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

Fwd: RE: PS/URL to S.D. Joshi and Saroja Bhate [1984] on stratification in Panini (Re: [Indo-Eurasia] **Stratification: Methods & Dating

Expand Messages
  • Jean-Luc Chevillard
    Message 1 of 11 , Feb 1, 2007
      >Subject: RE: PS/URL to S.D. Joshi and Saroja
      >Bhate [1984] on stratification in Panini (Re:
      >[Indo-Eurasia] **Stratification: Methods & Dating
      >Date: Wed, 31 Jan 2007 19:44:43 -0500
      >From: "Deshpande, Madhav" <mmdesh@...>
      >To: "Jean-Luc Chevillard" <jlc@...>
      >
      >
      >Dear IER friends,
      >
      > I am currently traveling in India and
      > don't have all the bibliographical
      > references. However, I just want to say that
      > SDJ's and Bhate's views on stratification of
      > the Astadhyayi have been extensively discussed
      > and contested by Cardona, and somewhat less
      > extensively by me. If I am not mistaken, Jan
      > Houben has reviewed this debate, especially
      > between Cardona and Joshi on this issue in one
      > of his articles. Recently I have met both,
      > Joshi and Bhate, in Pune. Joshi is planning to
      > write a separate volume on Panini expounding
      > his own views on all these matters. I hope his
      > health will permit him to continue to write. With best wishes,
      >
      >Madhav
      >
      >
      >
      >-----Original Message-----
      >From: Jean-Luc Chevillard
      >[<mailto:jlc@...>mailto:jlc@...]
      >Sent: Thu 1/25/2007 1:53 PM
      >To: Indo-Eurasian_research@yahoogroups.com
      >Cc: Deshpande, Madhav; cardonagj@...
      >Subject: PS/URL to S.D. Joshi and Saroja Bhate
      >[1984] on stratification in Panini (Re:
      >[Indo-Eurasia] **Stratification: Methods & Dating
      >
      >Dear list,
      >
      >here is a post-scriptum
      >to my former messsage
      >giving the URL
      >to the review in the BSOAS
      >(available on JSTOR)
      ><<http://links.jstor.org/sici?sici=0041-977X(1986)49%3A2%3C401%3ATFOA%3E2.0.CO%3B2-K>http://links.jstor.org/sici?sici=0041-977X(1986)49%3A2%3C401%3ATFOA%3E2.0.CO%3B2-K>
      >of the 1984 book I mentioned in my former post:
      >"The fundamentals of Anuvr.tti",
      >publications of the Centre of Advanced Study in Sanskrit, Class B, N°9,
      >University of Poona, Pune,
      >by S.D. Joshi and Saroja Bhate
      >
      >Best wishes
      >
      >-- Jean-Luc Chevillard
      > (CNRS, Université Paris 7, Laboratoire
      > d'Histoire des Théories Linguistiques)
      >
      >
    • Steve Farmer
      The key point of the note from Madhav Deshpande forwarded to the List from Jean-Luc Chavalliard this morning lies in these lines, which aren t found in the
      Message 2 of 11 , Feb 1, 2007
        The key point of the note from Madhav Deshpande forwarded to
        the List from Jean-Luc Chavalliard this morning lies in these
        lines, which aren't found in the slightly different version
        of Madhav's note that was posted to the List a bit earlier:

        > However, I just want to say that SDJ's and Bhate's views
        > on stratification of the Astadhyayi have been extensively
        > discussed and contested by Cardona, and somewhat less
        > extensively by me. If I am not mistaken, Jan Houben has
        > reviewed this debate, especially between Cardona and Joshi
        > on this issue in one of his articles.

        Does anyone have a reference to the Houben article? In any event, it
        has long been clear that lines have been drawn in the sand on this
        issue between Panini researchers, which is why we would like to
        examine it anew using the methods that Michaqel and I have noted in
        recent posts (using reasonably objective markers of stratification).
        Maybe that will lead someplace and maybe not, but it is better
        than trying to settle the war by balancing authorities.

        For me, the strongest argument for stratification in Panini has always
        been the simplest one: why would this text out of all Indian texts
        from approximately the same level of antiquity be exempted from normal
        stratification processes? Or do people claim that there are other
        texts from the 4th century BCE or earlier that are supposedly
        "integral"? It will be interesting to hear Saroja Bhjate's views
        here, once she is free for the discussion.

        Best,
        Steve

        > >Subject: RE: PS/URL to S.D. Joshi and Saroja
        > >Bhate [1984] on stratification in Panini (Re:
        > >[Indo-Eurasia] **Stratification: Methods & Dating
        > >Date: Wed, 31 Jan 2007 19:44:43 -0500
        > >From: "Deshpande, Madhav" <mmdesh@...>
        > >To: "Jean-Luc Chevillard" <jlc@...>
        > >
        > >
        > >Dear IER friends,
        > >
        > > I am currently traveling in India and
        > > don't have all the bibliographical
        > > references. However, I just want to say that
        > > SDJ's and Bhate's views on stratification of
        > > the Astadhyayi have been extensively discussed
        > > and contested by Cardona, and somewhat less
        > > extensively by me. If I am not mistaken, Jan
        > > Houben has reviewed this debate, especially
        > > between Cardona and Joshi on this issue in one
        > > of his articles. Recently I have met both,
        > > Joshi and Bhate, in Pune. Joshi is planning to
        > > write a separate volume on Panini expounding
        > > his own views on all these matters. I hope his
        > > health will permit him to continue to write. With best wishes,
        > >
        > >Madhav
      • raonath
        [Mod. note. As noted repeatedly in that original discussion of stratification of Panini , back at the end of 2005, the issue of whether or not you accept the
        Message 3 of 11 , Feb 4, 2007
          [Mod. note. As noted repeatedly in that original discussion of stratification of 'Panini', back at the end of 2005, the issue of whether or not you accept the *specific* stratification of the Astadhyayi by Joshi and Roodbergen is independent of the question of whether the text is or isn't stratified. These are two quite different issues, and they need to be kept separate: No one on the List in fact has ever even discussed the particular arguments of Joshi and Roodbergen, so any attempt to refute those arguments is irrelevant to our discussion. If you can come up with specific evidence for why the Astadhyayi alone among all ancient Indian texts of the same approximate age *isn't* stratified, these reasons need to be stated (or restated) explicitly, if our discussion is to be have any value. If you claim a particular text is supposedly immune to normal processes of stratification, you better have an explanation for why. - Steve.]

          --- In Indo-Eurasian_research@yahoogroups.com, "Steve Farmer"
          <saf@...> wrote:

          > For me, the strongest argument for stratification in Panini has
          > always been the simplest one: why would this text out of all
          > Indian texts from approximately the same level of antiquity be
          > exempted from normal stratification processes?

          In message 2567 (posted Dec 12, 2005), I pointed out one way
          in which the aSTAdhyAyi is rather unique among known works from
          5-3rd c. BCE India. I don't think that the particular points
          I raised then have even been adressed. Also, if we go through
          your checklist, marking 'yes', 'no' or 'not applicable'
          and then count the NA's, we can see the difference.

          There are two other issues I did not want to go into back then.
          One is the assumption (very obvious in Joshi & Roodbergen
          translation of aSTAdhyAyi) that the taddhita chapters are
          unnecessary in a Sanskrit grammar, and extensive use of taddhita
          suffixes is indicative of later age. The first, not to put too
          fine a point on it, depends on drawing the line between grammar
          and lexicon the way it is done in Western grammars, and I fail to
          see any reason why Panini or his peers must have done it the same
          way as 19th c Europeans. Anyway, if we are going to put rules in
          to derive kumbhakAra, why not rules do derive aindra?

          The second assumption, the one that use of taddhita affixes
          expands with age, is highly questionable. One of the prime
          examples cited, the line from mahAbhASya ``taddhitapriyA
          dakSinAtyAH'' said about the vArtikA ``... zAstreNa
          dharmaniyamaH, yatha vaidIkalaukIkeSu'' fails to note
          what Patanjali says latter on, namely the word ``krtAnteSu''
          is to be understood. In other words, vaidIka and laukIka are
          adjectives, in contrast to veda and loka which are substantives.
          [And the pUrvapakSin does not seem to realize the irony of his
          saying 'dakSiNAtyAH' instead of, say, 'dakSiNavAsinaH'.]

          This point needs to be appreciated in its historical context.
          The general tendency is that in the Brahmanas, substantives are
          demarcated from adjectives, and unlike in medieval Sanskrit,
          substantives are preferably not compounded when needed in an
          adjectival role: `grAmyAn pazUn', not 'grAmapazUn', `hiranmayena
          rathena', not 'hiranyarathena'. [In some registers of later
          Sanskrit, we see 'grAmyakukkura' etc, but that is not relevant.]
          Given this, we must accept that the regular use of taddhita
          suffixes to form adjectives out of substantives [and the fewer
          ones to form abstracts out of adjectives] were an integral part of
          middle and late Vedic Sanskrit. It is the alternate (MIA
          influenced?) expression, using compounds with substantives
          functioning as adjectives, that is later.

          Joshi and Roodbergen also make statements which make sense to
          me only if I were to assume that Mahabharata and Ramayana (or,
          if you prefer, their ``core'', but that needs a historical grammar
          of Epic Sanskrit to be written first) date to 5-4th c. BCE and can
          thus be used to judge the Sanskrit spoken by Panini's peers. However,
          the MIA influence we can see the language of the epics (for example,
          Oberlies notes, in "A Grammar of Epic Sanskrit" p.XXIX, "In fact,
          almost all 'un-Paninean' forms of Epic Sanskrit are innovations")
          makes it impossible to accept either of these two assumptions.
          I would also stress that one should not quote the conclusions of
          an argument unless >all< the assumptions made are accepted.
          In particular, to judge Panini's grammar using the language of the
          epics requires an assumption about their relative dates as well as
          their dating relative to the Pali Canon (oldest parts of which seem
          to be near the same stage as Asoka's edicts) which must be defended
          before it can be utilized.

          Nath Rao
        • raonath
          [mod. note: all of Skt in Panini? Certainly not all of Vedic Skt. -- And, if complete, why then the need to add to Panini by Katy. and Pat.? -- Also, where is
          Message 4 of 11 , Feb 5, 2007
            [mod. note: all of Skt in Panini? Certainly not all of Vedic Skt. -- And, if complete, why then the need to add to Panini by Katy. and Pat.? -- Also, where is the pressing need to have rules for names of wells(!) in the Panjab (but not in Delhi?) -- And, if no rule can be taken out, also not those at the links between sections? Ideal place to have additions. --Just for starters -- MW]

            -- In Indo-Eurasian_research@yahoogroups.com, "raonath" <raonath@...>
            wrote:

            > [Mod. note. ... If you can come up with specific evidence for why
            the Astadhyayi alone among all ancient Indian texts of the same
            approximate age *isn't* stratified, these reasons need to be
            stated (or restated) explicitly, ...]

            OK, here is a restatement, but without the details:

            Why aSTAdhyAyi should have been able to
            escape extensive interpolation or layering, which would
            make it unique in that respect. The short answer is that
            it is unique in what it is: It is a set of rules that
            attempt to produce all grammatically acceptable utterances
            and only grammatically correct utterances in a language
            that existed independently of that work.

            A collection of essays or speculations (as brahmanas
            and upanishads are) is very susceptible to additions
            and expansions. This also applies to the Tripitaka.
            The corpus of mantras is an anthology, and can be
            added to at will.

            But aSTadhyAyi is different in that it gives a number of
            rules which are to be applied as needed. The ultimate aim
            is to output precisely those strings which are correct
            and conform to Sanskrit syntax when implying the meaning
            meant by the speaker. Removing, adding or modifying one
            rule can have global consequences on the set of strings
            produced [I am speaking of just the sUtrapATha, not the
            dhAtupATha or gaNapATha]. This applies to all phrase
            structure grammars. Show me a phrase structure grammar
            in which no rule is a special case of another rule or
            combination or rules, and in which you can remove say,
            10% of the rules without affecting the output set.

            On the other hand, no other work from India of that time is
            anything like a phrase structure grammar.

            Nath Rao
          • Jean-Luc Chevillard
            ... Madhav Deshpande might be referring to Jan Houben s 2003 article Three Myths in Modern Paninian Studies (a long review article of George Cardona s 1999
            Message 5 of 11 , Feb 7, 2007
              At 15:52 01/02/2007, S.Farmer wrote:

              >The key point of the note from Madhav Deshpande forwarded to
              >the List from Jean-Luc Chevillard this morning lies in these
              >lines, which aren't found in the slightly different version
              >of Madhav's note that was posted to the List a bit earlier:
              >
              > > However, I just want to say that SDJ's and Bhate's views
              > > on stratification of the Astadhyayi have been extensively
              > > discussed and contested by Cardona, and somewhat less
              > > extensively by me. If I am not mistaken, Jan Houben has
              > > reviewed this debate, especially between Cardona and Joshi
              > > on this issue in one of his articles.
              >
              >Does anyone have a reference to the Houben article?


              Madhav Deshpande might be referring
              to Jan Houben's 2003 article
              "Three Myths in Modern Paninian Studies"
              (a long review article of George Cardona's 1999
              _Recent Research in Paninian Studies_)
              which appeared on pp. 121-179 in Asiatische Studien / Études Asiatiques, LVII.1
              and especially to the sections
              2.1: Relations between rules in Panini's
              Astadhyayi: the issue of possible interpolation (pp. 125-129)
              (where Jan Houben discusses George Cardona's
              critique of Joshi & Roodbergen's theses)
              2.2: Variation in options? (pp. 129-144)
              (where Jan Houben examines the various reviews
              of Kiparsky's _Panini as a Variationist_ [1979],
              by Hartmut Scharfe [Indogermanische
              Forschungen, 1981], Johannes Bronkhorst [IIJ, 1982],
              G.V. Devasthali [Annals of the BORI], Rama
              Nath Sharma [Language in Society, 1983],
              Madhav Deshpande [_Language_, 1984],
              Pierre-Sylvain Filliozat [OLZ, 1984], Rosane Rocher [JAOS, 1986],
              before examining G. Cardona's comments on the same book)

              However,
              this article of Jan Houben
              does not seem to be discussing
              S.D. Joshi and Saroja Bhate's 1984 book,
              which was the topic of my initial question
              and which is IMHO a very impressive book

              So, further pointers will be needed

              Best wishes

              -- Jean-Luc Chevillard
              (CNRS, Université Paris 7, Equipe d'Histoire des Théories Linguistiques)
            • Steve Farmer
              Dear Jean-Luc, ... The short review of the book you pointed to earlier: http://www.safarmer.com/Indo-Eurasian/Joshi.review.1986.pdf Since, as Michael has
              Message 6 of 11 , Feb 7, 2007
                Dear Jean-Luc,

                Thanks for the useful bibliographical data. You write:

                > However, this article of Jan Houben does not seem to be discussing
                > S.D. Joshi and Saroja Bhate's 1984 book, which was the topic of my
                > initial question and which is IMHO a very impressive book.
                >
                > So, further pointers will be needed.

                The short review of the book you pointed to earlier:

                http://www.safarmer.com/Indo-Eurasian/Joshi.review.1986.pdf

                Since, as Michael has noted, Saroja Bhate has joined the List, we
                apparently will finally get our opportunity to discuss this issue.
                There is no rush, since so much else is going on, and the preliminary
                discussions have set the stage.

                I think that drawing up a potential "marker of stratification" list
                appropriate specifically to the Astadhyayi, and then discussing the
                potential markers one by one, would be the best way to get a lot of
                people involved besides just the Pune and non-Pune Panini
                specialists. In a sense, this issue is too important to be limited just
                to Panini specialists, who clearly don't agree anyway. We can at
                a minimum clarify in the discussion where the lines are drawn.

                Anyway, I think that making haste slowly here is probably the best
                procedure, which is what we've always had in mind for the threads
                that carry two asterisks (e.g., **Stratification).

                Best,
                Steve

                > Madhav Deshpande might be referring
                > to Jan Houben's 2003 article
                > "Three Myths in Modern Paninian Studies"
                > (a long review article of George Cardona's 1999
                > _Recent Research in Paninian Studies_)
                > which appeared on pp. 121-179 in Asiatische Studien / Études
                > Asiatiques, LVII.1
                > and especially to the sections
                > 2.1: Relations between rules in Panini's
                > Astadhyayi: the issue of possible interpolation (pp. 125-129)
                > (where Jan Houben discusses George Cardona's
                > critique of Joshi & Roodbergen's theses)
                > 2.2: Variation in options? (pp. 129-144)
                > (where Jan Houben examines the various reviews
                > of Kiparsky's _Panini as a Variationist_ [1979],
                > by Hartmut Scharfe [Indogermanische
                > Forschungen, 1981], Johannes Bronkhorst [IIJ, 1982],
                > G.V. Devasthali [Annals of the BORI], Rama
                > Nath Sharma [Language in Society, 1983],
                > Madhav Deshpande [_Language_, 1984],
                > Pierre-Sylvain Filliozat [OLZ, 1984], Rosane Rocher [JAOS, 1986],
                > before examining G. Cardona's comments on the same book)
                >
                > However,
                > this article of Jan Houben
                > does not seem to be discussing
                > S.D. Joshi and Saroja Bhate's 1984 book,
                > which was the topic of my initial question
                > and which is IMHO a very impressive book
              • raonath
                ... The point of the argument is that changing the rule set changes the set of strings output. For Vedic to have any relevance, we need to be reasonable sure
                Message 7 of 11 , Feb 8, 2007
                  Again responding to comments in a moderator's note [from Michael Witzel], not to myself:

                  > all of Skt in Panini? Certainly not all of Vedic Skt.

                  The point of the argument is that changing the rule set changes
                  the set of strings output. For Vedic to have any relevance,
                  we need to be reasonable sure that the understanding of RV
                  (which is where Panini is deficient) changed in the relevant
                  time period: After all, the completeness would have been
                  judged by Panini's contemporaries.

                  > And, if complete, why then the need to add to Panini by
                  > Katy. and Pat.?

                  The language continued to change. It is almost certain that
                  there was break in the transmission chain between Panini and
                  Katyana, leading differences in how the text was unerstood.
                  The consequences of these two facts, and how much is left
                  after they are accounted for is a topic of much discussion by
                  the specialists, but the residue seems to be much less than
                  one would one might guess by simply counting the number
                  of vartikas.

                  > Also, where is the pressing need to have rules for names of
                  > wells(!) in the Panjab (but not in Delhi?)

                  I assume that you are referring to 4.2.74, in which case the
                  parenthetical remark is simply wrong: A sequence starting at
                  4.2.67 deals with geographical entities named after a person or
                  thing. 4.2.73 and 4.2.74 are exceptions to the general rule. Well
                  names not covered by these two exceptions, including those in what
                  is now Delhi, follow the general rule.

                  On the other hand, the exlcamation mark seems to be a dig at the
                  inclusion of well names at all. But once the decision is made to
                  include words of transparent derivation, it is proper that all of
                  them be included. I addressed this in #6021 already.

                  > And, if no rule can be taken out, also not those at the
                  > links between sections? Ideal place to have additions.

                  The effect of adding or removing a rule depends on its specificity,
                  not on its position. Removing 2.4.85 removes a construction
                  that we know was grammaticalized by the time of brahmana prose.
                  The effect of removing 8.4.66,67 or 8.4.68 is extremely
                  far reaching. Their position at the end of a pAda makes no
                  difference to their effect.

                  ---

                  A general comment: I was trying to address a question posed by
                  Steve Farmer: Why should the aSTAdhyAyi be different when it
                  comes to layering? I tried to point out the way in which the
                  aSTAdhyAyi is unique and why this is relevant to layering and
                  why claims of layering can be taken seriously only when
                  the layers are delineated. [I now note that essentially the
                  same answer was already given by R. Rocher in the review cited
                  in a recent message, but not phrased in terms of general properties
                  of string generation/rewriting systems.] I see no point in
                  discussing minutiae of the actual rules without first
                  addressing the essential difference between discourses
                  and string generation/rewriting systems.

                  Nath Rao
                • E Bruce Brooks
                  To: Indo-European In Response To: Steve Farmer On: Panini From: Bruce STEVE: I think nearly all Panini researchers would agree with the claim that the
                  Message 8 of 11 , Feb 20, 2007
                    To: Indo-European
                    In Response To: Steve Farmer
                    On: Panini
                    From: Bruce

                    STEVE: I think nearly all Panini researchers would agree with the claim that
                    the Astadhyayi was stratified insofar as it was based on the work of his
                    claimed predecessors.

                    BRUCE: I have to doubt that it is helpful to use the word "stratified" in
                    this wide sense. Example: NT folk are widely agreed that Luke used earlier
                    sources ("predecessors," such as the Gospel of Mark), but this does not lead
                    them to call Luke "stratified." A stratification proposal does exist for
                    Luke, but it is quite distinct. It asks whether there was a state of Luke
                    earlier and simpler than our canonical one, usually called Proto-Luke, that
                    is, whether the authorial (or formative) process was itself accomplished in
                    one step or more. It would seem to be only the latter situation, the
                    multi-step process, that is properly described by the word "stratified." The
                    difference is whether the earlier predecessors are in the same line of text
                    formation as the author of record. There is no doubt that the various Dharma
                    Sutras make a chronological sequence, in which literary and ideological
                    progression can be discerned (Patrick Olivelle used signs of that
                    progression to date them relative to each other), but this does not make the
                    Sutras a single stratified text. It makes them a coherent and evolving
                    tradition.

                    STEVE: The issue as I take it is rather whether additions weren't made to
                    the text long *after* the time of the
                    assumed systematization of the work ascribed to "Panini."

                    BRUCE: This, to my mind, is more in the right direction, but I have some
                    problems here also. For instance, I am not sure if the composite vs integral
                    question, in advance of evidence, can be fruitfully framed this in this
                    rather narrow way. Additions *long after* a text are not usually regarded as
                    part of its formation process; they are usually considered extrinsic to that
                    process (though they may be very important in themselves). Example: the
                    Pericope Adulterae in the Gospel of John, which, in one manuscript, also
                    turns up in the Gospel of Luke; we have here a later independent pious tale
                    of Jesus that is looking for a home in one of the established Gospels. The
                    presence of that pericope in John does not, at least not in what I would
                    think the most helpful sense of the word, make John a stratified text. It
                    makes it a corrupt text, or if one prefer to say it in another and nicer
                    way, a text which later persons thought it fit, or convenient, to embellish
                    with pious additions. Those additions are nevertheless a form of later
                    meddling, and it only confuses any discussion of authorial intent or
                    ideology, in John, to include them in the evidence. Such confusion is indeed
                    well exemplified in the Johannine commentary literature, but confusion does
                    not become good method merely by being multiplied. In terms of what I find
                    to be the most helpful schema:

                    Authorial/Formative Process ...................P................Corruption
                    Process

                    The text, to be sure, remains more or less fluid throughout, in that it is
                    in principle forever available to vicissitudes. For example: even today, I
                    may decide to put out an improved and rationalized version of Luke (it is in
                    the computer, but so far I haven't distributed it; I am first trying to
                    contact Luke about the copyright, as required by current law, and am having
                    an unexpectedly difficult time of it). But the distinction between a Beta
                    version of a company's own text and an after-market, third-party patch on
                    that text, seems to me to be an analytically useful one. That is, in
                    sufficiently many cases to be valid as the basis for a working terminology,
                    there is a definite point P when the text leaves the custody of its series
                    of authors or its institutional proprietors (such as the house church which
                    may have nourished the Gospel of Mark over many doctrinal additions and
                    suppletions, perhaps over more than one generation), and is first given to
                    the copyists to multiply. Additions may still be made to it, and it is the
                    whole point of text criticism in the usual sense to detect and adjudicate
                    those additions. But those additions stand on a different analytical footing
                    than the evolutions and adjustments which are often made to a text while it
                    is still closely held, and available to changes, not from a wider public,
                    but only from its custodians.

                    Methodologically suggested,

                    Bruce

                    E Bruce Brooks
                    Warring States Project
                    University of Massachusetts at Amherst
                  • Steve Farmer
                    ... The question of evidence in this case has been has been proposed via an empirical test, Bruce. I earlier suggested that we draw up a list of potential
                    Message 9 of 11 , Feb 20, 2007
                      Bruce writes:

                      > STEVE: The issue as I take it is rather whether additions weren't made
                      > to the text long *after* the time of the assumed systematization of
                      > the work ascribed to "Panini."
                      >
                      > BRUCE: This, to my mind, is more in the right direction, but I have
                      > some problems here also. For instance, I am not sure if the composite
                      > vs integral question, in advance of evidence, can be fruitfully framed
                      > this in this rather narrow way. Additions *long after* a text are not
                      > usually regarded as part of its formation process; they are usually
                      > considered extrinsic to that process (though they may be very
                      > important in themselves).

                      The question of evidence in this case has been has been proposed via an
                      empirical test, Bruce. I earlier suggested that we draw up a list of
                      potential "markers" of stratification appropriate to the Astadhyayi,
                      guided by what we know of stratification in general, and then test to
                      see how many of those markers can be found in the work. Unless we try a
                      test like this, given the fact that Panini scholars have disagreed on
                      this issue for many decades, I don't anything can be settled.

                      When you have two models in conflict, it is standard procedure
                      everywhere in science to try to think of simple ways to test those
                      models, unless you are willing to take the word of one "authority" over
                      another -- and I don't think that is a legitimate approach.

                      My own *working* hypothesis is that no text ascribed to (say) the 4th
                      century BCE is ever "integral", but that's just my working
                      hypothesis, based on trends in philological studies in many other
                      fields. Others take the traditional view that some privileged texts,
                      including the Astadhyayi, became "fixed" at an early date. I think
                      that claiming that unknown conditions preserved that work from
                      normal processes of stratification is difficult to sustain -- why
                      this text alone, as often asked, most recently of Nath Rao? -- but
                      this aside, you still need to test the working hypotheses.

                      The "marker" idea seems to me to be a promising way to test the
                      conflicting models as a group, but if anyone can think of other
                      proposed tests, that would be great.

                      Best wishes,
                      Steve
                    • E Bruce Brooks
                      To: Indo-Eurasian In Response To: Steve Farmer On: Stratification in Panini From: Bruce STEVE: The question of evidence in this case has been has been proposed
                      Message 10 of 11 , Feb 20, 2007
                        To: Indo-Eurasian
                        In Response To: Steve Farmer
                        On: Stratification in Panini
                        From: Bruce

                        STEVE: The question of evidence in this case has been has been proposed via
                        an empirical test, Bruce. I earlier suggested that we draw up a list of
                        potential "markers" of stratification appropriate to the Astadhyayi,
                        guided by what we know of stratification in general, and then test to see
                        how many of those markers can be found in the work.

                        BRUCE: The more we know of other situations the better, and any lists of
                        markers appropriate to any text would be welcome. I am compiling one myself,
                        which I hope to publish presently, and I would rejoice to see any others.
                        But I am wary of an approach which would limit what we may find in Panini to
                        what we are confident we have earlier found in other texts. I suspect, on
                        the one hand, that what is "appropriate" to Panini will also be potentially
                        "appropriate" to other situations, so that there is not really any Panini
                        subset, apart from points specific to his language; and on the other hand,
                        that whatever the nature of our list, Panini may contribute something new to
                        it: may present unanticipated new features of which we would need to take
                        due account for any future investigations. The prosodic integration in
                        Panini, at least as far as this non-Sanskritist is aware, seems to be
                        somewhat unique, and seems also (I have heard this asserted by some expert
                        Sanskritists) to guarantee a single authorial hand and conception behind the
                        work.

                        That point I am not competent to judge, but the marker, in this case (if you
                        will) a non-stratification marker, surely should be considered along with
                        other empirical evidence.

                        STRATIFICATION

                        As to detecting stratification in general, I would still prefer to reiterate
                        my earlier and simpler principle: We should first look for internal
                        differences in the text (any text, whether Panini or the Federalist Papers).
                        Differences of any type at all, whether of rhyme or the lack of it, form,
                        ideology, implied dialect, implied audience, ideology, technical
                        presumption, vocabulary, style (the collective use by a writer of the
                        lexicon available to him, which was used by Mosteller and Wallace on the
                        Federalist, and has been developed by myself as a tool applicable to any
                        text in any language; I have used it to distinguish Swift from any others in
                        some unsigned political editorials in The Examiner, and between Thurber and
                        White in "Is Sex Necessary," and also in analyzing various mysterious
                        Chinese texts; a similar method has with seeming success been applied to
                        distinguish genuine from spurious Epistles of Paul in Greek), or whatever
                        else has been found good. The list of what has been found good will vary
                        somewhat with the experience of the one who makes that list; here is where
                        collaboration would be very valuable.

                        Having noted differences, as it seems to me, we must then attempt to
                        determine their nature. Do they represent the use of one source rather than
                        another in an essentially integral but also derivative work, like the Gospel
                        of Luke in the opinion of many? Or do they (to keep my previous example)
                        betoken editorial second thoughts within Luke's own authorial process; do
                        they show that Luke, though consistent in its way, is after all not integral
                        but complex, as some reputable NT persons have argued?

                        Here is where judgement comes in, and also a principle which, the NT people
                        tell me with some asperity, I have no choice but to call the Brooks
                        Principle (which is the Tischendorf 1849 Principle generalized out of text
                        criticism and applied to all philological determinations). I had been
                        calling it the Metzger Principle, since I first learned of it in Bruce
                        Metzger's book The Text of the New Testament. I have been since informed
                        that (a) it goes back instead to Tischendorf 1849, and (b) it was by
                        Tischendorf limited in application to the adjudication of manuscript
                        variants. The extension to other areas of philology is my fault, or anyway
                        my responsibility.

                        I accept the responsibility. I think the principle is vital to all sorts of
                        determinations that philologists habitually make, whether in making them
                        they feel supported by established theory, or not.

                        STRATIFICATION MARKER

                        In conclusion, but still on the subject of relevant methodology, I don't
                        think that, over the long term, the evidence will show that there is such a
                        thing as a distinctive stratification marker, as against a mere difference
                        marker. Rachmaninoff countenanced not one but three versions of his Second
                        Piano Sonata, the second being an abridgement, and the third, due to
                        Horowitz, being a partial restoration; a middle version. Then length per se,
                        and indeed shortness per se, are not marks of stratification as against
                        another kind of formative process. All three versions mentioned are,
                        demonstrably, part of the coming-into-being of the Rachmaninoff Second
                        Sonata, but just such alterations (abridgement, extension) are also
                        encountered as post-authorial (post-P) changes, made by other people for
                        their own purposes.

                        Holst's "Planets" was extended in the late 20th century to include Pluto,
                        which had not been discovered when Holst wrote his original suite. Now that
                        Pluto has been declassified as a planet, perhaps that addition will again
                        fade into the background. But the difference in compositional style which
                        may easily be detected between Pluto and the other Holst Planets is not to
                        be attributed to the period of text formation; it is a later vicissitude, of
                        the kind to which the Acts of the Apostles was also subjected in the text of
                        which Codex Bezae is our earliest exemplar.

                        Another example: I have seen a version of Beethoven's Moonlight Sonata not
                        in C# minor, but in C minor. This can be shown to be a later publisher's
                        concession to the imagined limitations of the beginning pianist (a white
                        key, not a black key, as the home note). But it can be shown no less
                        conclusively that Bach and Mozart, in cannibalizing their own previous
                        works, not seldom presented them in different keys than those they had
                        earlier found most appropriate. So key change as such is not distinctively
                        either an authorial procedure or a publisher corruption procedure. Of
                        itself, it is just a procedure. Each instance of the procedure requires to
                        be evaluated separately, with whatever local resources of information may be
                        available.

                        The above cases may illustrate why I am dubious about the idea of a
                        distinctive "stratification marker." Most of the inconsistencies we can
                        detect in texts are capable of explanation in more than one way. The first
                        thing, as it seems to me, is simply to find the indications of internal
                        difference, or contradiction, or inconcinnity, so as to have something to
                        talk about. And then see where the talk might lead us, in the case of
                        Panini.

                        Or anything else. Each case handled by philology is by definition unique,
                        but philology itself is general.

                        LISTS AND PRECEDENTS

                        Another sort of list that may be helpful to those doing these kinds of
                        investigation is a list of possible text types: the possibilities are not
                        exhausted by the terms "integral," stratified," and "corrupted." The Text
                        Typology list which was earlier available on the Warring States Project web
                        site (http://www.umass.edu/wsp > Philology > Text Typology) has been
                        slightly corrected and improved, and I would welcome further corrections and
                        improvement, also any new and cogent examples of those basic types in any
                        language at all, which viewers of those pages may care to share with me.
                        There is an E-mail link on every page of the site, by which such suggestions
                        may be conveniently made.

                        Thanks in advance,

                        Bruce

                        E Bruce Brooks
                        Warring States Project
                        University of Massachusetts at Amherst
                        http://www.umass.edu/wsp

                        Bruce
                      • Steve Farmer
                        Dear Bruce, This thread began as a discussion of the stratification issue in general -- not one on Panini, into which we drifted. Whenever we ve tried to focus
                        Message 11 of 11 , Feb 20, 2007
                          Dear Bruce,

                          This thread began as a discussion of the stratification issue in
                          general -- not one on Panini, into which we drifted. Whenever
                          we've tried to focus on that issue before, things have gotten heated
                          fast. So can I suggest that we return to the original question
                          in this thread?

                          Are there reasonably consistent "markers" of stratification that
                          can help us identify layered texts, no matter what the texts in
                          question?

                          You write:

                          > The more we know of other situations the better, and any lists of
                          > markers appropriate to any text would be welcome. I am compiling one
                          > myself, which I hope to publish presently, and I would rejoice to see
                          > any others.

                          It is good to see you publishing a list of such "markers." No
                          such list has ever been published anywhere, I don't believe. Here is
                          the starter list I posted on February 1, which included 23 markers
                          of stratification (there are sure to be more) found in many texts:

                          http://groups.yahoo.com/group/Indo-Eurasian_research/message/5980

                          A later post by Daniel Stender (on the Mahabharata) added at least one
                          more potential marker, involving systematic shifts in metre:

                          http://groups.yahoo.com/group/Indo-Eurasian_research/message/6016

                          You write:

                          > But I am wary of an approach which would limit what we may find in
                          > Panini to what we are confident we have earlier found in other texts.
                          > I suspect, on the one hand, that what is "appropriate" to Panini will
                          > also be potentially "appropriate" to other situations, so that there
                          > is not really any Panini subset, apart from points specific to his
                          > language; and on the other hand, that whatever the nature of our list,
                          > Panini may contribute something new to it: may present unanticipated
                          > new features of which we would need to take due account for any future
                          > investigations.

                          I'm in full agreement.

                          > The prosodic integration in Panini, at least as far as this
                          > non-Sanskritist is aware, seems to be somewhat unique, and seems also
                          > (I have heard this asserted by some expert Sanskritists) to guarantee
                          > a single authorial hand and conception behind the work.

                          Well, the "single authorial hand" issue is what is in dispute in this
                          case. But I suggest that we bypass it for now. I know that Michael
                          wants to pick it up later: We talked about that by phone tonight.

                          > That point I am not competent to judge, but the marker, in this case
                          > (if you will) a non-stratification marker, surely should be considered
                          > along with other empirical evidence.

                          I take it that you are referring to the level of "prosodic integration"
                          in the text. I agree with you that that is important to consider. But
                          I'd also like suggest that integration of this sort is also often the
                          result of scholasticization of texts over a long period, e.g., in Euclid.
                          Often systematization of texts falls to a series of exegete-disciples. This is extremely common in medieval scholastic
                          traditions, for example.

                          > As to detecting stratification in general, I would still prefer to
                          > reiterate my earlier and simpler principle: We should first look for
                          > internal differences in the text (any text, whether Panini or the
                          > Federalist Papers). Differences of any type at all, whether of rhyme
                          > or the lack of it, form, ideology, implied dialect, implied audience,
                          > ideology, technical presumption, vocabulary, style (the collective use
                          > by a writer of the lexicon available to him...or whatever else has
                          > been found good. The list of what has been found good will vary
                          > somewhat with the experience of the one who makes that list; here is
                          > where collaboration would be very valuable.

                          I don't disagree with you here either, Bruce. All I've proposed is
                          that systematically *listing* these differences might be useful in
                          giving us a fairly standardized way of detecting stratification.
                          I don't think the differences in our general methodological
                          positions here are all that deep.

                          Best,
                          Steve
                        Your message has been successfully submitted and would be delivered to recipients shortly.