Re: Kaccaayana: introductory verses (2)

Dear George,

> The English word 'letter', so far as I know, refers (whether in everyday
> language, traditional grammar or modern linguistics) to graphic symbols,
> and never to phonetics.  Thus we can say 'people speak using sounds' but
> not 'people speak using letters'.  The latter might possibly mean that
> people speak by orally spelling each word; then perhaps it would not be
> nonsense, but simply false.

Many of the Pali grammatical terms we will be using do not have exact
English equivalents and it appears that "akkhara" is one of them. The
following definition of "latter" is from Webster's New World Collegiate
Dictionary (online):

1. a written or printed symbol employed to represent a speech sound or
sounds; character in an alphabet: in some languages, as English, some words
contain letters that are no longer sounded.

This corresponds to Cone's definition 4 of "akkhara": (m.n.) a written
character. This meaning is also recognized in Abh 1063 (akkharaṃ
lipimokkhesu). However, it is the second part of Cone's definition 2 that is
relevant to Kacc 1: (m.) a vowel; a letter, phoneme (a, ā, etc.) which is
also recognized in Abh 348 (vaṇṇo tu akkharo). There is nothing in Webster's
7 definitions of "letter" that corresponds to Cone's 2. I see no problem in
taking "a letter, phoneme" as referring to something that can be spoken,
written, imagjned, or remembered. We could also include letters or sounds
that have tactile representations (e.g. braille).

> Thus when U Nandisena gives the English translation (b) of sentence (a) in
> the vutti of Kaccaayana 1, his translation is referring to things like a,
> b, c, ...
>
> (a)  Sabbavacanaanam attho akkhareh' eva sa~n~naayate.
> (b)  The meaning of all words is only known by letters.
>
> On the basis of English (b) one might well infer that we cannot understand
> anything unless we read it in a book.  Not perhaps what U Nandisena takes
> the Pali sentence to mean.

I think "letters" in the translation is fine as long as we take it in the
meaning of Cone's defn. 2. We might also substitute "letters or sounds" for
"letters" to make that meaning clearer.

> It is of course true that letters are (by design) related to speech
> sounds.  Jim asks:
>
>> When we read written letters and words, don't we normally have to
>> transform (consciously or subconsciously) these into audible or imaginary
>> sounds in order for us to comprehend or relate to them? >
>
> I think perhaps we do, and if so, I would take it as evidence that sounds
> are primary and letters secondary.  But we do have to clearly distinguish
> the two,
>
> (c)  I readthat book yesterday.
> (d)  I will readthat book tomorrow.
>
> (e)  That book is red.
> (f)  That book is often read.
>
> In (c) and (d) illustrate two words which have the same letters but
> different sounds, while (e) and (f) illustrate two words which have the
> same sounds but different letters.

Pali doesn't seem to have words with the same letters but different sounds
or vice versa. However, there can be confusion with different words (in
their derivation) having the same letters and sounds---much more so than in
Sanskrit.

Best wishes, Jim