Khuddaka Nikaya - Nettippakarana - Author's Introduction

Khuddaka Nikaya - Nettippakarana ( The Guide )

ACCORDING TO
KACCANA THERA

TRANSLATED FROM THE PALI BY
BHIKKHU NANAMOLI
Pali Text Society

 TRANSLATOR'S INTRODUCTION
1. THE GUIDE
The book here translated—as it stands, it was perhaps composed
more than two thousand years ago—sets forth a method intended
for the guidance of those who already know intellectually the
Buddha's teaching and want to explain his utterances. I t is not a
commentary but a guide for commentators: it deals with scaffolding,
not with architecture. Its name, Nettippakarana, means 'guide-
treatise', but the translation has been called, more simply, the
Guide.
There are two recognized aspects of language, which are comple-
mentary, that is to say the Normative and the Contextual.
Of the normative theory of language the familiar characteristic
product is Grammar, which fixes the forms of words in a given
language and the rules for their formal association. Subordinating
ideas to words, it tacitly assumes that the correct meanings of known
words can, if not yet ascertained, inherently all be defined un-
equivocally in the Dictionary.
1
 There words (with their rules)
figure as the keys to ideas regarded as named by them, and contexts
are thus relegated to a mere matter of style. But the Guide disposes
of Grammar in one sub-heading (§186) and the Dictionary in one
heading (Mode 11, 'Synonyms'). I t cannot be called even a
'hermeneutic grammar'. Grammar seeks to govern all ideas
through words.
'Take care of the sense and the sounds will take care of them-
selves': but sometimes to be 'sense-wise' is to be 'sound-foolish'.
Words have, in their use, a notorious ambivalence which remains
always beyond the power of the Grammar and the Dictionary fully
to control. However precisely defined, they still retain that
element of inherent ambiguity (not vagueness), in virtue of which,
for a specially outstanding example, metaphor is not only possible
but necessary for a language to live, and a language—even a 'dead'

1 While Panini fixed the grammar of Sanskrit in the first centuries B.C.,
the earliest extant Pal i grammar is the Kaccayanappakarana, traditionally
attributed to the Buddha's pupil Maha-Kaccayana Thera but according to
Western scholarship composed sometime after the beginning of the 5th century
A.C. The first Pali dictionary is the Abhidhanappadipika compiled by
Moggallana Thera in Ceylon in the 12th century A.C.


language—lives when it is used. A contextual theory of language
recognizes use in contexts as an essential (ad hoc) denning element of
words, as representing ideas and here subordinate to them. But
contexts, theoretically unlimited, are limited arbitrarily in fact by
the necessary conventions of usage in communication; and usage, as
well as being thus arbitrary in the limitations it has to impose, is
subject both to caprice and to change (this aspect of language must
be essentially statistical). Properly the Guide—as we evaluate it
now—belongs here; for, as it will appear later, the task it has set
itself is simply that of showing compendious ways of eliciting from a
given individual textual passage the implications that the very
bulky contexts of a teaching as a whole both allow and require.
Therein it is concerned with a particular usage. I t seeks to promote
correct wordings of known meanings and to prevent deviations
creeping into such re-wording as a commentary in that usage must
involve. And the devices it employs for this purpose are, in fact,
nothing other than 'contextual types' chosen to suggest the desired
implications, these 'contextual types' being set out in the form of
the 16 Modes of Conveying (a communication) and the 5 Guide-Lines
(to the verbal expression of the Pitafoi-teaching's meaning-as-aim,
namely 'liberation' in whatever way expressed). The Guide seeks to
control words through known ideas.
Two kinds of use of language can be distinguished. One extends
to describing, gathering evidence, exploiting, commenting, drawing
conclusions, and so on; it is oriented to the discovery of some-
thing new. The other seeks to exercise the ideas so discovered
while at the same time preserving them intact and preventing
their change and loss. I t seeks consistency and is averted from
what is new.
Regarded in this light, the Guide belongs under the second kind.
I t deduces nothing and concludes nothing. I t simply draws from
the Suttas (more or less directly in each case) 16 contextual type-
situations and 5 verbal lines converging on the Suttas' meaning-as-
aim. From such an apparatus nothing could, in fact, be deduced
or concluded. I t can only facilitate re-wording of the ideas already
expressed in the Suttas. And that is what it sets out to do. All the
many passages of a commentarial or definitive nature in its
exemplification are incidental to its main purpose, which is one
reason why its definitions of words are always ad hoc and often
multiple and variant. Its elements (Sect. 7b) and how they are
exemplified must be clearly distinguished.


Commenting, like translating, consists in the re-wording of an
idea. (And words are related to their meanings and to each other
by minds.) The ideas to be re-worded should be known clearly to
the commentator, as they should to the translator; but when the
grammar and the dictionary have supplied all the aid they can, the
new wording, whether in the same language or in another, still needs
controlling in the matter of the new contexts and the trains of
thought it follows and suggests. With Metaphor (and the Pseudo-
metaphor or Pun) possible, not only do words tend, in new uses, to
suggest and allow to infiltrate after them their own habitual families
of contexts that may bring about deviation of the ideas they are
intended to re-express, but also it is as well known as it is often
forgotten that an individual set of contexts, each with a cipher or a
gap—a strange word (an i
aj>arimana-j>add>
. see §49)—in it, will
prompt the mind to supply or form an idea as a 'meaning' to fill the
lacuna. So the re-wording of an intellectually known idea without
due precautions in regard to this aspect may make a communication
different from what was intended, and it can even mislead the
would-be commentator, or translator, himself by undermining in his
mind the ideas that were intellectually clear to him.
Works on the contextual aspect of language are few.
2
 The
present one is a special case; for it is concerned with a particular
branch and specifically with the Buddha's teaching. Normally this
aspect of language is left to the native wit of the individual to
manage as best he can. Its nature must make it much less easy
than in the case of Grammar for the theory to emerge from the
welter of material in which it is embedded. How far the book—
especially the detailed part—has actually succeeded in its purpose
must be for the inquirer to decide. I t is only not easy going if
mistreated as an actual commentary—as an attempt directly to
explain ideas in the Buddha's teaching taken as not yet sufficiently
clear, or to give demonstrative proof of some conclusion drawn—
instead of taking it for what it is, namely a guide for the re-wording
of ideas already known. If so mistreated it must seem not only
quite procrustean at times but also a capricious mixture of the vague
and the obvious, by which almost anything can be proved. But
that would be to take it wrongly. Its instructions are, in fact, such

2 Roget's Thesaurus is the classical English example of this approach to
language, with its 'Tabular Synopsis of Categories' ('Plan of Classification')
and its avowed objects of finding a word for a meaning already to hand and
of suggesting trains of thought.


as the character oi its intended readers makes proper. (For details
see sects. 7a, 8 and 9 below.)
2. EDITIONS OF THE TEXT
There are four well edited editions, one in Roman script published
by the Pali Text Society, one in Sinhalese script, and two in Burmese
script. The work has a commentary by no less authority than
Acariya Dhammapala (late 5th century (?) A.C.).
3
 There is also a
sub-commentary (Tiled) written in Burma in the 15th (?) century
A.c. So the text can be regarded as established and presenting no
problems of this sort.
Neat and well rounded off as the work is in the matter of its text
and its commentaries, it does nevertheless pose a number of other
unsolved major puzzles. There is no evidence of when, or where, it
was composed, though it is assumed to have come to Ceylon from
India between the 3rd century B.C. and the 5th century A.C, and
it is older than the works of Acariya Buddhaghosa since he quotes
from it.
4
 Its authorship is in dispute. Its relationship to a very
similar volume, the Petakopadesa ('Disclosure of the Pitakas': these
two books are alone in Pali in presenting for its own sake a method—
the same method—for the wording of exegesis), is far from clear at
first glance and is indeed the opposite of what has been supposed.
5
It contains a very large number of untraced quotations. . . .
6
An attempt will be made to examine these problems for clues to a
partial solution, though some must remain open.
3. HISTORY AND AUTHORSHIP
Here, to start with, is the traditional history of the Guide as Acariya

3 The authorship of the NettiA is discussed at length in the PTS Netti Pal i
text Intro. , pp. ixff. See also E. Lamotte, Histoire du Bouddhisme indien
des origines a VEre &akay Louvain, 1958, pp. 174, 207-8, 210, 356-7, 468.
4 E.g., MA. i, 31: 'Ettdvatd ca yvdyam
"Vuttamhi ekadhamme ye dhammd ekalakkhand tena
Vuttd bhavanti sabbe iti vutto lakkhario hdro" ti
evam Nettiyam Lakkhano ndma Hdro vutto, tassa vasena . . . ' ; this answers
Prof. Hardy's question (PTS Netti pp. xiv-xv). This version and that at
Pe. 90 have tena where the printed Netti texts have keci; also a few other
minor differences.
5 This will be argued below after comparison of the two books (sects. 5 and 6).
6 The Pe has a similarly large number of untraced quotations, but of these
only 1 verse and 5 prose are common to the Netti.


Dhammapala gives it in his commentary: 'If it is asked,'
7
 he says,
'How can it be known that the "Guide-Treatise" is what was
uttered by a principal disciple and approved by the Buddha ? (it
can be answered that it is) because it is a text ;
8
 for there is no other
criterion beyond a text, and any text not in contradiction (when
examined) under the four Principal Appeals to Authority (§120) is
the criterion. And the "Guide-Treatise" has, like the "Disclosure
of the Pitakas" (Petakopadesa), come down (to us) by way of the
unimpeachable succession of teachers (see DA Introduction).—If
that is so, then why is its source
9
 not given ? For a source is
given in the cases of the Subha Sutta (D. Sutta 10), the Anangana
Sutta (M. Sutta 5), the Kaccdyana Samyutta (S. iii, 9 ?), etc., which
were uttered by disciples.—That is not always so in the case of
disciples' utterances and even in the case of some of the Buddha's
utterances; for no source is given in the cases of the Patisambhi-
ddmagga and the Niddesa10
 or in the cases of the Dhammapada and
the Buddhavamsa, so that is no criterion; and that is how it should
be regarded here too. And then any source is itself always the
utterance of the custodians of the Sutta and Vinaya, the Elders
Upali and other principal disciples,
11
 and so that too is uttered by
principal disciples. And anyway why this investigation about a
source, since there is no one else to whom to ascribe it except the
Elder ? What needs investigating here is only the meaning (in
order to see) that it does not conflict with the texts. Besides, as a
method of detailing (samvanriana) the meaning of texts, this work
has no more need of a separate source than have the Patisam-
bhidamagga or the Niddesa9
 (Netti A., p. 3). The commentator,
however, did make certain minor reservations elsewhere.12

7 'Etth'aha'. Prof. Hardy (PTS Netti p. xi, note) seems to take these words
as introducing a verse, which he then suggests comes from the Pe; but what
follows is not a verse or a quotation. The words 'etth'aha' are normal
commentarial usage for introducing an objection or query.
8 Meaning presumably that is accorded the status of a canonical text (but
see below, next para). Doubtless based on the words tha t conclude the
Netti text .
9 'Source' (nidana): see Netti §§184 and 192ff., where the term is given as
a sub-heading of the 6th Mode. But no 'source' of the Netti itself is furnished,
which would introduce it at the beginning if there were one.
10 Both traditionally attributed to the chief disciple Sariputta Thera and
both regularly included in the Sutta Pitaka of the Pali Tipitaka.
11 See Vinaya, Culavagga, penultimate chapter, also DA Intro.
18 See translation below, n. 1/6, n. 2/2.


This —still in the main the tradition accepted in the East today—
sets the work, generally speaking, on the same level of authority as
the books admitted within the Tipitaka itself; in fact, in Burma
both the Nettippakarana and the Petakopadesa are included (along
with the Milindapanhd) in the Tipitaka, both being printed in the
Burmese Chattha-Sangiti edition of 1956. Nevertheless, none of the
lists of Tipitaka books given in JLcariya Buddhaghosa's works men-
tions either book, and in Ceylon the two—like the Milindapanhd—
have never been admitted to Tipitaka status, and remain outside it.
Modern European scholars have rejected outright the ascription
of both works to the Buddha's disciple, the Elder Maha-Kacca(ya)na:
'In ascribing the Netti or, strictly speaking, the Patiniddesa-
portions of it to one single author the Buddhists are undoubtedly
right. None but one could have planned a work of such harmonious
unity as the Netti proves to be when closely examined, notwith-
standing much seeming incoherency, which is mostly due to the
strange mixture of exegesis and analysis in it. . . . Yet in ascribing
the Netti to a disciple of the Buddha they are undoubtedly wrong'
(PTS Netti p. xvii).
13
 Again 'According to tradition they (the
Netti and the Pe) were composed by Maha-Kaccayana, one of the
most prominent disciples of the Buddha. This is however certainly
not true. The author was probably Kaccayana by name and was
hence identified with the renowned disciple of the Buddha. The
same was the case also with the grammarian Kaccayana' (PLL
p. 26),
14
 who 'belongs to the age posterior to Buddhaghosa .. . he
is to be distinguished from the author of the Netti' (PLL p. 37).
15
The only extra source available is the contents of the two books
themselves. So discussion and appraisal of these opposed views—
how far they are justifiable and whether they are as opposed as they
seem—will have to be left till after an examination and comparison
of the contents (sect. 5).
4. DATE
Tradition places the Guide as a product of the First Council (in 483
B.C. by the usually accepted Western and 60 years earlier by the
Sinhalese reckoning); but modern European scholars reject this too,

13 For the critical remarks made here see section 1 above, last para, also
note 49 below.
14 W. Geiger, Pali Literature and Language, English trsln. by Bhatakrishna
Ghosh, Univ. of Calcutta, 1943, p. 26; cf. PTS Netti pp. viiif., xvii-xix.
15 Cf. also PTS Netti p. xxxii.


along with the traditional authorship. Professor Hardy says 'The
Pet. seems to presuppose the Netti,
16
 but acquaintance with its
doctrines on the part of the Pet. taken for granted, it would not
help us much, because both works are still floating dateless on the
ocean of Indian literature' (PTS Netti pp. xix-xx). But he adds
'Unless future research prove me wrong, an approximate date for
the Netti will be the time about or shortly after the beginning of our
era' (p. xxvii). In another opinion 'a work which is probably as
early as the earliest portions of the Milindapanha, and which also
possibly dates far enough back to have been written in India, is the
Nettippakarana'.
17
 But this really tells us nothing since we do not
know how long before the 5th century A.C. the Milindapanha was
written. I t has come to light since Prof. Hardy wrote his Intro-
duction (see note 4 above) that Acariya Buddhaghosa does quote
from the Netti, and apparently from the Pe.
18
 So the Netti is proved
older than the 5th century A.C. There only remains what internal
evidence a comparison of the two works can furnish.
5 . THE NETTIPPAKARANA AND THE
PETAKOPADESA COMPAEED
The problem of the Netti'a relation to the Pe needs a comparison in
some detail. First, however, an important fact has to be noted.
It is the singularly corrupt state of all the available Pe texts.19 In

16 Cf. M. Winternitz, History of Indian Literature, English trsln. by
Mrs. S. Ketkar and Miss H. Kohn, Univ. of Calcutta, 1933, where the Pe is
called 'a continuation of the Netti and most likely not much later than this
work' (vol. ii, p. 183); also A. Barua: ' Netti-Pakarana, however, is older than
the Petakopadesa' (PTS Pe. p. hi). But see note 5 above, since this view
cannot be maintained.
17 Winternitz, vol. ii, p. 183.
18 Vis. p. 141 = VinA. i, 143 = DhsA. 165 = PsA. 181 (Sinh. ed.), cf.
Vis A. 194 and 874 (Burm. ed.). The quotation is not now found in the Pe
and may be from a passage lost from ch. vi, or from another version. Cf. also
Vis. 690 where similes slightly rewritten from the Pe are ascribed to 'the
Ancients' (Parana). Also the NdA (Sinh. ed., p. 224) quotes verses as from
the Pe which are not found in it now, and are perhaps lost from the end of
ch. vi . See also sect. 5 below.
19 Roughly all the more important corruptions including even copyists'
errors appear consistently in all the printed editions. PTS is a valuable
document of the present general state of the MSS as a basis for a restoration,
since it gives a good cross-sample of readings, absurd though many of them are.
Burmese eds. : Zabu Meit Zwe Press ed., Rangoon, 1917; Chattha-Sangiti ed.
1956. No Sinhalese printed ed.


this respect it is unlike any other Tipitaka or main commentary
texts in the confusion it exhibits (in which it is the very opposite of
the tidy Netti). Its only commentary is a modern one.
20
a. The state of the Petakopadesa texts
All editions of the Pe are packed with mistakes. There are countless
crude copyists' errors (as the modern Commentary observes), often
hard to rectify though some are corrected by the Commentary and
others by the second Burmese printed edition. Verses, etc., are
sometimes quoted in partly rewritten as well as corrupted form (a
fair example is the version of the Uddna stanza 'duddasam . . .'
etc., (Vd. 80) at Pe. 56). There are displacements of sentences, as
at Pe. 208, line 7, or p. 217, line 6, and so on. There are confusions
of schedules, as at p. 6,11.16-20 and p. 60,11. 2-5 (not hard to restore
from the subsequent detail which they are intended to summarize).
There are some apparent displacements of whole palm leaves, as at
pp. 137-42 (corrected in all printed eds.), and pp. 188-93 (corrected
only in the PTS Pe), and an exchange of two whole sections at pp.
231-41 with part of one section interposed in the other at p. 234,
1.11 ('sa nibbuto . . .') to p. 235,1. 28 ('. . . dutiyena padena'), which
belongs to the middle of p. 237 (noted in the commentary but
uncorrected in all eds.). Further, there is the intrusion of an
extraneous palm leaf from a Sumangalavilasirii MS (— DA. p. 971,
1. 25 'dutam . . .' to p. 973,1. 4 '. . . mahS) at p. 239,1. 8, to p. 240,
1.19; how this passage should read when the intruding matter—
which starts, as it ends, in the middle of a sentence—is eliminated,
can be seen at NettiA (PTS Netti p. 261,11. 27-8), but this intrusion
(in all editions) has not been noticed at all.
21
 Lastly there seems

20 By Narada Thera, known as the Jetavana Sayadaw (who died within
living memory): published Mandalay, 1926. (This is the Pal i edition, of
419 pages. There is also a Pali-Burmese word-by-word transcript published
in 3 vols, totalling 1220 pages.) The Gandhavamsa, however, mentions
'a Tikd on the Petakopadesa by Udumbariya Acariya' (J.P.T.S., 1886, pp. 65
and 75, see also Index to Gandhavamsa, J.P.T.S., 1896, p. 56); but either
the work has been lost, or, which is more likely, this is a mistake (like the
Gandhavamsa's attribution of the Apaddna commentary to Buddhaghosa
Thera, and some other errors. Cf. also PTS Netti p. ix, note 6).
21 The Pe passage (PTS Pe. 239-40) as found in all the printed texts, when
its Udana quotations are put between quotes and when the absurd DA
intrusion is emphasized by capitals, comes out as follows: ' . . . "Kusalo ca
jahati papakam" akusalo na jahati. "Ragadosamohakkhaya sa|DUTAM


undoubtedly to be a displacement of the end of Chapter vi (Suttat-
ihasamuccaya) backwards from p. 153 to p. 140 along probably with
the contents of p. 140. This involves the worst confusion with
apparent loss of one or more palm leaves (which might account for
the fact that certain quotations from the Pe
22
 are not found in the
text today). Actually a close scrutiny of the material belonging to
Chapter vi must show that it cannot but continue (with some minor
confusions) properly from p. 139, end, skipping p. 140, on from
p. 141 to p. 153,1.11, ending with the word '. . . mettd' and perhaps
followed, after a gap, by the contents of p. 140 to terminate this
chapter.
23
 The material belonging to Chapter vii (Harasampata)

PESETVA PI NA PAKKOSAMI, SO SAY AM EVA MAHABHIKKHU-
SANOHAPABIVABO AMHAKAM VASANATTHANAM SAMPATTO
.. . (p. 239, etc. , down to p. 240) . . . SABBAPHALIPHULLO VIYA
YOJANASATIKO PABICCHATTAKO PATIPATIYATTHAPITANAM
DVATTIMSA CAKKAVATTINAM DVATTIMSA DEVABAJANAM
DVATTIMSA MAHA/nibbuto" asekkhassa natthi nibbuti.' Now the DA
passage begins in the DA text with ' "Yena bhagava ten'upasankamimsu" ti
(D. hi, 207, 11. 14-15) bhagavato dgamanam stdvd/DUTAM PESETVA
(see above) and ends with ' . . . DVATTIMSA MAHA/brahmanam suriya
sirim abhibhavamanam viya virocati\ By removing the extraneous matter
from the Pe text and placing the two halves of its split sentence together
again, we have ' "Kusalo jahati pdpakam" akusalo na jahati. "Bdgadosa-
mohakkhayd sajnibbuto" asekkhassa (sic) natthi nibbuti\ Turning now to the
NettiA (PTS) p. 261, 11. 27-8), we find ' "Kusalo jahati pdpakam" akusalo
pana na jahati. "Bdgadosamohakkhayd sa nibbuto" tesam aparikkhayd natthi
nibbuti\ In a restored version of the Pe, therefore, besides throwing out the
intruding matter, read for asekkhassa either sekkhassa or as in NettiA. The
Pe Corny, passes this intrusion without remark. (The intrusion contains
virtually no mistakes!)
22 See note 18. Prof. Bapat suggests that the 'Petaka' so called there may
have been a different book, now lost, since the Vis uses this abbreviation of
a quotation not found in the existing Pe texts (see P. V. Bapat , Vimuttimagga
and Visuddhimagga, Poona, 1937, p. xliii). But the NettiA uses this form in
its prologue (see note 25) and both of a verse not now found in the Pe texts
(Sinh. ed., p. 3) and of a lengthy series of quotations from the Pe which (with
some discrepancies) are in the Pe texts (NettiA Sinh. ed., pp. 40ff.; Pe pp. 46
and 44). So it looks as if the more likely explanation is that the untraceable
passages are from the old lost palm leaf or leaves of the Pe MS.
23 The sentence as it stands in the Pe texts (PTS Pe p. 151,1. 11) is 'Tattha
katama jhanaparamita? Suparamita metta "kamesu satta kamasangasatta" ti
(Ud. 75) yamhi sutte desandya voharena (sic) dve saccani nidditthani: dukkhan
ca samudayan (sic) ca' makes no kind of sense as a whole (though the com-
mentary does try to explain it as one). But if a break is made between
1
 . . . metta* and ' "Kamesu ... " ', here we find where the material of ch. vi
properly ends and that of ch. vii begins. The matter ending with ' . . . metta*


then properly begins with the word 'Kdmesu . . .' at p. 153, 1.11,
perhaps preceded by the stanza quoted at NettiA (p. 3; cf. PTS
Netti p. xi, note), which is missing from all Pe editions. (The
Commentary senses something seriously wrong earlier though not
here but only attempts to rectify things by making a separate
chapter called 'Pakinnaka' (see p. 146), thus increasing the number
of chapters from 8 to 9; but neither does the material justify this
nor is it any solution.)
Except where stated above as corrected, the mistakes, including
a striking proportion of copyists' errors, appear the same in all the
printed editions. The explanation is doubtless simple; for it may
well be assumed that the MSS used by all the printed editions all
stem from a single ancient original containing the common errors,
omissions and intrusion.
24
 This is indeed not at all improbable.
The book was always regarded as overshadowed by the Netti as is
shown by Acariya Dhammapala's treatment of it
25
 and by the
absence of any old commentary. On this view possibly only one
MS was ever imported into Burma (from Ceylon) in the first few
centuries A.c.—the MS containing all these common flaws (including
the extraneous Sumangatovildsini palm leaf)—, which was then
copied (without being studied) and so gave rise, through diverging
lines of subsequent copies, to all the MSS used by the printed
editions. Meanwhile in Ceylon any old Sinhalese MSS (themselves
doubtless very few) remained uncopied and were eventually lost,
perhaps in the turmoils of the 10th or 13th centuries, or later
through neglect.

(which seems to be the incomplete detail of the last item of a schedule on
p. 146,11.4-16, ending with the words 'jhanabale thitassa ayam pdrami (pdrami-)
pattassa imani jhanangani') has nothing whatever to do with what follows,
beginning with ' "Kamesu ..."' , which is the first Harasampata quotation:
they belong to two different chapters.
24 In effect PTS relies only on Burmese MSS since its only Sinhalese one
(called '#' ; on paper and therefore modern) is, from the nature of its readings,
no more than an inaccurate copy of one of the Burmese MS types to which
that called l
BP belongs. PTS's two best MSS (called 'BT and W ) are very
close together, to the two Burmese printed editions and to the Commentary.
The only Sinhalese palm-leaf MS the translator has been able to trace in
Ceylon was found to be of no great age and showed roughly the same character-
istics as PTS's & MSS 'S' and 'Bl', pointing to its being a recent copy of a
Burmese MS of that type too.
25 In the prologue to the NettiA: 'Petakena samsandetva'.


6. Comparison of the Netti and the Petakopadesa
In order to compare the two books, a start can be made with their
chapter headings. Here they are:
Nettippakarana
i. Modes of Conveying in
Separate Treatment (Ha-
ravibhanga).
ii. Modes of Conveying in Com-
bined Treatment (Hara-
sampata).
Moulding of the Guide-Lines
(Nayasamutthana).
Petakopadesa
i. Display of the Noble Truths
(Ariyasaccappakasana).
*ii. Pattern of the Dispensation
(Sasanapatthana).
iii.
iv.  Pattern of the Dispensation
(Sasanapatthana).
iii. Terms of Expression in the
Thread (Suttadhitthana—so
read),
iv. Investigation of the Thread
(Suttavicaya).
*v. Modes of Conveying in
Separate Treatment (Hara-
vibhanga).
vi. Compendium of the Thread's
Meaning (Suttatthasamue-
caya).
*vii. Modes of Conveying in Com-
bined Treatment (Hara-
sampata).
*viii. Moulding of the Guide-Lines
{Nayasamutthana — not
'Suttavebhangiya', see Pe
p. 259).
From this it will be seen that the names of all four Netti chapters
appear among the eight Pe chapters as numbers 5, 7, 8 and 2
respectively. And, in fact, also the contents of these corres-
pondingly titled chapters are roughly equivalent. Further, nearly
but not quite all the material—all merely exemplification of or
introduction to the elements—contained in the remaining Pe
chapters numbers 1, 3, 4 and 6 is, on close examination, found
represented in some form in various parts of Netti Chapters i to iii.
This will need going into in some detail.
It is convenient to take first the four chapters whose titles and


general subject-matter are the same and directly exemplify the
elements (see sect. 7b). I t is these that exhibit the Method.
Netti ch. i and Pe ch. v have the same title and are parallel in
shape and general treatment. Much detail corresponds, though the
Netti chapter contains far more, being much more fully illustrated
and nearly three times as long. A point to note is that of the 16
summary verses (one for each Mode at Netti (pp. 3-4), the appro-
priate one being repeated in ch. i at the beginning and end of each
Mode) only ten appear in Pe (in ch. v at the beginnings of Modes
1-7, 10, 15 and 16). Now in the Pe these verses for Modes 1-7 and
10 often differ considerably from their counterparts in the Netti:
the Netti versions seem better. Those for Modes 15 and 16 in Pe
differ widely. Certainly the Netti's treatment of these two Modes is
clearer. In all 16 Modes the Netti is much more successful in
eliciting the distinctive aspect of each Mode. I t may also be noted
how the quotation 'nissitassa calitan ti'
2
* is used by the Pe to
illustrate Mode 16 while the Netti uses it for the 12th Mode.
Netti ch. ii and Pe ch. vii have the same title, and their contents
differ only in detail. The Pe here applies the 16 Modes to 16
different quotations in succession—in nearly every case to the one
verse and one prose quotation for each of the 8 types of Thread laid
down in its ch. ii, 1st Grouping. This repetition is avoided by the
Netti in its corresponding ch. ii, where it applies the 16 Modes to
only a single quotation in two parts. The Pe, with many of its 16
quotations, gives first what it calls a 'Thread Demonstration'
(stUtaniddesa: e.g., p. 187) or 'Thread Meaning' (Suttattha: e.g.,
p. 192), sometimes as long as 6 pages (pp. 182-7), before applying
the Modes. This device (used by it to introduce a range of exegetical
trains of thought) distracts from the effective demonstration of
how the Modes work. The Netti only uses the device twice, and in
its first chapter, but exceedingly briefly (p. 10, lines 1-3 and p. 40,
last 8 lines).
Netti ch. iii and Pe ch. viii have the same title (the title 'Sutta-
vebhangiya' appearing in all Pe editions is an absurdity resulting
from mistaking an epithet applied to the author for the proper
chapter title, Nayasamutthdna, see Pe p. 259). The subject-matter
corresponds closely, but there is a notable difference not only in the

26 Netti p. 65; Pe p. 110, line 12 (also 1.17) in the garbled form 'Nissitacittassa
ca mattiho ca nissayo tanha ca ditthi ca? (so all texts), which should read
i
Nissitassa calitan ti ho ca nissayo? Tanha ca ditthi co?—a good sample of the
crude readings common to all eds.

order of the Guide-Lines but also in the allotment of the material
to each. The Netti has greater length and detail. The last five
verses of Pe's ch. viii are found at Netti p. 4, w . 17-21 (3rd
Summary).
Netti ch. iv and Pe ch. ii have the same title and both are con-
cerned with grouping types of Threads. While the Netti has two
Groupings the Pe has three. The First Grouping—that based on
the four types of Threads beginning with 'that Dealing with Corrup-
tion'—is nearly the same in both, only the Netti has 16 permutations
and elaborations against the Pe's 8. The Netti's Second Grouping is
the Pe's third. In this each has 9 triads and one single heading, and
of these only one triad differs, though the order is not the same.
Pe in its schedule of this Grouping includes (without explanation)
two extra triads not in the detail. In the detail of the Groupings
the Pe gives in each instance one verse and one prose illustrative
quotation, but the Netti, while mostly giving the same quotations as
the Pe, adds many extra for each type. The Pe's Second Grouping
is not in Netti ch. iv; but the six terms beginning with 'gratification'
(assada), from which its 13 instances are built up, are all found in
Netti ch. i, Mode 1, while the Pe does not give them at all in Mode
1 in either its ch. v or ch. vii. Again, in Pe ch. ii, the Perfect One's
10 Powers are sandwiched rather arbitrarily between the First and
Second Groupings. The Netti, however, puts them in its ch. ii
under Mode 2 (where their presence might be a little more easily
justified under the heading of 'Investigation'). Also the form in
which these Powers appear in the Pe is much further from the
' Pitaka version than that in the Netti and the order differs still more.
As to the remaining four Pe chapters, nos. i, iii, iv and vi, they
are all subordinate and introductory to the other four. Examina-
tion shows that nearly though not quite all the material in them is
represented in some form in the Netti, and also that what is not in
the Netti is, however weighty in itself, not directly needed to
exemplify the Method intended to be set forth.
Pe ch. i is introductory to its ch. ii. I t has two parts: an Intro-
duction (pp. 1-5) and an Exposition of the 4 Truths (rest of the
ch.). Half of the Introduction (pp. 3-4) appears almost verbatim
at Netti pp. 1-3 (in the 1st and 2nd Summaries). The contents of
Pe pp. 1-2 are represented briefly at Netti p. 8, and the main contents
of p. 5 appear in slightly altered form at Netti pp. 8-9. The second
part of the chapter does not appear in the Netti, but the pith of it—
the teaching of the 4 Truths which it exemplifies—is, in fact


abridged in the Netti's treatment of Mode 1 in its ch. i (as a teaching
of the Truths, see p. 8 especially). Three of Pe's quotations in this
chapter (that on p. 5, and the first on pp. 13 and 17) appear res-
pectively at Netti pp. 72, 11 and 24 (repeated p. 53). I t can be
criticized of the Pe here that the second part of this chapter, as it
stands, does not directly illustrate or elucidate the Method which
the work is intended to set out, and that the form it takes, besides
being rather incomplete, is something of a distraction. In this
respect the Netti's handling is better.
Pe chs. hi and iv introduce its ch. v. Ch. hi in its name (Suttd-
dMtthana: so read) reduplicates Mode 14. Its contents, though, fall
into three sections: (1) Expression in terms of the 6 Eoots (this
covers the two triads in the 18 Eoot-Terms), (2) of the 3 kinds of
Action, and (3) of the 5 Faculties beginning with faith. The first
is devoted to establishing how the 6 Eoots function as causes for
outcome in this life and for fruit in future existences. This chapter
is also open to the criticism that it is not self-evident why it is set
out in this form and that it does not directly exemplify the Method.
Quotations, mostly followed by a discussion, illustrate the 14
headings. There is nothing in this chapter of which it can be
definitely said that it is not dealt with in some manner adequately
in the Netti.
Pe ch. iv is the shortest and perhaps most consistently difficult in
style and treatment. I t is redundant in its title (Suttavicaya)
against the 2nd Mode (Vicayahdra). I t deals with the injunction
given in the four Principal Appeals to Authority (cited on p. 77).
In the Netti this general subject is covered more neatly in ch. i,
Mode 3. This Pe chapter has 3 parts. The first tests for com-
patibility under the profitable and unprofitable; the second does so
under conditionality (treated also in ch. v, Mode 15, cf. Netti ch. i,
Mode 15); the third does so. under 'what is agreed by the Buddha'
(anunndta: the title reappears in one of the triads in Pe ch. ii, 3rd
Grouping, Netti ch. iv, 2nd Grouping). A tail piece (p. 80) discusses
how the incompetent can confuse the basic types of Thread laid
down in its ch. ii, 1st Grouping, and is called Suttasankara. This
too contains nothing of importance not dealt with in some way in
the Netti.
Pe ch. vi is a kind of 'omnibus chapter'. Its position is unexplained,
though it can be taken to introduce ch. vii. I t is the only one which
contains some exemplifying material definitely not found in the
Netti. In its restored form (for reconstruction see sect. 5a above:


taking it to run from p. 113 to p. 153, line 11, word '. . . mettd\
21
with the end displaced back to p. 140 and some portions lost) it falls
into three main sections. (1) The first defines the three com-
prehensive and complementary Pitaka classifications of phenomena,
namely the Categories, the Bases, and the Elements (pp. 112-4,
1. 12). (2) The second deals in turn with (a) the Truths, (b)
Actualization of Truth, (c) definition of Defilements and (d) their
Abandoning (the 18 Root-Terms again), (e) the Noble Planes and
(f) Fruits, (g) the Extinction Element. A terminal 'schedule' for
this second section appears on p. 136,1. 4, beginning with the words
'Iti saceani vuttani ... ' referring back as far as p. 114,1.13. (3) The
third deals with the 9 Attainments of Concentration (p. 136,1. 9 to
end of ch. as restored). I t is in this last section that the major
confusion has taken place, with possible loss of some material.
As to the short 1st section of this sixth chapter, while the three
classifications appear with others in Mode 12 as 'Ways of Entry'
and so are redundant here to that extent, the definitions are not
given in the Netti. But they are incidental normative material not
essential to an exposition of the Method.
Section 2 is also largely concerned with definitions. Some but
not all are represented in the Netti. Compare the definition of
ndmarupa at Pe. 116 with that at Netti. 15 under Mode 2. Again
the definitions of the members of the Dependent-Arising formula at
Pe. 117f. appear in altered form at Netti. 28f. under Mode 4. The
definitions of the 3 Unprofitable Roots and 4 Perversions at Pe. 118
are also represented at Netti. 27 under Mode 4. Pe's definitions of
Craving, etc. (p. 121), Quiet, etc. (p. 127), and Faith, etc. (p. 128),
appear mostly in altered form at Netti. 27 and 28 under Mode 4.
The treatment of Quiet and Insight as 'medicine' (bhesajja) for
Craving and Ignorance at Pe. 123 appears at Netti. 140. The rest
of this section contains some matter not in the Netti, notably the
passage on 'actualization' (abhisamaya), which Acariya Buddhaghosa
considered important enough to reproduce at Vis p. 690, though
attributing it there to 'the Ancients' (Porana).
The contents of the 3rd section include a long detailed dissertation
on Meditation (jhana) not found in the Netti. But this material in
such detail does not in itself directly help to elucidate the Method,
rather the reverse. From the strict point of view of the purpose of
the works, its omission from the Netti is justified.

27 See note 23.


So much for the individual chapters. As to general handling
several things emerge.
(1) Nearness to the Pitakas. Without the question of its un-
orthodoxy arising, the Pe's handling of some subjects is sometimes
(with no reason apparent) much further from the Suttas than the
Netti's is. Note especially the '9 types of Axahant' on pp. 31-2.
Some of these are apparently found nowhere else in Pali ('pative-
dhanabhava9
,
 (
sace ceteti na farinibbdyi no ce ceteti parinibbdyV and
its pair) and two are called by names used only for Meditation-
attainers (in the Puggalapannatti: 'cetanabhabba9
 and 'rakkharm-
bhabba'): compare Netti's '9 persons of the type of Thread dealing
with the Adept' (§946), which retains the number '9' but discreetly
substitutes other names more in line with Tipitaka usage. Also the
Pe's use of the term 'farihdnidhamma' (p. 32), applied to a path-
attainer though not to an Arahant, may be compared with the
discussion in the Kathdvatthu (p. 69ff., trsln. pp. 64ff.). Again the
Pe lists four 'noble planes' (bhumi: p. 135), namely dassanabhumi,
tanubhumi, vltardgabhumi, and katdvibhumi (so read for Jcatdbhumi,
cf. Pe. 66 'akatdvV) for the paths of Stream-Entry,
28
 etc., but the
Netti only admits dassanabhumi and (for the other three collectively)
bhdvandbhumi (Netti. 46, etc.). Again the Pe's handling of the 10
Powers of a Perfect One is oddly further from the Suttas than the
Netti's (Pe. 32ff., Netti. 92ff.). In this, too, the Pe includes an
exposition of the 4 patisambhidd (pp. 33-4), which differs considerably
from the Pitakas (and even the later commentaries) and is absent
from the Netti. And the Pe sometimes uses Sutta similes in a way
not found in the Pitakas, which the Netti does not do: see the
similes of the man making a spark from wood, drawn from M. i,
240-2 but applied quite differently (pp. 1-2), or that of the water-
lily pervaded by water, from M. i, 277 where used for the 3rd
Meditation but here applied to the causing of determinations
by ignorance. The Netti has none of this (except perhaps in
§136).
(2) Use of quotations. The way in which the 'type of Thread
dealing with Corruption and with Penetration' (Pe. 25-6, Netti.
153-4) and that 'dealing with Corruption, with Penetration, and

28 These four terms are not found, apparently, elsewhere in Pali in this
setting, though the individual components occur. The pattern does, however,
occur, for instance, in the Astasahasrika Prajnaparamita in its list of ('Hina-
yana', probably Sarvastivadin) stages 'surmounted' in the Mahayana concept
of the Bodhisattva. (But see PTS Netti p. 257.)


with the Adept' (Pe. 26-8, Netti. 156-8) are differently illustrated by
quotations in the two works favours the Netti. Also the different
choice of quotations in the last Grouping of the two Sasanapatthana
chapters (Pe. 48ff., Netti. 161ff.) and elsewhere may be noted.
Note also the use by the Pe of the wrong half (if not a copyist's
mistake) of a verse at p. 48 (kdmardgappahandya instead oisakkaya-
ditthippahandya) in its section Lohikam Lokuttaran ca. The Netti
in that section gives another quotation (p. 162) but gives both
parts of the verse at p. 146. ('Sensual desire' is abandoned by the
first Meditation, but that is not 'dissociated from worlds'; Embodi-
ment-view' is abandoned by the 1st Path.) Where they differ the
Netti's choice and handling is always better. The Pe has a rule
for using one verse and one prose quotation for each heading, which
the Netti with advantage ignores.
(3) Handling of subject-matter. The Netti is economical, neat and
not uningenious in the marshalling of its complex exemplifying
subject-matter, never redundant though sometimes elliptic, careful
to avoid tangled discussions, and successful in differentiating the
individualities of the 16 Modes. The Pe on the other hand is some-
times redundant (see above), does not always subordinate well the
exemplifying matter to the elements, fails to exploit the 16 Modes
fully and so has to multiply its chapters (see also above), sometimes
rambles into distracting and incomplete combinations (e.g., pp.
30-2, where three of the seven types of noble person (see e.g., M.
Sutta 70) are omitted from a complicated argument—n.b. the
absurd but rectifiable mistakes in the long para, on p. 31 are due to
bad copying and must not be ascribed to the Compiler). How
much clearer the Netti can be than the Pe is instanced in the res-
pective expositions of Mode 15 in the two Haravibhanga chapters
(Pe. 104-9, Netti. 78-80). Other examples could be cited. The Pe
is in general much less successful than the Netti in making the Method
emerge from the material, and this is especially evident in the Pe's
larger number of chapters and its handling of its two chapters on
the 16 Modes (chs. v and vii): for example in Mode 1 (ch. v) it
merely defines what is taught, namely Suffering (already done in
ch. i), whereas the Netti takes trouble to show what it means by
teaching, which characterizes this Mode. The Netti's more difficult
handling of the Guide-Lines turns out to be justified.
(4) Lay-out. The Netti is severely tidy, regular and formalistic in
the details of its compilation: the beginnings and endings of its
paragraphs, the placing of the verses labelling the 16 Modes, the


settings and presentation of quotations, the elaborate portico of its
three initial summaries with the verse pancma, and so on. Its four
chapters follow a natural order of development. The Pe is some-
times so loose and inconsistent in its paragraph-endings that the
reader is puzzled to know where a paragraph or a section begins or
ends, its untidy and inconsistent introduction of sub-headings is
particularly noticeable throughout ch. vii, and it is often very scrappy
in the way it presents quotations (e.g., p. 26, 4-6). I t begins
without adequate summary or introduction after a short prose
pandma. Its eight chapters are in no particular order. A curious
feature is the 'elegant variation' in its chapter-terminals, no one of
which is worded like another; note also the addition to the terminal
for ch. i.
(5) Style. The Netti'a style is even, clear and economical, though
with one or two rather abrupt switches from one phase of an argu-
ment to the next (see ch. ii, Modes 1 and 2), which is a fault of its
tendency to over-terseness. The choice of words and idioms is
regular, severe and never strange. I t has peculiarities of its own
(see Sect. 12). The Pe's style varies markedly. Mostly it is
crabbed and hard even when the copyists' mistakes have been
removed. Sometimes it is unlike any other Pali work in this
respect (e.g., ch. ii, pp. 33-43, or ch. iv), but sometimes it is quite
clear and straightforward (as in ch. viii).
I t is remarkable that the NettiA takes upon itself (PTS Netti
pp. 251ff.) to 'improve' (there is no other word) on two of the Pe's
expositions, namely the application of the 16 Modes in Combined
Treatment to the verses 'Manopubbangama dhammd . . .' (PTS
Netti pp. 250ff.; Pe. 163-70) and 'Dadato punnam pavaddhati . . .'
(PTS Netti pp. 157; Pe. 237-41 but including the displaced passage
at pp. 234-5 and excluding the intrusion at pp. 239-40, see 5a
above). Acariya Dhammapala's re-presentation of these two
expositions, so very close as they are to the original yet in part
markedly rewritten without comment, seems to imply a tacit
criticism of the Pe's handling of them. (There is no doubt that
these versions are rewritten and not mere old textual variants.)
Far more detail could be unearthed, but there is no point in
doing so if this survey has succeeded in showing two things: first,
that from the strict point of view of the Method both books are
intended to display, what in the Pe is not represented in some way
in the Netti has not enough direct importance for the Method's
elements to justify the Pe as a 'continuation' of the Netti; and


second, that the Netti is an 'improvement' compared with it.
Both points seem clearly enough established. All this considered,
then, neither book can be called a 'continuation', but one of them
must rather be a rewritten version of the other. In fact, the Netti
is in all ways so much more polished than the Pe (as the Maha-
vamsa is than the Dipavamsa), and so different in this particular
way, as not only to preclude their being by the same writer, but to
make it inconceivable that the Pe was compiled by anyone who had
written, or read, the Netti.
I t seems always to have been taken for granted in Europe
29
 that
the Pe was a continuation of the Netti and therefore younger, even if
not much. On this theory the facts that emerge from this com-
parison are quite inexplicable, though they are natural enough if
the Pe was written earlier.
There is actually no real evidence at all which obliges us to assume
the first theory: everything, in fact, favours the second. The whole
internal evidence points unequivocally to the Pe's greater age and
suggests that it may be quite considerably older—and from that
point of view it is the more interesting work—with the Netti as an
improved version.
Only one argument against this has to be considered (leaving aside
the tradition, which does not directly affect this question, namely
one of different presentations of a method whose origin tradition
ascribes to an earlier period). I t is put forward by the editor of the
PTS Pe edition as follows: 'Netti-Pakarana, however, is older than
the Petakopadesa. The latter presupposes the former as will be
evidenced from the text (Solasahdrd Netti, etc.) and also from
various other quotations' (p. iii). But actually this argument begs
the question; for 'netti
9
 (the purely European conceit of the capital
W is quite unjustified here) is primarily a noun, not a proper name.
In this verse 'Solasa hard netti panca naya sdsanassa pariyetthi
atthdrasa mulapadd Kacwyanagotta-niMittha' (Pe. 3; Netti. 1, but
ends instead with 'Maha-Kaccdnena nidditthd') the word 'netti' by
no means has to be assumed to refer to the Nettippakarana, it being

29 No European scholar seems to have explored these two works much
(see n. 16 above). The Pe is 'called by Hardy a most obscure text and by
Fuchs'—Specimen des Petakopadesa, on the Pe's 1st ch., Berlin 1908—'as
offering insurmountable difficulties' (PTS Pe p. i). True, there are difficulties,
though they are not 'insurmountable' except doubtless the lost paragraphs of
the Pe.


there no more a proper name than the words 'pariyetthi' or 'naya?
are. The Pe being taken as the older work, the noun (
netti\ making
its first appearance in the Pe's verse, simply means 'guide', just as
'nayd' there means 'guide-lines' and 'pariyetthi
9
 means' search'.
The Nettippakarana can then be regarded as taking its title from this
word, perhaps from this verse, which it reproduces (p. 1), and then
'Netti
9
 as terminal-title (Indian books have only terminal titles,
initial titles being a desirable Western addition) appears in its
proper place at the end (Netti. 193). This is consequently no
evidence either way in itself. As to the Various other quotations'
(see above), presumably such as the passages at Pe. 3-4 (= Netti.
1-3 roughly) and other passages common to both, they establish
nothing in themselves as to which work is quoting which. But the
rest of the internal evidence all shows that the Netti must be quoting,
and often improving on, the older Pe. Two further points support
this view: first, the Netti commentator's treatment of the two books,
leaving the Pe uncommented and dismissing it in his prologue (PTS
Netti p. x), and second the fact that the Pe makes use of the old
device of mnemonic verses (udddna) for its contents (pp. 12, 21-2,
43, 57-9, 72-3, 80, 258-9), while the Netti does not. This habit
belongs to the period of oral tradition before the commission of the
scriptures to writing. The Pali scriptures were first committed to
writing in Ceylon at the beginning of the 1st century B.C. (Maha-
varhsa 33, 100). Perhaps the Pe was earlier than that and the
Netti later, but before deciding from this, the commencement of the
use of writing for such books on the Indian mainland would have to
be considered in this case.
From Acariya Dhammapala's account (6th century A.C. or a
little later: see sect. 3 above) it may be assumed that the origins of
both works had become a matter of fixed tradition, and that he
regarded both as setting forth the same subject-matter, with the
Netti overshadowing the Pe: 'Petakena samsandetva9
 ('Having
collated it with the Petaka': Netti A Prologue); but he says nothing
about either work being a 'continuation' of the other or about their
relationship.