Vol.63 , No.3(2015)210張本 研吾「マンダナ・ミシュラ作『スポータ・シッディ』に対するパラメーシュヴァラIII世による『ゴーパーリカー』註における原典批判」

Jborrnal

_oflhdian

andBudtihist

Studies

Vbl.

63,

No.

3,

March

2015

(129)

Textual

Criticism

in

the

Gopalika

ofParameSvara

III

on

Mandana

MiSra's

Sphotasiddhi

HARiMoTo

Kengo

1.

Introduction

Many

of us, students ofpre-modern

South

Asian

civilizations, strive

to

learn

more about

the

_past

based

on written sources.

In

order

fbr

the

infbrmation

we

_gain

from

them

to

be

ac-curate, we would

like

to

have

reliable

texts.

For

that

_goal,

we would

_prepare

critical

edi-tions

and

_practice

textual

criticism.

It

is

_yet

diMcult

to

surnmarize what

textual

criticism

is

or should

be;

i)

I

would

hence

like

to

make

do

with citing a

fragment

ofHousman

(1922)

:

".

. .

it

[textual

criticism] comprises ofrecension and emendation."

Conversely,

ifwe

see someone

does

recension and emendation, we may consider

the

_person

a

textual

critic, a

philologist.

We

do

not always associate authors of

_pre-modern

South

Asia

with

the

idea

of

textual

criticism.

Still,

since

textual

criticism

is

"purely

a matter of reason and of comrnon sense"

(Housman

1922:

68),

we should not

be

surprised

to

find

an author who wrote

in

Sanskrit

and

_practiced

textual

criticism.

What

I

intend

to

do

here

is

to

draw

attention

to

a

commen-tator

who,

to

my eye, appears

to

have

_practiced

textual

criticism

by

comparing manuscripts and

_proposing

readings

that

he

thought

were

_preferable.

2.

The

Gopfilikfi

of

ParameSvara

III

on

Mapdana

MiSra's

Sphotasiddhi

The

Gopalikfi

(G

in

the

fbllowing)

is

a commentary on

Maipdana

MiSra's

Sphotasiddhi

(S),

and

its

author,

ParameSvara

III

(P)

is

a

member

of

the

celebrated

Payyur

family

of

Kerala

who many consider

to

have

flourished

after

₁₄₀₀

CE.2)

One

feature

of

the

com-mentary

is

its

very

frequent

mentions of variant readings. 3)

Then,

when

I

read

the

S

with

the

commentary,

I

started

to

be

_puzzled

by

the

fact

that

very

often

readings

mentioned

as

(130)tcxtual

Criticism

in

the

GopalikE

ofPararneSvara

III

on

Marpdana

MiSra's

Sphotasiddhi

(K,

HARiMoTo)

3.

Preliminary

Observations

3.1.

How

ParameSvara

Introduces

Variants

P

uses

the

fbllowing

expressions when

he

introduces

variants:

.

.

.

iti

va

_palhah

"or

the

reading

is

.. ."

(pp.

12,

15,

20,

45,

101,

116,

165,

167,

174,

192,

232);

...iti

palhe

"ifthe

reading

is

...

[fo11owed

by

an

interpretation

according

to

that

reading"

(pp.

15,

34,

87,

92,

96,

118,

150,

151,

178

×

3,

191,

218,

244);

...

iti

_pdthah

"there

is

a reading..

."

(p.

29);

tathaiva

kvacit

_pMhah

"exactly

that

reading

is

fbund

somewhere"

(pp.

57,

244);

...

iti

kvacitpaghah

"there

is

areading... somewhere"

(p.

66);

kvacit

tu...

itipa(hah

"some-where

is

the

reading..."

(pp.

73,

89,

126,

154,

191,

237,

240);

...ity eva

pathah

"there

is

_precisely

the

reading

..

."

5)

(p.

74);

in7acitpa(halt

...

iti

"semewhere

_js

_the

reading..

."

(pp.

106,

131,

152,

161,

162,

163,

166,

175,

190

×

2,

191,

195,

196,

198,

212,

224,

236,

242,

246);

kvaeitpunar...

iti...pathyate...

"but

somewhere

the

reading

is..."

(p.

123);

tathaiva

va-

_pa-!hak

"or

the

reading

[should

be]

exa ¢

tly

that"

(pp.

138,

200);

anyah

_pathah

...

iti

`Cthere

is

another reading:

..

."

(pp.

158);

kvacit...

iti

_pa(hah

"some-where

is

the

reading..."

(pp.

165,

241);

kvacit...

iti

"somewhere..."

(p.

166);

kvacit

tu

.

.

.

iti

_grantho

na

dlr:!iyate

"hewever,

the

_passage

_.

_.

_.

is

missing somewhere"6)

(p.

171);

tathit

ca

kvacit

_pa(hah...

iti

"similarly,

somewhere

is

the

reading .. ."

(pp.

67,

176);

...

ity

eva

kvacit

_pathak,

kvacit

tu

...

iti

`Conly...

_is

_the

reading somewhere,

but

in

another

place

is

the

reading..."

(p.

176);')

aayatha va

_pathah

...

iti

"or

the

reading

is

different,

narnely., ."

(p.

246);

kvacit

tu

,..pathanti

"they

read

..,somewhere"

(p.

254).

The

expression

typeset

in

bold

typeface

above will

be

mentioned as

_possible

typical

ex-pression

to

introduce

suggestions

fbr

emendations

below.

32.

VVhat

ParameSvara

does

Net

Report

Certain

kinds

ofvariants are never reported

in

the

G

although

P

most

likely

encountered

them.

First,

no completely meaningless readings are mentioned.

It

is

a

fact

that

certain

types

of errors

in

manuscripts

_produce

completely unintelligible series of

letters.

Such

readings are never reported.

Nor

have

I

fbund

a mention

that

he

could not recover a

mean-ingfu1

reading

from

his

sources.

Major

deviations,

such as a chunk oftext missing or

displaced,

are not reported, either.8)

The

S

is

relatively a small

text.

So,

this

could

be

a

testimony

that

the

transmission

of

it

was more or

less

stable.

-Textual

Criticism

in

the

GopalikE

of

ParameSvara

III

on

Mandana

MiSra's

Sphotasiddhi

(K,

HAntMoTo)(131)

9)

In

addition, although

I

treat

the

first

reading

P

mentions

as

his

constituted

text

and

cal1

the

readings subsequently

introduced

variants,

I

am not certain

ifhe

in

fact

_preferred

the

first

reading.

Sometimes

the

interpretation

of

the

first

reading

is

long

and seems

fbrced,

and

subsequent variants are clearer, requiring no commentary or only short commentary.

Those

"variants"

are at

times

_preferable

(hence

the

suspicion of coojectural ernendations).

He

dismisses

a

reading

_only

_one

time,

as

far

as

I

could

find,

by

calling

it

an

_apapatha

"cor-rupted reading"

from

another reading

(p.

1g1

)

. iO)

4.

Classifications

of

Variant

Reports

in

the

GopZlika

I

classify

P's

variant reports

into

fbur

_groups.

4.1.

VAriant

Readings

of

Ilexts

Mapqana

Quotes

P

mentions variant readings oftexts

_quoted

by

Marpdana.

He

reports

that

he

knew

t

ent readings of

Kumarila's

Slokavarttika

twice;

those

of

Bhartrhari's

Vakyapadiya,

lst

Kai

_da

twice,

that

of

Sabara's

Bhasya

on

the

MImEmsastttras

once. ii)

In

those

cases,

his

sources of variants

do

not

have

to

be

those

of

the

S.

He

could

have

been

looking

at manu-scripts or commentaries of

those

texts,

or

fbr

some,

he

could

have

already committed

them

to

memory.

He

was at

least

looking

at

the

NyAyaratnEkara

and

the

Kfigik2

commentaries on

the

Slokav5rttika

(pp.

207-8)

.

I

do

not

discuss

those

readings _any

further

here

although

I

believe

comparisons of readings

Mapdana

_quotes

against various other sources will

yield

interesting

results.

4.2.

Variants

with

Sources

We

may

_presume

that

variants

P

reports with regard

to

S

came

from

S's

manuscripts or

commentaries on

it.

Still,

I

would

first

like

some confirmation.

Here

are a

few

examples of variant readings ofthe

S

that

P

most

_probably

saw

in

his

sources:

(1)

We

find

the

fbllowing

on

_p.

89:parve

'pi

varpak...nlipratyltyaka

iti;

kvacit

tu

pratyayaka

iti

pa(hak.

P

considers

that

the

S

had

nlipratydyakjih

but

notes

that

he

finds

the

reading wnhout

the

double

negative somewhere else.

The

editor,

S.

K.

Rfimanatha

Sastri,

adopts

the

first

reading

in

his

S

but

notespratydyakdk. mtilam

in

a

fbotnote.

This

means

his

manuscript of

the

S

had

the

reading

P

says

is

a variant.

That

is,

there

indeed

was a

manu-script

lineage

with

that

reading.

Since

there

is

very

little

semantic

difference

between

the

two

readings, we may eliminate

the

_possibility

that

the

reading

P

first

introduces

was a

(132)Textual

Criticism

in

the

Gopalika

ofParameSvara

III

on

Marpdana

MiSra's

Sphotasiddhi

(K,

HARiMoro)

(2)

On

p.

13

1

, we

find

prathamasamadhigamnd

anupakhyatam

iti

vylisena

kvacitphlhah

for

_{prathamasamadhigambnupakhyatam}

of

the

constituted

text

of

the

S.

Again,

S.

K.

R.

Sastri

notes

that

his

manuscript

in

faet

reads what

P

says

is

a variant.

P's

variant

is

corrob-orated

by

a manuscript.

The

difference

is

whether

to

read

two

words or one compound.

TXvo

readings

do

not cause much

differences

in

the

understanding of

the

text;

if

any,

the

variant reading

is

easier

to

fbllow.

Thus

it

is

unlikely

that

the

reading

P

first

mentions was

coajectural.

Additionally,

one may compare many examples

like

safigatisampvedanasamayasyeti va

pathah;

arthas

tu

_ptzrvavat

"or

the

reading

is

sangatisamvedanasamayasya;

the

meaning

however

is

the

same as

[explained]

befbre"

regarding

the

first

reading sangatisarpveddna-samayatah

(p.

12).

The

word

in

question

is

read with

the

_preceding

word

_purah

"befbre"

that

can

take

the

ablative or

the

_genitive

case.

The

meaning, "befbre

the

time

of acknowl-edgment ofunderstanding,"

is

the

same

(as

P

says).

Since

there

is

no reason

to

introduce

a variant as an emendation,

it

is

more

likely

that

this

indeed

was a variant.

This

would

be

a

12)

sign ofthoroughness ofP as a

_philologisL

4.3.

Variants

vvithout

Sources

(Suggestions

of

Conjectural

Emendations?)

Altihough

P

most

_probably

had

various sources

for

the

text

of

the

S,

one expression

per-haps

signals

that

the

altemative

reading

P

mentions

is

not

from

a

source

but

from

his

thought.

The

expression

types,et

in

bold

typeface

in

section

3.1

is

the

one

(pp.

138,

200).

After

a

_paraphrase

of

the

word

P

considers

is

in

the

root

text,

he

says

the

_paraphrase

may

be

the

reading.

The

reason why

he

did

not adopt

the

reading as

the

reot

in

the

first

_place

is

probably

because

he

did

not see such areading.

Here

are

two

more exarnples of references

to

alternative readings

that

might

be

P's

sug-gestions

to

emend

the

text.

(1)

samuccayavis.

_ayatvad

iti

va

_pdghah

"or

the

reading

is

`samuccayavis.

ayatvat

(since

[the

knowledge]

has

a collection as

its

object)"'

(p.

174)

in

reference

to

the

reading

samuccitavis.

_ayatvat

"since

[the

knowledge]

has

collected

_phonemes

as

its

object":

the

lat-ter

(but

introduced

first)

reading seems a

little

clumsy

to

_yield

the

desired

understanding

(samuccita-

var4a- vis.

_ayo

_yasya

tasya

bhitvas

tattvam)

.

Still,

from

the

context,

the

preferred

reading

is

this

one

that

was

introduced

first.

[[he

expression,

iti

vapdthah

("or

the

reading

is...")

attached

to

the

altemative reading might also suggest

that

he

did

not see

the

read-ing

anywhere.

-Textual

Criticism

in

the

_Gopilikh

of

ParameSvara

III

on

Mai)dana

Migra's

Sphotasidchi

(K,

HARrMoio)(133)

(2)

kvacit

tu na

te

'sactharanariipegeeti

pathah

(p.

240)

in

reference

to

the

reading na

tenasadharanena

riipepa:

The

whole sentence reads

bhede

'pi

pratyuipatti

var4andpa

na

tendsadhararpena

riipe4arthasya1'n-'jipaklih

("Even

ifthere

is

_a

difference

_every

_time

pho-nemes

are

_produced,

[the

_phonemes]

do

not

let

the

meaning

known

by

that

dissimilar

na-ture").i3)

The

sentence

is

incomplete

in

that

it

has

no

subject

but

only

the

_predicate

(as

seen

by

the

fact

that

I

supplied

the

subject

in

brackets).

P

says

the

_{subject of}

_the

_sentence

is

vavea-h. and

it

is

understood

by

context. i4)

The

word

tena

is

not

doing

much, either.

In

the

alternative reading,

these

two

clumsiness

disappear.

This

is

a case where a superior reading

is

mentiened second.

4,4.

Multiple

Variants

At

times,

P

mentions several readings

fbr

one

_place.

For

example, we

find

kvacitpd(hah

junyopahangatve

caJ'nNopakasamketakala

iti

_.

_.

_.

kvacit

pathahjnNopakEingarp

ca sahketakiila

iti

on

_p.

250

ofthe edition.

This

is

in

reference

to

the

readingjnnylipanahgatve caJ'nN4pahah

sanketahale

that

he

first

introduces.

P's

statement can

be

visualized as:

jnnyapanahgatve

cajfiapakah sanketakale niyogata

aSriyeta

jfiapakahgatve

cajfifipakasarpketak51e niyogata

aSriyeta

jfiapakafigam

ca sanketakale niyogata

asiriyeta

Immediately

obvious

is

the

repetition of

_graphically

similar

_J'n"opana

and

inNopaha

in

the

first

two

readings.

None

of

the

readings appears

to

be,

despite

P's

explanations, very

com-prehensible.

What

the

context requires

is

something

to

the

effect

of"Also,

if

[the

cause-ef

fect

relationship or

the

inherent

difference

ofphonemes] i5)

is

_part

ofthe

informing

pro-cess,

it

must

be

relied on at

the

time

of

fbrming

a convention

(what

sequence ofphonemes

denotes

what meaning)

."

Then

the

most

desirable

reading

becomesJ'n"tipandngtxtve

ca

san-ketakale

nlyogata

dsrlyeta,

i.e.,

the

first

reading withoutJ'nNiipakak. i6)

We

experience

this

sort of

_possible

multiple readings when a correction regarding erroneously repeated

text

is

present

in

a manuscript.

Sometimes

_the

nature of

the

correction

is

unclear.

There

are cases

when

such

confusing

corrections

_produce

different

readings

in

descendant

manuscripts.

I

suspect

that

P

was

looking

atsuch a case

in

his

manuscripts and

_proposing

different

solu-tions.

His

in,acit

does

not

have

to

be

limited

to

real sources

but

it

could

be

his

thought.

(134)Tbxtual

Crhicisrn

in

the

_Gop51iki

ofPararneSvara

III

on

Ma4dana

MiSra's

Sphotasiddhi

(K,

HARiMcrTo)

5.

Conclusion

ParameSvara

was aware

that

the

text

he

was commenting upon was

fluid

and

that

various

readings were

_possible.

He

_probably

witnessed

that

in

various source materials regarding

the

text

of the

Sphotasiddhi,

including

manuscripts and commentaries.

In

addition,

he

ap-plied

his

thought

to

determine

how

the

text

should

be.

We

may very well call

him

a

philol-ogist.

Unlike

a modern

philologist,

however,

he

did

not

decide

what really was

the

text

Mandana

wrote.

It

is

left

to

the

reader which reading

to

choose.

That

is

what

Biardeau

(1958)

did.

Still,

readers should

_practice

their

common

sense

and

reason

to

detemiine

which reading

to

choose.

Better

_yet,

we should

_go

back

to

(and

find

more) manuscripts of

both

the

Sphotasiddhi

and

the

Gopalika.

This

seems

_particularly

necessary

_given

that

the

first

one-third of

the

text

of

the

Sphotasiddhi

was never edited

from

its

manuscriptsi we

only

have

an extracted

text

from

the

Gopalika.

Notes1

)

For

an

influential

view on what

textual

criticism

is,

see

Alfi'ed

Edward

Housman,

"The

Applica-tionofThought to

[[lextual

Criticism,"

Proceedings

_qf'the

CiassicalAssociation

18

(1922)

,

pp.

67-84.

For

why we

_practice

textual

criticism, see,

for

example,

Martin

West,

71zxtual

(:riticism

and

Editorial

fechnigue,

_Applicable

to

Greek

andLatin

7laxts

(Stuttgart:

B.

G.

Tuebner,

1973),

pp.

7-9.

For

a view

on

how

a critical edition should

be,

see

Harunaga

Isaacson,

"OfCritical

Editions

and

Manuseript

Re-productions:

Remarks

Apropos

of a

Critical

Edition

of

Pratnaeavinis'caya

Chapters

1

and

2,"

Mbnu-seript

Cultures

(?Vewsletter

oj'the

MCAA?

2

(2009),

_pp.

13-20.

2

)

Its

editioprinceps

is

publjshed

as a

_part

of

S.

K

Ramanatha

Sastri,

SPho;astdcthi

ofA-ca-,

_ya

im"-duna

MiSra

with

the

Gopa-likii

of4siputra

Paramesivara

(Madras:

Madras

Goverriment,

1931).

In

its

introduction

the

editor

S.

K.

RamanEtha

S5stri

has

a

discussion

on the

Payyur

family

of

Kerala.

3

)

Inthe

fo11owing,

I

will obviously

be

relying on

the

_published

textof

the

G.

I

de

not

_put

_particular

faith

in

the

edition,

but

I

hope

the

_general

course of

discussions

will not

be

altered even when

I

_get

ae-cess tomore raw material.

4

₎

Cf.

Biardeau's

edition thatwas

based

oll

the

editio

_prineeps

<Madeleine

Biardeau,

La

Demon-stration

du

5ipho(a

_par

1vai44tina

MiSra,

introduction,

Ti"aduction

et

Commentaire

[Pondicherry:

Insti-tut

Frangais

d'Indologie,

1958]).

She

adopts many "variants" mentioned

in

the

G

in

her

constituted

textofthe

S.

5

)

Should

this

be

iti

vapdthah?

6

₎

In

this

_panicular

case,

the

"passage"

in

fact

_missing

isjust

_{one word,}

niravaclyam.

Or

we should

emend

the

textto

_granthe

"in the work"

instead

ofgrantho.

Puses

the

word

_grantha

jn

the sense of

book

in

other

_parts

ofthe

G

(e.g.,

pp.

262-65).

This

would mean

that

he

leamed

the reading with

the

-Textual

Criticism

in

the

Gopalikfi

ofParameSvara

III

en

Ma4dana

MiSra's

Sphotasiddhi

(K,

HARJMeTo)(135)

word niravadyam

from

somewhere.

A

commentary that

_predates

his

is

a

_possibility.

7)

These

two

cases on

_p.

176

involve

readings ofthe

gabarabhasya

_quoted

in

the

S.

8

₎

The

only exception

is

the

one about

just

one word noted

in

note

6.

9

)

This

appears

to

have

been

the

editing

_policy

of

S.

K.

R.

Sastri.

First

efall,as

the

Sanskrit

intro-duction

says

(pp.

ii-iv)

_, the manuscript ofthe

S

was available only after

_p.

₇₃

_{ofhis edition.}

The

_root

textwas extracted

from

the commentary up to theend of the

Vptti

on stanza

8.

Even

after

that,

Sastri

often

_goes

against the reading

in

his

sole manuscript of the

Sphotasiddhi

and chooses

the

reading

ei-ther

P

interprets

or

he

mentions

first.

See,

e.g,

_pp.

82,

83,

86-88,

90-94,

99,

100-105,

108-110,

112,

113,

116-18,

12zF34,

136,

etc.

10)

aayahpathab "spastamppaclabhecienapratiyete"ti.paclabhedo

nety

_apapbthah

(p.

191,

1.

22).

11)

Slokavarttika:

Sphota

71

(p.

116);Sphota

120

_(p.

178);Vakyapadiya

1.87

(p.

158);

1.86

(p.

r

r

r

163)

;

Sabarabhasya

(p.

176)

.

The

numbering of

the

Slokavarttika

fbllows

that

in

711ie

MTmdnsdi-Sib-ha-vartiha

_ofKltmtirila

Bhatta,

with the

Commenta,),

CZiiled

ACydyaratndhana

_by

Ptirtha

sarathi

MiS-ra, edited

by

RfimaSastri

Tailanga,

Chowkhamba

Sanskrit

Series,

no.

3

(Benares:

Chowkhamba

Sanskrit

Series

OMce,

1898-99)

.

For

the

Vakyapadiya,

VZiklyapadlya

ofBhartrhari,

with the

V}"tti

and

the

Padtthati

_of

P?is,

abhadeva, edited

by

K.

A.

Subrahmania

Iyer

(Poona:

Deccan

College,

1966).

12)

Variants

ofthis class may not affkict

the

meaning ofthe

text

in

_question

much.

I

believe,

however,

thatreporting them as

_thoroughly

_as

_possible

is

important.

_For,

as a whole, we

_gain

knowledge

about

the

_quality

of

the

text

transmission:

if

it

was more

fluid

or rigid.

13)

The

difference

being

referred

to

here

is

of

the

same

_phoneme,

for

example, ra,

in

rlZi'aory'ara.

The

opponent's

_position

is

that

there

are

differences

between

the

rli

in

rlli'aand

that

injara.

14)

`Y'n-Eipaha"

ityatra arthad "varua"

itisid7iyati.

(p.

240,

ll.

12-13)

15)

For

the supplied subject,see

the

next note.

16)

The

masculine

_gender

ofthe word va-cakah

is

_problematic.

In

the

next sentence,

the

same

thing

is

referred to

in

the

feminine

case

<endm).

It

is

not

impossible

to

interpret

it

(the

first

one referring

to

anupfirvivisfesa

and the second

anuptirvi)

from

the context or

by

fbllowing

P's

interpretation

(the

first

being

kliryakiirarpabhjb,a

and

the

second

haF:yakEirarpath)

.

Still,

from

the context,

the

singular subject

(indicated

by

the

verb

dsrlyeta)

should

be

samutthCipahacittakaryakara4atd _tatkrtova svabhavabhe-ciah

in

the

preceding

parallel

_passage

(p.

248).

Since

these

are two

_{possibilities}

ofwhat dnuptin,i

is,

the

feminine

_pronoun

in

the

next sentence

is

reasonab]e.

<Key

words>

Sphota,

_philology,

textual-criticism,

commentary,

Payyur

family