partial ordering",
      not a "total ordering", can be imposed on the complexity of theories.
      It's like comparing apples and oranges. Unless there is a single
      standard that both can be measured against (usually arbitrary, e.g.,
      number, price, weight, volume, etc.), who's to say that 1 apple is
      "more" or "less" than 3 oranges? In a partial ordering there are some
      comparisons that can validly be made, even though there are some
      comparisons that cannot validly be made. For example, we can definitely
      say that 2 apples are more than 1 apple, and that 4 oranges are more
      than 3. We can even say that a group of 1 apple and 3 oranges is more
      than a single apple (with no oranges), that this group is more than
      just 3 oranges, and even that it is more than 1 apple and 1 orange.

      Comparing the complexity of synoptic theories is a little like that.
      We can easily compare the two-source theory with the three-source
      theory because the latter theory includes all the hypotheses of the
      former and adds another hypothesis: that Luke used Matt too. But,
      in terms of comparing Farrer and Q it is more like the 1 apple of
      Luke's use of Matt compared with the three oranges entailed by the
      Q hypothesis. Who's to say that one is more complicated than the

      To move forward, some reduction to an external standard has to be made
      and the most common is simply counting the number of extra entities
      (Occam). If those extra entities are documents, then we can compare
      3 documents (Mark, Matt, Luke) versus 4 (Mark, Q, Matt, Luke). If we
      follow Kloppenborg and also count utilizations, then we have three
      utilizations (Mark->Matt, Mark->Luke, Matt->Luke) versus 4 (Mark->Matt,
      Mark->Luke, Q->Matt, Q->Luke). Note that documents are incommensurate
      with utilizations, so we would have to reduce them to a single standard,
      e.g. "hypotheses" if we want a total ordering. If we keep them distinct,
      however, then only a partial ordering can be imposed, but in the case of
      Farrer's Mark-Matt theory versus Kloppenborg's Mark-Q theory, under any
      combination of these criteria and orderings, Farrer is simpler.

      But others might feel that Farrer's 1 apple is somehow equivalent to Q's
      three oranges on some yet-to-be articulated standard, so let us go to the
      level of methodology, which you do:

      >At the level of methodology, both theories at first seem equally as
      >plausible. But Mk-Q theorists can expect Farrer theorists to have to
      >demonstrate Lukan dependence on a known document (Mt) which is inherently
      >more difficult (than on an unknown/vague set of "Q" materials) and it is
      >therefore harder to make one’s case sound plausible. Thus unless one accepts
      >Kloppenborg’s Q, the burden of plausibility unfortunately lies moreso with
      >Farrer theorists.

      Methodologically, we have a problem in how to evaluate competing
      but incommensurable hypothesis. The problem here is not so much
      that one is clearly superior to the other. If that were the case,
      I wouldn't find the synoptic problem so interesting. No, the
      problem is how to assess two different theories that are both
      plausible, and it is here that the allocation of the burden of
      proof/persuation/plausibility is crucial.

      As to where this burden ought to be allocated, I would suggest
      the burden of persuasion ought to be placed on the theory that
      has the less relevant (i.e., verifiable or falsifiable) evidence

      In terms of Farrer versus Q, there is more evidence available
      to evaluate whether Luke used Matt: we have MSS that allow us
      to reconstruct the respective texts and we can examine those
      texts for evidence in favor of Luke's utilization of Matt and
      for evidence against that utilization. On the other hand,
      have any direct evidence of the text of Q, so we have no idea
      whether it contained the same kind of features in it that we
      would have found problematic for utilization in extant documents.

      Yes, we can attempt to reconstruct Q in accordance with the
      Mark-Q theory and compare that reconstructed Q text with the
      (reconstructed) texts of Matt and Luke, but the key difference
      is that the texts of Matt and Luke are (or can be at least)
      established independently of a particular solution to the
      synoptic problem and would therefore contain evidence that
      could conceivably falsify the utilization of Matt by Luke.
      This is not the case for the reconstructed Q text--without
      an actual MS witness to Q, the reconstructed text of Q is
      at best a fragment of its true self and there is no principled
      way to reconstruct Q at present that could possibly contain
      the evidence that would falsify it. (In fact, any such
      presumably falsifying portions of a reconstructed Q would be
      viewed as falsifying only that reconstruction of Q, not of
      Q's existence.)

      Thus, the Farrer theory and the Mark-Q theory are not on a
      level playing field. There is simply a lot more evidence
      available that could conceivably falsify Luke's use of Matt
      than Luke's use of Q. Because of this, the burden of
      persuasion ought to be put on the Q proponent to show
      that available evidence actually does falsify the Farrer

      The burden of proof/persuasion is about who deserves to
      get the "benefit of doubt" when there is not enough evidence
      to subject both to a thorough, rigorous, critical analysis.
      In the case of Farrer versus Q, there is an asymmetry of
      available evidence. Any benefit of the doubt should go to
      the side having more available, relevant evidence.

      If the analysis of the available evidence clearly falsifies
      Luke's use of Matt, then we can go on to explore less evidenced
      theories, putting the burden of persuasive on the next most
      evidenced theory. One alternative would be Q, of course, but
      other possible contenders could include any special pleading
      used to "save" the Farrer theory, such as presumed but uncor-
      roborated or unevidenced motives on the part of the Third
      Evangelist, overlapping oral tradition, differing texts of Matt
      and/or Luke, etc. It is only when that these ad hoc hypotheses
      to "save" the Farrer theory (if it needs it) becomes so
      unevidenced that Q may then be entitled to the benefit of the
      doubt. (Note that the Mark-Q theory has its own set of ad hoc
      hypotheses too, including presumed but uncorroborated or unevidenced
      motives on the part of the Third Evangelist, overlapping oral
      tradition, overlaps between Mark and Q, differing texts of Matt
      and/or Luke, etc.)

      However, if it appears that arguments on both sides seem
      to be about equally plausible, then there is no need to
      hypothecate Q.

      >P.S. Stephen, How does one post a brief comment concerning your updated
      >Synoptic Problem Home Page?

      I haven't decided whether to enable general comments on my new Synoptic
      Problem Website, but in the meantime you can leave comments on my
      Hypotyposeis weblog or here on Synoptic-L.

      Stephen Carlson
      Stephen C. Carlson mailto:scarlson@...
      Weblog: http://www.mindspring.com/~scarlson/hypotyposeis/blogger.html
      "Poetry speaks of aspirations, and songs chant the words." Shujing 2.35

      Synoptic-L Homepage: http://www.bham.ac.uk/theology/synoptic-l
      List Owner: Synoptic-L-Owner@...
    • Ron Price
      ... This is only part of the method. One also has to incorporate into the equation what I would call explanatory power . We then get the following
      Message 2 of 4 , Oct 18, 2004
        Stephen Carlson wrote:

        > To move forward, some reduction to an external standard has to be made
        > and the most common is simply counting the number of extra entities
        > (Occam).

        This is only part of the method. One also has to incorporate into the
        equation what I would call "explanatory power". We then get the following

        Two Source Theory, with one hypothetical (Greek) document

        Explains the close textual agreements in the double tradition
        Explains why Matthew and Luke have many doublets whereas Mark has hardly any
        Gives the best explanation of several cases of Lukan primitivity in the d.t.
        Explains how Matthew did so well at preserving so many primitive sayings

        Farrer Theory (no hypothetical documents)

        Explains the close textual agreements in the double tradition
        Explains the vast number of minor agreements in the d.t.
        Explains Luke 1:1 (reference to the gospels of Mark and Matthew)
        Explains other indications that Luke knew Matt, e.g. "Sermon on the plain"

        Verdict: inconclusive because, as Stephen says, it's like assessing apples
        against oranges. But now add:

        Three Source Theory (one document known to Papias, but none hypothetical)

        Explains all that the 2ST can explain, plus all that Farrer can explain. And

        What is there left to argue about?

        Ron Price

        Derbyshire, UK

        Web site: http://homepage.virgin.net/ron.price/index.htm

        Synoptic-L Homepage: http://www.bham.ac.uk/theology/synoptic-l
        List Owner: Synoptic-L-Owner@...
      Your message has been successfully submitted and would be delivered to recipients shortly.