Mnemonics in Ancient India

[George Thompson]

It is good to see that there is so much interest in this topic.

A couple of points:

First, the discussion began with Steve Farmer's comments re pre-modern
European mnemonics, which developed the use of the visual cue into very
elaborate 'memory palaces,'  from mnemonics in ancient India, which appear to
be primarily a matter of strenuous rote memorization.  Though it is clear
that later, post-Vedic Vaidikas may have used mudrAs and something like
proto-maNDalas, it should be noted that there is no mention of these devices
in Vedic texts themselves.  Or perhaps I should say that there is no use of
these TERMS in Vedic.  Is there reference to such techniques [i.e., the use
of visual cues] using other terms?  By the way, both Malamoud [mentioned by
A. Sandness] and Staal himself have mentioned yet another device: that of the
rather forceful, 'rhythmic' steering of the Veda-student's head back and
forth at the hand of the teacher.

Steve Farmer's point that the Vedic vikRtis may well be byproducts of rote
memorization of Vedic texts, instead of mnemonic techniques in the strict
sense, is interesting and deserves further consideration. As Staal has
repeatedly emphasized, these vikRtis function to prevent the unconscious
substitution of a synonym or an approximate homonym in the slot given to a
particular element in the transmitted text.  The obligation to observe sandhi
alternation in the recitation of these vikRtis seems to serve the same
function.  That is, the vikRtis are a backup device intended to inhibit
unconscious substitution.  They are not simply tour de force displays of
skill, as the memorization of a text 'backwards and forwards' would appear to
be.

This suggests that in Vedic there is in fact a kind of ekAgratA [to borrow
the term used by S.R. Mehta].  The vikRti, as a mnemonic device -- if that is
in fact what it is -- is just one manifestation of that single focus which I
think is characteristic of Vedic in general: that is, focus not only on the
received text, but also on the language of that text -- in particular on the
ANALYSIS of that language.  This involves not only the analysis of its
phonological shape [in the prAtizAkhyas], but also the analysis of semantic
features [as in the nirukta tradition, and even in the brAhmaNas to some
extent], as well as syntactic and morphological features.  These late Vedic
activities are the predecessors of the astonishing Paninian tradition which
was the first systematic linguistic analysis of a language ever attempted.

I believe that metalinguistic operations like these can be traced back even
to the Rgveda, where metalinguistics and poetics are thoroughly intertwined
[of course, the metaphor *speech* = *weaving* is widespread in the RV].

Well, there are other things to be said about Vedic smRti [memory], but I've
probably already exceeded the acceptable limits of a single post.  In
closing, I would agree with Arlo Griffith and Dominik Wujastyk that Harry
Falk has done the hard work of examining the evidence and should not be
dismissed by reference to the unpublished remarks of Patrick Olivelle and
unnamed others.  Reference to Patrick Olivelle in this context is just as
unfair to Patrick himself as it is to Falk.

At least so it seems to me.

Best wishes,

George Thompson