Professional Documents
Culture Documents
FOUNDATIONS OF
D HARMAKI RTI 'S
PHILOSOPHY
John D. Dunne
Fil'lC Edilion
0, 01 07Qj)OS 04
ISS4}11
Li.,." .rc.-.g,m
Dunne, Jolm O.
c.uJ.si..,..j,.-~ 0.,.
w..tGm PubIleationi booIu an: printed on acid·1m papa and meet the
........
guiddiner for pmnanmct and dunbiIil)' tel by the Council on Librvy
Preface I'
Abbcniarions
A Note 0 0 Ibe Sanskril and TIbetan Tnmlarjoru III
"
I PUMAI:fA Ttu:ou:
DHAIlMAItIllTI 'S CONCEPTUAL CONTEXT IS
1.1 TIN l'rocf!U lI(Knllwirt,,znJ lu Instn""4'rIt IS
I. S Sumnwry p
79
1.)
The Two PrIlmf1i14 The Two Rc:a1irics
MQrt D1I P.rtit:14ltm
The PC'rcqnibk as Ultimacdy Real
.
79
8.
The U\cimardy Real as Inaprasiblc and Momentary
Do Paniculan: Have Sp:atw ExtC'nsion~
"
,8
1.1 Utliwrs;t/s II}
Swbh411t1 as · Propeny- IU
SWlbhillfJU ·Naru~· 158
Naturc-Jl!/tbhilllf and the Dusal Complex 161
.. I NSTRUMENTALITY:
JUSTIFYINC THI1. SOU RCIlS OF KNOWLEDC E 22}
CONCLUSION ]t9
Nlfturr. P""!'tUn. ifNi RLfi_mr )10
~ CO NT ENTS
INDEX 4'9
Preface
"
PREFACE
'"
divergence of TIbetan opinion by understanding the history of the inter-
pretation of Dlwmakini's mouglll in South Asia iuclf-:a shift encour·
aged by my gradu;nc work with Charles Hal.I~. A grant from me
American In5(irulc oflndian Studies enabled me to spc:nd (W() years at the
Ccmrallnstitute of Higher Tibetan Studies in Samath, where I read com-
mentaries on Dhannaldrti's works with Prof. Rim SaJ'!l1w Tripi~i. and
this pt:riod in India was crilical lo my resorch. E~n in the Sanskrit works
of South Asia. howcvtt. the inu:rprttation ofDhannUim's thought devel-
ops and diverges (0 a wide alent; hence. with the mnaned hdp o(Tom
Tillcmans, I setded t'VCl'Ltwdly on a focused accoUnt of the c:arlie:st South
Asian interpretations of Dharmwrti as the subjm- of my doctoral d~r
alion (1m), which is dftttively the fim draft of this book.
As I was completing my doctoral work. Gco'ld Dreyfus's !WopW'"
Rt~lilJ, an eltmsivc study of the Tibetan interpretations of Dharmakini,
appeared in print. ThaI work, along with numerous conversations with
Dreyfus. aided me considerably in my research. While tremendously help-
ful , Dreyfus's study of the Tibetan interpm::ations also highlighted the nttd
fo r a similar, historical KCOUiU that focuses on a specific South Asian iiUer-
pm::adon. This book contributes to fulfilling that nttd.
My dissemtion built on Int work of numtrous sc:holan, and in tht
course of the substantial revision) that led to mil book, many res:pond<ed
with helpful comments and suggestions. Tom TiUmwu conrinued to p~
vide the son of advice wh~ pel'$picuous practicality is matched by the
keen philosophical insighu on which it resu. Ernst Steink.ellner, whOK
work figures prominendy at crucial juncrures of my argumem, took the
trouble: to go through the enure text- His critiques. suggestions. and encour-
agement have added greatly to this book. ShOryU Kabura likewise provided
a number of luggcnions, some through an attnded and entertaining
debate about paniculars. Brendan Gillon's careful and detailed respofUCJ
wen: esptcia.lly hdpful for darifying my analysis ofDharmakini', ontology.
Eli Franco provided a comprehensive response 10 my discussion of justifi-
catinn or -inslrumenCiliry- (p,..lfIIirrJNV that heipM me to clarify my inter-
pretation. HcJmut ~r dirr:ctcd me to some important passages and
provided welcome encouragement. And Richard Hayes's pithy remarks
proved especially helpful in reconceiving the overall COnten of my inter-
ptetatlon .
Many others who work on Dh:um.akini and related iuuc:s aided me in
varimu ways. A few ,hal enme ~ily to mind are T ab.d!i Iw:ao. Birgit
Kellner, Horst Lasic, Parim.al Pam, Ernst Preu, and Mark Sideriu. In this
PREfACE
uu
Abbreviations
I
NPOJ:L\NT PA$SAC E.$ fROM s...NS"JJT and
OUt this book. Most of these umsIations arc also included in the appen-
dix. where they art ~mOOided in the largtt pusages from which I have
extracted th~m . For convenience. th~ titles of most Samkrit and TIbetan
texa arc abbreviated in accord with th~ tabl~ of abbrrviadons, which is also
a key to th~ various roitions of Sanskrit and Tibcran tatS that I have
employa:l for the translations. As with most philosophical works in San-
skrit. Dharmakini's taa often employ II dialogic modd, whereby Dh:uma-
kini argues in response to critiques apressed in the voice of an objector
(Pjjrw~), whether actual or hypothtt:icaL To represent this convention,
I have used quotation nudes to indicate the beginning and end of an objec-
tion in a nanslated passage. Another feature of thu textual tndirion is me
interweaving of tats, such mil( a commentator's prose often includes
phr:ues from the veiW or commemary mat he u discussing. [n some ca.scs,
it is especially hdpful (0 know which phrases in a commentary a~ sup-
plied from a verse or anomer commentary, and in such instances, I have
italiciz.ed the phr:ues in question. Finally, a.s a plaincd in me inuoducrion,
I have avoided to me greaten at~n t possible me use of square brackeu to
indicate irucrrions in uaruJacions. When brack.ru remain, th~ inscmons arc
particularly lengthy, or mey arc less dearly supponc:d by commentaries or
grammar.
Introducrion
B
UDDHIST PHILOSOPHUS often spcU of beginninglcuncss. It is
daimN. thai !he minds o f livin g bcoi"Sl> for e>ample, have no Mgin-
ning. and mat OUf current universe is only one in a beginninglcu
cycle of expansion and dcoy. Some Buddhuf minkus would claim that
even me mon mundane tuk can have no true beginning. That is, if a
beginning OCCUr$, there mUSt be some moment, some -now," in which it
occun. For the present to exist, howcvu, there must be a pm and a fuNfe,
for what wo...kl - now" mean if there WlI;re no lime o ther than now~ And
of COU lX, if there is a pUt, then how could now be a bcsionin!j? Now
mould instead be the end of the past. Each beginning. in short, must iudf
have a beginning.'
In a more a:mcrct:e senK, this book also sam ITom bcginningle5$neu, for
it arises from a need fOf a point of dcparture----a pl:llct: from whtch to
begin-in my work on the thought of Dharma1c.lrti, a South Asi2n Bud-
dbul philOlOphcr of the KVCnth "011.11')' (c.a.).' ThaI Dharm.a.ltirri U wor-
my of OUf attention SttmJ sc:ucdy ncttUaJy to justify. Following upon me
work of his pmf«:euor Dignlga, Dharmwni addressed at length numer-
ous quesrions that are ofcentral conct:m to Buddhist thought and practict:.
The impact of his views on Buddhist throries of ~roeption, infettnct. and
,
, FO UNDATIONS O F DHAlMAKIRTI 'S PH ILOSO PHY
3 1 am .mrnns 10 eM J."l tip" tWt, the ~ lup holiday of which L..obans GyIIIO PI.,.
vida • fucin,rilll and mo¥in& account ;n bia mtmOin (I".). Gcorp Drqfus pva an
=ended...d ~riw iiIIXOlIIIl of lhe ewtIt (1OO}:'Wf). and he libiK diJcw.a; eMO¥e\"'
&II pbtt of DlunnalrJni in I>F lup cduaOOn.
.. G.W.F. H.-F. 1Oc CDmpk. ~l\Ilhe nWn body orhis IcctUICJ on rdipon of 181, with
thil art:U (r9IBa l}):
The qtlCJ{ion with wIUch _ haw 10 bcpn is: · How 1m.." 10 ~ , IwgiMins?" FOI
il is of CIOUlX at Ieaa , IOrmal requilCnlml of aU .oa-.rific: knowkdsc. and ~
phiJo.opby. thai noct.illl should O\XW it. it that hal noc fCI been po_cd. AI the bcsin.
nin" • .0., <a . _ haw noc fCI pro_ed l~ytbin&i and we annoc fCI appcaIlO an,.·
thins anlrc:nlcnl.
, l..aCapra h !JIj:}I).
INTRO DUCTIO N J
~ coming to the point where we ask how theories about an entity's nature
relate to orner issuC$, such as the questions of rational jwtification and
authority. The central aim of rnis book iI thus 10 conuibure 1Ow:mi the
devdopment of a COI\SMSUS by presenting the foundations ofOharmakini's
philosophy in terms of a consciously coruUUcted starting point.
In speaking of the "foundations" of Dharmakini's philosophy, I mean
rnose issues that repeatedly surhcc: throughOUt his work: they arc the fun-
damental demena of his conceptual system mat, on my view, make all of
hi, argumenu pos.sible. I have otg2Jliw:! mose demenu under thrte broad
categories; (z) ontology, (1) me "natunl rdation" (swiJhilJllprlftihllnJh.)
in inference. and u)
the issues ofjwufication and authority, which I place
under the rubric of • instrumentality" (prlim4?tJ11). These broad categories,
which structure this book. encompass all me demenu that enable one 10
understand and appreciate any argument made by Oharmwni. At the
same time, these categories include what is most difficult-and hcna:, most
controversial--in Dharmakini's though!. Thus, somewhat to my surprise.
I have written a book that is both an in-depth introduction to Ohar-
makini', philosophy and a detailed intcrprctluion of certain difficult poinu
in his work. t
m~ntafors, thor practices and ccrain fearurc:s of their texts likewise posed
a set ofhismrica1 problems thaI compdlcd me to construa a scuti.ng point.
In sum, I will focus on only Dharmakini's earliest tUts, me PrIlm4!U1"
INirttilt" and SVtI"!'fh, and I will resort only [0 the earliest commentators,
IXvc:ndrabuddhi (c:a.. 67S C E.) and ~uddhi (ca. 700 c.E.). To under-
nand my rea501U for restricting this srudy in the aforementioned f.uhion.
I should first aplain why I was kd to rely on commmwies.
In dlC: several yean: of research that went into this book. twO reasons
compelled me to resort frequently to commentaries. First. in practical
terms, any reader of Dharmakini's Sanskrit (au knows mat his dliptica.l
and intricate statements often remain im~nttrablc without commentarial
elucidation. Spcalc.ing in general of Dharmwni's style. Rkhard Hayes
rttm to - the tortUow writings of this highly compla: thinker.'" And refer-
ring specifically to Ihe SlIIJurtti (PVSV), a tat that is especially imporuru
for my analysis. Hayes and Bundan Gillon together nOle:, -Dtwmakirci',
styk is 50 rerse mat it is nor always immcciiatdy clear what philosophical
points h~ intends 10 make...• I would add dUll . leave alone th~ question of
its philosophical conl~nt. evm th~ straightforward meaning of a sentence
sometimes seem utterly obscure in Dharmakirti's sparse 5l)'1~. The f'CIuh is
that. unless one wishes to argu~ from highly conjm:unl int~rpreutions,
one mwt rcftt 10 commemaric:s. wbm: missing phn.scs:ll~ supplied and the
dcgandy IOnuous relations of Dharmwrti's grammar ar~ plausibly
re$laled. Thus, for purdy practical reasons. comm~nwjes become an
jnevi(2.bl~ companion on any foray into Dhannaldrri's texu.
Beyond pr:r.ctK:al conUrN. however, lies another compcUing reason for
my rdial\Cr: nn oomm~ntaries: my larger airn-.n~ dut extends beyond
Ihe prescnt work-is not 10 understand Dharmaldni's thought in and of
itself, bul rather th~ 5ubscquenl '"' of his thought throughout the hislO!)'
of Mahiyina Buddhism. Thus, cvm if one could somehow undentand
Dharmakini's works in a manner mal ignored the history of their intc:r*
prctalion, such an approach would thwart me in my goals. In part, an:abir
.nrical ~ing would be useless becl.U5C it is a fanrasy masquer.Wing :IS
truth: my a.ssumplion h~rc is thar my own undemanding is historica.lly
condilioncd, and thus, an ahistorical reading of Dh:umaldrti would be at
besl ddudcd. 8uI setting uiclc questions of ddusion, onc of my central
aims in attending 10 the UK of Dhannakini's thougiu in particular hislOf'*
7 HJ)'ft (191")19).
8 Hlya iUld Gillon !10K (1991:1).
INTROOUCTIO N ,
ical momCf1ts is to creat(: an awattMSS of my own location by rdkaing on
me way omen ITom a diffeunt time and piau appeu ITom my perspective
to be conditioned by meir circumst.:lnccs. I do not C-nvision that such an
awattllCSS wiU awaken me from the nightmare of history so thai I mighl
move beyond my own ·contingent arrangements.- InStead, by gaining a
greater awattncss of dut contingency, I hope to ante a more dkctive
agency therein.·
Thus, in consulting commentaries, I have not used them as unc;:onrcsted
rtStliltermnts of Dhumwni's tau. Instead. I have sought to learn and
even to employ me style of reasoning dut they bring to the interpretation
ofDharmakirti's thought." In doing so, I am nO( only able to USI! this ....-ork
as a foil to the styles of reasoning available to me in my own milieu. but I
also hope to have offered an interpretacion of Dharmaldrti's thought that
readily enables ooe to more deeply apprmate its relevanct in the history of
Buddhism.
Having chosen 10 rdy on commentaries, I eventually encountered thrtt
featurtS of their style of reasoning thai are especially relevant here. lbc first
is the S)'Stetnattaty that we have already mentioned. Put simply, for modi-
rinnal commentalors. the hen interpretation of the mailer at hand is one
that allows for the grearest rohercn~r al lean- produces minimal ren _
sion-with any and all other issues addressed by Dharmakini, One: upshot
of this sysrematic :approach is that it inadvertcndy highlights the pieces Uul
do not fil readily into Dharmaldni's philosophical punic, Among the
str2rcgics used to cope with such inc;:onsinencies is the second rdevant fea-
(UJ'C of this style of rc:asoning. namely, that meaning res.idcs in the author',
infennon. noc in his ccxu, and. mat in most CUCII the author', intention
remains the J2/T1e over the enrirt corpus of his ....-ork. Finally, the third fea-
rure of the style of reasoning employrd by tflld.itional comrnc:nt.:ltors is
straightforward: Dharmakini, to pur it Mundy. can never be wrong,
, I dnw tht notion of "coolingm :ur:ancanmu" fi-om Quentin Sk.inna (I~), Oapir.r
SkinntT'l probkmaric mlplwil on authorial inrenl, M doqucndy ~ III IV .tIow
laa &om ochn- timet ro di'f'l,« OW' own _ptionl. In tone. Skinner'. ~ mill
taembIa 1..aCapn'l (191,), and it coouva with the monolithic: (and --'vol hubrisric)
notion of one'l own lUaorialloacion due is implied by the "re-«fIlClIion- m:jwmt by
Ridwd Rony'l mnbod of"nlional -oon" (1914),
10 In uans tht notion of • ...yk of lftIOIIi"" I om rdrrrins 10 the - " of Ian Hadins
(1911), In brief, Hadcin" poine,...nidl mill\! Ix c:onaiveI! •• micldk way Ix..« .. incom·
~ity and iD<klaminacy, is dw OW' COIItti II with a oryk of lUIIKIincdoa not CO&-
urn uulh Yaluc: ' " N, but ruMr Ulal which II\IUJ • proposlllon · up for pabI. ill a
·c.ndidat~ for brin& ttuc: or~" (19141),
6 f OUN DATIONS OF OHARMAK IIlTI 'S PHILOS OP H Y
prove problematic in a way that has led me to place: strict limits on the
prescnt srudy. The attention to SYSlematicity, inasmuch as il is coupled
with an appal to Dharmakini's intenrion. pennia commentators to move
freely among Dharmwni's tc:xU, and since: DharmaXini composed eight
philosophical works, Il a contemporary interpmer would encounter signif-
icant problenu if she were 10 uncrilically accept the commentators'
approach. In pf1lCtical terms. the shc:tt size ofDharmak:itti 's written corpus
would require a contemporary hinorian to apply the ament, highly focwcd
standards of historical interprc:t:ation over an unmanageable :amount of
matc:ri:al. BUI more imporwu, a hilurc: to attend to the: diffc:renc::es in Dhar-
m:akini's lau c:ffc:aivcly leads o ne [Q :adoptlhe same: Stance: :as the tndi-
tion:al commentators, namc:iy, th:al meaning resides in Dharm:akini's
inlemio n. nOl in the tau. In orner words. if! choose: to aplain rne mean-
ing of:a passage in Dh:armakini', Pra1Ni!fllwirttW" for c:xample. by recourse:
to any other passage: in any orner ofDharm:akirti's tau, I mWI da.im th:at
something beyond Wt tCXt itsclflinks it ta thO$C othe:r tc::lU. For the: tr.;J -
dilaon:al commentators, that link is provided by rne: intendons in Dh:ar-
mw"i's mind, and unless I affirm:an cvc:n more: obscure: link:agc:, I tOO
will evcnru.:ally rtSOn to the notion ofauthorial intention. ~nd the pmh..
lent.. anc:nd..:anr upon any a tte mpt" to u ncrlVc:r inrc:tuion , oJ thc: main diffiallry
here is Wt, in constructing my own versio n of Dharmwni's mind, I will
fail to Itt the version presc.nled by the commentaries :at hand. In olher
words, as the particularities of the tau themsdves fade from coll5iden-
cion. 50 too will the distinroon bt:two:... my own imagined Dh:armaki:ni and
the: commennu ors' venian.
My ~pon~ to mil problem il not to rejc:.cr m e: c:omme:nr,uiCf; in favor
I 5 h "f'pc.oo.. th.. in cad. of hi. ...... n <>Ommen" on .he "".- o r ...... p..,wJ'!"'u, .......
Manoc-adwwldin (o:wdfrl,.( ~l «Drury), theautbor olf'VV. recotdr1lUb.cim INIl)' olJX¥en.
IN TR.ODUCTION ,
uraily repeaa Dcvendrabuddhi's commentary. since part ofhU work is a
subcommenwy on Ikvt:ndrabuddhi's tat. But Sikyabuddhi also expands
upon Devcndrabuckihi's work by adding his O'NO insighu. When we then
come to later authon such as Santaralqira (71.S~-788) and Dhannottafa (A.
ca. 800), we find that the ideas and often verbatim phrases.----of both
Ikvt:ndrabuddhi and Sikyabuddhi are repeated in their worla without any
indication as (0 their origin. " This trend continues even in Tibet, where
ideas and phrases of South Asian commentators are repeated without iden-
tification by Tibetan authon."
It is worth reiterating that the Iayaing of commentaries does not come
about bccau.sc commenta(Ors, in some slavish adherence (0 tradition, fail to
be original. Indeed, the layering of commentaries indicates exactly the
opposite: each maNm rcpresenu a new set of innoV2tions and insighu
brought to the issue at hand by that generation's commentaton in response
to the various developments of their times. Considt:r. for example. the fol -
lowing comment or Sikyabuddhi . Here he summariu:s a passage from
Dharma1cini's SVdl!J'ffi that addresses Dhannakirri's philosophy orlan~
drabuddhi'.commma. And nm PrajMbncupu, ...... won iI: Ilriki.. for h.iI ~dy
ddilMnu dtricion to ~ ..... prmou. (:(IIJUnmwil:$, _ ~'I -0. from lime
10 rime. Sec, r...r insona, hili . . of doc eampk of doc twlns.I PVJ.n. (~u·
lor'l"u...., 9).lj; PYP:U9aI) and me nocion of Wlivnais as _1Ur.f ( h'-~U·
~,,. .); PVP: I )l~} in flil commala on PV).I9-l.I .
16 Slntanktita. for eumpk, dttm::J hil: notioa of tho! three ~ of constrvin& me tmn
IUIJi1 , rrs" OO:HOO)J from ~buddhi ~ bdow. 1)11f, and abo PVr".2b-I4D -
K:l~. Dharmo«an (PVIIIT. 5mnkdlMt and Krua.tr 1919,,).)11) adopu ~'.
notion of mIriaM" (_r./!J and auinPc ~) inamunmtaliJ}' (10, In IICICOIlIII of lhoeK
norioona. _ bdow. ~f~lI). ~ ..... (PVi"T. S...<nkdtn........t Kn- '~:'.I If)
.dopa, albric wUh _ rnocIifiaUon, ~nbuddhi'l noc:ioN of"....."... and "y.M
(for ~i'I.,jew, I « bdow, 166ft). n- I « Oldy. rn.. 01 numeroou Qampk$.
17 See, for Qamplc. ~ Md.os Idan'. dioOnction beno«" Il\ISfwonhi,- (.""''!'..u.)
in IcnN 01 AIh;m and ob;ea (Drcyfuti I~). This dioOnaion b in fJo: 6", prescnltd
by ~ (PVP"~ , , poilU dw Siky. Md.os &dan doa noc: niK. Of count"
wbm Tabaan coau:nmOll)($ repeal !be WOfdI oflhrir South AlAn praLl - s, Ihry do 10
in librtan tnndadon.
18 PVT(78b41.JPVSV .JPVI .6~ l""lti!hJi' plM" J.u J.,,.i",", "",.,.,,. j-,
10 FOUND ATI ONS OF DHARMIIKIRTI'S PH I LOSOPHY
Tht: basi< poinl here is dul2l1 objca (1lnJu, ~)of 211 apress.ion (iIlbJll)
such as "cow" is actually a spt:cific type of nt:glUion dut Dhnmakini calls
an "olhcc-o:dusion" (Imyiipohll). Words, in short, have ncg:uions as meir
objt:ru.
Wim Ihis p:w;age in mind, let us turn 10 the much bttt commenttry' of
l<an)akagomin (ft. ca. 9(0). Al is so often the c:ase:, hc repc:au vt:rb:!.tim Ihis
comment of Sakya.buddhi. bOl hc makes 211 important ch2l1~:
N. 'P ..... /IV.,."". ~,. tk'/ !hJir [roj l pIN.. ,,{ IM i "" (... ,,"" J. tn4,. ~
......,.Ji"_.
.y.,;m,..
. . .
iii,. ";,iI,i.,.,.,..,,.*' sMJIw",
19 K(11P7-1I) MJ PVSV PVI.4 ~uU (i",..~~ ,,;w,.,...,.'~
in ;~ ,,~.
"'..."
N()(c WI me, nnphasir in the Inrul..
lion .. mlM.
20 In ~ Irrms. his plObkm ir .... one fim tUtcd by UddyoWwa {NY .... NSl.l.66;
617.1-4; lruubud below, 1)7}-namdy, WI the conlmi of an a:pra.>on Of c:oncqx....m
U -CIM'" iI JUbjccriYdy apcricnad U an affimuolion , 10 how can ,he ob~ of sud! an
a:prcstiort or C""lXpt be • MPlion1
21 Ginn mar hinorial i",pon1IKC, " ir ,,,,nw; ....... Inn aIonc any deu.ila of II".;. Ii..,.
or Uutitulional afIil.i:olionl, _ CilIInOI na.1U the pr«iK dales of theM: twO aucialIy impoe·
.... , conunen<a<on on ~m.akin.i '. ~ ••,,,, .~ w. can only _ e ..... ~.
.m
mooco.er, wnMC COIT\mCIItarioon (and ~lsrudicd dirccdy under) SinW'llqit:J. Siner
T.bcun -.n:a allow ... 10 pbWbIy daim Wt ~waIctUa _ aai~ in the mid
nuy. we m.... hnc • rdati¥c duin& thai pbca Dcw:ndnbuddhi and $&y:.buddhi noc bdOn::
em-
1M ICWCIIth cmnuy, b.n noc afia' tIM- atfy cipltl. This type of rdaUOOf. appro1illUte datin&
it I)'pieaI in !he ~ ofSO.JlD Alian thinkrn. but ro. ow pu~ a hiMorieaI uW)'$U Mcdo
only !he rrYliw duos of !Mae thinkcn in rrlarion 10 ada odw:r.
22 In addilion 10 {"" "~oric1I ~Iio", ,hat underlie dw intnp~rion of DIw-
mMirti·, work. one can also poinllO a pnaical CIUIaIt"M of approoKhins his wOOt in me Wh-
ion. Sp«ffiaIIy, ;1 mabIa one {O pba timiu on !he- wcondaJy mamUl lO ..nid. one: rrkn..
I. .. pn:ciody b.ru. n::uon thaI we ...w.,.y rdari¥dy KUlI .'~Iion 10 _ _ I. bud·
olok or' .." vb ... .t- Q{ 0 •.,.,... to,,1) .ond 1<1"_.
(.".)-d,..o, """'. oohuwiK be
CONidm:d rdcnn .. WCtC '"' 10"udy Dhumaki"f. thou&h. io an a/Wroric:tJ fashion.
11 FOUNDATIONS OF DHAlMAKIRTI'S PH I LOSOPHY
Because: this book has various ainu. it also has various audiences. My over·
all aim is to make the content and ityie of Dhannakini's reasoning-u
interpreted by Dcvendr.abuddhi and ~buddhi-available to all my
readen, and this goal thus applies to alIaudienco: of this book.. I also hope.
however, to speak directly (0 specialists in an attempt 10 encourage a his-.
lorica.lly focused consc:nsus on at least some centr.al issues. Hence, I aim to
proem an argument Wt is of sufficient rigor and deW! 10 maint:ll.i.n a spe-
cialist', atttntion. In doing so, however, I wish to avoid the risk wr, in its
technic:ality and minutiae. my presentation will become impenetrable to
readers not activdy tngaged in this fidd.
Balancing me nccd.s of Jpecialisu and more broadly inleltSt«i readers has
iu dangcn:. Those tngagcd dirtttfy in rtscaICh on Dhacmakirti or Pramitta
Theory might occasionally ask for (Yen greater abundanee of deu.i.l and
citation, while thost: nOI dircaly tn~ in such research will find addi·
tK»nal dttail supttfluous or evtn overwhdming. To allay some of this risk,
I have taken ~ral st~ FiBI, I haY(: plaud all innoduaory materia.! in
the tint chapter, when I prestnt an overvi~ of the style of discouI'U
fOcused on Pram~ Theory. Second, when we tum in the remaining chap-
ten 10 a deuikd cuminalion of Dlwmakini's own views. I offer a sus-
tained bUI no, overly rechnic:al argument in the body of the lat, while
providing greater technical detail in the nottS. Third. ,0 provide all read·
en with ,he most nlevant primary source material, I have included an
aleNive appendix of cr.ansluions from key passages in Dhumaluni 's
works. Finally, I have auempled to avoid 10 lhe greuesc extenl possible the
use of sqU:lll! br.ackeu in my tr:rnsl.arions of primary ( att. In the academic
study ofPram~ Theory. the u.sc of (square] brackets has bewme a stan-
dard pr.lctice as a means 10 indicate words or phrases in me translation
mat. while implied by the source text or supplied by a commenwy. art nor
actually presenl in the soura: lal. This pracria leads to an atmndy lit-
24 Paul Griffiths ",...:,1). in hia wry dUawion of &Buddhisl: Hybrid English: PUff rho:
iJIue in mrlt ICmII: "'l1loert is abtoluldy no fQJCn why8uddholocr sIMKLId bemI,~ &II her·
mail: trldition, JaIed from rho: Wlinilia~ and pwed down from maNn 10 pupil by mpli·
.... II i .. . ......, --'1 lica cati.."ti.." . .... ~. I.c.oM od(·t.....u.lUlIC'I" r,urn the: ....;.k, .:.da..it:
mmmuniry." .5« aOO c"bnOn (' m ).
1 Pramiil)a T heory:
D harmakirti 's Co nceptual Context
I
F WE A1E TO ENGAGE wi th Dharmakini'l philosophy in;ll manner that
enables us to think through his style of reasoning, then we must learn
to speak Dha.mu.lOni's language thai is, we must b«omc skilled in the
discourse WI makes Dharmaldni's philosophical choices pou.iblc. Since
thai philosophicallangwgc is highly complex and prccisdy inflected. some
rcadcn may find it helpful to have a primer of sora. With those readers in
mind. I have provided in this chapler a basic overview of Dharmalcirti's
conceprual conte:n .' Tn do so. the du.prer cmphnittS some JigniflClnl
poinu of convergence among South Astan philosophers of Dharmakirti's
era who participated with him in a styie of discourse 'MIT call apramil)a
Theory.- ThUl, in a secondary sense, this chaprer will also alert readers to
some of my presupposirions. for any :lUcrnpt at a synoptic overview
inevitably ~ea.ls at lean some of iu author's assumptions.
1 R.c:adm. who Kdt • more cxtetW.., inuodunion lIl.lIy I1nd Jorwdon Ganon', ill
a.maJ lNii4 (wo. ) 10 tM, ap«i:ally bdpNl. A 11..., inlrodlKlOl}' work fOcwnI on 1M rdc-
PIn_I'"
'+'Iftl 8l1ddhi" phUotophalndilioru il raul Wil1ianu ' B...uJnn T1».pr (1000). for
........ >11:1 . _ _ hill • ...tHl'a"n• ..-noicW" ut ..........iuno oha.m! ...-., Pn.m~ 1bcvriau, KC
.j
16 fOUNDATIONS O F DHAlMAKIRTl 'S PHILOSO PHY
bngo:r ('996:164-)66)·
PRAMA~A THEORY, DHAIlMAKIRTI 'S CONCEPTUAL CONTEXT 17
orists do not simply note what had been thought in the past: rather, they
anmlpt ( 0 justify a particular imerprttarion by rC$ponding to the criticisms
of others, whether within or outside their own traditions. Each ~nm.tion
of philosophers thus represents a n~ layer of interpretation formed by
new criticisms and rebuttals. Already by Dharmakini's time, me debates
bctwttn various traditiON had gone bacIc and forth several times, and his
war\- "- thus thoroughly .. ng;n~ ;n rh~ cnn laT formed hy eI.rlier criri.
cisnu and his own an:empt to justify what he sees as the Buddhin view. One
up$hot of all this is that, in some ways, DharmalUni shares more with
thinkers from other traditions than he does with Buddhists such as Sthira-
mati or Candralcirti, who do not en~ in pramli,!" discourse.s
The general contoun of PraIJW:la Theory that delimit Dharmwrti's
own thought find their fint systemni( er:p~l;on in the NyliyllJiitTllI!> of
Gauwna (ca. I SO c.!!.).' Even at this early stage, a notable charactaistic of
Pra.mir)a Theory is the development of 1 technical vocabulary that aliialer
Pra.milp. Theorisu inherit and share, A cenlral theme in this vocabuhuy i.s
the usc of what J callihe ~ "rd.
system. ~ a formulaic way of analyzing the
"functional e1emenb~ or It.raltll5 that contribute to an action (1triy4J.' Fol-
lowing Gautama'r lead, Vitsyiyana (0. .. n ), the eul.iest oeomment:;ltor on
the NJ'liytUiitrta, applies the /u1,iJk4 system to the verb P'IIma, " 10 know
indubitably, ~' Of the possible Itir.us or elements in an action, thltt are
panicularly relevant (0 the analysis of the act oflmowing: the agenl (Itim!)
who acts on :ut obj«t or ~pati~nt~ (ltllfflllln) by means of:ut instrUment
(U'II'!"). Adding to these three: the action (mp) iudf. Vitsyiyana and all
mind that all cognitions occurring with those emotions are necessarily dis-
toned. This way of apprmching cognitive distonion--4fld numetOld othtr
ruch wues dearly indicates [hat an accoUnt of the cognitive cyent or act
called ~knowlcdgt" (prllmiti or pwni) is ooncemed largely with the proa:ss
of producing that cyent. And the mood that we have cited- involving the
imeraaion of agmr, object, and instrument- provides the ~15tructure
fOr Pramil}a ThMriu..' :anal"..is nf that pmcr:n;. 11
When Gaur:ama, Vitsyi}'Ula, and subsequent Pramil)a TheoriSIl used
this modd to give an accoum of knowiedge-f:Venu, their works address
especi:illy me "rll~ or "insu'UlTICDu ofknowkdge,· and it is for this rea-
son that Marilal and omen!'der to this genre of philosophicalliu:rarure as
Pr:uniI)a Theory. Bur why tlike an analysis of the instrument ali one', the-
muv: foc1ls 1 Why nor foell.( ;n..,e:ad on the agent. obj~ , o~ evenr itJdf?!J
To answer such quenions in a somewn:ar speculative manner, we might
give a historical argument mat borrows a principle of Prami .c:ta Theory
iudf: if two persons arc to have an argument. they must fin:t share many
poinu of agreement, TIut is, if any TWO discussanu arc to disagrtt mono
ingfully on some poim. their discussion must be framed within some uo
of ag~ment . " When d i$CUiling rhe acquiJirOon of indisput:oble knowl-
edge, Pr.unil)a Theorisu geMn.lly agree o n many basic noriofU about the
instrumenu of knowledge (prllmJi!'llJ. wherC2S they gc-nerally enoounter
fewer UC2I of agrttmcnr on other aspca.s of that proct:SS. Since they (end
(0 agree more readily on issues related to the instrument or means in the
II Mari1al h !1i6:los) wccincly poinu 10 lk proca.s in qual;on as aUA.!: "In the tam
o o
~ the notion of ·a llJC and °lxalUC ~ inlO OM. "
12 For lhole already funilw .nIh p~ n-y. om m.ly ~mply uk: Why doa
r~,lh • ."..tId Iilu: a ~ moniker for mit atyk of discount. ....... UJ ,,--ri-
tiJ,btr, and 1".,,"~,.:IUnd ridicuIow!
IJ On I p~ 'Th.roty 3C"«IWII, Ont CItI only lfI\H" :aboul 1M Inllh of. p"""",rion
{yttdjUJ If om bepns by accq>tinz (II lax provisiotWIy) W Cl:Uimu of dill piopo5i·
tion '• ....bjm " . . , tIM",..·,,).
FOUNDATIONS OF DHARMAKIRTI 'S PHILOSOPUY
15 Sec N8h (4)0-4.4 j) and NY (4)0-4.4d .. ms 1.1.11- 16 .nd CIp«ialI)' NY (16-10). Vic»-
patimim'scommcna (NVIT: I6-IO)'~ UJdi.oI Jo,m" a1d.ougl dwycomc mud. 1011:1" in rhc:
hiscoric:allkvdoj:llllft1l olNrtra- A nujor COl"""" of thrsc pM 8' is dw conranulity of
the u....t. and rhc: ddinidon of :an inatntrt>ml (u"ruJ as d..- " _ prominent cauAI &c.-
un" (...o,./u~
16 1M arl"mml for rhc: primaqo of INN!", th~t I ht"" summMiud ho:K if !'rom th.:
PIlAMA~A THEORY: OHARMAICI I.T1 'S CONCEPTUAL CON T EXT
17 Indeed. IIx Nlliyiyilw. ar InA. aplicidydia:wa thr noOOn Uuol dim: ~ am.in philo-
me.
tOpIUaI principls dial an: shared by philolophcrs: i"of Ihtm• ..u of principLes an: aspecu
of p....w:,... n-ry (_ NSu.1l wilh NY ud NBh If4 rit. (l6))).
18 Thr:daim Uuo. COITCCI knt:noicdp: iI indispcnAbk for !he awnmcnl ol llbttation iI ma<k
by a number of authors, indudinS Pnhsupkb cPOS:!). Cauuma (NSI.I .•). Vluyt~
and UddyooIan (NShand NY;~ ..Ji¢). Dlwmakini (PVu7)-l741. and 100ft.
I~ S« F",nco ('917) for C)n~ of me, fey, In-&plh WOfU on Ihl' form or Sou,h ~an
philwophy.
PRAM.J.~A THEORY: DHARMAKIRTI' S CONCEPTUAL CONTE.XT 1)
II A1tbou&h Kwnirib ill Sv I.!~ 169) mainwru Wi dxrc an: only five ItIlK organa.
JIu (" .1.:.'"'41) nota Ihll in !.smuli,iIt41lw mind iI lila poxilcd ... I IenIC. Gal,IWI\I
(NS:I.I.,. ) mo spoke of ooIy Ii." ICtUa, bu. UddrouJan (NYuJ MiNS:u .... ) and Mlbtc-
quml Naiyiyiku aeeqxed me mind ... IItIlK (1ft Di~'1 cnficiun oflhil illoCOlUiSletlC)'
ill PS:.,&I:I94 and 195: Hattori 196I:}l-J9).
2J 5«, for cumple: YO (1:11°, 161; lhu). NV <9• ...,.,. "" N$:I.I .• I. and Sv liNt.
J'I'Ik;Ioc}ll n lb:and l I1.a1-1!J)· ~rti docs I'lOl offn any al.....vr c::ommmu on Itw
Ihco.y of KnK faculty conlXt (iMiri,.,.,."iu".), bu.... ir cvidm. in other coo.au (i.e..
fOU NDATION S O F DIIARMAKI RT I'S PH ILOSO PHY
(4f11.".). They also agree: th:U physica.l (i.e.• m:attrial or fubstanti:a1) defecu
in th~ ",nse: facu lties can contribute to c~rt:ain types of errors in pt:rttptual
awa~nw. as when a person with ataracu appauntly sees small hairs or
bugs in front of mdr eyes." Another important point of 2grttment is mat
perttprual awarcnw is either the most vivid or thc least mediated form of
awarenw. and that in this x nsc it wes precedence over oth~r instru-
ments of knowing. luch as inferencc.11 Most of these philosophcrs also
agree Wt me basic building blocks of matter ar~ itreducibl~, pmkss aronu
or "infinitesim:al panicles" (plmIWll~II). According to the philosophers
who accept this notton, infinitesimal panicles:m t OO small to bc pt:rttiYed
by ordinary pt:1'$01U; in5lcad. the maner perceived by ordinary pt:noN
consists of particles that ha~ somchow bccn aggregated into:an entity of
perttptible size.·
Although thc:sc points of agrccmcnt :m ccrt:ainly significant. it is impor-
tant to noc~ that Prami~ Thcorisu often disagree upon m~ prccisc con-
tcnt of perceptual awarcness. ~imcr because mdr ontologies conRia, or
because they differ ov~r thc d~ to whKh perceptual awareflCSS is deter-
r··...,. ,~). loio Il ...... T.,( ""'_I""~.lo ""welT ~ ........ v...t-..Jl ....·......... ..Iuo'li ..,11•
po.,_ ... ""
..... (qui.., cig>ifica,,,) rnodi6a ....... ~ (_
PS,). lf _........., WI. in rho:
awe.:u wbm, DhannaIdni admiu D.femal JC!\R obj«u. M IOIIowt Vuubo!.ndhu·1 wofk
",t.c,( ... il " nor $UprncdnI by D~ hi< tMory of _ orpn QOtIIK1 would bo: simi.
br to the one f'wnd in V.ubandhu·. A~~ (AK:J.)O and AKNo M tir.).
24 Sc... PV (p,J). Sv (~SH4). YO (l1:P.7). and NY (114 MlNS:I.I .41.
25 For Viuyiyana and UddyoaIwa. pctc<ptaullWU'mCSI is what finally puu all doub! 10
1ft( and dimilWl:I an)' lUnha "dcJift 10 know" (fijtUslJthal objea (NBh and NV:'1-9J MI
}.IS:......). Fu. DIwr"..Io..Eni. unlr po:n:qot .... ~"-""'1OO" -.;.;u" (~_d"""ft .. jO .....
n.,.). in t(lnlfUC to infnmcc and other OCNICCplual copoilions. l'hU: iIMK bc:cumc. paniao-
luIy Alieni lOr DtwmWrti in his diK'IMiM of,.ope pclutpUoo (PV,.zll- 1I7).
Kwnirib doa nof mdonc an)' lIOlion ol..mdne.. pctMps in tuppon ofhU ftj«Uon of
~ pc'ccpliQn (SV. "'~16-17) • ..+.ich ~d ~ SUppiUlI the VcdM u a
InC2N ofknowi",.o--.. H~ doI:I maintain. ho..( ... . thaI Qt/xr i/Quummu of~
(such ill in(nma) an MutIA.riIy Ph:a ded br ptl utpt,w lwarmat (SV. """"""',,-n). 1
ID: IhiI notion of j>i(Cdc...... whido if; Wcm kwpnlfll by aD ~ 'J'h..oNu{Mohan()'
1 "1:1)1-~J. u an epUwnlC panJld 10 mot'I: ~IC conom\I with VlVMinea.
28 Many South AAan phiL. ,Nn 0II'tft awa.ft that if 01>1' a;ouId lIOI ~ an adequau- aa:owu
of~ tbt arWrur.m1 of opirirual fi«dom (~moaplKil pi of nearly aU ~
South AIiln pftiIoIopbm of this period, would bt impouibk. To • pnl C'%ttD1. tbt c:rucial
rolf of~ in tbt analnmenc oflibaetion raG on ia \lie u. tool dw allows one to 1Up-
pU.n1 faIK bdir& ( ";~"""" m.I]tI. ctt.) wid. indubitabk knowkdsc.
16 FO UNDATIONS OF OIiARMAKIll.TI 'S PHILOSO PHY
The former is simply an infCmcial cognition: one looks at a smoky room, for
C'Umple, and (with other conditions in plaao). one infm that fire is present.
In COntrast, an infermcr-for-othm is one that is stared verbally 50 as to indue:
an inferential cognition in another penon, In other words, this laner ~infa
encc" (which is aauaI.Iy a series of stalCfTlcrUS and not an inferma:) is meant
to result in another penon having his own infcrenct-for-oncsdf with regard
10 the question at hanci. In this sense, infettna-for-oncsdflies al the core of
these: thinkers' inferential throry. But ironically, the I[rucrural dements tholt
are neccssary for one to have an infertfltt-for-onesdf are primarily aplorod
in di5Cll.S5ions ofinmncr-for-others, To avoid w confwion that this over-
lap incurs, bdow I will often speak simply of "inkrena:: with me under-
standing that OUf main focus is the enminarion of the conditions no "pry
for a correct (as opposed to a spurious) inferential cognition to oc:cur.
S is P because E
The hill (S) is a Iocw offire (P) because of the presence ofsmoke
(E).-
J611w: un .....ldr tranJiuiO<\. " pot;iti~ ~ ilo.ncc" r.nd "MpIM ammmiWK:c" ha""
become lruldard (or .....H]olII· .,..'''l and .,..rimi.tl.-.,..,,,l.
[)upit .. I~ ' irw:kpn«. I
haw ct.o.cn ,0 .. mploy m..... ,nn,u,tionl hm:: 1'0 as to.ovid dw IIftn«aAr)' confwion of
introducill,_ talnS.. For ••...,.. Oberiwnmer n aI. (1"1:61) ~lMIId "Gcmcin·
......... Vorlwmmc:n [von Grund lind FoIlCl." 00, il is IlOI II all dnr how thiltum--'d
~ diMinpilhai from
.0 trW ••
I." ..'. .(co-oa:uncna). Althoupl ......,.. docs it>dcN amount
""'P"-'-' (..JuMw-) in i .. "tI .... .... in do. ...... ,.... of inr.r-c,. (d'. Ober-
hammer. eI al . ,,,,,61), this intcrp~rion or .......,. iI applic3bk 10 .da<ivdy kw IQII.
~ il is Rjcc!nI by dw: Pnmf.J:Ir. l'hcorilUofDhannakirti·. lime. orr'lCft bcfoft (Itt below.
ft.)I). My own prcktcna: (or ••....",. whm unckmood to mun .....NJII-.,..".. would be
" (nWlmau." Thilla'm apturct both dw: mtI:ophorial ~ ("folIowin, &\oa,") and .1x
JosiaI JIm'" (• •ict or nrcc:.a'1 impliationl of tlx Ia'm as it ... UJnt by Prami.qa 'The-
orim of Dnumakini'l ,ime Ind mn. For ".ri"br (whm UKd ift Ihe KlIK of
.".,ri.. ~,tiJ. I -.Jd. rccomnwnd "l'C$Iri(tion; ';nee dx ill.ltfltion b= ilto 5bow tNl
" " ' .. """ ... 0 ur <100: ",cd"""e ,",c <>C<.QO;Uil,. n::... i<:.nI ' u ","w,a..:a uf <101: noidaouo;. O ne
of u.... pl'tlblnn, with Inllubtiont: lNt involve tlx EnKlilh word " 1qI'1~~ (u ift "ncp-
PRAMAf:lA THEORY, DHARMAKIRTI 'S CONCEPTUAL CONTEXT :'9
concomiunc.e IperiAo that the evidence (E) is present only in the pres-
m tt of the predia.le (P) and not in any other circumstances. Dharmakin i
often States the nqarive concomirana:. or "restriction; in an affirm:uive
statemmt (i.e.• a statement that involva no grammatial nqacion). In our
cwnple. me positive statement would read. "There is smoke fUJI] where
there is fire."" Most Pramil).a Theorists. however. formul ate the ncptive
cn ncom it:...a (I f re<tn <;t i(l n in n"V,iV<': 1000fm~ : fo llowing our 6:11 mp10!:. :II
negative statcmem of this concomitanc.e would read, "wherever there is no
Arc, there is necc:ssarily no smoke." For Dharmakirti and Kumaril:a-and
probably also for Uddyotakara and PraWtapida-the positive and negative
concomitance: arc in contraposition: if smoke: is n«C$suily presem when
fire is present, thm in the absence of fice. smoke: is abo necessarily absent.-
A.o:o.-d.ing 10 th.,... philosophcn, in ortkr 10 h:t_ :an inn:ll n ~ of inf~f'
m rial knowlodgc: one musl be aWl« of the pervasion-the rwofold rc:I:IItion
consisring of the positive and neg,uivC' concomitance. So too. the pervasion
must be general: it cannot be restricted tO:ll singlC' case. but must pertain to
tift conc.omiWlU") II dw IIJWtirr•• il noc nec:euarily Ailed» I Mplion. Sft. fo, earn·
pI~ Oh.,.... lrl"i '. fn'mlll~.;"n n( "7 ",;,.,tll in PVSV .J PVI.I (Co ,s..)) , "J"i~ 1M
"'",UN ihl"",~(" H 8:1' ,]-I),
)7 Stt m.e pt'c.ioua notc for roo« on OhanrWiini', pOliti'" formulalion of Inc n'Sl'M
c:onoomiW'>C.(.
:J8 1 ha", choKn dK ~ Imll "n«.CSAriIy" and in tda~ forms 10 aKI'I'C)' fWO rypa of
Sanskri, conRnKt~ lhotc thai nnploy ~ rc:Rricti", ~i<:k tw, and tho.c: thaI nnploy
an ac!-.nb such as iIttwM", (e". Sv, ~1f""'-n.4b) Of """""", (c.". PVSV PVI.lI:
G:l9,'). 1M usap;t' at "'" rjwc: "onIy1 in me fotmulation of ,ho: mdcnoe-predicalC rd.-o.
*'
'I,Oft • do: I
" an•Im~"~"P"'<n. .on P~ ~-- . 0'nann .......
. ' ...... T 1ft , ...I """"
' . ...
' . , ..
' en ___
- ~ '- -
Ibctc pbibopbcn '0 uno,kqw,d ,hal relation as • nco;aJaI}' tdation, ",rbft lhan • ~
c:opr_nc:c (*Mno.. etc.) , In hil dir,"ina of the hiKoric:al II"I.IUition from throtiet tN,
potitl mct'ecopt'-.u of c.idcnoeand pmlian: 10 rhoK'NI poIil a ,'" "" '')' n:lacion, Poe_
Iei' (1977-"'- ' 94) has;upN thai pbilolophcn ouch ill Uddyoabn. and PnbmpL.b rq>-
raenl an inlc:nncdulC", btt.w, me rdation »wpiUt,U and ~ rebtion » ncccauy,
While i. is uuc lhi, Pnh'laptda', work (pru ' 147-~. UI) ahibia ooIYI modeM IIIcmpc
10 $!'lOW beyond tIlCft wpmmoo, UddyoaIwa. in tw aiDqlK ofDi&niP (NV:l6)-167 .J
N$ ••IJ; ci. H.ya ,,&0,,..,11) &nd eI, .. Iw;..... "PI'"'" 10 ..ndoc........t p<Ni'' "'
And rocy,i...,
oonmmiWltt ill mnuapoliOonal. 1M implication here il tN. the evidence-predicate rda-
lion is NCn"')'oo hi:!: tboory. UddyoaIan.', rcc:op.ilion ofdK~,c ldation
ill nrusary ill Wo augmed by the faa that. e¥m in cue when; only the MpM con-
oomiWlOe 0' rauiaion can be acmplihed. dK cvidenc:c-Pftd.ica.c .mtion nill c:onai...
both (NV'I .... - LU).
Ohartn:akirti', RCOS1Iilion of ~ ~pred>ca,c rdltion as rw......my io abuncbndy
dar, and his diat:inaiw contriburion lie in m.efotmubtion of. 1N~"UhII (5«
SfCinkdlna' '971:101-104 and &110 below, chaplcr J). Dharmaklni nnpIoytan no:n ~ ~
cite UK ai tWI (e" . in hil initial praocnano. of.,qti in PVSV ..tpy,.I; G',-I1-IJ). His
,0 FOUNDATION S OF OHAIlMAKIRTI 'S PHILO SOPHY
pofoiIion haJ _ aimibrities with KwnitW.'l, who« detaiprioru of ~ jQiln.e and rxp-
Ii", OOIKOIliiW>Ol' ind,,* ~ 1OUowin&:
If ~ pRKOOf of ~ wac: pnv:oda! by ~ prama of fin::, ,hm non-fi~ bet..
ndudtd fivcn f.II:>OU, would be ~I only in ~ c::uc: of ncH\·unokc. Thw.. ill ,his
--1' 11.. , .. ioU, io "",.-...Jed t..,. .""............... Lilo.c..iK, ;..... ",..;1...
II""-r.,~ I-,u,....
non-fin:: is ~ by I'IOn~. IIrIOkc: g. adlldN from non-fin::: at auch. i, is
N ( p';ly rlbno-.v
~ by fin:: bc:a.- " Iw no pouibillry of aisrinl U11CN1W
odw:r non-f,", kKus.1""'"-iIMw pi.l:Wo._ ~/'k iupir III"'1"~ 11.Jhii_ftIf
"*""J ....~ ~""- III,.." III~WJ"'"'
-J<6,..._...JtM.tNJ ",."." JInw ..",
",0,;;_ "'/'k ~ III ... '1"~ 11
~i_ Sv. . ...."u,U,I1w1-I141bj.
J9 WhelM GOt 01' maR a::unpia nmiloo be- ("ed in an '1Ikn:nu-IOr-odont. r,.n',
rIM
_~ NI_ ...t. in PO" .. ,....,. ...... 'YI'" ~ ... ;.t~ "',•. ho-i"'.ddt ......nd in PO" "fW'" m., vV-wot
of tilt- pniloJopha in qucaion.lJddyoW;ara (NV:!+4-'"'s), KUlIIirila (S\!, ..... .........,,,'j,
and
DharmaidRi (1'V,.J6) all n::cosniK INoI in _ infcn:nco:t only me, poIiln.e W"CO",iOOOf
an be- OftIlplihed bta. . lhoe domain 01 ho;,tlOS"_ CIHI iI nnP')'O !lUI iI-. thn-e an:: 1'10
insana:r. of mClies thai do I\OC poACIf tho! propnt)' 10 be prowm. Such cua an:: prim.ily
thoK in wbKh cURal« is impoMibk widlOullbe pmiig.IC in q..-ion. For a::unpIc, (Or.
pboo.opha- who !!Wnw", We all raJ tIIinp an:: fM'O'SRrily impm!W'lCnl. an inf=nc:c in
which impmnanmct' is. pmticaw< haJ no M~ ~ ilrC" no aiKml, ptt-
m&roal ..... n5'-
In...do caoca, j, .. . - ... 7'0 p<CKf". C01oIn'en::><ampic. F_ Dh.,..,.·
kini Ind Kumi.;b. lbe supnfluif1 of ~ coun'~mpk JWnf fivcn lhoe ~Inr-i,ift
PIlAUAI:<A T HEORY, OHARMAKIRTI' S CON CEPTU AL CONTEXT }I
nuu~ of positive and ncpUve concomiWI('.C (,,~IW)'" and ",.nrrhj. Uddyoulun ""'yalll)
oIurc ..... "'- (NV"+t-'.s). A1thouf!;h Udd,..,.:obn·• .-m-;.. en.""'"
no< dar. ;' _ _
likdy thaI for him poIicift concomilVlClt is not I malin of men: copl'CKna, !.ina a theory
thai pmniu -'I·formed infnmca when only lbe posilive coneomiuna is aempiific.d
......w.J n(K be pMXft4fu1 if thaI relation _ DOl ,........,...'Y.
UddyotKan (NV"4<I- 141) ..wnta.iru thaI in JOIDe infermca only dw n<pIM
concomiuna ("",Ii.mvlCln be eumplifitd and thl- poeitM _ i l i n a (-.....,..JQlllDOI
be.. Heft, me prob&tm is Ihat lhe domain of hotmIop.u il\lGflCG (..,."",) ~ atlp!1. How-
rva , ullliu CIIa wile", me ~~ domain;' ~mpty, the alJIcn.:~ ofbomuLasow
cumpla doc. DOl .... ve to do with an incomp..tiloility btl . c... dlC ~ 0( d.. pmlicalC
in quenion and th.e ............. of ..... ..,b;ec.. I....-t, in ,hae ~ aU poooi&k i....,."",.,. ....
indudc..t in die: wbjcn IlII<kr diJpuOtion.. In .mn. then: H~ no noncOIItl'O'f'alial c:ua in
which die: potilive «N\COlIIiunu could be dllfl(>l\$lt'lled,. FOt difkrHlI ttUOfU, both
Ko,unlrw. (SV, ....,,""'-I)I- In) and DlwnWtini (MY ""PVI.I)Ib. G").I- II) rqea this
type of atp;umnll.
In .adi.ion to allowins I .... ' tilt counrnaatnpk is not ,M 'Y in aU cuoa, Dh.unWtini
(pVl.l7-tJ, PVSV ..., ~ PI as &r as to Ay that 011< 1'211 diJpmx with aampks alto-
~ if lbe pernsion (~,D) iII"'ready &milg, ,0 lhc in,c.!ocu.on: in...do cuoa, .... JIC"
...-......... doa.- .to<d '''' be ClIplicidr _cd. P... funh.c. __ oua ..... .....Jpu .... "-...d
00wr rda.«! ......... _ TIlInnanr {I<nO) .
.a Mohamy (1911 and 1991:101- 111) rmw'u ~I lmgth on ~ in Soulh Amn the-
oriaofin~.
41 The claim!lw the!t philotopMn ·a", not i.,em,t<:!" in fOl'mal rnsooinl ""',. KCTIl a
bit n trem., bu, siven the abidin, conccrru wilh praclical application (1,,,"!'1i) Ind
pUfpoK r,r~"") in PramlQ-l Theory, dcvdopilll a J)'Mcm of (<<nW reuonin, would
prob~l,. .ppal poinllCSl, inwnuch II I"orrm! '1'1C1TlI dclitlCntd,. divola IhemlClva
fivm thoK c:on«tnJ.
FOUNDATIONS OF OHAkMAKIRTI 'S PHI LOSOP HY
"necessity of cnmples in an infere:nce-for-othcrs. According to many South
Asian philosophers, me (W1)fold reladon betw«n evidence and predicate
ClnnOt be $t2ted in :a.bstncrion from the substances that bear ~ predi-
cates. When a dispuram (In us all him - Devadaml attempts to indUtt
another to infer the prew:nu offire: on a mountain from me
smoke on that
mo untai n, Dcvadana must dcmonSU;Ji U," 10 his interlocutor that the pres-
ence: of smol«: is neassarily concomitant with the presence of firc. BUI M
cannot do 50 by appealing to the case at hand----the smoke and fire on the
mountain-preciKly bcc:ausc rhis case is under dispute. Of coune, Dcva-
data migtl! simply stal e' m;u rdadon in abnracrion from any given locus
or substance. bur many PraJTlil.u "Throrists, especially those from non-Bud-
dhist traditions. resist this approach. This is duc in part (0 the nOlton that.
if the p«d.iCI.[CS in quesrion are tttI, they must Ix insWlriucd in some sub-
stance or locus; and if one CUlnOI appeal to any such undispUlcd inst2mi·
arion, mm dK rWity or rhost ptedicar(S mnains dubious. Hence, fu r some
Pram1r)a ThroriSt5, on~ or the reasoru for insisting upon examples is that
they savt [0 demonstrate the reality or the emines adduced as predK::ate and
n-idena:. This ontological mJuiremem also has a ee:rrain resonana: with an
epinemic rcquiremem-namdy, th:u the relation in quc:stion must have iu
finaJ appal in sense pe.ctption itself. In this smsc, ev~n ir one can logicaUy
adduoe rQSOns why some stat~ or affiairt must hold t~. one's argumenu
arc generally considered unreliable if one cannot appeal to sensory opcti·
enee: to suppon that rc:asoning.·l
A RuTATEMENT
With the above discussKm in mind, In us resate the basic demena ofinftt·
~ aa:ording to PrarrW:ta 1bcorisu. This resamncnt combi.nci tht dementi
of both infen:nce-for--o n_lf and infercnce-lOr·othcn, and it .. malnt ....
heuristic 0VttVit'W of inkrmcc, r.athcr than the depiction of any philoso-
pher', theory:
45 Conum;nl ...... ieh clemenu m.... be: aplieidy n:ltw. Uddyo.ah.., and his fdlow
Noiyiyikao stand ac 01>( o:nd of dv: ~, ....t.lk Dtwm:oJdni and ..... foIknotn take ~ dia-
mctriaIly oppoKCl ¥ioew. Aaotdilll.o .be sundard NyI,....-kw- dc;fwdcd by Uddyotabra.
dv: poopwirion mid< Ix _mI DOl only ac mr~nnin" bu. ;. m.... Wo Ix tqlt:lml l ' ohm
mil ... conc:huion Of -1Um1N00n" (NpMlUl4). Hma. foo. Naiy.iyiba, .I\ill.~ in/ff.
ma-Ior-odlm Iw fiYe dcrnmu Of "limbo" /.,;p).
I. The mountain q.1ocw offiR (the propoai. ion: ,~lijU,
1. Beaux ;. iII.1ocw of srnoU (the ~ J.m.)
J.. WhcRO(IIben: il lInlOIu:. then: ill fin:. »in ~ banh; without fin:, then: is no omoIu:,
.. in • t..b: (~ion and annplificolion: "JtI.......,....)
,...,·4,
.... The rroounwn is a Iocw of .mou (!tN: ,pplic uion: "I"''"''Y4 or IUtenlm' of
I. Tha&'OfC,
.toO
me, mountain it ~ Iocw of fin: (oondwion Of swnnwion: "ip_,..,.
PRAMAIiA T HEORY: OHARMAKIRTn CONCEPTUAL CONTEXT J5
W e haW! now owerM rhi!' mou Io:Ilii!'nl vill'WS IMt Pr:a.mil)a Theorisu
share aboul the two ubiquitow forms of instrumenu of knowledge: per-
ccpt:ion (prtl~)and inference 6"'II*nIl). Let us now rum to lO me basK:
views concerning the instrumental objcct (pramry.t), the objca of an
insuumenc of knowfedgt:.
Al nOled pln'iously, the term prafflrya refers 10 the object of the indu-
bitable knowledge derived from an inst rument of knowledge or pf'llmi!f4,
and to clarify that a pra",lJ" is specilically an object of this kind ofknowl-
edge. I will pmly manllali!' ,ra",~:as ~inJt rumen~ objecr ...•
For P~a Thcoruu, an inn rwncntal object is ncccssarily what we
might call "rt"a!. in English. I am thinking here especially of the Sambit
tCfm SIlt, a participle fonncd from tht verb ~ tO bdaiu· (111). The: connoa-
In irs fuJIac forni , this ~ indudel both 1M poei~ and neprivc conc.omiunoe in the
aanplifialion, bul as _ iu.,y 1«11 , Uddyoubra:Wo:&dm.ia inrerenas thai invohoe only
poIJlM concomitance (Le., I '-'IboMJfI'l) or on.,. nqaUw conoomiu.na: (i.e., a '-''''"'r
.ri,mflj. For additional mru.rb and ¥ieM on the Nliylyib approach. I « ....,....;.lI y Mati.
bI b~:n-7I). Pomr (1977:110-111). and Mohanry (1991:101-106).
In COrIttUllU Uddyoubra. Dhumakini, on his m(-.f rnarun..,ft. prensl panimoniaw
approKh. This panimony Menu &om his rommlion (iMerilcd from DignJp) tNl an illfn-
a-ott thouId OftIy contain the "mcana" (IIIJ}",M) for ..,-..:ruil'll:an inkrmtial WF'irion in the
-41 ~ ( ,~)6i) has DOted: "As · krIowabk· li~,,.. .,.. OI'.J'tIo;:IooI .. ",;dUn_
d-ococia of,,-.h:w- cquivalmt IG 'e:aismu,' ~t and dnmnin.Jtion cqlUlly Im't II
airma fix o:ziormor.." In hi. ..... on dUs I'nIWk ('996:n.I-4), he c:ommcntI !hal • on this MJb.
.ita, ~crisu I YUf lilmlNft from both Indian and WaKm IUthon .. ..· 'J1w; poinr Mre ;'
dur if 'p""'(IIIOOXUII, if '. ,urily hM I rollhint; II ia objttt. Hmor,lIIl)'thln& known by
a ~ iI _ lily raJ.. l'hi5 is indc-td a ubtq.u1Olal dWn IlIlOClK PYun1qa lhtorUu; in
Dtwmakirti', tIIW, ic;' rd'.a:tu:! in Ihcdaim. "To aile illO be paca+<d" { _ .,.1,6".
I,;~ .... PVSV ..,/PVI •.J>. For _ on Dtwmalcirti·, ..w:w, in thi. rqard. _dupm 1 (t.m.
481 rekrMre fO thcnocton oliUusionorth. probkmofoloo_~ (u.r.;,.ti,
.y.mtlljU,.., " • .,tM'M, Oc.). N.5. Drmd (t~17) noIet tNl "a11lndim ~ilomph
kaI JdIoob••• , fO;,.. all thrir doarinal diffcrmcc:t, IK in JClUiliOIt on one pOUII, namely qN-
1CII'OkIp'a" n:aliIm. llw obja:r of cop>itioa (an " " ' " be an unrtal m ilt)'. - To Aly thlr thit
it tturol"a11 ~ achooI." is an owasaremellL For ownpk. on CandtUinj', nrw,
lhe WIlI",1 of an ordirwy penon', JIftttPlion il Mo...,. an unrnl ",tit)' (M.J},.
JW.-.uow"""~'O?-loI" M~_w.."""I5.lJ; cf. rdlcmalll l!J9Ol'4S.... 9). Nev.
~ OnYid', basic point orrt:a1nly applies 10 p~ Thtorisa: C'Im in _ of ntOr,
...... connor cb.im ,har ~ occun..;thoo.il an objea. For,....dy of _ problems
!har an inlmrional theory of ~ (!Wet for Buddhiat pbilotopMn. I « Griffi lh.
h9l6l.
49 See Pon~ h?61 '9) for _ on thit iMtJe..
PRIIMAf:lA THEORY, DIiARMAJ(IIlTI 'S CONCE PTUAL CONTEXT }7
refer 10 a single obj«t. These argumena rest on the daim dul the gramnur
of a:prasions corresponds 10 the rca1 propmies of objcca. More sp«i6ca1ly.
!he singulariry of an exprc:uion corresponds ( 0 W singulariry (dAt.l, mttJJa)
of me object ( 0 which il is applied. In shon , in Ihis ~rd allan, grammar
and Ontology stand in a relation ci isomcr:ric correspondence.»
When combined wilh the nolion dut the real is simple or singular. Ihis
alleged isometric correspondence: })(:fWttn gr.tmmatical and ontological
number leads to Ihe second type of argum{'nt as an imponant corollary,
namdy, that a grammatically plural a prcssion musl in fact rder to mulli-
pie obj{'Ct$ that ar{' ontologically singular or simple. Thus. ifit is meaning-
ful , a grammatically plural expression or concept must correspond
ontologically to numerous, ontologically simple entilies.
This condngcncy of grammarical pluraliry on ontological singularity
poina [ 0 thc third set of argumena in F.tvor of the reaJ as simple. These
arguments rest on lhe we of reduaM analysis (vibh.if., vicar.. etC.) and the
principle rhallhe real is irm:lucibk That is, when we apply the appropriate
form of analysis to a real entiry, we should nOI be able 10 break or anaIyu il
inlOsmaller partS. since a real enlity is simple. If that seemingly real entity
can succmfully be fu nher analyud--broken into para. as it were-then
its simpliciry is only apparent; it K'ttIU 10 be simpl{', but in faa il is com-
pia. and as such. it is not truly real. In this way. onl~ simplicity cor-
responds [0 analytical irm:lucibiJiry. Thus, if any real thing is n«emrily
umple or unitary, il is also neccuarily irm:lucible under reductive anaIysis."
Many Pramif.\a Thcorisu usc (or at least allude to) all mree fonns of vgu-
tnc:n l to c:scablish a real thing as simple. but throughout these arguments. d~
notion thaI the real mus! be simple remains uncontened for Dtwnukirci
and his principle opponents. Their un:mimiry on the issue of simpliciry,
hownn. lcads them to a shared problem, which we can illuslfa[e in [eons
of me aJlcgcd whole du[ is a water-jug. We may daim Ih.U, wbcn we sec: a
Wolfer-jug. we are I«ing II single thing. but we mwt also admit that we can
readily I« its paru---the base. me rim around me top. and so on-in the
same fashion. We thw encounter an apparent antinomy: the water-jug is a
tingle: real ming looud in a paniculartime and plllCC and consisring ofa cer-
rain IImount of mattc1". and ~ in milt very same time. place. and maner. we:
also sec (and can meaningfully speak of) multiple real things such as a base
:lnd rim. Thm. we mll.n Hk: are we sccin& one thing o r many thinp~
The possible rcsponso; :lie perhaps obviow: one can either choose to
defend the simplicity of things that presuppose the aisrence of real pans.
or OM can intist that the simple is ncccw.rily pardcss. The former position
is ch:aracteristic of those South Asian philosophers such as Uddyotakara
who StfCSS the pe.ctprual and linguistic approOKhes 10 simplicity: for these
rhinkenl , :lny :lOXIunr mun r~rve the onf"logieol inmir;.,nJ rhar sr.....
from the way we perceive and speak of things such as a water-jug. If a spa-
tially atendcd object such as a water-jug appnrs (0 be onc thing. and if we
can speak meaningfully of it in the singular. then our ontological account
of the Woller-jug must likewise show how II single. real. uniruy w,:uer-jug
docs not 1ose ilS simplicity even though that single entity is disoibuted ovo"
multiple paru that are themsel ves sim ple and 1e:U. With this issue in mind,
phllO$Opherr fum u Uddyot:akar:. ,pc21c of a real "whole" or "part-posses-
sor'" (Ilvayavin): a rttI substanct instantiated or participating in iu real parts.
and yet entirely distinct fTom thcm.
A theory of substantially ainem. unitary wholes that are distiner from
~it pans may satisfy somc intuitions about perception and language. but
even on the view ofiu proponcnu it leads to some difficulties. For narn-
pl~. given these (hinkenl' view of m:lrter. they must admit that a whol~
water-jug should weigh more than thc total weight of iu pam. That is.
before the twO halves of a water-jug are conjoined. th~ haVC' a certain
wcight. and whcn they arc conjoined. a new. additional substana--thc
W;lrcr·jug-comcs into !xing. Since thc conjoining of thc halves creates a
new substanct oVC'r and beyond the halves of the water-jug. one would
eo<pe<:t thu~ to be some :addition:&! W<l!ight from the p~nOl! of (1.:1.1 new
substance. Uddyotakara, in a ramer undistinguished. attempt to deal with
this problcm. dainu that a whole docs indttd weigh more than the total
weight of all iu parts hut that the difference in weight is undettttable."
s.c Uddyocabn nukts thioo:him ~f NV:.9J ("'NS~ I .H) in hil atemivediK'\Wion ofthc
..,.,..,'" (NV'4~I} ~ NS;1..' .JJ-J6). forDhwrWtin:.-, rdiuallOO of tIua YJcW. 1In'. for
cnmple. PV• . lfl-16,: 1ft Wo 1Ji&rUp., con:mmu in PSV (4;rt.-47)).
-40 FOUNDATIONS OF DHAIlMAKIIlTl 'S PHILOSOPHY
In contrast. South Asian Buddhist thinkers utterly ret«t the teal exis-
tence of wholes; indttd. a mereologic:a1 critique of wholes is one of the ear·
lieu and mOil paradigmatic forms of rrouctive analysis in BuddhiSt
thought. In their crilique of wholes. Buddhist thinkers mainwn thai enti-
ties such as water-jugs may Sttm 10 lK simple. but in fact they are not
because il is not possible for;ill real end£), [0 lK distributed over or panici-
pate in partS that are themselves simple. Many of the arguments that they
adduce for this critique fall into a genre that Tibetan thinkm later called
the " ndthe:r-one-noNnany~ argument. This nyic: of critique rdics on milK-
lUI tUi llJm.nJllm to demolUlrate Uu.t it is unu:nOlble to maintain that;ill whole
is identical to its real pans Of that a whok is dislinct hom its real pans. And
sinct any rm thing must be: either Ktentical to or distinct hom any other
mtl thing, if the pans are incked real. then one mUSt conclude: that the
whole is unreal." Hena'. on the view of Buddhist thinkers, only panicss
things can be simple. which is 10 Ay that simple things cannot be disuib-
uted over or instantiated in other simple things. And since they agrtt mal
only the simple can be real. they must insist that only the pardeu-the
undistributed-is real.
Ahhough they reject [he existence of real wholes. Buddhist thinkers
unde:rsruJd thai they mUSt also account for our perceptual and linguistk
practices. whereby ~ believe oursdves to be: pcrcriving and spealUng of
wholes such as water-jugs that are distributed over their parts. This leads
Buddhist philosophers to discuss two "iffirrnt typn of"Illity: an appuent
reality in which things can only Ix called "real- (or -rrue") in conventional,
contingent. or nominal terms (u'!'''.rtiut or PfI'jiillptwu). and a highest
level. of rea1ir:y in terms of which mings are ulrimaldy real (/M,..",.nhtu.I).
This fundamental flOtion of the ~twO realities- or ~ twO truths· occurs
throughoUt Buddhist texts, and the works of Dharmakini are flO acep-
cion. Within the Buddhist context that informed Dhannakini's thought.
the: mosl relevant sratc:mtnt of chest twO levels of reality occurs in the
Ahhitlhttrmttkoill (and bINi.tpt) ofvasubandhu:'"
55 For tht "ncidw:r_,*' lIWIy ~1 ' ln ~. 1IfC rdkrnant (I~l and 1~}). 1Up
.mn (JOO.:."-'IOo4l disa.laa _ o f tht ~ ~ of .bit")'k of ana/ysia.l. is
WORh nocintI m.u tht "1!X'ftI or IUdI arpmmu Pmllfl90Kl a IIOOOM&lU IOrm of1or;ic.
56 Thecommmt;llOf'1 Dewmdnbuddhi (q.. PVJ>o,b "'1'V",U-'I7. and • • .JPV).I94l
,....1 hJ..yJ ....1.IJ,i (...... t'Vr;.....t.) r.~lOCfl.IT";'" V....L-..J.h .. _ .............u: uf ....... ul
Otwmaklni', thKNia
PRAMA~A THEORY: OHAFlMAKIFlTI 'S CO N CEPTUAL CONTEXT 4'
That of which one docs not have a cognition when it l'las bttn
broken [into parts 6nl"JWw)j is conventionally real (Ul'!'lJ!1isat);
an example is a water-jug. And ilia, of whidt one does not have
a cognition when other [elemenw qualities (dlMmsa)j have bttn
excluded from it by the mind is also conventionally real; an
example is water. Th:l[ which is otherwise is ultimately real
(p'wlwui~t).'7
S7 AK6.4' ~"" un'flN 114 ".II.. ·.lJl.ir tI1I]"i,.N Jhryi (II lIf' I
~~ ~bnckl:tcdpbn$Ct · inlg puu" (II ...... ,,{.~) and "dnnmwq.w.
"",.Iffionw,
III,!,,,,,",,'
iti"j"'~ .
04.1 FOU NDAT IO NS O F DHAIlMAKIATI 'S PHILOSOPHY
rd:uion ~ to bt' ultim:udy rt:a.I, then it [00 must bt' a simple. unitary
entity. If a relation iJ hypomsiud in such a £uhion, Ihe mereological style
of analysis applies because the relation mwt now be conceived much :IS a
whole: a single thing that. while existent in iudf. is sommow dintibuted
over IU pans.
At various poinu in the SambanJh.1pttri~ Dh.armaIOrti relies on a Mod_
rh er-l)ne-not_mllny" ,..g,,~nr In m:lke hie rooinr. :Inti hie :lrgumenr mnves
b:ack and form across a cc:nrr:U question: if II rel:uion is II real thing. then is
it one wim iu relata. or is it different from tbtm? Noting that II relation pre-
supposes the pr~nce of al lusr twO rdatll. Dharmakini dismisses the
notion that ,he rebtion could be a real thing thllt is one with (i.e.• identi-
cal to) the rebD over which it is distribured. In orner words. if me relation
aM the rdllr:a ue """. then how can ~ intelligently fpe<tlt of tum n!'bta?""
And in response to me daim mat me rdation could be different from iu
rdaD. he offers a verse that is panicularly helpful for understanding
Dharmmni's ontology;
60 I:>twmUini makes rhiI UJIlffiCflt at variow; poinu in Ia.t. btu pcrNpi lht most obvi-
_ ; . in the t«Ond >aX: "You may think thu a..mUon is a tomminpins of na'-1,lf'f:L Bill
if the ..mea I~ twO, then how can that be'- (rii~ In _ ...... m_,. u ltI~
iIM_). )n hia commmwy. DharmakIrti notcr. "If the mac.a wen: to be one, then una:
~ would nor be: two n:bta. wlu.1 rd:nion would tbcrt bet w~ ask ttW; bea ...... relation
prnuppota [Ill leastl two fdac:.- bs6a: tdt,.O'" M:JII"l 'mI,. til" prJiI-J,.i fhlir
'mI"I""l ~,."", iii";".,.. 1M,..,., # rI}.
s....,• ,.",..,,..u,. """
61
",,u;,. liM_tIM t il M _"""""'_tis
.u.yn.~ 10..4'""9" M~"'~t
IItdM.
~f U!t ""'-
4<t FOUNDATION S OF DHARMAKIRTI 'S PHILO SOP HY
relau, or different from mem? The infinite regress from this point should
~obviow .
I have cited. Dharmakini's argument by infinite rqrm because it $0
dearly points (0 a theme within his ontology, namely, the rejection of me
notion that an entity could ~ at once one (and thw a simple real) and yet
panicip.m: in what is many. Such alleged. C'mities includC': a wholC' partici-
paring in its pans; a univenal participating in iu particulars; a perdurant
entity participating in its tempo~ insunces; and a rdation participating in
its rdata. Whatn'C'f motives ~ might attribute to him," it is dear [hu
Dharmakini uttC'riy rejecu any possibility of unity wimin plun.1ity, and as
a result. all such entities mwt ~ u1tinutdy unreal for him because they all
can ~ reduced (0 me entities over which they are allegedly distributed.
The argumem by infinite regress is also particularly belpfuJ for under-
standing Oharmaldni's omo\ogy in iu wider context. In pan, the argu-
mem is hdpful because it presupposes a fundamental area of agreement,
n:undy, that a real thing is simple or one. At me same time, howtVtt, the
argument by infinite regress also points to an especially crucial point of dis-
agreement. That point becomcs dear when W1;: rcc:ogni:te that the regress
suCCtt'ds only under a cenain condition. As Stephen Phillips n()(cs:
62 ~ likdy mQII;iv:l [ion;' Dhannakinl'l n«lIlO okknd the Buddhitt nocion of....~
(MJI1IUII or IMmtwf]tl). which is anna! 10 hi! _crioIopcaI project. Flp«ia!ly alta
N~ Ikoddhist ~tI apinst ~ Idf (m-"j ro.:u. DOl 0lI the: imp<l"ibility of •
....f ~ N , but "'ther the impoAibility of I whole cWJ of entities, dx Idf brin& within WI
cbsa. AI~ Dlwmakini hi!Nd( t&n no dar critique of the: d , it IftmI' likely WI he
100 would _ the td( as akin 10 l..-hok tWa of muties, rwndy, dIOK WI ~ d$ribuud.
For an iOlctprtUlion aionA!hac lines. 1ft Fm-ooo hood.
63 PhilliPI' h !r.n :I) .
PRAMAt::IA THEO RY, OHAitMAklRTI'S CO NCE PT UAL CON T EXT 4S
604 I un rdminc hrn: 111) tho: LUC of the: ramJ - . . . "';IM, ~ Uri... by variow NyJya-
Vli'qil.. thinkcn.. Ser: Poun (' 97'r.140-141) for an excdkm ~ of mil u.-. For
Ohannakini. thinp mUN be raJ (1M) "ill the ~ Wly" ill mal any lUI nlliry mUN _
the:_criurion of-1M dlicaq-" (~Manyofhi.-OI'f gftCIIU, !.ownu, aR' wiD-
ina; 10 apply di-..: uilaia: _ tb.inp aR' lUI in dUl: they an: dir«tly COIIlaClro by tho:
KIlXI, odlCI1 an: tal beaUlC, for Gampk. !hey an: dw objeco 01 ]in~ Of tonapfWl
mpWons. For ~ dnaiIed cfUr:waon of rd-ucd -.-. '" DunM (1999:b -70).
.. 6 FOUNDATIONS O F OHAIlMAKIRTI 'S PHILOSOPHY
ttnnal topic of his work. he larer defines it: "the purpose is the ll'f1htJ aim-
ing at which one actS . "6) In other words, it is with some purpose in mind
that one ittlu ro act in a manner informed by the indubitabk knowledge
that an instrument of knowledge provKies. In this selUe, purpose is a au-
cial conrat within which ruch knowledge occurs. Gaucuna'$ definicion,
ho~er, is 5OmeYIM.t difficult to understand, fo r it employs the ambigu-
ous term IInh4. whose many meaninltS include ",oaI," "thing: and
"object." This ambiguiry often awes confusion, but it abo allows one to
make :I. point: when :I. "thing" is being raken a! an "object: one docs 50
bc:cause that "thing" will serve SOIYlC "goal" We see this in the commencuy
offered by me earliest Naiyiyika commentator Viiayiyan:l.;
.-
,.", ",..", ~ 1M",.,.'!' "" .) "')or ~", .....tiP,IMui ~
rM .$",.,.... ,....""u.m.tMfII J i"..", .nJJn. .,.,...i 1MsJJ"'; Wli .".~
68 NY (m ) _NSI.I.1<f:
,...,;~ ~.oip.,.,.. ~~ ..."J"Ih-.7i HMJ..";11 -H'
", 1M "~Iip'" jli I ~ no~.,. ,,,,,.,;._ ...... "",. .yffluJ
iUllto*
~rtU/t.tr in.
.,s FOUNDATIONS OF DHARMA"IRTI'S PHILOSOPHY
Even a fool doc:5 not act without being directed toward a pur-
poK. If he we~ [0 act in that fashion , what would he nced his
intellect for?n
A£ wilh Naiyiyilw. Kumirila ties purpose with action. One acts so ali to
obtain a purpose. and the role of knowledge is 10 enable 011(: to determine
both the purpose and the means [ 0 Dixaining it. We find much the same
sentiment in Dharmaki"tti's ph ilosophy. but in his cue, purpose takes on a
distinctive role in the determination of what COnstiNteJ an ilUtrumcnt of
knowledge. That distinCtive role is indicated by his usc of the term
arth41tri]i.. Below. we will have an opporrunity 10 examine this lerm and its
meaning in Dharmwtti's philosophy, but here we can note that one ofia
meanings is simply the · accomplishment" (Irriyd) of a "goal" (arth.), or
what I call · tdie function ." Of course. for Dharmaldni to speak in these:
[erRU is nothing new. The commenu cited above dearly suggest that
Uddyotalcm. also $2W efficacy ali a crucial component of knowlcdgc.." In
69 His two IUCI ofd~ In m "-for'" in PDS (1S4 and p6'J ~ &r &on. helpf1U.
700:PD51: ....,...... ~~r't~·+I1r ....".
Mj* ....".urltlllljU_'!'''~ AJthouch lhc meana of obtainina this knowkdcr
coma from !~ 8 D/w,.. nunifC:sco:d Ihroup lhc ill.jllRl:tions 01 the: Locd-
(I/N,..u,..~ tiJMndJ ~ on .raWtapida" ""m!I tha! [)N".. (or. moft
ptopc:. l, opnki,,&> lhc matd _ UMS 10 obWn IcnowIecIp- ihroup tNl IN:M, tj . ;' iudf'
~, ..-!'14
71a .OW( ......... !)-....).
noW. ,.",_Jtti4i.,.,.rihJ.. ss:
,. • .,uiJ m'lrIi".""...,. iii". ~ H.
,..w- _~ _ -.ttl. .,., ..",.",. f ...... ....
73 Indml u Halbha: Iw no«d (1991:7'0). Udd)'Vl2bn. and lid (dkIw NaiyirikM abo wed
PRAMA~A T HEORY: DH ARMAKlilTl 'S COf'lCEPTUAL CONTEXT -49
agree on how that causal process OpeDtei. Nevenhdcu. sintt the action of
76 In ~'. work. chc ultinwc idcntiry of !Tt'''''~ and ,nmtili' '''''''!'i9f.d" is di..
CIIDed II in PSV .J PSI.I.Icd- IO (Halloo 1j/6lh8-19; d . Hlnori'. not:5, I.U-67,
PP.9]-I07). Dtwmaklni diJcuaell the AIM \auc II ICYetaI painu, lhe moa Alient bans
PV).}II-}19 and PVHH-})9. Se... chapccr .. (16IR).
n ~ for aampIc. PDS (141-141), SV (~..u...7) and YO (Ill-IU).
78 Stt NY <697-n:} UNS}.1.I-17). lJddyobkan.'. lcnpr ~ on U-fI ao:nWly
cncOO. Myond NSU .17, but chc portion rckrmllO here lOt"" chc (OfT of the: di Kl lllioo.
7'J For IrtdI of analysU, _ chapter 1 (s,ff). TIw; moo: impoltlnl toUKG for me reduction
ot .... ~ .... oIo;.a ... "... ................. ..... PV'. ~I)-a-I f and PV,.~J1O. t d* , -
dolt in chapccr .. (,aff).
PRAM: Ar;.rA T HEO RY, DKARMAKI RT r S CONCEPTUAL CONT EXT SI
1.5 Summllty
W
H lN WI. EX.UoIIN.l Dharmakirti's ontology. we immediatdy
encounter a problem: DharmalUrti appears to emptoy ",flilipk
and C\'m w-,.nbkontologies. It is only by undemanding how
this multiplicity ~rves his rhetOrical and prulosopnical mttbod mat wt can
hope to avoid a dismWivc and incompltlc interpretation of IW thoUsflt.
Hence, we will 6rst discuss mal method, and with that diV1!f,ion in twwl.
we will then study the dct:a.ils of his ontology.
"
S4 FO UNDAT IONS OF DHAut AKIII..TI 'S PHILOSO PHY
rance: (allidyi) mal pc:rperuues sWftring.' BU[ a ueuiS(: that aceeds the
abilities of its audiCDct would not ~move their confwion; hence:, a com-
position that is superior in its anaJytical accuracy may be infirior soterio-
logically in relation to a particular audience.
The notion of a sliding scale of philosophical views is found in numer-
ow 50Urccs before DharmaJdrti'. time. and it was particularly imporwu to
Mahayana thinkers. Early Mwyinists such as Nigirjuna dearly wished to
idcmify tbenudves as Buddhists, but they also sought to introduce new
views or descriptions of realiry that they considered to be more accurate
than those promulgated previously. The problem is that. if a Mahayinist
supplants earlier views with allegcdly more accurate ones, other Buddhists
might argue that the Mwyinist is effectively ~jecting the Buddhist tra-
dition as a whole. The solution to this problem is to accept earlier views pro-
visionally: although they may be inaccur:lte in some regards. these views or
descriptions help beings to free themsdves from lheir confusion. That is.
if deeply confused beings were presemed with only the most accurale
descriptions of reality, mey would reject them OUt of hand. since: those
descriptions are 50 countcnntuitive as to seem prepo,nerous [0 most ordi-
nary persons. Hmo:-. to be most dkaive, Buddhist teachings should be tai-
lored to the audience at hand. In sud! conlan, more acussible descriptions
may be caught. if they help beings to eliminate at least 50me of their erro-
neous beliefs.'
2 For tIk JIO'ilion Uw liM: putpoK of a ualiK iI to dispd oonfuaion. _ PVLsb (Uur.1fI
"""'-i_..). Confwion (",.J,c) ill aplicidy idmcificd .. ir;nonnec It PVI. lll.-UJ I.!Id
...-...
PVSV .J ril. (G:III.IIft), ...bot it ill also apWntd thaI i&oorancz ill the- und.trlyinsaUlt of
3 Nipj ...... whom I take 10 bot liM: au~ of the /WJtbJi (_ WWe.. lOOJ.), ..as prolNo.
blr liM: lira to o;kfmd an apli(il hio:nrchr of news in Il)'IInnatic oontat.. Sec his RAnwINil
_,.i'riiU...,.1I
Hi'; o.J'.,_~~,~ / "
(}bhn 1911.:111-0)0)].
IJdrUJnjll/t~~
pbtue:I piOyidt • • Kalaln pua it. "~rimebonored ddWtion of error in India b94-u,j.
Oil 6c:cmcJ1t1 abouc rhiaddin.irion. when one inqui... into 1M buU for ~error. For
I>IwmWni•• conotpnW urorlftlWl copition ilOM thai "aupuimpota" (_ J II~' type
.,(~ unu> . . ubj<n ...... Oua """ On r.......... u.... r~'7 (_ .......... .,."
a..d alto PVSV MPVI., I " b; G:.+M1). SolIdo APa phlIo.ophc:n &om ochu rndi .......
often IIIWDC aNI cmM" iI bucd upon • lUll similari'1 c:orIMilvtcd br • uni¥ualW+-1oca-
cion. in ipKJt and Iinx hill tomdIooo b«n miIconnniocci. For dw: variow approachu.. _
Sdunitha_ (l96s) mcI. Frv.oo (1914).
I 8C.A ~)-43Ir. u_"J.N/JM tIn!-,.,.'n
"'4yllr U 'f,('r.llrJlli~"';_
~
~ 1.
,.,.., _ ~ ..H " •W
NOfC dIac Pnj6ibrunati pro-
.... Iiw P- of I.'.. _
. ~ . r;.-J.
--. to be rooced itllhote: uncbbomcd, CO!IUIMXIKIIIC bdic6 dw BuddhiIu take to br; mit-
Weal. See Drqfua U99T-SI-f9).
II ~ mist\! br; wnpced 10 c:onsidu !he olt.ma!ipv:d ViWputrip.;l .. an uccphoP, bul
only if _ ......,toeq...re'v''-wid! ~ s...:n Buddhia dUnkcn. 1,4 C'd , . .oduiCUCid
ds.mucha 10 br; re;e.:w,. ~"and InrfiJtI. To r•
KnIC: 01 d>ci, ~ ~
in dIia ~contidcr.n.- __ of praiIo: 10 the B.ddhaeom~ bran unk~ a ...... ~
When the mlnd CDnmiru die _ _ ", am" (4: . +f.(/. the condouicy olbinN will
noc br; f*ified. "' am" doc. noc Ica¥C the bean what chm: iI bdid" in the rdf. And
anor no otbcJ rcac:hcr in the world ill proponcnr ofldi'm theft iI no pull otbcJ
dwI)'DIItvicw, theWrJIOpa«. (d'7+ m _ · ... U-'!"Jllii-
! · M y:1tb.I.w.n.,.,..~" .... 1tIlJIlfl n ."".; Ibu j.tpti
. .......,. 16 / - , . . M IJ~_-«'''_IIl.oIlZU+ III . (II.cuHded in
u,..
• f • .IL.
URi
13 Fill' J lucid and rcfrahin& approach to dw; term -rnIi$m,- teo: ConCOI (1000), In £.ala"
naI R.caIiNn, cbf Icnn -rcaliwn" mm:;we to cbf local 00""''''' .1 c:omnwncrl! Wt lOme per-
uptu.al obj«tJ c:Ut1 oulliOc the mind. Hena, when rpaItin& from thil pt:!$p«tiV1:
Dlwmaldrri CItI be. -ralia" about am _mmw tbinp: and yn I "roominaliR" in that hoe
~;ea. the a.iunI« g( u.niYnWll. Sumc lNy claim mal _ ahouId jUlI UK the Icm\
"s"lIuinlika," .... ich 1M a.rIy COnIlIlmtilOf ~ docI we 00 I< Ic.t _ O<;CII-
Don (u9b), 8m wine "Swuilllika" in dW Whio.:wI CItI lad Ig both hiMora! and philo-
IOphiaI coaf'uaioa. In a hiauric:al c:onteu, me Ierm INy ubIcun:!he c:ompla td.otioruhip
anxK\I dvtt diIIina bnnds ul ~t: the aplicidy · Swutntib.- thcoria aptewd in
VuubandhU'lAlCBh, the ~ RcaliIc Ihcwiag(Dlwmakirri 'I". ' [)f Su~pta,
and Dtwmakitti', own bcerrqJ RaIlM dxoria.ln.bun, it lNIy be hewimaIJy Ilnwilc to
cquarc -Sautdntib- with all puII-Vuubandhll F..amW ReaIiam (1J/s;JtIl riM.wMJ, eV1:D if dw
&tcmaI Ralian if qualified all 1\01 SarriWrida and noc Madhyamaka. TItii bc:k ulhiaOf-
ical datiry ~ 1M pocmti..al lOt ~ c:onfwion, lOt if Eacmal RaIism an-
nat be caoiIy ~_cd wiIb aplio;idy SaUlrIn.iIta ~ an inlnpl'mltioo g( DtwmaIdrti'I
&tcmal RaIiIt qumma in lermt g{ Sauninrib phiboph;w moQnUonJ will be mil-
!cadi,... II iI ~ IW thiI teaJ(ltt that bo:Kh ~ and Sik,.abuddhi dearly
p.Jcr!he term ~
14 Hm:, dw; ICt'Ill "qmmnic" iI meant to reinforce dw; nocion WI DhmnaJdrti'l critique
60 FO UNDATIONS O F DHAi.MAKIRTI'S PHILOSOPHY
(- .~ .
n.
$ikyabuddhi LUa
ram.,,,,. ....
knghy di..·.;o., of tM inruwncnw dfea ~II me end oflbe IhUd chapen,
farti,. wim tM pro&opc at 1<).4If. Conamin& the ti"II,. __ ...
it in ~ pbca; uamplai include "Jt;oNJ and "Jt
w6as. Vinitadeva &lao u.eet the' term ~I>tndy (~IM:~. The: San-
skril -..rj"",... 54 iI a. rMj. The cquhoalcnl term -..rj~ 0C0ln in ~
(:uo.,!l '" which..., -..Id ~ me libnan tq1.linkn, "- ..... -... ~~~y tJ",J.:
unfortumcdy, the lalla' pM.: is COCI'Uf" in PVf (1 L4bl). N lOr ObarmaIdni'llIKof ~
Ii.....,.,..., PVin·D (l66aJ- 1.).
15 Sec PVI.l1lo- U J and PVSV" cit. (G" u.ufF, ~Icd in the: appmdiV; PVa. I. ,....,Ub;
PVu "cd- ",; PV1..JO,cd- JOub. 1M m«twWm or sWi"crinl iI IWnmariud U
PV1..:u]Cd-I'~
One who bdi .u (lilcnlly, "au"] thc IClf' will always dillJ to it ,.. "I." Duo: to dUI
dillli.. Of!( WIRI for pL no. and mal minliUdu the f"aula [of tho.: dUnp mal
aft imacined to 00", pl.euuft1. Seana:!hoIe dUnes .. havina poUU'I'e qualidu. OM
yarN ' " them, and doinllO, _ ~optiaIU" "m.iM" thoK IIXInI of ac:mm-
plilhins the dui,cd pIalu~ Thmfore, q ions as OM iI .1U(bed '0 the «Ir
(bltUl1IMti~ _ wW mNin i ll wpdra. And when "'- it I noOon old, dine
if llIIXioon 0I1Mhtt. F...... thcdiJrinaion bdu&i. Jand omaCDn'Oel anxhmanand
neniotI. IJr4' ~ ~ ..mi.", iii ~ nwIM9 n mJ.b...MJq. tr1JMi
'!1~ Mrr' ~ / l"~ ,.,.~. --m UI~ ~ H __
~IIiw»""'t"_,. ~ / invIr; _',.,~ ..,.,.,,;6401p1,,.nrr..
, ",.... .01.
Note thai tho: idtnUfic:aclon of,.,., fn!i .. the' f;,,~metu:al eaute 01 d-din&in& it 0ftI1
impliciL Sec, for cumpk. PV1..IJS- I)6ab:
11sc: anridol:t for u.. C&IUC: 01 RLffcrin& it QbbIitbcd by ~ the _liaI
nllllU of du, ClUK. .5uIf"ain&', ClUK iI dinPII8 ~ taka oondirioned Wnp '" iu
ob;a:u; WI dinr;in5 ;' a_cd by_'J fiDtion on the lIOliont of IeIr and miM. Tb.:
-.I;.arion of ~ wftid, motndM:u !hole notiorw. . IGpI the dinr;in&. I ~·
~ "Jr,"'''''''. ''-rO:, '
........i,
•
& ~.,.. , 4s"", "'.
1+ "t l .-woIJ1' + +r>4-'-(r"'r4,~ U
DHAR-MAKIR-TI 'S METHOD AN D ONTOLOGY 6,
ing about ignoranc:z (avitiyti) in lemu of me primary role it plays in tbe gen-
eracion of suffering." Thus, when one elimin:ues I4JkliyaJmi, one is elim-
inating a form of ignorance. If we atlend carefully to Dharmakini's
approach to ignorana, we sec that on his lcx:oum ignorance is an ingrained
cogniri~ habit that imputes a unity Of sameness omo thinfrS [hal in bet
bave no such unity or sameness. Allhough Dharmakini does nOI PUI il
quite in these terms. ignorance thus amounts to tbe belief WI a re:a1 entiry
may be disuibuted (alluiw) OVtt space and/or time. 11 Dharmakini proposes
thai ~ eliminate ignorance (and, hence, 14tltiJa4r!.tt) by realiz.ing thaI its
alleged objecr-an enciry distributed over space and/or time-is unrnl. If
this interpreution is correa, then Dharmaltini's 5Oteriology rests upon a
progressively subtler critique of all d.i5tributed entilies.
If we apply Dharmalcini's .KlfC:riologial madl.":1 m (hI.": hil.":r.lrchy of Vlrwll
discussed JUSt above. we sec wtthe hierarchy reflectS the IOteriology. In
short, Dharmakini ainu (0 elevate suffering bein!? from (IJ the completely
erroneous belieh that arise from ignon.nce 10 I-4J the elimination of all
enor, which enables the elimination of all suffering. The progression from
111101-41 is characterUed by a refutation of ever subder fornu of disuibured
enr:iria, lind it corrHpondJ to a sotmological p r~ion from w seate of
an ordinary penon 10 the state of one who has attained spirirwl lii>cnlion
(_}q.). Thus, along the lines of Sintideva's nOtton of luccasively morc
refined levels of undcntanding. the scale of analysis is also a scale of pro-
gression toward spiritual perfection.
While d-.e 1MUllllfit", for moving up the scale is 5Oleriological. the mt41U
16 Dlwmakirti'l .,;xuio!otJ edUbia. ddi~ turnudopell imp!ecision dw " ..... &om
an manpc 10 aa:onunodace at laM twa dUfwm .....,.. of ddinll1l* ClIUK of suan."" ,~,
cid>cr .. ~ or .. conNaion (,.J,.), wbac both 1m)' ~ in'nprrlnl • ~
(~. The e:rpIicit l'qUItioCI in qllCAion (II"~ - ",.;,. _ • .v,.J ia: mack.,
KYa'll
pbas. indudin& PVUll_ll).
for moving up the scale is a panicular form of reasoning (pKtI) that enables
one to critique ever subder types of distributed entities_At each lcvcl, this
reasoning is used to analyze some enuties that were taken to be ultimately
existent according to the ontology of the previous level; the result of the
analysis demonnrarcs that some kinds of entities that were thought to be
ultimately existent at the prnious 1C\-eI of analysis arc found to be ,.ot ulti-
mately existent at me next, subtler level of analysis. Generally, the entities
mistakenly thought to exin are thOK mu scc.m to be given in perception,
or in some cascs the existence of such entities is required to explain what
SttmS to be given in ~ption_ For example, at the first level of analysis.
a water-jug exisu as a whole in distinction from iu ~ fo r it seems that
when we see something that we call a ~ watcr-jug. " what we see is a single.
whole, entire thing. But when we move to the next level of analysis, we
diKovu that. in filet, all that truly exisu arc the components of what we
would call a "water-jug"; no "water-jug" exists in distinaion from these
componcots. Rawning (pin) mables one to move from the first level to
the second in that one arrives at the rtjccrion of truly existcot wholes by sys-
tematically employing a specific form of JUKU. namely. the mercological
anal~is diKuucd above. What ill most important here is that this lame
form of reasoning moves one not just from the fint to the s«ond level,
but indttd .u the way to lIN top ofllK Kak. That is, when one moves from
one level to the next. in each case the style of argument used to aitique the
entities that were thought to exist at the lower level is the "neither-onc-
nor-many" analysis mat lies at me core of the critique of wholcs.
To sec how this style of rationality moves one up the scale, let us brieRy
rcsate that movemCDt. In moving from me first to the $«Ond level. one
refutCl the real c:xdtence of"wholcs" and similar cotities (such as the .twum).
To do so. one argues that it is impossible to demonstrate the existence of a
whole mat is either truly distinct from or truly identical to ia parts. This
very inability to specify whether a whole is identical to or distinct from ia
pam is thought to be sufficient grounds for rejecting the existence of the
whole. In moving from the second to the third level of analysis, the para-
digmatic a.sc is me universa.!, which a.ccouna fo r conceptual extension in
that it is allegedly distributed over the individuals that insWlOate it. Simi-
lar enocics include a ttmporal.ly persistent entity that is meant to be dis-
uibuted over ia temporal instances; also included are entitics that arc
thought to be distributed over a spatial extension. Here again, the inability
to specify whether a univt':ral is idenTical to or di.snna /Tom iu insrann:ues
is sufficient grounds for rejecting the rrue existence of univcnals: the same
D HAR.MAKIRTI ·S METHOD AN D ONTO LOG Y 6,
meroological critique of wnoles. any one all the w:ly through t~ levd.~.
The style of reasoning employM. however. is not the only continuity in
the levels. We have altcady noted that at eac:h level the entities whose real-
ity is critiqued are pm:Udy those that were taken to be ultimatdy real at the
prrnous level of analysis. Thus. the Icvck of the.scale exhibit continuity. in
that an ontological commitmem at a prrnous level is taken to the next
kvd. with the oOt':(Ki"n thu the d~im i. IIffi""",~t the p~oull~ . but
tinJiuiat me nat level. An additional fC2turt of this continuity, however,
is that even when critiqued at a higher Icvd, .somc aspects of the omologi-
cal commitment of a previous level mAJ JtiU IN m4ind, but only as con-
tingent or conventional. For eumple. at the second level of analysis (me
Abhidhamu typology) the existence of a water-jug will still be allowed. but
now only ~I ('f"'tJntrimuu, erin~t . "'nher man "lti"","lyaUtcnt. In other
words. the opression "w:ltcr-jug" may continue to be used, :is long :is we
undersW\d that it is merely a oonvenicnt term for a bundle of parrides. And
at the third level, me ultimate existence of univcrsaU is rejca:ed. but one
may continue to speak of them as conventionally aistent. This way of
relating me lcvcU-wncre ultimate ontological commitments at one level
are carried over to the nat all conventional commitmenu--rdlects the
doariM of the twO truths. a Ufltr.l! motif of Buddhut thought that ~
have discw.Kd earlier."
While the doctrine of me (Wl) truths helps us to unde:rstand the conti-
nuity among the levels, that continuity still involves .som~ significant com-
plaitic:s. One issue is immediately probltmatic. SpccifioJ.ty. it is sometimes
difficult to separ:ate ontological claims that are being relegated (0 a provi-
, ional n . nu &om those tha, are being rejected outrigh •. 'In;' problem ;,
encvbated by the fact that the same belief. such as -a water-jugs aim,"
II Set !he uansbrion of PVJ.I~-l14 and abo tf". cU.a.aion on 'f»rlaI ataItion t.a--, 981[
"'-ha UJUmonI: in tf". ~rion from ExremaI RaUsm 10 EpistmUc Idcali$m ;. doe notion
m.1I an ExItrlW Ralia cannoc specif)o wbMcr I ~ objcn is idcnOcal (0 or dUtina
from tf". JUbi«tMry a:IIniIin& it. bu. dU!: ItJUIl'rfI' appc:an 10 be KCOndaIy.
19 Stubo¥'t. p .
6, FOUNDAT IONS OF DIiARMIlKIRTrS PHILOSO PHY
nuy be preserved wid! one meaning of"w:uer-jus- and "aim" but rejected
with another mooing. If we consider this bdicfin terms of the movement
from me fil1t hcl (the beliefs of ordinary ~rsons) to the K'COnd level (the
Abhidharma cypology). we find that at the second level one will compkuly
rtj«t me enstcnce of a water-jug ...ncn, in accord with the nalVC ontology
of me first kvcl, it is concrivro to aist as an entity in distinction from its
pam. However. the existence of a w:llcr-jug can also IN provis;tmaDy lI«~uJ
at the second level. bill only ifone construes water-jugs as conventionally
w tent entities thaI arc in fact reducible to their pam.-
Discriminating continge:ndy preserved ontological claims from those
that arc rtjccttd may at limes be difficult. but we have JUSt hit upon twO
principles that in all cases should guide our interpretations. We have nOted
du[, when w(' critique but uiU renin an ontological commitmem to some
entities thought to be ultimately aistent at a lower level, we only commit
[ 0 the conventional existence of those entities at a higher level. Our first
Siky2 Mchog ldan', analysil iJ useful because it covers the gamut of posi.
tiom found not only among Tihmm interpreters, but also among more
reant interpretm and possibly among $Orne! South ~ian comment:ltors.
Speaking of the first three levels, Dreyfus goes on to note:
FOUNDATIONS OF DHARMAK I RTI 'S PHILOSOPHY
Th~ differt:nt levels (t), h.), and (J) function at differt:nt levels of
analysis. When discussing epistemology. the lim level is wually
prckrred for it is the closest to the way ~ conCC'ive of things. At
this level, ~ conceive of ourselves as perceiving objccu such as
jugs. Such a description is nOf sustainable, however, for com-
monsense objecu are not lindable under analysis. Hence ~ oem
to move to a higher level, at which our epistemic p~ica art:
redescrilxd as involving only phenomenologically available enti-
ties such as colors.. But funher analysis reveals dm even these
entities, which we usuaUy think. we: pcruive. are fiction.a.l. Hence,
~ need to move to a yet higher view, according to which only
momentary particles are rnI (the Standard interprtt:uion) fol-
lowing a nr.uqy of asanrung scales. h is important to r"W.iz.e
that Dharmakirti does not believe that these levels are equ.a.lly
valid. Rather. for him , each level has iu own limited validity
within its own proper context of usc. Ultimately, none of these
thrtt levels is valid, for they all assume the aistence of atemal
objeas, a prcsuppruition mat Dharmakini ultimately rejects,'·
y-
5undard Interpretlltion'
l S DrqoNs (lm:a,).
26 ~ • dar 1tW. In hJi: view, Dhlrmaklni .uows for conf\iain& IICClOUlIIS within thot
Saurttnrib (Le.. &:lcrna.I RalisI) mw: IhfOU!.h mol( ofhil......,m DIwmakini prelmU
72. FOUNDATIONS OF DHARMAKIRTI 'S PHILOSOPHY
The most significant insigiu her~ is that our intttprtution must not
oonJfrue Dharmaldrtj's philosophy as fully un ifi~, if such unity would
mean that, for example. his entirt epistemology and his entirt ontology
must fully support each other tU1f1S1 kvds oftlnAly1is. In other words. what-
ever Dhatmakirri', Anal position might be, he need not argue in a manner
thai is fully consistent with that final position. so long as his arguments-
through a kind of sotcriological syst ema t icity-I~xi one up the sale of
analysis, Ncvt:nhclCS5, if we h«d the int~rprtt.illtion of the earliest com·
mentalOn, we also must not ign o~ the coherence or systematicity withi"
a levd that we discussro above, And if we do admit such systcm:nicity
within a Icvd, thcn we will be hard put to find any clear instance in which
·spatially enended ohjcctt arc 10 be included within ,he purview of thc
rtaI.· Why then wou.ld Drqfus hold this pos;ition? He has two basic rea·
sons. The first concerns a tension that he 5C'CS in Dharmaldrti's thought
The dichotOmy th:u Dreyfus highligha is one that runs throughout Bud-
dhist thought: it i~ the b.:oPc ~n~ion ~ " n ~rpe:al m the evidener: nf the
senses, on the one hand, and me mgcncies of reductive reasoning. on the
otha. This tension c:m:ainly is relevant ro Dharmakini's work, but I would
maintlin that, based C$pecially on the earliest commenwies. Dhannakini's
thought confomu prinwily to reductive reasoning. and not to me evidence
of the senses. ~ we will ICC in greater detail below. when confronted wim thls
r~)c," in ~ ro ~ ~ ~tiry mch :as" w:a~_j ug. Dh:arm:a.kini ~
flO( choose to somehow defend me commonsense intuition that water-jup
exist. Instead, he gives a detailed cxpbnation that can account for me causal
functionality required fo r the successful reference of the tcnn ~water-jug"
willxnathc c:x.istmcc of any single mtity that 15 the ,d"aem. Rather than sup-
poning a single, spaoor.cmporally c:xtmdcd ,deem, Dhannakirti's apIana-
tion :accounts IW the ieU:iehce of" term such:as "w:ater-;US" only by ~in8
to infinitesinul particles. Moreover. when fpedcing of en:ended entities such
as a wa£ef-jug, Dharmaldni MltTcaIJs them parricuWs (tm4t",,'(IItS). If. now-
~. we c:n.mine the problem of "foella to ouended entities through the
lens ofLatcrcommenta.ton writing in Sanskrit o r libecm, then we do find ~
eYer increasing tendency to mab: cona:ssions to commonsense inrerprea.tions
of the ev1dence of me senses, to the point mat even entities such:as water-jugs
will be :ulmirted :as p:atticubn prior to any reduction ( 0 infinirsim:al. p:atti-
cia. Since Dreyfus is mgaged primarily in a study of the libean intClprea.-
tions ofDharmwni. it olMou.sly makes good senJC for him ro heed dosdy
Thia conclusioo lconcernin, .,.acia! aunaioa] ;1 certainly ... rpriJin" for if aoa
apl..... 0N.t....JUm'. ~ •........d.""";mal Oftt060po.i ........ ry. h ........... ·
,ndicu 01:00 puAp, ~ DtwmalUni _ .... fo imply thaf mafm.! rcali!)' ia
mluc:ibk co .tomic compoomu. . ..This lppam1f inc:l)lUisctney ia dlK. U l:argutd
e:arlic:r. fO, f~ within DhannWrti'. rhou&hf ariAn, OUf of the double pmpcc-
riw: thf" orime. hia thoupt: the ontolopcal and the cpuu:moIop:oL Onm1opa!!y •
... awuion ia mIuud co inl1nimimal panidcs. EpUrnnolopcaUy, he :c.el , a.do
I rnluaion is Iwdtr to 1\UWn.. Ac:tordin, 10 hia Ii.ondaa.entallhaia dI.u pu«ption
accurardy rdk.:t.. rcaliry, e:u~ would ttenI to m.... b:u:ruion 'ppc:m, anti" all,
10 paLCption. Thi. oeemJ 10 lad Dharmaklrtl 10 ag;cpllt _ kw:l • minimal
notion of awuion.
76 fOU NDAT IONS Of OHARMAKIRTI'S PHILOS OPH Y
this realist trend toward [h~ accqawa: of ClteOOtd entities. And in doing so,
h~ provKlcs us with a ~ apprKiation for the need ro coma rualiu his-
torically the commentarial im~rpm:l.cions ofOharmakini.
Beyond dirttting our attention 10 Tibcran commentators' appeals for
the rality of cnendtd emitio, Dreyfus also raises an issue based squarely
on Dlumukirti's own way of speaking. Specifically, he noto that Dharma-
kini himself appc:lrs to speak of atcnded entioo such as water-jugs as if
they were indttd truly ral. Whm speaking in this fashion , Dh:umaldrti
onen uses the tcrm "baWl, which we can translate as "cntity," -cr.iucm
thing, ~ or JUSt -thing," We will not undcnU~ here a detailed cnminarion
of this tenn, but we ClJI agree with Dreyfus matwhen Dharmakirti uses the
term, he seems 10 mean that "hdllttS arc causally efficient. And since (as we
will sec) only ultimately ral particulan are causally efficient, it v.-ould cer-
tainly seem that "hdvlI5 such as water-jugs should be COUnted as ultimately
real particubn. This would, howc:vcr, contradict our omologK:al dictum
that ultimatdy real entities should no< be reducible to other emities. Turn-
ing again ro the Tibcun thinker SaJcya Mchog ldan, Dreyfus remarks:
35 11Nl
78 FO UNDATI O NS O F DHA1MAKIIlTl'S PHILOSO PHY
J7Thi1quia bul arikillllCimiaion iI die INin poilll ofDtwmaldni', mponac wan ot.;ec.
Don t.c PVJ-V WI all production is j_ CGnwnrionai. SIICh an ob;eo;tion deady undc:f-
mina 1M wrimacy of ~icuIats, bu, ntb.:. du.a ~ goualiry, OburrWJrti bronlcalIy
rQPOOCb. "-.. J6M wJtj, "lOr which David Edd in • priYlK( communication (,,;, T-shin)
.........
Iw: propooord the hipdy au aM:~. ~t(nf." Sn: PV).I- Io, ,nNlarcd in thc
}8 s.:..the di.."sfton ~ in chIpm I (J7ffl. I. Mldidon 10 AK:6-4 and AKBh ... nt.. aor
abo VMUbu!dhu'nrJUmall from simplicil)' in dx V;~ (.., ..... LII- I 'l. whnr: it
it; UKd in I Yopdn COftIalIO rdUlC the: reality of alengJ ob;ecu. Note dial it doa IMK
If"PW that Vasubandhu·. dil(!'qiolu ol"_ ~ pvticulan" (.,.u_w,.~
~:JO-fJ ...../\Xl.IO) and ~COfIIIomcntcd panida- (~II~fII)(.AJ(Bh: IIo'
AlU.uab) '" 10 be: I.km .. c:hananUcK ol the Sauulnrib pOODon ~ in AKBh. It
80 FOUNDATIONS OF DHARMAKIRTI 'S PHILOSOPHY
U lhil UII~ poaition thaI appean co be tM primuy poinl of dcpanun for Dlwmakirti'l
w
En... Realise onlolop.
39 I>hannUlrti mmlioN tM p;utkunat 01 poanialU.........m.: 10k ulrilNldy rul lhins--'n
K¥rn1ol pbcu. indudins PYSV .J PVl..... (G:17.1). The ul!.~tabiliry of particulan is
...Ied, {oHumpl<-, . PVI.7Uband PVSV ......t. {G: ". ~l. eo.-nuns
moma,~
f'ot- a div'lMion in Vuubandhu'. work.. _ W p«KIIulion of the: Saulrlntib arpmmu
apirut the four ch.m.c:ttristig of ~~nslO dot VlibhiPb (AKBh:1S?ff
. . AK1.~) and abo his commmlS in AK.4 (a lai h,. Dreyfus 199;.61). Dlwmakilti dis-
cwta mocnenunnat al ~ points in dw: S... ",m. inciudinS PVSV "" PVI . I~t.-I96
(G~..6-'UU.£4) "lid lOll Dlmai.... "'SUII""''' ..JP'VLdX>II. TIE ba.t L ........ ·li. IIM i.., ...........
in HR. wt.n.. the argumenl il one: &om amena: (- J . For an account focwi", on thaI
qumcnl , _ esp:ciaIIyOetkt (199) .
"0 Diptlp dairns WI particubn are inaprasibk in ....0.1" JUCh u PSV . . PSI.I.wI
and np«i.Uy PSV .JPSUff(17I7ffi,
41 DtwnWtini IUIIllI up lhiI basic pmniJC in.be phtue •... individual entities nefti' pro-
dUQ1: • di.uribu.lcd aWUt:neM wilhoul dcpoetId.ins on , "ninnal' 1M UJlriJ IIJtI!"'Jt&!I
sJttti1rJ'V'irqrltfl ~ ~MI!'~-""'; II'VSV ~ I'V I. I04; G:SJ.-4-1JJ. ~ _ will
dioa.uo f'unt- DharmaIurti'. notion thaI" univmoal. beins, discribu.red mory, c:annoc be
ultim.ndy rnl. Oharrnakirti abo =narka:
DHARMAKIRTI 'S METHOD AND ONTOLOGY
"
from every other entity (J41V.dt6 hhmNL J41V.dtil vy4vrtliL U4ntll"1'illJ1liL
etC.)." Exprused in mis fuhion. uniqueness points to one of (he rwo basic
constraints on Dharm.alci'ni's ontology, namely, that the real mwt conform
(0 the dictates of reductive reasoning. That is, if an entity wen: distributed
and rnw reducible. it would not be unique, sinCt' it would shaI'(' a spa-
riote:mporallocus or semantic identity with the entities over which it is dU-
rrlhurM. AI [he .lame rim .... Ihe crilerion of uniqu~eu:l1Mo me:lru rh:lt Ihe
u1timatdy real has a specific {niJAtIl}or restricttd spatiotemporal locus and
nature: it occurs in a specific time and place with specific causal capacities.
Expressed in this fuhion, uniqueness satisfies Dharmakini's s«ond basic
ontological consmint: mat the real ahibit a kind of regularity (implicit in
me tenn niJ;ttll) such mat our perceptions do nOt appear to be simply ran-
dom (tlltllmriJ.II)Y
In terms ofOOm irreducibility and regularity, me uniqueness of the ulti-
mately real is reflected by 1Vd~?U'> the term mat Dignaga and Dharma-
kirti USC: to refer to ultimately real entities. In its earlier wage in
Vasubandhu's Abhidharma works, II. 1Vd1a~~ is II. way of dwacr:crizing an
...tIcft Arap CO<\SU\IQ du. IUImnnlt in rom. of ~ 1"hiI b...ac noOon !hat IQJ mu·
ria ~. tptcif.c: If"" , ., iiploc:ut abo ~ IaUrnmD in MV, such • ...tP'Vt.nd,
(G'44' MH'(" ,..' IIw ( ow Witfe.,...,.,. -,..,., u.,..'!"l and ..... N1oJfl, ocw: olioe>'·
en! pbca: what: DIwmikIrti ~dy ~ w poISibility of randomllCM (G:99.ilff. ...
... , .. f~-'"'1 ""~ \J.J ""' , + .... ~ I _¥ ,;.,.. """+>1.+-,,,..; ?J.sJ<'.
Anocha _ of du. iAur in PVSV iI ..... P'Vl.n (G:ll.wft). Sa: also the dltcuaion by
Dnyfu&. foIIowina 1M Tibetan Ilithor Go ~ pill bSod nanu ""'" F hm:6~).
+t For 1M ddinilm 01 rii,.
II ~l'iU"'!Uo .... AXBh 44~ ! . AI for the UAF o(
~!Mo"" thc detcripl:m 01 the ~ fonN ol ~.ttmdin&· (_«1Mr.)in AKBh
()70) ..... AK1.7l.c:: "IJII _M~ ,...,,~.!i"-~ w,..thi ~iU""'IJII Nft-
it] nM~ I .......) ,,~ __ ~ ~".,NJtduIJ l .Jhiwu.hi~
_ . ·MIIo' ",•• I!b"""",,lto!lJi.t...
~s""".-..u..-u.
P F (,\.10".... ....tAKR.. ~N. (,01» .. ill.-rau-of ..... W1yin ....Oct.
~ and ,."..,,~ I « IIied at p1i«d ~ in thc K.I.:.
particulars}, and since only paniculars arc: ultimately real, anything mat
bib to produce an efJ"«.t is not ultimately real."
_ ]i" ,. j . , . , , . i roflll Wi" ",....," ...." .,rJ M I pi- iii . . " ' . '" ,....,
"lIP" .;.,,.' _,." -.,,.1 pc:.tfo.-:n
PV).JO: A uni¥en;aI (:&/\IV)[ nom
~r,.",.,..u 'dwI,.",u ""1,
tIM, tdic fwv:rion of Amply prodOOnc i
ImIOf)' lIWUCIIeIIofiadf. ADd p.eciody (tIW) IxaUK i. it irK'lpobkof.tw. illw no
narure (...v,.J [i. .... i. iI untmL d . PVP:l.4J1.sJ. lOt ohit incapecioy lOt rdic fUnction ill
dK dcfjnition of unrc:aI dUnp. (ju.........mJ.d.r.!'" ."~,,, .u nM 141 I tM
~ ,,.. -,.,.""',u.,-..~ ..I.
S3 Noc~ dul pan of wiuo, it II JOb " - it tIK ~ iNuc: of ""'- it IIKaIIt by · p".Q<pti.
bk" (~), fOr which Itt Kellner (1m). Tbt noOOA dw the poa:pobilily it tIM, millimal
relic cffiacy requiml of. FWtiwbr QHrIG frwn DbumakJn;', oom~tI ill PVSV .J
PVI.zll (G : 14,.U-lJO.~, w!I<'tt hc.aIJIIC' that an ~I thins'. lion it "'" aUlCd, but
ill ;~I 10 dul W", iud( The I sr mds:
We Qy the foIIowir,c: FOC' tho.e eLitrml Ihinp that. Iu.YinI beat knowrI (i.e. . per'
ca-I] II _linK by IOI'iXOIK, arc IlOl !lOW known. their OISAtion ill invariably COIl'
oomiwlI with their existtnce. They, 00"1 produoed, art aabli&b«I UI be
impermanmt. The teaIOI\ fOr lhif it dw !he foIJowi"l is "'" poAibk: hi !hal tho.
thinp ~ dK ~ of produeinc an nnsmc::. of thenuod.a ya: bter (,.~), IIOC
Ita...;"' ......oed no 10"1" ptodooe such an - . OC' (1) that dwy -..Id dqKnd
upon ~ eke ill ordo:r UI produa: dw • ....aIffM:II. NadKr of thac it poMibic
becauK dwy an: embIithed to 1u.V't the ~ of procfucins thai i W U - " And in»-
on,,", at ....... thi"", ..., ........... ckpet>cko>.. 0<> od.c... to produce ...d:o. an _
no ~ of them wou.Id ~ ........ Apel dlClr is II!! 'I'cb real 'binl baz'lK cyqy
rqI rbinl is known h:r mint knam u IOlQC riQK and pin If dK thi", irl quarion
~ IlOl "'polk of ~ <he: .c& funaioa of pcoducin,an _ !ben il wouJd
IMX bn ral thlllI (-'"tl For. all I will czpIain ri.ll dK ~';"rf itt. real chinp
lint dul (.hiliry 1(1 prodooe an iwatnKII) all their Odini", clwxtcristic. \)or It.t/kil
......, ~ jMI4 _ ... pu,-II ~ ..rth....,JM ..M.;'; ~ I .. ,..
~fJI!I~ / ... ",~ .......... itiJtU/It~~,_
....,.14 ... ~ ~ "',.nf1ff ( ~...~.,.". ~ I ,.. ... rqtt
.~ ~ /uMIkU It;I!'rij jUUI!' ,,;""rwg I ... ujN"."""'!' It;~ tin I
,.,...,.~"..n, ~ jUu ' jU-..k#"J.d"h'" ~ ~ _ ...... Nt
~"I4liM" ~ _in Nkgi"';.1 .
Sikyabuddhi (PVT, ..,..:sa&If .. K:J)4.19ffl imu p'cLi dK undo:rIincd ph..... III mpotue UI
<he: ob;eruon, ~1IKfI would QOC be the: a.ae.".;m • produ«d u OIi ty mal it IMX an objta of
kllo...tcd&c- ( ... "~,. ... .,.. -JUt" * ... ",;
Dw Ill! K: ". ,."" ~ .,.,.",
.'r ' ". OIJ_ .......". ~.l . H. do ... ",""......0. "oJ... io 1>O...d:o. ...... thin.s" to
mean "there iI GO such I Wr., due it IlOl all ob;ra ofk"o.. k.,"
[PVf (..,..:Ja6): /u *
DHARMAKI RTI'S M ETH OD AND ONTOLOCY
H Cono:ponul and I\OClClOtoa:pcual puu:pc:.w iDlUion arc indudcd IInde. tht rPJU&I ate<
pyof-5pUriow paa:pOot,. r,..~1.r/Ms..). dacribcd by Dhannaldni M PV).1U: "'Thuc
ate lOut kiDdI of ~ pclapcioo: three kinds ofllClOOtJKual awuenca and one thaI is.
norw::oncq>n.W I~ thaI an- I'rotn cfurutbanca in rk physial bua. ,i.c., tht: &c •
..Icy]." I """~ ...".....,jU_... .u~~ If ! _'" I . ...-'. "...... ~~ ~.
~ (_,.uu...- /4.
no., mree rom. of ~ 5pUriow pc:la:ption Il'lUlDoncd in tM _ arr. "uroncow
awumCII" (Mri"tijUu). "ronvcnoon.alllincW-ie 1W:lm1CA- (~vrujU-). and " in~·
entiallW3lmCU and .., on- (.,,-U;). "l"bc. (l.tCJOOa m: bMcd llpon DiI"iP" brief
IfGunmt of UfO. (PSI.I.~). We m: Q)named MfC only wid! "fflOlICOW ~
DQl," whido if IImIIiIr. form of pc:ra:pnW iUwion. All
Q;XS of nonconcrprual spuriow
or Mthis is a mirage- must occur after the perception. In the most rypical
form of a conceprual illusion. me image in perception arisa in such a way
that an unschooled person is confused by the similariry bctwcrn the per-
ceived ob}«t and some olher object. He thus ·superimposes· (I<Imi-4nth)
some aspca of the similar object onto the perceived object. Hence. when
a penon unfamiliar with mirages sees a mirage. the similarity between the
mirage and water confuses the perceiver, and he superimposes the fact of
being "wan!r~ onto the mirage. lbc resulting judgmcnl. "this is water," is
a case of ronccprual iUusion. For our purposes. Ihe key issue bere is that the
nonconceprual content (i.e., the imagt') of that perception is not itself
Rawed. II is the fX:Icci\'tt's inability to correctly interprtt the image. and noc
the irmge itself, that is cawing the enor. On the basis of the same kind of
perccprual contan. a ~n with the correct mental conditioning--f.unil-
wiry with mirages-would not make that error. In other words, it is po$"
sible to distinguish a mirage and wata mcrdy by sight.II
In contrast, in a nonconttptual error the mechanism of the pen::eption
itself is fbwed in such a way that it becomes impossible to intcrprtt the
image corrtttly through a pcrcq>rual judgment about the issue in qu~
rion . Ooc cnrn~e cired by Dharmakini i5 the ap~(r ofhain in the
visual petctpOoru ofa person with cataracts; his basic point is that. in regard
10 me question of whether one is smng hairs, a judgment twed on per-
cqxua1 conlenl cannol be corrcct.!O A correct interpretation through per-
ceptual judgmenl is impossible because, with regard 10 the issue in question
{"Art. these hain~', a judgment that is COl1.linent with the image will not
cnablt' tht' peroen.-tt to act upon an objco that can function in a manna
consislent wilh that judgment. That is. in relation 10 whC'ther me perceiver
is scc.ing ha.in, a judgment based on the image itsdf will lead the pcrcciva
10 condudt: thai she is indeed smng ha.irs. Even if the pm:eiver determines
through other mam Ihat me is not s«ing ru,in. the pen:qnual image will
still Rp~IIT UllHhain; hentt, a judgment 1w«I JUSt on the image will inter·
pm it as an image ofha.in. Nevenhdcss, while the content of the percep-
tion always scc.ms to be hairs. if the percei\'tt cries (0 act on the judgmem
s s Dtw-makini dclcriba concqxual mor in PVSV .J PVI.,. H,b (G:4HI). For the
ncxion NI habitlWion condirions !he dctt:rmiNotiont m.: follow upon • pcttt'ption.. tee, 100-
cumpk:. PV •.fI; and PVSV M m. (G :)LI R) , u'&nA:aud in die _ d!apler (n.,,).
56 Thil caw it dx mot< fmjumdy cUtd. 001 other aiCII include: the pcrccpOon ofalip.-
cirdc made by a quicklywhiricd Iotdt. me paapilon ofllftl as morin& &on. the pasp«-
,;"'" of. pc_ '"""'ins _ . "-" and ........... ~ drccco....-l br dw: ;",L.'. _
of!he body'l hwnon. Sec NS •., .
DHAJl.MAKI RTI'S MET HOD AND ONTOlOCY 8,
that rhcy 2rc indeed h.airs. mc perceiver will not be 2blc to act on an object
th.at has ~ causa.l durwcristia apC'Cud of hairs: any attcmp[ to brwh
those "haiB~ 2way will &.il. It is in this seruc th.at 2 judgment consistcnt with
the imillt will nO( enablc thc perceiver to 2ct upon an J,jtct m,at can fune--
rion in a manncr consisrcnt with m.at judgmcnt.
In citing a case such.as thc appearance of hairs to 2 person with cataracts,
Dh2rmwtri retnforttl tho!- ;n(~biliry of nonoonttpuul crro.., th~ du..
toned inugc (which lculs to misleading intcrprct2rions) comes from a p:u.
ticu1:u flaw, and .as long .as thc flaw remains inlact, it will always produce a
distorted im.agc.'7 Without removing the Raw. no othcr factor- increased
acuity. h.abiruation. graduatc studies in optometry- will prcvCnl thc dis·
tonion. To PUt it anomcr way. evcn if an optometrist with catar.aCU gives
an honest rq><)rt of w6.1 ,h~ sees, th~ optOmetrUl will ':ly that .h~ tees
hairs. Thus, an a.pen in dcsctt travel can distinguish a mi~ from water
JUSt by looking ae thc spot in question; but a cat:u'aCtOus optometrist can·
nOt distinguish bctwcc:n hain and certain cfkcu of cawacu jllS( by look-
ing at thc hair·Ukc images appearing in his visual perception.
To rerum. thcn. to thc notion of thc pc:rccptiblc and thc uhim.atdy real,
D6.rmaIdrU', theory of nonconcqmW error rugsau ~I. despite the pl:au.
57 Thil point is made in PVJ.JS9-)61, whm:- Dharm&Itlrti diJaweJ the IUION fix daim-
in, dw the oubjcctIob;.a duality in our P",cw:pcio;u: .. ..:nWIy un-.l. Noc~ hen:: chat tM
kind of topition mat
aNa: from · inmnal dlstonioCl" (.~ is comJNn::d 10 _
t'amiliar ilIl11iona (i.~, the QW'XI-hain iIIwioa), ud mal distortion ml$ tbcOlCf(M~ amounl
coa dcf«t in the mind or mental f.cuky ir.df, norher th3n a II>M: miKorumw ol'iu ob;ca.
The _ in qlKlCion tad II follows:
Thinp do . - lllrinu.tdy luve W MNn [fIlCh all bans an ~ obj«t
~tMIiI~ dw they an ptiai....d to Iu-.c ....... mc they hive neithct a go.cuW'-
a m..Jriplicil_ ~ [PV},)S,].
- SuI ~ 10 -wJy COO'tUIUon (WI), dlK 10 upaicncilll_ Jimibmy,
an illusion an.. becaUK one demmiMs tNt ...tUch does noc ha¥t such-and.....dt a
nawn all luvilll that iWW'C.. Bw in tho. aoc, that if noc poMibic becaUK on )'OIl!
..ww noc C¥m 0fK mti.,. in the -'d iI ot..a-i 10 hive llul illtim' [or ba"l an
...... cl... ..Jo:,J "" ......... ..a..;.:........t ...d.J. (rv,.~,.t.J.
11lft'e iI challtind «illusion. bul theI-e iI alto the: one dw br its 1lI~ an..ibwo:d
ill dw it onpnalo::a; ffom an inu:ma/ d41onion (_~ ~ if W c:uc wilh the
pou:ption of luin by a P"1OII wilh CIW'IICU and 10 on, it conwru a hlK ~
appcanna withouf ckpcndinl upoa W ~Uon of Iimibri.,. and wc:h.
[PV).j6tcd-P'V).j61] IM.IN'"... ..mifJtt". ~,!",..". ~ I J-lrttU,.v.
~ w ~'!' tqJJ!f _ ~ U ·14 1M) { "",Vi ~ IJn.Jllti, ~ 1
~"i WU""""",.OIlJ#j<_ ..",. ,., III ...n.1f4j j4p1J _i"" d.".,i
filMr-"'; I .fiJUf .",. ,. .",."..,.,w_IlilillM.. If 41, ,~,.
I .........
_II
" cmain timc:s dw to ba:ing q>istmUc&lly I'UIlCn in fttmJ of rha:ir narun: and 10 0<1,
and beoux Oft ~ orha hand. rtw IIOftpaa pcion is tqU&lIy I(l in dw: ax oJ nona;.
iRan lhlnp aIio. IG:lol.llff. _ _ ". ",",?k""Jll_ .."",,, 1 _ _ qi
_f& ' ·Ji..,.,rd,,'1iI ~ "",."u-t&" UlJhAlDqi n.9wtlthl.
91 FOUNDATI ON S O F DHAlMAKIRTl 'S PH ILOSOPHY
individual or entity that we aU a · cow" would be one thing now, and after
2 change in which no sameness continues, Wt "ww" wouJd no longer be
2 cow, In short. mere mwt be some continuity rllnvlIJ4) th2t accounts for
twO f:acts: since we can corrccdy call that individual a ·cow" over the entire
dun.tion of ia existence, something about the individual must remain the
same over time such that it can always be: called a "cow," And sincc ~ can
also corrc:cdy rc:oogniu other individuals as being a "cow," then: must be
something the same about all those individuals, despite their m:any differ-
ences, Wt allows w to corttCtly call c:ach of them a "ww:The alleged
entity that ac:c:ounu in both ways for this samencss is a · universal" (s.lm4n-
Jil. Urn', erc.), and it is the fact thar c:ach cow-individual somehow insran-
tiates the universal· cowness" th.u allows us :always to call such an individual
2 "cow," If ~ consider a universal such as "cowncss" to sommow exiu in
distinction from the individuals themselves, we can easily allow for the
many differences among individual cows, md we can allow for the m:any
changes that:an individual cow undergoes over time, Despite such ch:anges
:and variations. the univusal "cownes.s" that each row inswulares remains
the same. and inasmuch as each oow-individual alwaY' instantiates the uni-
versal "cownen," each one can always be coruoc::dy callffi a "cow," As long
as we can assert univcrs:a1s in this fashion, the cb:angcs:and variations among
individuals will not vitiate the sameness uquired by thought:and lan~
Suppose. bowcvcr, that me universal · cowncss" iuelf were to change in
such 2 m:annet that what was once ·cowncss" now became something omer
than "covmess," In that case, an individual which a momcot ago inst1nu-
ated the universal "cowness" would suddenly instantiate some univtnal
that wed to be · cowncss," bur now has become 2 non-cowncss universal,
such as "honencss," Hence, what was a moment ago ealJed a ·cow" should
now be called a "horse," Neither Dh:mnal6ni nor :any of his opponents
wouJd maintain Wt what was oncc corrccdy calJed a "cow" could in the
vel}' next moment be suddcoly and correcdy called 2 "horse," For mese
and other such m.sons, a univusal cannot change precisdy because it is
mrough direct reftttnoe to a univtrul that mought 2nd language can indi-
rectly rekr to me individuals that irutantiate it,"
Implicit he", is the auumption tim any changt: in me univetUl mWt be:
a change in its idenu[}'; that is, if the univl:'rsal ·oowness" changes in any
way at all, men il has changro into something other than the universal
·oowness.- We do nOt have Ihe luxury here of claiming that me universal
can change in some of irs accidenral propenies and remain unchangro in
ia essential properties. We cannOt Opl for such a solution b«:ause: univef-
ull on he:ir nn pm~rTi~ ar all , wherh~ we cmuider them in ternu nf thl:'
l ul>stantialisl ontologies of mosl non-Buddhist Soulh Asian systems or in
temu of Dharmwni's nominalist theory of universals.O: Hence, since it
cannot manifest change through some alteration in accidenw properties,
if a uni~ changes, it essentially changes, which is to say thai it becomes
50mething dse, In shon , if "cowness- changes, it becomes a univerul that
~hou ld ~ insr.> nfi ~ ~ in indi"id,,~ll .har weff' nncl' .. nWl'. hll' nnw aff' nn
longer cows.
Thus, in order for the mcrcntw function of language and thought to
suc:eced, univcrsals cannOt change. On Dharmakini's view, hov-ner, if uni-
versals do nOt change, then they cannot have any causal efficiency bc:cause
in order 10 function cau.sally, an enrity mU$( change. That is, if an entity
cannot change, then ifit i$ not currently producing an efha. it will n"",~
produce an e/Tect. It could not produce an effect becaw:e an unchanging
enu')' that is nOt currt'ntiy in the stllte of producing an dfea cannOt change
from the state of nOt acting as a cause into the na~ of acti ng as a cause:.
Alternatively, if an unchanging entity were to produce effects, it would
i".,,_
......,Id """ boo 1""",;bIo!. i"" ....... (IJ,.,U) ""n....,. ................. induu""'" *
copition. r,,~ iii ~'" igtftr I tI"u,.tfIt ~"."" INul/Jil ~
tUf'tlfJtIJ>IJ"th II.
l'he ~ dw a uniwn:aI ""'" in _ _ bo:CO<L$~' or undw.P"t: nw duor'gMo' ( tht
diffMnl rtnndo ofSouch Alian philo.ophy. It appean 00 lqin wid. m. gnmmarianl:. as in
K.il)'lyalu" wdI·known phruc ITom !he fiflt v",,,,;u, "If tht rdarion bmoccn aptaAon
andob;m (tInhtI)is ....w.!;lhed (...:.tJhrt ItIWrn.I....,"''''''''M ... • (d. 8Wdcau 1~'S), In dx
variotu inlcrpmacioru: eo<Uidcro:d by Pau.tlj,Ji in hisfUbKqr.lm( commentary. m. common
!heme: is !hal 'ia eaobIisbcd" (~ m LUI <nean 'is permanenl' ("i".). In tcnns of m.
ob;.a (nrJ.), Paafljali ra:.onh two aide. u. 1hr drilatc: Vippi)'VLA. for whom a c:bu prop-
my (~) is m. ob;ca. and Vyi4i. for whom dx object is a .oo.12IlU (.,.,."..) (Sdwf
' 996:4111). h depicted I I Inc Oillln of tht M"""'~ (_
I ?"n~ffl. each A.x suppoIU iu cbim by noanS dw the cntiry!My iddur;e as
m. uanslat ion in Biardt.iu
m. object "
pcrmanml, while- me «her cn ary is Unpcrrn2llUlt, It" thus cka. ,hal. ir PaWljali', drilala
rdka early pwnnucial conmru. IIx ~ 01 dw ob;ea u permanenl _ oenmil<) me car-
liat layc:rof phibophy of~;n South Alia.
62 Sec below, 116.
FO UNDATI O NS OF DHAIlMAK i RTI' S PHI LOSOP HY
67 Dharmwrti mda mil point II I numba of pIxa. indudillB th( discussion of momm-
",,';MiI al PVI .)) and PVSV "" ti... ...t.m: he mn.arb.1Or aampk. · 11 ill'IOI com'CI thai a
qw1iry wIUch ia I'lOl csublisMd whm tofM chi ns is nabliJhe:I if 1 ",oper!)' oi dlal thi~
nor can _ ay mal • q...... ,ty ~ a ..- ~ dilrnml /tom 1M auaa of _ thi", is a
"'<>pUlf of dlll dUns.. (G W.l~U: ... hi _ ill.ug,.1fM Rig-... .,.,,~ "" Will>-
.Mb.~ .
1lw: faa mat aU 1M F"openia: pmliabk of. thin,OOfnC: into ~ with thai thinS
is tM on rolopal buillOr Dhmnol.k.im·, daim tNt one: proP..,I, an t>:n'f M ,nVlliabk m..
den« for anotha plOPC:II,. and ;1 if implicil in 1M oft tcpnted phflS( MtiNIlliITilo",.JJ,j1l
(i ...... ..... dw po P '"'1 .0 be p ... . ~ .. io "l .. YUiably _ , mi ..... wid. dw...- pracncc" of
1M proputy addllad iN utOc:rcc). S« 1M ~ di.tawion below in ch:.pI~ J.
DHARMAKI RTI 'S M ET HOD AND ONTOLOGY 97
have ~rttptions of mem only at spC'C ific times and places (and not at all
times and places). S in ~ thOSoC' thinS' change. mey mU$t cease; and Ji n~
those things ~, they must nOl iK: perduram. Moreover, if those things
arc not perdUl'anI, they must arise:u nOI perdun nl (i.e. , not unceasing).
And since: things arise as not ~rdur.tJn, they cannO[ exist over time, which
mC'aJU that me')' mwt immroiatdy ceue. Therefore, all petttptible things-
which means all cau..tally efficient things-mwt Ix: momenury (/q4l1iltll):
they endure for only an instant (~!"l).M
If ~ construe these: arguments about momentariness with the require·
ment for the constancy of an a prasion's object (lItbdliTtIut), we an Stt
that. if particulars mw t Ix: constantJy in flux Ix:c.awc they are causally effi·
cient, men particulars cannot ~ the objecu of thought and language pre·
civ:Jy keaLLY. they M nnTperdu","" Th L L~. ,.lrhnugh Oh,.rm,.kim 1'I'I:Iy Ix: l...u
than explicit on this point. the awal efficiency of particulars underlies the
claim that they are inexpressible.
68 a. PV5V a'PVl.1b (G: •.,.ufJ). ln thU Kaion 1 1u~ presented 1M IrJUn'lmt for
lIK>mCIlw1nao in a way maL ;, meant 10 bridsc m.afJ"mml a)nenning m. cawdco.nca
of caution (..m.u..)with d~ afJ"mcnl for momenw1nal from cximnoe (i.e., m. to-aIkd
" , _..---. _ ~.J1y ()"d,., '99) r- an .,...,."i- t~ .........) . ~n"01I ..... (,' ") ....
~ed mal the IfJtUnlm l from ainma: ill mort prominall in DlunmkIni', Iala faTS,
bul be dearly don rIO( mean to vitia(1: rh.c: strong roruinLLiry mal 1'UJlS 1:oo:1 .. « um. rdi,.",.
rion of any QWC' for a:waUon, on 1M OM lund. and m.
arcuma ll from cxlRma:, on m.
other, l1w COfLrinWty oprillP from Dhannakirti'l rheory of ptoperrics U;I n:U.ICl IOawaI·
iry and pm:cprion. 1I is ill w.
~ dw _ _ borh arsummtl in PV and PVSV, ~
Dhanna.kini fIOII only rdilla aLiX'd assaOOn , bul he also maka a.lcnsi~ UK of !he uxu·
menl 110m aillmU (PVSV .II PVI.I,l.-I9'r. G:'7. t8- loo.Ui). Orhcr.1as a lmsr,." P''''V''
indude:
/'VI.II(>: ~6ItttU". ... _~.tmrt. ·" "" 1 W'JI'~ riA7,wU'I]4nIM'!I df;t
~1WfI1Wl Hand PVSV .J til. {G:,,,,lt--I.J}. ..~ ~~n.. tw lOI_
n.n....~~.. JJ.r-~~~ ..1Ni..,,,..m!I. Stt We PVSV iii
PVI.I17 (G:".17): ,.Ji umw'lfl ""i"._ "":!If1Tll oJ /,nvf. ~I }, and PVSV '"
PVI.ISllb (G :U,.l)-lS): _".". ,.~u'!l P-"'I"1"'!' b "-!t _ w i ~
IrJ ""j".,.. ..".m_ I!tt.WtiJ M.nwi I ... . ", lbJ",iulW JlfmWIJII MWi..;,.__
"uj,, ~
We have juS! Sttn that, , intt paniculars an: o,usally efficient, Ih~ mWI ~
inexpressible and momentary. Tbis ontological rtquin:mcnl emlxds par-
ticulan in Ihe KgUbriry of causa.lity, and it mw gu;tranlcu tnal pc:. ec;prions
arc: nOI just random. Paniculan, ho'NC'o'u , mUSI also conrorm to ,he: dictat~
of reductive reasoning, and they thus mwt ~ irrc:ducible. Moreover. on our
rtlding of the commennricr 10 Dharmakirri', work, thc itttducibiliry of
particulan requires mat they lack spatial exlension. In our di5CWSion or
Dharmakini's philoJophical method, \W: n(Ked thaI other interpreten do
nOI accq>t lhill view; Drcyfw., for aample. maintains ,hat what he calls the
"alternative vi~. " which :mribul~ sp,uial cxtcnsion to panKulars, is a
poIition ( 0 which Dharmakini mak~ an ontoloeical commitment in at
least lOme comexrs. We will now examine this issue in greater delail
To recap some points made: ClItliu, we should n(Xc that for Dreyfus, the
"sundard interpretation" is what he calls the "Saurrintib " position: rttI
external objt'CU (bihytinhas) exist, and thOK objt'CU have no spatial exten-
,ion; they are, in shon. infinit~ima.l particles (p4mm4!'I4). This is what we
haVl: called "Exlernal Rali.Jm.· In OOOlr.uc. , Dreyfus::oho idcnuflC:f an "alter-
narM view." whtrmy external parriculan: may have sp.aci.al extension. Drey-
fus arguts that, while the srandard interprtllllion may be primary for
Dharmakini, Ihe alternative interprtt:uion nC'Venheless forms pan of his
onmlogy.MA curiow aspect of Dreyfus' approach to the "alternative view"
ill the cla.im that he is "not arguing thai this is Dharmakini's vi~, but.
ramer. thai this vi~ is prescnt in his work." Dreyfus does not mean mal
Dharmakini ,imply memioll$ Ihis vi~. fo r DharmalUrti frequently cil~
views specific:ally for the purpose: of rejecting them. Instead, Dreyfus means
that tht" alternative view is acccptnt (at least implicitly) by Dh:llTl'lilini in
lOme case where Exlernal Realism (the ~'nndard view") is IOmehow inad-
equate, and thai in such cases, Dharmalcirti employs the alternative view
without dearly rejecting or embtacing it.
In the section on method. we noted thai Dreyfus across levels of analy-
sis reminds w to resist any tempt:u ion to formulate a systematic uniry in
Dharmakirti's work. Dharmakini's method is such that he: is not only per-
mined, but is indeed (lMp!o argue from positions that he wiU CVC!nrually
aNndon, if his 5CXeriological project is 10 succeed.. Neverthelc:ss. we also
argued thaI DharmalUrti'. thought cannot be: so disjointed that. wht"n argu-
DHARMAKIRTI 'S METHOD AND O NTO LO GY .
ing from an ontological srance that he will abandon. that same ontological
SUJlce itself exhibiu .uch:oJ, drgrrr ofinCOMrencr thai il is no longer clear
what onr should critiqur when making thr transition to a higher Irvd of
analysis. In otmr words, even though Dharnukirti', Exlrrnal Realism dots
not providr a fully unifiro and systematic ontology. he nevmhdess is quite
dear on thr issues that COUnt when we mo~'r from External Realism to
Epinemir: ldel1ri.~m.
And what issues do COUnt ? A full reckoning would require a close c:xam-
ination of what Dharmaldrti and his commenr:uors mean by Eputemic
Idc:alism. and we will not attempt that prodigious task hert:. Ncvmhdcss.
we can fint naIr that , when Dharmaldni applies an Epistemic Idealist eri-
dque to External Realism in his PrlfmA!'avdm;kll, he does so whrn pre-
~nrins the Exurnal Rral iu rellpon ~ 10 a problem in ,he theory or
prtuption. In brief, thr problrm is that thr objrcu of prrcrption srrm to
exhibit at least some spatial extension. but mrrrologically rrductivr rn,-
soning suggesu that only infinitesimal partides, which lack spacial exten-
sion, can be truly real. We will shortly Stt that in this context. Dharnukirti
as Extrmal Ralist has every opportunity to resort to the alternative inter-
prer.llion to solve th:u problem. Nevenheless, he chooses 10 avoid the
alternative interpretation. What is mon important here is that the very
concerns mal lead him to avoid the easy solution posed by me a1trmative
interprer.uion arc prrcisrly thr conerrm that lead him to abandon Exter-
nal Realism ahognher. In other words, WI:' srr a Strong dtgrtt of consis-
tency in the move from External RcaJism to Epistemic Idealism: nOI a
consistcncy of ontological commitmcnt , but rathcr a consistcncy in stylc
or 'CIl$Oning.
Oreyfw agrees mat thc alternacive view becomes rdevant in mc context
of peruption. That is, on the: one hand. External Realism (Dreyfus' " St:U'l -
dard view") oprtatCS on a principle of ontological parsimony thar rtduces
all u1tim:udy real physical cmilies to infinitesimal particles (pttntmA!'I4),
which have no spatial cxtension. But on thc other hand, DhannaJcjni', the-
ory or ~ption mainrai.u that indivi.iu.al inlinitO!$ima.l ~ides are nOt
pm:cptiblc to ordin:lf}' persons; ilUtcad, only "aggieg:ucd" (l4'!'OI4) infin-
itesimal pan;des a~ prrttived, and the mntcnt or appearance in peittption
is therefore extended. According to Dreyfus, Dharmmni dcili wim the
problem by implicitly introducing the "alternative view," whttcby Dharma-
kini concludes mal spadally extl:n<kd, rc:aI physical cntities do exist. Drcy-
fw nOtCii that th is issue becomes most acute in an argument that appears at
PV}.194- 2.1.... although we will $CC mat mc Swvrtti on PVI .lj7<'42 is aho
100 FOUND.... T10NS OF DHAkM ....KIII.T! ·S PHILOSOP H Y
A& ~ h:l.vc: not«l.. Dharmak.ini and his prcdttcs50rs maintain that infini-
tcsimal p:l.nides arc not on th~ir own perccptibl~ (for ordin:l.ry persons);
instead, they mu.u ~ -aggreg:lted- (J4,!,ril4) in orde:r 10 ~ ~rceived. And
since: the: term kim4ll1" runiymaJ") an ~ :l.pplied to :l.ggreg:lted p:l.rti-
d es, th~ objector points out that Dharmakirti Iw fk fonoadmitted thar per-
ception an ~ conceptual, for Dharmakirti himself maint:lin$ dm
universals arc: nca:ssarily associ:l.ted with oonce:pts.
Dharmakini responds that "aggrtgation" hcr~ does not mean that the
particles arc: forming a singie: whole:: rather, "aggregation· refers to a par-
dcular state of those panicles, namely, Thar their proximity ~nablc:s th~m 10
ausally suppon QCh other such lillit they can ause an image in th~ pet-
ttiv~r's mind. He: remarks:
76 PVJ" 9J- '96: .nhlllUrrlbhiu",J,.,Nihi) " " " ~ ~ ,.N / ..1t14l "~'!'ri14l "hi
.;.'iU'!" ju-j---!o /I "~".'!" '"
~ ,,, _'~,..r'" "!"''' I - ' r~.;"'''.Ij
jU-.. ~ "~_ Il a . the mnsbtion olin- ¥tna by Drqfw ('m17-&1).
77 IXomdrabllddhi (PVP:II,b.). commmrins IlJlOIllhe wnICI rileV above. cbriIia how lhe
Illim ~ pin a ddtinaiw: !,>lOp" rt·nwWMw by ";rt\K of the prmimity of II", prm.
ow partida: in.he same Q)<Irinl,l.1.. Hf mrwb (WOIdr from the w:rx:s v.: iQ!irited):
Dw .. " rrt..,... wi,..,Jw,. ""Ift1-i.f .. dIM: 10 II", procna: of COndiriOM which en-
Ilf "'" p'opclry-_Iobrl"" dial iI d", apac;ty 10 produu Ill\ a~ j"fi"j·
1rIi...J ~ INI ""..., the capocity 10 produ« an IWII <not. .n. &om
m..; , • •hu.n,i.1 ""•...,., n• .....Iy. ~ .. inl; .. i•.,.i......1 f"'rrit:1eo Ii" ,r... .. ...., """" in.
uumlmal do ...... ha..... mar apacity. n.., word ·agrrp.lnI ' ap~ m.:- putick.
that N..., their ..spo...,~ apxi.ia which an: lItaincd ....nm lher arc in proximity
wim dtia and mal od"'l" putidc. 1*" pMn """f ,.; '"lIt''' iIrtI phJi, I ""'''' ,., sbn
1'" ~ I"r ~ I" i ",., '" j ",.., khi" Ii,M'" 'j ,.,.,..." ,,~ "" i # I
*",.,"'' 'pm. ,...
", I """,hru pM" Mt *
,,~ ' - k>t", i T1" "'l*'" " .... ", __,.. ..., 1M "..,. ",.. u,."",
Dw' "" 1M J..,pJ- J..., pM" """ "J'I k;-,- ~.,
. . . ,., OO" .. ~ _ _ .._", ...... . ....,."" IfNI IoJ-J HI.
~1*; ,,'"
From ~buddh i 'J commel1l cornes the ~ution that he and Dcvcndra.-
buddhi attribute to Dharmakirti in response: to me problem of numerical
cortespondence. To be specific, the singularity of the perceptual image is
not congruei1l to (i.e.• has no oomorphic correspondena with) the singu-
";.,, ....." t.. /oil .N/oil ~ ~ l "wI 1M u... r.- .. ,,;" III R 'J.u 1M . . ~ I ;. ' ahit'
m .,. .. "i IIlit Ii ru". '*' >ht. ,. ..... I ,.",Ii""'" pJM" "..., .. ; trfI i:JN,. III" I
tit ur >ht. ,. ,.,uI ?i ... tI" ,.;". .,. .. IIi P '"' P. -"I" "'" ",. . ,.,.,. ..
"' I tit III ...itt" ......". ,.,,J,n /II J. ,.,.. Hr ~ .. P'" ""Ji" ,. tU "'" iii fUI hAr ~
i#
,., t." oi !Shit tis ~" .",..,. din I?' III ,• .., ti"" ... p-" .",..", .. i W i:JN,.
P"f,no 116M,.* "",III iii.,.,. """I «.r.,.,.;". ,""" .- /,..",,,,. ill", _,. t.."
,. "'" "i ;J,i" hi II. "",,. _ ,n. _ I tU "'" . ." J.,..,. III.",.; p rriI tis ji Iur
f"j./rr ... 1*;"" "'" pM/I "'" I.u ,.. "",. i lIJI' ,.,uIJ. eM .,. .. J. .,. ,. _
" 1.... pI- ........ -',.~J..r,. -',.MtJ. ..... "'*',.P"C,.. ,..~
""'Ii",. Ji" u I rNlfII r..t J" "'" I.u u.,,. ,.,uI iJi """ .. I PI' ~ " .......
DHARMAKIIlT I'S METHOD AND ONTOLOGY '09
larity of its physical causes. Instead, the singularity of the image' ~"&"t1
with a singulariry of cauu! function : multiple external causes ue produc-
ing a single effect, the image. In other words, the singulariry of ,he image
coITCliponds to the f'act that multiple infinit(Simai particl(S af( working
togmer to produce a single dkct. Thus, the fact of some atoms being Min
aggrcg:arion ~ docs 001 signify any actual, physical unity; inncad. it signifies
only the singularity of their effect. And indeed. this is not a singularity of
all their drecti, since each individual atom n ill produces individual dfeco.
such as iu own suint:quefl[ moment. Thw, as is strongly implied by
Dharmakini and spttincally St:lt«i by Sikyabuddhi," the aauaI objccu of
a single pcn;:cption arc mulciple infinitesimal partK.lcs. Hmce, jfjt is a phys--
ial obj(Cl, the panicular (1JJ/I14~!I") mwt be an infinit(Simal panicle
(pII1'IJlm,i'!'tJ). lind "n "88" g:a'" <:::a nnor bf." nNI"'~~"
In her excdknt dissm2tion on the panicular. ~ mnarb upon Dharma-
kini's sU'atCfJ' oflooking to the nature: of the correspondence bc~CUI image
and object in order to accoum for numerical corrcspondcnc:e.- But although
Kcyt a.ppe::us 10 have a fairly dear norion ofDharmakirti's solution, she~-
89 5«. for eEUIIpk. Kq-t h911o:190). 0rryfUs (1,"-13) adopu Kryt'l lnalysiJ ofpWcuWs.
bul ncilhn ~ not Kq-t ~ addmICI adoeq_dy tho: conmrdloiorl in rnainfZinin&. 00 dw:
OM 1wKI. WI ~fed mlfnr do IlOl form a whole. and on ~ othn . Wt an agrtpl~
;. a fin"", uhimaldy real panicub r.
90 II an bo. arp..d dw in dw:1 .. undtrconoidtmion (PV).woiJ). Obarmakini dos 1\01
adoeqIWdy .ddmr dw: mil. ofUddyolahno', ~ of whoks. rwndy. dial a wbok docs 1\01
in fact ronuln Of erw;omput puu. bu, i, rI'h,n a dwint."l: n"ioy crea,ed by il:J pan.. For
UddyoIabta. il iI mIlS i11q.i. illli<C 10 poinc ro 101M incompatibility bo......... tht ~ilJ'
of the whok and the mWripliciryofiu p.ua. Sec. for 0tlmpIc. NY wn-,IJ) .JNS:u.p.. BUI
Dharmak.ini'. upo<nen' """"' ..... upon bit -t;.,.. m.o,- of ~MJw (PVSV ..J
PVI ....)CdfF. G:69 ••6ft). whlch undttmi .... Uddyoubn ·. dd'enw..
DHARMAKIRTI 'S METH OD AND ONTOLOCY III
n:al RcUist argumem is ,hal, when leveled again$( the Exlern:al Re:aliJl's
nolion of a cognilive image. lhe critique of variegated singularity pointS to
the same problem in his own system. In mon, while me Nyiya.yaH~ib
theory of perception restS upon me supposition of a variegated singularity
[hat iJ material, the aternal R.ealist theory resa upon the notion of a var·
iegated singularity that iJ mental,
Dh.:umaki"1 maireE this point at PV),108, where. ~pealr:ing in ,he voice
of an objccror, he says:
~ I f singul:trity
is not possible in the case of objects such as a bur·
terfly's wing that have vari~a[ed appean.nces, ,hen how can
there be a single cognition whose cognitive appearan« iJ vuit-
ption?"'"
94 nUll il. I>IwnWdn; would no IonF haw _ 10 panic:ulan. the ioIc ukimardy ruI
ml;tia in hi. Iy1I:nn. And wi~1 tudI aa:aI, that- wou1d be no maN to diJrincuisb (.(III.
<><pu tho. rdc. (i...trccdy) 10 ~ ...d thooc tha. do .,.... foo. ..d. • deoc" ..; .....ion io
u1.illUld,.. I:wM "poll the Idie ~/hcx-, thai dwxtrri_ puticubn aIoM..
DHARMAkIItTl'S METHOD AND ONTOLOGY HJ
commrnt:lry. mis V'a'SI: specifically lhows us trun the External RaliS! argu-
ment against variegated singularity rests upon a rejection of any such entity
at w physial levd, for it is pKCiscly this rejection mat makes the argumerll
contradiaory {and, hentt. inferior (0 the EpiSiemic Idealist theory) al the
I~ of the cogniti~ image:. And sintt me External Realist critiques the
notion of a vui~ted singularity at me physica.llevd, he cannot defend the
8uddh;~r view of pr.IO:l'rion by f'M iring ~m~ ~ingl~ ~nriry Ih::If is an ap;g~
gation of infinites imal panicles. Keyt is thus incorrect to claim that
DharmalUni's argument against variegated singularity requires him to
accept an aggregate as a single particular.
Iu we conclude ow discussion on parriculan, we should now confidmdy
be able to answer me quc:st:ion mat gives me section its tille. -00 pan icu-
hrt 6.w: ~f>'ln:ol <':l<tomJi.on? ~ If Wl." I'C'ttncr ourwlw:t ( 0 Ohum:Udni's own
texts and those of his earliest comment.aton;, our answer must ckarly be
-no.· And sinO! only particulars ~ causally efficient, we an aI.so conclude
mat any entity mat arises from causes and produces effects mUSt not be spa-
ci.aLly o:tendcd. This point. however, raises a qucstion mat we touched upon
earlier: how do we interpret th~ pa..uagc:s in which Dharmakini appears to
.pal< ... if <=<Iended e nlit>e. such ... . molt.! 2rf! produced by c:auus :&nd " lIVe
effects~ A complete aruwer to mis qU$ion would rcqui~ mo~ '~ th:rn
we have here. but in brief. we can mum to me passage from the Sliflupti
mentioned earlier (PVSV .ra'PVI.t}7-L41.). lbal is, we should simply under·
stand these: as cases in which, for convenientt, Dhannakini speaks of, for
enmple. -me causes and dfecu of smoke- ntther than - the common ausal
origin :rnd potcrltial for dfccu rd~ t to multiple infinitaimal panicles
that, when consickred in ~nru: of those common co.usaI chuaaeristK:a, may
be called~. ~ Indeed. as we have noted Dhannakirti occasionally uses the
cerm Maw ("ming; as:rn expression for material things such as mlO~ lim.
although extended, an: actually reducible to unextended p.:tniculan;.9!
2. -I Un;WTSI:Ils ·
96n.., inf£l tll(U w., IUI~ ill thtitobjau" rYidcnllTom PV).I-). II - 0 11 ~'I
'P.o:<nmt (ciu:d by I)I...nnaldni in PVSV .J PVI.I. G :1.11-Ld: _ .....,. .. .......,y-
1U<-,.",.11IIith. h ' fUJi'~ ~",."M.*- Noel du. hcK htUJ". rmn ro I
conapc.w cosni1ion. S« abo 1M di'''''Miont al PV 4.80 ;&I'd PV4.176'.
'17 The.- conciIc lIilancnl of dw ~ indficicncy-q.nd hcna irroli~ lUIiYCr-
_ .. '-'dindws,.".pn(-'PVI_I66,G:l+IO): . . . . _'!I~_' 1.~
~ . _ 5« "00 PV"'- J .owI .... Im""n! met .ph<>.- of ,he: aonuch (PVSV . .
PVI ,2.lG- m . G ,loti.l7-IO?\I: Ir.. ...t1lN on chape .... 4. , 10).
DHARMAK!RTI'S MET HOD AND ONTOLOGY us
m:a.inwns dut we USC concepa nm: limply out of some pernicious habit. bm
rather with a specific purpose or goal in mind.'" We might, for I:UlTlple,
seck [0 heat ourselves in front of a fire:. and ~ might then usc the concep-
tual knowledge: of fire acquired through an inference in order to obtain a
real, particular fire that has the capaciry 10 fulfill the tdos (art"") that we
seck. On this understanding of why we usc lingwn ic and conceptual cog-
nitions. Dharmalcirti is obligM to , hOY! how univeruk, ~ though unl'Cl..!.
can nn'Crthdcss yidd information about particulars. That is, he must show
how our words and concepts make seosc, even without the prcsc.nce or
"affirmation" (vidhi) of any ulcimatdy real universal.
To gain at least some sense of how Dharm:akirti dcaI.s with the problem
of meaning in the absence of real universals, let us brieRy uamine his
JljW.:w-theory. which for him ('Xplairu how we arriVf':1.I meming through the
amstruaion of universals dut are rul only conventionally (not ultimately).
We will begin with a summary, and then we will also inquire into two
aspcctJ of his theory: the nmion of particulars having the same effect, and
the question of whether conventionally real universals arc permanent.
FinaJly, we will raise some problems that the 4PO~fheory must address,
and we will answer them by referri ng to three senses of the term JlptJIut.
Overall, our aim hen:: is 10 raise: Ihe central issues and problems in (he
apoha-thcory, and to avoid surpassing the scope of that goal, we must fo~
any detailed o:amin:nion of the other analyses by modern interpreters,
despite their importance to my undmt:lnd.ing of Dharmakini's philosophy
of language."
99 n... work of Sidcriu (a~s. 19511 and 1995I) ... '-n Qf>KialIr hdpiUl. dopil~ my dis.
"V-n>Cft• ..,.;.h _ _ Df ... - I ........... 'lVlU1e f'ocu.inll p,i_ily on !an•...!.ti•• and
Kamablib. Sidniu iI MYCnhcIe. quil~ 'ist" 10 nOt~ that mud! of thn. approKh I<) W
•...,.w.thtoory is 10 ~ IOund in DlwmUini. apcOally u rt:ad mroup. thr: ~nwy of
$~. 0tMr mockm 'II'Ofh of impooutlCC to my inlapm::lrion etperiaIly induck
1'- ofTIllnmnt Crml and Ka~ (1991).
100 PVSV .JPV,.6I-7f. n:and..la! in m. appc:ntia. «NIWn. I numba ofw ~II
dl,cwwd in dw remainder of !his chafm:r.
101 On Ihr Ihm: rypn of unlvrrul.s, ICC PYSV . . rvl.l!ll (G :91019ff).nd rV}.f ,l;d:
"1M",.". "j; ·'In,. _UlIthhiWM,.,~ pJ/s-.rJl. Conc:aninc I'CICIDp1ition ~u..).
DHARMAKIRTI' S METHOD AND ONTOLOGY "7
in the next chaprcr. the range of dTccts that the particular is capable of pro-
ducing is determined by the cawes from which it has arisen. A particular's
uniqueness thus amounu to the faa that it has arisen from specific causes
and that it therefore is capable of producing a specific or restricted range of
cffcru. '11 If we COlUidcr an image that arises from what we would call "red,"
that i~ (a mental paniOJlar) is unique or "excluded" (1IJd1J!TUl) from :ill
other paniculars in that no omcr partiwlars :uUt from exactly the same
causes or produce aacdy the same effecu. The image, being the unique
dfca of the unique paniculan that produced it. thw setvts :u the b:uis for
acluding the: imagts produced by other paniculars.....
The: f..C(. however, thaI each image acludes:ill other imagCll by vi nue
ofiu uniqueness is not in itself adequate to account for our usc of concq>U
",n,! I",ngll"'se wr: require '" nOtion o f J:lmencu. ",.nd n O I jU!t d ifTe:lellOt!. We:
must have lOme notio n of lameness bc-causc we ntt<! [0 :account for
.nlNlJ4. the "repcarabiliry; "distribution" or "continuity" applicable to
any cognition that seems to rder to multiple iruranccs. The conceptual
cognition of "rcd," for example, appean to assume a "redness" that is pres-
ent in multiple irutanccs. and in thil smx the concept of-red" has anwrya.
Dlu.rnu.ldni mainninr that in order to conrrruct the nmc:rleu required by
.rn""J4. we pia« "Iim iu" (av.rJh.) on the c:aU5ef and efTceu upon which
we foo.u. In other words, we have apecudolU in mation 10 what we wish
to obtain or avoid, and our concepa-which arc indininguishable from
universals for Dharmakini-are constructed in relation to those ap«ta-
lions.... In the case of the concept "reel; some set of interests or other
DhutnUlro', ""'* dIorou&h accounl "found In I'VSV " " Y' .9' 97Jb ,G~s"'6-1 •.I!I.
n.r notion tN. an imase iud(, brin& tu>d4triOO.rd. annor be, J ...nm....J it ~~ II .
for oampio:, PV,.16j-.67 and MY "'PVC.7.aI (G:40-)-.1) •• ruuIamI in the appmdi;L
TI.t notion that ~ Qltitia may be, conPdomd thc AmC in tNlChey JR all eqcully dif-
fm:nt f'roon oWr nltiria: oaun rhl'OU!houl PVSV. 5c'C apm11ly PVSV MlPVL7J (uans-
"ted in thc ;oppcnclm).
IOl5c'C, 1Ora:ampk. PVSV "'PVI.I6(; (G:l4. u.~j.l) and thc nGl chapen (I',), wht:.., ....
;.....io~ .
• 03 ".., norion tNl ...niq""...,. of J>Mticubn: it ... ltilllJldy the ~ 101 CONINainl uni·
¥aaU rhlllUl!ih adusion is aprcsscd ill, for cx:a3pk: PVSV '" PV' .70 (G:}I.•?if; _ thc
appendDJ; PV'.71Cld and PVSV "'til. (C~9.1 6ffi_~: PVSV "'PVc.&! (C 'JS.1-Jr.
and PV)..I69.
104 n.r ~ role of ap«ettioru in cht ronwllCiion or univcnah it indiattd by tht
~ . . of thc ~ dbi_ r apeacd1 wid! rdaion 10 dlorClIIk'I and dkcuon thc
basis ofwbldl J unl¥tTJ&lls CQf\IIl\Kttd. oS« I'VSV . . 1'\'1.9' (d ied earlkt. n.98). $eo- alto
dw: ~rma 10 dbip~ in PVt.6I-70 and MY '" til. (G:" .'; _ thc appendO: fo. a
III FO UNDATI O NS O F D HA1,o.tAKIRTI 'S PH ILOSOPHY
oornpIctc tnrubtion). Tlw: notion of dIC ncpiVl: - limit" (,lIwJllij in oppoaition lo wbid! ~
"elution il CI)I'lHruClrd appnn to OClaIr only ona in dv s.-rm(i.~, PVSV MiPVI.I' j)•
...t.eR DlwmWni rmwb (G:tl.J.H}'):
'J"hndor~ it iI CSlWliahrd that aU apralioN and COIIttpU ~ a d wio.,. (~)
u mnr objtcu. bon lhou!h 1N:s.t CG-rJemui&l aprcssio.,. and ODn«pu.:kpmd
upon lhe _ rnI,hi"" ,hey ~VI: di~n' obj«u; lhey haft diff'ercnl ob.iecu
~ apt->on..-:I conupto or~ ~ (wf" ' ,... ' 1, r",",-) in teO/'1IU of
acJwions .ha, appnr in copilion all if they ~ o:llffi:rcnl; and .no..:
CIIdwlon,
appear diff'erml due 10 those: adusiont· diff'~ tilal an:' eoNIn.octrd \hrougb die
dilkiCi'ICCI in thdr dditni,m (• ...Jhi). [uuM' siJJh.", twI urw yUJ viwbllilql
-i1u4JI ('ff 1 M m ~tiYnI~.,. ~_ .-'hijJ"J.~;"'i~ ~~
Mt;,,1tt1" ;"'P"';~" r,Yhs .. .,;,orffpi,.t.y..... w.;",...~ .... ..,.
~i (rvr:1,~PVT..,PaH _ K:}47. ~jO) drfina "ddimilcr~ (nMihi) u follows:
lui adU;SO(ln ', dduniln IS IIIaI from wluctl one emblu"," lhen: 10 br ~ adUSlon,
For aampic. die ddimi'cr of die acL.uion a1kd "pmdua" jo "I\On,produa. ~ Likt-
'III'ik, I'ot dv cxdwion drfinrd u " impcntWlCfl«" .M delimite. is "pcrmaneD\".C." and
on on {i.e.• the same .. appliabk 10 all odlcr COIIttpUI. IJoI PJbt"i. ,.. ".,....u!J".w
Ii ,.,., ••Wh# I Jf'lhi -'1vM' V-,. .,.~ I"v", .,.;".,..t,~
,?-,,,,,,,,,,,,u,. j~ .
I ~ TIlC ~ tt... o:rWn !:.lIil;'" nuy bcc:ocuidercd nondilinull bccawI: dwy arc &II differ·
6 ...... .oJI...u.:. cou...... crnphui-I .. lCV'Uoi pi-:... indudint; I"VSV .Jrv..;ryd {G ...... t;Il),
CIl<
MV MiI"V,,9!Cd (G.......l and csp«iaIIy in I'VSV MiJ'VI.IJ7-'41 It..e m.
appmdix).
OHARMAKI RTn METHOD AND O NTO LOGY
".
rh:u n~rin n :appl i~ In :all I.he inst::llnces in quesrion: :all me instances in
question exclude wh~t is not red,
Through mis :approach, Dharmwni arrives at a theory of universals
(s4tn1lnytllalq.a!'4) that requires both the image :and me exclusion. That is,
$[rialy speaking, a universal is ~ combination of mar which is nor dismb-
Uled (i.e., lacks Ilnvaya) :and mat whim is distributed. nil: image, as a men-
Ial p:trricub. , is nor distribtlt~ . bu' ,hI! 6dLls;/:>n (IJ)I"JvrrtiJ. :1_
' a n~';n n
applicable 10 all the images in qucstion, is distributed, Lacking distribution,
the image alone cannot be the universal. But on Dharmakini's theory of
qualities, a negation cannot exist in distinaion from that which it qualifies;
mef(:fore, me negation alone also cannot be the universal. The universal
must merefo re be an image mat we construe in terms of a particular ~
of neg::ation. mundy, the exclusion of th:at which does nee: h:tve the expected
effects. As we have secn, for us to apply this negation to all the images in
question, we must admit that all the images in question have the same
dfttt. Let us c::umine this imponant issue in grearer derail.
106 FoIIowin& OIwmakIlti, ~ will5ptU; of"'hinp" ( ~_) IUCb III . "w;ucr-jUJ" rp.,u j
in. ~ Iha sIoaa OYa"tM lOOK prcciK 1R:almcrn ..nutbt tIM: nuclo»c:opK - '""ius
iI in faa ncxhin, bul mimwnpic infini!t$inai panicIcJ. M ~ haw alrndy M"nI/.98I1l. !hit
ptooedun puu,. limpl:il1a DharmWrti '. aalt. in tha. M ~ no! m(. GOnJtandy 10 inhn·
ilC:simal pmlcks.
107 ..... though DI>:armUl rti tpcQfia1Iy ditcWKS 1M C'ONuuaion olllnr.tnaJs in lmN of
mfiria linin, tIM: AmC (ypcIof "*,... b.r: tmds to foo::w upon IaInm$ of dkct. PM '"
thai mmtion both wa)'l of wllJInKlm& u,mcneu Include !,\lS\I "" t'\lI "}7-141.
(G:A .I.4-69,l ):
110 FOUNDATION S OF DHAItMAKIRTI 'S PH I LOSOPHY
Somt;purlculan haw tIw GIIV cffea. In order to aprai INI thry haw that dI"ea.
onr: ~ thnn.nth ~rCS$ionlwchu -.mct+ilIIi" pt u. id«! thai tM Ii~
"'.oed.
............ tiuI, tko, JV""'1l> ...... 0. . . . . . . . . . . . ' - " "......1 t",;c.
...royi tIfAtA parricubn in <hal &Won beeN·... dwy _ ~ t'rom purio.Ibn IN,
0- o:ut liodl-
are Othtt than thnn. Liltn.iloe. in tennl of havins tnc AII10e cawc. one can aptcll
... tw ia naru.i~ ..ith a "n~ ap........ in onkr to fxili t.tc prxlial aaion.
F.Prnpks _ "H uo;f",d. " "Jcrxy c-," or ". ~ an.e. irnmuIiatdy after dIOn~ or
".ound ia cawaUy producul" tJoolJ\&i~..,. utU.,. bllliJoi.,. _,.,.Mc~
p.jMid.u.jfJ ~ f ~Itm~ • ...,."thu~ttW f
,,"1M r'4"'':rt N~ ~""N1"Url]il~ M~ irua _ 1.
Other aampla include: PVSV .d 1'V1-40-4l (G:II.It-l) : .,. ~ IJMM ilili I.,.· ,.
dt~ ~ ,....,.,; ,-~. ,...., ,.. ~ ~ I w.s..w-;..
lxausc: thty allle.td to th~ sam~ judgm~nt . then one can also gy that th~
individuals that pnxluttd thov: cognitive images in the fim place arc all th~
gm~ because thty too produce th~ sam~ dfcct. Thm, th~ wam.1\I for th~
sanl~ness or the individuals is that thty produce the same cfkct: the cog-
nitive images. And the wamuH rOt the gffiCncs.t or the cognitive images is
ag:lin that thq produCt' the 5:lme effect: a a nain ~ or judgment.
/u ~ hav~ dcscrilxd it so rar. this d)(ory leaves itsdr ~n to an obvi-
ous rebuttal: what warrants the sameness or the: j udgm~nts ~ That is,
Dharma.ltini', initia1 problem is that individuals arc unique, so the same-
nc:ss required by language: and concrpu must be accounted ror by samellCS$
or df« t. Bur ir he tUfns to th~ cognirive im.ages produced by those indi-
viduals, he has the same problems because those cognitions, like the indi-
viduals them.sdvC$, arc unique. Ir he now turns to the claim that those
cognitive images arc th~ sam~ b«.ausc: they produce the same judgment,
th~n h~ appean to havt fallen into an infinite regress. In other words, it
would appear that we nttd, once again. to warrant the g mencss or those
judgmenu by appealing to the sameness or Ihrirdfcca; and or course:, me
wnmcss or the judgmmu' dfcca will once again tC'quire the: same warnnt,
:;& nd JOon.'"
Dharmakini's response: to this problem is c:xprc:sscd. ir somewhat ellip--
rically, in his commenral)' on rhe verse: cited above. Nore mon here he uses
the metaphor or an ~ overlap-or ~ mixing" (141!U'l'1.4) or individuals whereby
the nature or one is somdJow partially prcsc:nt in the nature or the other.
For Dharmakini, such an overlap is impermissible in the as<! or causally
effickm things. since causally cffidem entities are paniculars (or arc rcdu-
cibk to panicul.ars). and thl:)' must be unique. With this and other such
issues in mind. he co mm~ms on the aforementioned verse:
pJM" 1M I.t,. i .,. .. '" rtyitI tItnt ..",,.,.,, oi "Y''''' ...... ,.i -p",.,.J,.
rAn 1Ihit ....... ,., """'~ / ...... ,., mlffil"; ~,.,. "1If1i..usJM" """ J.t,.i
.or - Mo,r,.. .. fI.fo ..).
AJ it dear hcre. nwMnIl'''''' is I2km to ~!ht: pmicu1u (_!.r""~ -HNM iI bctc wed
moch :II -~ jn Gjridlioill such:ll ~ ", JWticuIu kind oftfft," or~', par'
bctc rakm 10 ~ JynOIIym of ~ 11W iI the inltt-
Dcular tree.." In short....1IhiH is
pm2oUon takm by Scftnkdlfter('97I:'90 and ,-;",j. Hownft. since the mon:alfItttI of the
diW' ..ion fOa.-on iIwIIIu diliadK%. andNnu the next 'I'ffK ItrtSICt ROC!ht: ~
bul.hI. unreal ad"......... ~., inleljri'ftalion doeI ROC Re'JI WUrlllled heu.
II) rvsv a'rv','Q7 (G :16.1t-S7.7):
"iwtli_ NUI,.thl """ MiMI....,. "",WM_1fUII7P u;ri 1""'" Ill~ 1w.tIJj,i,
.......,;, .... 1 uI!' ", WWitu/! ~ ....~ I';u!;wNvw WM_~,;
"Mi_
_."""_ iii til / ill , . qb1I iIwIII I't} ~~NII/.t!I AwJtri'd.cr)or u,.qW
"1Il1.hi""'~"""II<t~ 1 /iJ .,; p",;J,,,,,,,I!' M+wWu", ",i ,,.q,,.;u,~'"1-
...."""'.w,.w,.,.JJd• ..Ji1N hmo, M.NJ dJ,;"""I!' ft~ri / 1llm.~Ill"""'''-
_,J.d..., ,,~Mi.si,.. jU~,,~ hmotf1U ",.It,.,. ii lIl'fU!1f4U"'I!'
...H ...,,11. hI' __ I -,. 'M'~ '. ~ ,...,..,..,.~,.,,?....J.r.I ..J.u... ....,
l uu1JtMi ~w.. Mth..... ,,~
DHARMAKIRTr5 METHOD AND ONTOLOGY
'"
COnttm~ with thtir coloration. wt igno~ that diff'tttnct in light of tht
samtness constructed in terms of color. And of courst, our ust of "applt"
and "Str.lwbcrry" is dtpendmt in pan on our habituation to certain lin-
.. .
gwsuc practl(:cs.
Whilt Dh.atmaldrri's appeal to o:pc:ntnct and dispositions rdltcU mt
mind-depc:ndent aspects of the o:dusion procas. it is coupled with somt-
rhing mnrr:: an lIp~llo Ih~ narure {prllftrti or WIlhhtivtt,J of thinV lh~m _
sdvcs. That is. when ~raI individuab produce cognitions that in turn
product the Amt judgment. "This iI red," it is not just my own apcaa-
tions. conditioning and other relevant dUpositions that go into the con-
mucrion of that exclusion. Ramer. beyond my own subjectivity, the: entities
in question., Ihrir ""tllrt (WIlbh.i1h%1/IfI. prllJrrry4. ttc.) product cognitions
wf,"'<e C(lnrenr i~ (:lIp" hI ~ (Of hr.ing mnsrn,,~d :15 " r~ ." More rh"n 5Om ..
appeal to c:xpuienct. it is this assertion of the nature of mings that putl an
end to any infinite regress. We can pose the question. "But wh] do tbose
individuals all produce cognitiolU that can lead to the same judgmcm?"
And Dh.atmaldni can answer, "Bca.usc it is their nature to do so."'"
Some interprcws ll1lIy fed rather dissatisfied with Dhannakini's appeal
to nlllUl(,. In cfk.ct, he i.s Qying um wh~n wt can all :ill red things "red;
for eumpl(', it is not that they :ill imtanti:lt~ th(' univerul "rcdneu"; nor
that they all possess some real. specifiable similarity; nor even that tbey all
have the "same" dfca in a way that 'Nt! can ultimately specify in objective
temu. Rather, all those things arc JUSt different from non-red things, and
tht reason for meir diftcrence is simply that ., thrir ""tJlrrmcy appear that
way to us when wt attend to what 'Nt! mean by "red." Even me seeming
objeeUvity of thu appeal to mature lll:IIy diQPpoint some, for:ll we will see
in the nat chapter, a [hing's Mnatunt (SVIIhhtillll) is also ronceprually con-
muct~ through the RfHJI»-dt4Jry. On this interpreflllion of what Dharma-
kini means by nature, Dhamukini's talk about the naturt: of things that we
call Mred" is besl understood as a way of saying that, in ultimate tcnns, there
is no mcuphysically defensiblt reason for the fact that we call them "red. M
•• " 111;' lppn1 to the natliK of tninp b«oma I~Cnt wMn, in a rfl.alni (Ontc:..,
Dharmakirti rcmuica:
Indeed. it ia /lOC COIf'C'Ct f,.. ... IIrlMui) 10 qllCltion r,.."."""".; <he nan,uu of dUnp.
II in -.;cIhy doer fire bum? Why ia it hoot, and water ill DOl!" ~ should jllK ask dW
m..m, "From wn,.t cawc don 1 dUnt;...nn thia Rarurc comc?" IPVSV "PVI.I67ab
",,,,,...-.v..r .. .,._.,. ,-,.....,.,.'"
(G :....1,...u):
1M 1 M'.", iii I nhtft twlJit ...... "..,. _~ iii 4.
*1M"
MIMtui ti", "PI' ~
116 FOUNDATI O NS OF DHARMAK I RT I'S PHILOSO PH Y
they did nOI clearly cxplain how DharmakIni would avoid his own criti-
cisms.III Fortunately, an interprtt.ation lwt:-d on Dhannakini's tCXtS alone
relieves us of any tedious n«d 10 confront Dharmaltini with his own
altaeks, for Dharmakini clt:arly rejeas the nOtion thai the core of his ver·
sion of unNcrRls namely, cxclusions--are petmanenr:'"
A ",.,,,tim. IltH"n ""tillY: IImu. tint' CIZ""nt amui", nJ;t lIS ""u-
int )trJuranCt' ~ Dr ~nDn-ptrJuranu."
[PVl.I69abj That is.
there is no such thing u all as an ~Olher-cxcl usio n ." And con·
cepts of that eJ:c1usion's ~rduring or ceasing by its nature,
which would follow from it ha\'ing a nature, do not make $Cnst:
(nA ItlZlpllnu).
remaining after its instances cc:ase, fe r the universal is also not pemunem,
and thus cannot pc.rdure. We have scc:n, however. thai Dharmakini admits
n.., <hen: is. purpu.c. To ~ opccifoc. "'" Mis (,.a.J I" t1N-ffi'"'.' i ll BIIPis.
M tn_n..u for ,,,- 1M ~.IIS (i.~.• f1"Iiu" "'Ims;.ou fWiI /U "CftI1IItII·.NI
,..11m ~III ".m ......... i ;, 1Iu,INf-wr ,r«U.Jn ~ Jiffrmo,..ti"f r-/j.
rin (~. wink IN I."" J.n 11M. T'bU Jijformu i ll _"M r."_~,, (~
r.ttforwu u IN tq«IIUHN.J 1M JInS'" ....". ;,,. ~ , ,.,; IUInIInII [PVI.'I].
Wbm. ~ [i.c.., aliicencrJ who Iw II() apccolioNconcaning the p«ICII« of
OIhttcadusionswishs ro know.-hnhc Of noc 1M _adusion ol'horw' " .ppli·
abk ro • $p«ific sro- obj«t, 1M $JICUcr, in otdn 10 inform 1M listener m.1I thaI
It:ind of mft~ odUJion ill appIObk. rru.ka it known wid> an aprtaion fOr which
the- XIIWIlK: COIlYUlUon ha& b«n consauc:rm in thai bshion.; that is, the ~ "")'$.
"N",,·honeneu pcnairu 10 this" (.!WcwtlWllllU]MJli). On 1M..mer hand. when . . .
~Mf who is "", wilno..l ap«UlioN c:ooccrning acluaioru oth« than non ·hone-
...,.. wi.1oa 10 know Wt Wicabi1ity of Ihe adwion of hone, then lM JpcUtt
emplorl an apresaion for thc adusion ofhone: COf\SII'\JCd in I manner dlat does noc
disalloorl)lhn caduaiom: in nnJer.o....uthar ob;m: koown in duol way. T1w is, lhc
apukcr iI)'S. 1'hlI is I noo-honc: Haw:c. sina in d~ fomxr C&K the \IX of lhe
apreNion diAllows other differmca. MKh an cxplwion can be wrd 10 apml Mi·
tMr c:o-mel'ftllialil)', nOI lhe qualil)'/q......tifird relarion. For cumpk:, om ealtn(t{
pl'Of'C'lly "'y, "lu arwnen ill whi~: A prc<iicalc-ctpfUlion such II "CO'NIIOI" 01
"hono:na." ClllII(l( GJlft$I c:o-ttrmnti;&[jl)' becaIUC the imace in I~-' ..WIll
.. 1M qualiflCaOOn oilhe object lIT the nine aduoion, °non·hotxDcta,· does noc
appear 10 be the: locus of Other pra;iia,,.. by vimao: of being .be JIm<' . . the: $Uhjca
il qualifin. And pra;iiale cxprcuioru &Iso Clllnoc npr_ thc qualil)'/qualimd rei..
...... \ ~ ........ ;. .... _,.-...lon 01 ......... _d....t-. a... in ..... _ nd Iolnd 01
a~, rwndy, a wbjca-apreuion, thcrt is the- poMibilil)' of C:O-"{UUltiaJicy
and the qualil)'/qualilied malion. A Nbjta-aprnt.ion may indicale c:o-~enrial·
icy bcc::aux. in acrord with the ..:mantic COIIYUlrion JO¥mIi", thc aprusioo. it indi·
ala thallhc q.....!ilia, whic:h a.re aduaioN, a.re conjoined 10 a qu.allI)"poAaIOf or
oubjccL Sin« the .... bjca thus -=nt 10 boca W1iury dUng WI ~ nriouo puu.
it thetdQrc appean thaI way in «>snition. A oubject-aprc:.ion QII abo aprra the
qualiry/qualified malion bcawt; OM apecu cxhcr adwioru.. IIrllJljtutillM ..., hi
..~_~ ... "., I,..." _" ~ l ,.tIM hi I ~1IlInJ'~.ri~""riJtn4 ..
UJ-~ / ~~iu.rwIV...,,-o.;-!t IPVI ·"Jr '"~"·
,.IU ...J.",..?"_ ...J"ItM..... ~ ,,'!ffU1Ii/qtt ....~"..,.,.". j iju..u
"'tlMMti"iu,.~,!, ~~ J.Unt.".., ""; .. • -n.m..,.. ."..nri
1)IIfIitI "....~ ""_i,"J.ilUMl"i~~ ..". iU"'-
ialNtri •• a,m,,",.~
,u""," r.ur.iMiN.,-.n1ltwk ~u,. ,,..,..A~u '-n. ~,.. iri f .", nc
,.,.". ",~~n,.U.il'ilJllj ... 1hthMhiJur.~,!, .uq.~.
11M" N I fMN'" ..,.1tJtI._
iri I "",,,Jtn~ iwIJhn 6"'1l]l rlt.·
twd"..~lr 1 "inltlu,.tak tw 1 'ririyr". MIt ...ri I """"" ,.~,,~
""""If)
,.~'!.ttJ--i'" &1..,.,..;, .,;~ ...,. ~.t...::rnw ,.....~ ".".".,...... 1
.,.._ ... ..",~,a,.tNc tw 01·
~i !7.f,J7fI) and ~;n (\.49.uH) confum that in the obow: ~, the
aample, ° lu COIlO...,.. is while: is mun! to iIIustnlt an absurd statement• .sa, alto
M:onoradwwKlin (PVV .. rir.), who rcrnarkI thai in this CIfIe -In INoI cue [i.e" when I
prcdicale<ll~ is to be LUcCI], ON: ... ys, 'lIS CO"h'.,' bvlOM don IlOl fly. ' Iu wwu ... ,
is ...rule~ [""',.,...,...,.. if) WJtI,, 11ttI "',.,...,Itf.,.
Mt._ itr'j.
DHARMAKIil.T1·S METHOD AND ONTOLOGY n,
a cognitive inugt: or appearat\tt:as the positive content in conccprual cog-
nitions. Can we not say that mu image u eimer pt:tnWlent or imperma-
nent? Dharmaldni responds:"·
An im:l8'" in " con ..:rpn.o:al cogn irio n, w h,.n co nJtruM :;oJ qu:al if1«1 by an
cr.dusion, seems to be an entil}' repeated in multiple insCll\Cel. Coruuucd
in thaI fashion , even the menw con.tent of a conccpt'ual cognition-thc
*
pb.lIlMtfU Ow" p1If]lll,. iii "';i W. 1# . - . , .. ...., . C'4II W. ,..'" """"
..,Nt ... U',.~.,.,., ~
*
n", lImJ"]l1I .. I Nt... ... """'"
. * ,.,.,]N" ,.,i1ft
_]III IN. PVT·"
I /J M~ Ul ";..,."-.1 ..".,,...,.; {!
Thru W..".o!Gmstnti"IApoha
O ne can perhaps already set: mal Ihe interplay ~n image. adusion
and particular is crucial 10 me apo/Ja-theol)'. bUI despite ml! importance of
~ dU'~ :upecu of the I!XduJ;on proceu, Dh:rrm:lkirri himsdf did no,
offer perspicuous a plarullioru of Ihe precise roles thai they play. The task
of delineating the apol»theol)' in ternu of these th~ aspccu was left to
the early commenl2tor Sikyabuddhi, whose a planarion became a san-
dud for some later Buddhist scholars., A Sakyabuddhi remarks:
Cona m ing thi' theory. there are threoe kind, oP" other-adu-
, ion, Fim:. there u the t:l:duded (1IJIi,,!,11I) partiwlar. because it
is mat from which an: o:duded this and that other [pan:iculatl.""
"'~ ,. tf7iJ II. 1M" pM" 1M fM,,. ~ brU7I,. 'i 'l' __ 11M M IsM......, ... ; ,h,ir
" 11J4f,.;.,. ",n.-", ... J.s ~,.. ;~,,. I jn JIIUf'"
M'i "",;" ... 1M
M fur ,.",u",Sf'1"j ",rsJ." "pJ-,., t - ,.,;" .... ..,.
126 'IlUI i$ _ noubIy w ~ for SlnwUJiu :and lUmaWi.... Thftr a!XOW\l t/ W thr«
lOnna of.".e. appean 10 be: t.ued upon SiJcyabuddhi'. inlt'p'tution (5« TS:1oo1o-1009
and TSP -' nt:)90-)91). a . Ortyfw (Im ::}SfI) and Sidcriu (1m),
127 pVfoS rads, "mm: obj«u thaI = .. :
128 'IlUI ....lmtt IX*' • kw philolop::al problmu. The fine ptobkm ODCK'UN HI ..., ftJtJ,
,.-.,.,.4
which .....,..J,d "'118"'" :an ..m.....,..,.......mo. 0 .... u..-,·
-.I.d, h.o .. . ... __ , ..... odd, .,<>OC _ -w.I be ..... ul~ ' 0 _
in .he orisinal SaIuIcm. ",..
. ro.-. Ur.:r'" .....,
cue. I ""'~ rhw cno.m 10 inc.nprc1 thiI u a puai.... co...uuaion. Thc probable Suukril ror
mil sJou of ."JI'-" would ~ oW:: ~ _","."," .,w.,."(nQ(~ tN, thiI con·
flNCtion docs no.: !'alliin: a plunl conjuptionl.
The IC'COnd pbilolof;ical p!obkm iI tht .bUm.: ,.. ( 'Ji J.s phtI" ""'" pJu• ...J. Althoup.
Ihil radin, iI c.onfinntd by me Olber editions of tht Tibc:an ten. tht ..... ti~ ,.. iI fn:..
qumdy confWcd widllho: p.utic:k J., whid> would htre lUnd lOr a kK:am.: ~ mdill&-
Such. n:adinS would pdd tht Sanskrit sic.: lIS",;""",".""", _p,""'u. Thil sJou iI in
I()me _)'I ptcf~. abk 'i) me ""'" fil~ Ililo'o<:. for II roon: dc:arly UitIwI mar me p.utkIIW II
tht NsilIOr .be formu1acion of adwiont. ThiI bner sJou aIJo rdatel mort ca&ily 10 tht onE
I}l. FOUNDATION S O F OHARMAKIRTI'S PHI LOSOPHY
:dJj.., P"t nr,... oi "'" q." (P: "'" -J J.r P _1ItI -'
1M pJ,. till'" JUt
'" ft#fJ*' "" ~,."... "" pIMf,.yi""" __ rM"""'" _,. ~,..,
nut",
- ",. i ~, "" I" ,,,,,,.,,; 'Jis ph- .J.., ;,nI,.;,;,,;, "... ,.., ".,,.; "'" t.
,_"," i~'lI fP"f""" kif --ft'S ... ih;.J,,"~" ttyiJ ...
tiM ~ .. iIi,.",.. 1M!.p1 u,..,.,ti",Dba "]iJl4r,..i.a-, ... .". _ "?*'
'*",.
,. M",;"", II I M;..,....,.;4. Ow".,." ... M..w,,. PI" oM..., Dro,u l JIM
JM,. ,.,J ,.; ~, "" f ,..", ti IIUJh.,. ",..,.... , . P"I',." ,. J" """ M I.s pJ-
",... 'J.UlI l
Ub..,I..".... ....,... b\... ",' .4."'"....,
,..,dhhJ~""lIuW~;MoIt.i...-M ... ~tlIJdk!i;pi,~~.
_ I Jj ... .1UbJI ... """.+6.1-.
.,.~ /,..,
concept ~fire" have any rdarion to me pan icul:u s that we call " fire "~ O n
Sikyabuddhi's interpretation, the claim that the panicular is ~ th at from
which all others are excluded" jusrifies for Dhacmakirri the relation between
a cognition that involves a univcrsalluch as the concept "fire" and a spe-
cific instance of a fi re. Thai ii , the univma.l "fire" rcfen 10 all fi res because
all fi res are diff't.mn from aU otha entitia thai do nOI have the causal char-
acteristia o:pc:ctcd of .something we call "fire : This differen tiation of all
fires from non-fires is not a mere whim, fo r it is basI!d upon the uniqueness
of each mlity that we call a "fire": each one is actually differcnl in all ways
fro m IlUother entities. Although we consuua a universal by focusing upon
only .some of thO$(! differences, Ihis docs not vitiate the IX! that Ihe differ-
ences we do focus upon arc basI!d upo n the uner difference of the entitia
in question. ThUl, since the univcrul- which amoun ts to a sd ccrive dif-
ference or acl usion-is based upon the utter difference of each entity. the
universal has an indirect relation to thO$(! entities.OJ'
(2) Distribution
This second problem concerns IlIfIltlJll. the disuibution or continuity
that is nC'C"Pry for a universal ro apply ( 0 muhip\e instances. Here the
quation is: if there is no real univerul instantiated in all instances of fi re.
what sameness could acco unt for the faa that we can refer ro all of them as
fi re? O n Sikyabuddhi's interpretation, Dharmakini accoun ts for disn ibu-
131 TM nocion of a "Jdc:ai,(" difktrocc orcxduaion can ~ K'CfI in, for exampk. tM foI·
lowing pN " (PV5V ,.jPV' ,9' nJb: G,..9.1'--t9-1J). which rcfft1llOW di~anIOIl&
uhim.o,dy dis.,,,,,,
minp ~ "p...al tn"D' ("'!".~ i .. UK .... u{ I ~.i.iu ..:
H.";"I stnf tNt miJ11' {arthu) • • /J~,u.tfornl. ~",p/iJI tIN 16"'~ IIIk fo"m."
(. " bWiyl) ,..d. ~ tIN ,-'wri." .{ r.r.td..J"m, ."t rr"j.illJ 1M ,hi"" IlIilh
t:tprmi.1fJ II¥I ,n, 111 thri~ Htyrr tIw Ji/frrntn fiw", tbi"l' u..1 ." .tINr u.." 1"-
wINII ur,,,,p/iJllbt .Prmtnltilw ut.s. H..Ii1tf.", II, "., c." Ibnr M'llJ'liu II¥I
""., """';, tIN 16"., ~ 1M .fo"",nrti.NIi mi"l'- ........ ovbt:oI.", INf .",,,,," ,mil'·
.tuIJ ..~ t/ti1tf./tbt 16_ ,,~. [PVI.98 nib], II hat al~ been .aicl lu
PV •. 7S) duo ...... d-ash ........ ouch .. the ~ and 00 _. ""' dinjnct. they accom ·
plish lk.......: Idil: function. A prnon __ rNl amons things. tome aaompliah dv:
.......: Idil: function, JUCb . l k pi'odl.lClitn of an ~ U wth thaN: thinp arc
coocq1(u.al1y diHinguiWd &om 0100 things !hat do 001 do to. ThoJc, things thlll
produce by dxit vny ~t\Il( • &L.r: a~ in thai ptnon.; thai tlO'UalClJ iI asIOIi-
Iltd with apru&ionr W I l\aV( as Ibrit ob;ea 1M adusion of tho.c things from dial
which doa 001 pa rorm the afomncntionrd actMiy. 'J'b.iI &It( 11II2mI(SI iI tIIw 1M
rccocnition. I'ltUI iI m.1-' 'I'h4 rccocnitionaiawarcna:l arna beaUS( lbe: imprinl
pbad In UK mind by dull per""" pln10w apc, ichcc baa bca-o ...........ocU [loy wh...
1Ih... pruendyl«inlll . In Ib il aa of ~ilion lbe: differenc.. bctwct:n lhole
DHARMAK!RTt 'S METHOD AND ONTOLOGY 'l!
oonfiuioA. n.""UII'_
rlMt. fflfI tho'lff
. ....W••
Tbu -.: ~ i, .. ","-..ins an tskntiaJ n.aJUn (~~;" simply I eqe of oosniove
iI;, _ ..iii_*,. 'f_
M._
..i.ui,.,.dNnviu, tbnt_ ..._ .....it
n«Wi.,. ouIndI ;, • _ I /biltf 1#9'''~. 76it
h.J. tliJj"rrn«IN... "-t. " (PVI.n .. ] An eucnlial OWun (riiJl4) it whal a thi", is in
~ of!he u1umaK. lr .diffc,uu [....,;h. medilti:JUIIO< from non",fttl ~ 10 haw:
11\ CllmIW n.run. d.m il would rithct 1M. tnmlially 1M iruu.noe Uu.1 il qual iroes
[i.~ . tp«ific Uft]. Of dK il would bo: aKntiaUy whal ill not mal ~ (i.e., it
~ not IM..ha. ,ftt1. ,(;.-.... in - . . u... u..w- ludf. ohm;u.. dw u.-na.
would IM.diffamtial<d &om thotc od!erenriria Ii..:., non-ma]; llOocha inmn« [of
• Iftt] ~ br diltft.mtialcd l&om DOII·mal bra..... ...-hal Iw die nil"", ofbri".
thM ;NU-I>D: doa not ","ve dw: ....."'" ofbeing -Ihirts dk. Hmo:, me ...... tiaJ
owun of the diffucna.not to be-;we 11M. (5pCCific IIftl. and ill UKllIiaI .... run is
w bo: tornethint 0Ihcr Idw! tIw ,roe-Wunu]. Ru. in .ha, caK •• hi:.tkct, dw: dung
[i.e., tht I~~] would btc::rd....t.d "'"" ma. ldiff...."'a: from non-ln!Oi:]. Hmcz,
ON' could not ay du. tht inSlana (i.e., 1M ipt'Cific .ra:1 has I diffdtno< from thOR
! i .~, non· .m:t]. FOf dut whic:h is tUtudni fro... lIOmCthilll is p~ th:o. IaUft'
thin«.- "ThaI'" diflUaoa: !No: is one thin« c:annoo: pmain ro lOOK odItt mu. ........ ,K
.hey would N." no Ieb,ion bt""ecn than. I( thtrt W(f( I rdation. i. woukI ha." 10
be- mu:awddfm rda.ion !bcaUK tht rwo ctllitia: ~ difftmoc]. III m..: QIt. 1M di(·
f(lt,1IO< would be all -nI.i.&I ptDpUi)' tN •• ptoduad br w panicuIat. HmO<. slnot
lhing c:alkd "di/ferm(ft" and minp c:aIkd "effi.cu" woukI ..... br disrina. aU cffcca
woukI br adusioru thai pertain 10 dxir alUel.
Mooco.t•• if the diIk,.,oa: [from rIOn-lrCt] W(f( in UKnot ~ odItt Ilhan
!he lr=illSW"Oot it IttmIIIO quIIifyJ. ihtn the [differmc:c from non·,fttl abo would
h__ • ,Ii"'" . _.~ ft.or" , ..,.. [.-On........... ,,,-" i. 'l.... lir..,,). "",,,.. oinac (..... I;~
from non·.fttl would have as iu ddimitins quali'1 (..;i*i) I diffi:>woe [from tIw
uce-iMWIOt]. it would not br a difkrma pcrWnir\110 WI iMUnot, jwt I f an
u llirdy diff""," .. ,btlanor doa noc poruin 10 lhal ilUlII'I«. Bu. [the diffumoe
&om non-.fttl ...... ittd( ddimitcd by. diKC=>et" [from iQ in.rtancott). llUItad. one
qualify Itho:: Itcuayin&1, "'This is (mit .m' . ] diffuuoa: from tNl (non·Uft]: Honco:.
irwm..m as ;1 Iw no ddimilinJqualil)'. tht diffumoe doa not cUe. Th.:n(o«. tht
dilK,tuO<;" no! ditiinc [bum tht illlOl>tt mal i, ~ to qualify]. and ';noe. lhcK
it no third P' "il*;.,.I«. caJ ohi.. looM.- <han bciAlI d-.c --= "" .... diKcrcn. r-..
_ mriryJ, doe diA'.,tl\Clt ...... wrimacdy ral. (_ *" IJW" ;y.; I ~
DHARMAKIRTI 'S METHOD ANI) ONTOLOGY
IJJh. ~ iii III , - Mtl~ ~,. p"" tifft' in {II, - p~ rulsti 1Ii~ III . dMtm
pU~ .n. ~IJmW.t4tIt- I .MMIIIU ,., "" ,..!~ ,,.,;,-t,,.,,.,ti,.,,,.,
'i
~ PI "" pIltltJAlr.""ruU MJ.iw ,m,. "" "" ,.".".ri,.m!l l ~ f . ,wi·
f'U'1I ".,.,.~, "" u pMUM .Mh.I", A.Jrit ".n~". iii.
1),5«, 1'01' o:nmrk, PVSV "/PVI."-1o. cmub,cd in me appcnd.U:.
1}6 ThO. I'orm..L.ti.on oppIia only <0 ..... i,..,....J.., fo, mod on .Iw: buio of rcal .ninp. ......id. art
!he only type of univcnalt lNI _ an' comidmn& MR.
DHARMAKI RTI 'S METHOD AND ONTOLOGY 'J9
of image and the aclusion. The a:dwion on its own cannol be the uni-
versal because it lacks meneal. conlen!. But the inugc on itS own cannO! be
the: uni~rsal bta~, as a menial particular, an imagt: cannot be distributed.
137 In addilion TO PVSV UPV'.91 (cind atx-. n.,8) I « a11O. for aampk. tIM: IrptncnU
in PVSV PV,.,..«
iii againsc an "ilimatdy c:Itina1, c:bu-Iiyt (}tin) u tIM: object of aprcJ-
lions (ld MrdJ/I). 1M poin, oflheK ugwncnll il mal iflhe:acnuJ alntml 01_·1 00f;'I,.
lion ;,.~ . mm or.e willaa upon !he univnW, .nd no!: upon tIM: parti(uIar.
.40 FO UNDATI ONS OF OHAItMAKIItTl'S PHilOSOP HY
·WeU, through this universal which you ha~ dcfmed :IS dilttr-
enee, does one cognize me particular as me same as other panic-
ulm, or does one: cognize something dse [-namdy, a cognirive
inugo-J as me same as other [imagcsP If me panicubr iJ whar
one: cognit.es as me same, then how can it be an object of con-
ttprualiry? And how could thm be Idie function (~rth4Itri]i)
mrough ma, other objecr [i.e., the imageH And [if the inugc: were
the obj«t) men on~ would not cogniu: universals such as imper-
mmcnce :a.nd 50 on in paniculm. Thus, the panicul:a.rs would
nor ha~ impertn:l.llencc:a.nd such as thor na.,u~, :a.nd impenn:a..
nenee and such would nor be the qualitlc:s of rtaI things."'"
1)8 I'VSV "/PV' .7sd (G;.p..J-..1): Iti,.,.. _ _ 'hi &t,+,o- ~...",......,..- ....
~ iri,~", nIv"",;",. , Iti,. tiu; I,,*,,' ..t..~,. ~ ~ ~
I.r"'!"""
....._ "
1 • .,._ H It.tJ,.,. ."IMItriJI I -v~~ ti,.j".~,
Soon- objtt.... "Bu. lina one cu:l.uded !lu ",;' OWN upcalM in other mingo .......
(an iu other-adusOon be • ..... m,naI!.
And PV,.1,:
1lIc appc:arana of &II Glcmw 1U[1ft with rrprcI [0 a dau-Iip 01' the: ~
of an ob;m .. if the: un~ ~ ia _tUI nat\m it a ropIi~ mot'. lluo, mot'
it caused brme menw condi[ionins thai coma &om .mns minp thai war ~nu
bqinninpc. lime:. ,....."". .........". 'ar)VJ u~ .. tr~ 1 ~1I,q,
WtMIiHli-u,u,yJItj ' w";"";¥J I".
140 hl .ddition 10 the: P"OP cited juK bdow. _ alto MV .,{PVI.70 (G:)I.I]-U ). "{
PVI.71 (G,...,.1O-1I). and .JPVI.,I "ab (G:.,.u).
142 FOUNDATIONS OF DHARMAKI RTI ' S PHILOSOPHY
";ut,-Iu,,, ."."."",
n4j jUtW,!, _6IJh..rrlln!lil".. M.tNll<lhitJ'!' IIIW...,. iJrilJ'l
·wn,.,..", ",i wi";""''' iow uJn ..lth..-J,i"' .......·
,nou..· 'l'~ oJ".." j.. . ,4U~ ~",."+ h, . .
oJ · ,' ',..-;;_ ~" .
...~Jm.. tM.."",IJ>d"!".""...m.....u..nh,,,. IMN", 'niMlth-!> ,r",;MJ,;
DHARMAKIRTrs METHOD AND ONTOLOG Y
'"
To overcome [he discominuity bc:tween menral content and refttc:nce in
conceptual cognitions. Dharmakini here prc:se-nu what amounts to a - the·
ory of unoonscious error, ~ as Tillemans has called it. In orner words. A type M
10 T .tlcnuru 1I"'''!n).
144 Sec PVSV .... PVI.:uo-:1.I1 (C!l07 ....); lIusialfCi in chap.:" .. (j oo).
I<H fOU NDATION S OF DHAIlMAKI RTI' S PH ILOSO PHY
I
N TH!! PU:VtOUS CHAPTE" K'CO
Dharmakirti admit only two forms ofin.m umenw obj«u (prllmtyta).
namdy, particulm (JU4uA;uI!UfS) and univcrsili (JIi"'inJ«"'~!w) .
Both philosophers m
th<: existence of only two instrumenw objeas as me
WUTUlt for :admitting only (WO forms of awareness as instrumental: p:r-
cqx:ion (Prll~) and inference (."'111111,..). Their argu~1 resu in pan
on the claim thaI particulars are inexprestibk: and univcn:a1s:ll'e not: causally
efficient: ~ing in"rn!Uibi<!. I""rtiClll:l" eannor Iw: r:&1c",n :u the objKrs of
mought and l:lngua~ and .incc inference involves ctlnoepul2liry. p:articu_
lars cannot be the (direct) objeas ofinfttence. l..iktwise, since: univmals are
aprcuible. they cannot exhibit change. Hence. they cannot create dkas.
and as , uch, they cannot be the objectS of perttption. which is causal.'
While this -disjunction" (vipl.rnvt) bmvttn perct:ption and inference dis-
tinguishes Digniga and Dharmakirti from 8rahmanical philosophers.
admining o nly two fornu of irutrumeno oflmowlf!dse ~ nor n~l2J'i1y:ll
significant departUre from non-Buddhist South Asian tradiriofl5. II is true,
of course. that many other traditioru do admit more than pcrcqxion and
inference as in5lrumenw. Naiyiyiitu such as Uddyotakara accept ('NO
additional irutrumena: analogy (f4JH1miNl) and lan~ (iAbJmlAbJA).
a_a uN widl JeaminllinsuOOc con~rioN. III wbm OM is mid "dw mw-Iiu aaruu
ova' tbm: ill eaIIN a ,....,.. - S« aI.o Uddyouhn 'l commnlD (NV:I6i-'71), ..ruct.
do.dy ~kI V~ (NBh:uS'): " .jU_ ,."..",.. ~tliJtIsJt' riU,.lf4 ..
..,........ and especilily (1"""']0): ~ MwJ..,.".,..., ..1Ifl~,!, ~"w,.k
1IIM~",,"""""'~qili,~~..,.~iIy
. , ~thI h]Ol P"'''''''!'f ,...,. i"..,..tr
~ pM ..- ' r ell. , , _ .m,..
1JtJ>i]4> lw. . tI·~ -JH'd N _ lJor p~~ ..1f'itIni MIfIj/WM'!'ftIiM.... ..u...
~i#.
'"
1,,6 FOUNDATIONS OF DHARMAKIRTI'S PIULO SOPH Y
-~
Kurdrib (SV. ~tlil) pYU: ~";ju. ~""" IIhJtUlM w""wt f .fn1'D!'
..~...".".~ni'~ fl
4 For a ~nl trt.llrMnl. I « KdI"". (1997).
S Dignlga IIIIS mil dearly H rs,.1Ie£. fUya I ~:lst) . Dh.um&k1ni'l opinion is found",
nllmcrow points, indlKlilll PVl.laI-t. TM IpC!cific ~ of;nkrmot intr:ndcd hm: iI an
infdUICte from df«r (ta: below), lilltt that iI. callAl tdation beta" .. dlC ilnlSt in. fit.
rmer'1 mind and dlC IInonnu of 1M spnkcr. Stc-. for c:nmple. PVSV fill PV I.117
(G:UJ.lI- 17):
A natcmcnl is prompted by dlC ~·s inlmlion ("',"i~ to eonununicall: I 'fK'
cilic meaninJlob;m (lin!.), for. pnIO!I..no knows, ilIilsulUllCn t iscornilll from
WI immrion." th.Il II<lIcmcnl indica.a the mc:.\Jl;ns/ob;«t (MnJ..) m.r it the men·
raJ appanncc Inamdy, lhc intmtion,l..bich iI dlC su.nnmr 'l ClLIIC. 'Tb.tft il lhwl
rdluon of produotr and producrd pm:ainins bo: •• ca. I mental mtity [lWIldy, the
intmtion] and tbt ~.ICI (Npij~ri). (..m-~mihI,.m",,,,, M"" iM". iii
~ nwlUill."." hi...... linN,!, nit.,.';Ii htUhiy.vp;jlWplJe. jI~-"
~M,""..J.NM .
6 Di~ (PSV'4Q&1ff .. PS2....49) fPCCificaly diKw.a p'QUmption (tmItI,.m). and he
1f!\KS rhl! 101M fomll of praumption arc aau.ally inferm<:n. Dharmaklni (PVSV . .
PVu,&ab; G:II]'-1I1) mcnOoaa an ilIJW"Cf't for dw: ois<mo: of the ~ fKUlric:J (i,.;n,.J.
..... whac he doc. not opoocify i. ~ ....:h. ..... MJWDCD' .. dc.uty all ~.,( p,Q<U..........
(II"""niJ.
SVAIJHAVAPItATlIJAND HA: TIlE BASIS O F INf ERENC E 147
II 11K fine problem it .hal i( _ I5IUIM Dignip to be -aing " d>rory of in~ thai
Ic.b . 0 indiopuUlbk ~ in aU c:uca. <hm _ ..... ob!ipl'o inclu<k III ru. 'JS"fII •
~furm of inducrlw OHWtIpcion. dUI if. lht: _pc;"" w • ....t.;,1~ .datioN tw.id in lht:
induaM domaill aIin hold in lht: tub;«t cia... of tho inft"......• (Ha)'Q .988:.60). ~ tee-
ond poubLcm u thal Dip>.Ip does lIQ( proridt an Mkq\u'I' ma.ns of dcurmininJ tho nc:p-
1M rorK'OIniW>a: ("'JtUiI'rhl~,n) (S.ci nkdlnu '".:)1,,),
'I'M ~ of thcJe diffieulriet
if WI DiA"lP has acmwned fnt otLIr _ ~ eondilioru, b..al not MIfIkicn. ooca.
/Or dnwiAS . tru#W<Xthy inr(l("..... Haya (,,sI:.slR) arpn tha. Dipip', tw I'IIIC ihnd.y
"failed· ill lW thooryof infnmu:. bul racher mal the WKntain.,. implicit in Dipdp·l!ftc.
IIrf is an ClJIftlUon of his "'rpcicilm. Keprda of whethn Ibis if 1M caK, Dhannaklni',
a,
....... Ihrory is cbrIy ~med dimilQu"I m...c
unuruilllis.
12 Sn: ~ (I".:........l.1, ). Sidtri.. objeaa to llling "induaion" II a phibophiallool.lO
praml thor: issueJ .ha. prompc thr: formub.ion of.M ,...~ti..,wIM. bul hill :up-
menc it ~ on tho UlUlnpOOn thal "therc: uonly one ... uu~acl'Oi""..mOon in the
IopcaI -rem of ""y inkrmc.." booJ:Jlo). 1would. inltmd up wr.h= i I _ dian
_ ony to pac•• e INlh bca"vlhctt ill _Ihan one: ony 10 I!ndcnand th. ~ ma-
,ion (N.- """,.. pk, .. " ... "'u,•• c '" , N""'Llc~). WI ... """,,"," ot....." •..doJ"j'. poubk .....,
inducti..., ""'" Ulha. Iw -U . rdalion ....oo.c
trUlh dc:pmds l1li th. IQlUtc:I or idalli!ioos of
SVA"HifVAPItATlBANDHA: THE BASIS OF INFERENCE 149
lhc.dao thnnJeIoa. wtw. ..bt lhc R~ of u.. mation annot be di......:cd &om iu con-
_.....
tall. CO'm if _ undom.nd tilt' omla to be ~ thAt do not rdn to rW thinp in lhc
The ba dill lbinp lUnd in the rdationUIip r1 a.1ISC IOd df«t 01' that OM thing iI.
plopc:<l)._....... of IOIt\C odin thinl an: ,uc,iainl tdlliomhjp. (.. ~ theft,.
r....:, dfCICI' and ..
II ,",,~~"I...tifi.od Iorthc: rule (~ol ........... mpo..
niecl non-uiaing. Thai ruk i& MM cktmniDed from not &erina the evidence in
hetuOSU_ caxo and .mng i, in ~ QICI. Otbcrwisc, how could one
arrM " the Nk WI OM thinlo namdy • CItIK. ,....-april, aim beau« a ruin 01'11-
a'I, which aft lhc drecu. aiat1 Or how could _1ITi~ II dw principle if an Attrib-
lite mal i& I _Mad of the e¥idena: 11M. CIIlIe that i& dil'farnl &om the CYKkna:',
aLQd Thia; would be lih inftrrinl mat IOIDCthrnS ;.. mel. bccalUt i, i& • doih,
r~'" RW'''''''''''' ''~ I ".;• • r.-.ry.._ ","",Uri 1M 1M
tJ.n.1W H • ...,.'!'~"iJw~ U; ,.-,...,.*
,.,~ 1 .~ .. r.II/Mi,,, ..,
tIh.~ /IIls.ui~,,,,,
ISO FOUNDAT IONS OF DHARMAKIRTI 'S PHI LOSOPHY
I~ [I may be hdpfiallO think of IIXft CO-p'eRnCe in term!' of the MKalkd parMkut of rtu.te-
.,J.
rW impl.".nnn . TIu. io.. wfw.n _ ...... ,~ " .... ., on . r, III • _~ ........ ••
~ I.. <oun.mnru_
ilM reruI, ,ha[ ow c:ondilK>nal hokIa wbnt~ , iI fal~ H~n« , "if all appiel a~
(CI:~ tbm all appieI ~ INitl" iI • rnx: oondiliorW. bul il iI [I"W only bccaUJc _
ba~ red1Kal the rondirion.JllO a IXlII.junaion.. ThU rrukcs no _ foe- Dharmakim. ~na
M pbas I hipt 'IlIhw: on dw: intuition dA,. when _ UK oondirionaU in raKKlins. if-=
ploptlly understand dw rncaninl of dw condJtionaI'l aniKecknI. ,.., an compdkd to acapt
1M IOONftluall. if _ wish 10 I"ftnain "..ioaa!. Thw.. in raptltm 10 1M I~ cumpk.
otwmakirti ...",.w uk. "Why should I think WI Mins • IrtnhcdlOn has an)"lbi"lto do
..iu. btlnfi frui. 1" "'" ~--.... ~ P''''''P'''' w ,""'""-'c <l14' the ~"O"'''''''' ,d.tlu., ""',-
rtOI be ~'M .... ''''"eftal condi,ional. and liu-;... Ihll Oh.armak.ini', .~ of
_MJ."..ti"'''''' unot ~ I miner of cominc up.,.;th 1M q>i... nnic ptxtic:a dA.
allow W 10cktmniM..-hen. pervuion holds. For a OttnletUlory tra.unc:n, of !da,ed iaucs.
I« Sideriu (WO}).
161 n Pins thil compound. 101M phil. . .' i _ tw;.:omc apparm •. The meaning dAr
I ~ .. 1t ed.......wd bo-~ primarily from an inletp!da!ion of thecompound _1-'--
~ (.~,..{ w. • .u,.-.~ althoush :an .mIpis ofi! u :an inllrummw .., -
"'nqJI...",.w aJ.o yjdd dw dairal maninll:. It u impon:anl '0 nett, ' - ' ". thaI in dw
000. •.,., of the initial pracnllOOn of dw IJ"P'S of rt_ in HB r... ki..I • ... IriJJJ,a. ~ J
mlfiMi...,,;,.wUJJJ.m 'H' b,... ,.,.,JI).
DIwmalcirri pnwides a sIou lhac IlDOUn tilO
limili~ .." .. ~ (ni"'MJ-,. ~). Mo~ IptcifiaJly. hu inl~'P"lItioo rudJ
(HB:s' ·H):
ni~ meuu mal IN: predate pr:rvades lM anribu.~ of lM IUbjea adductd u
n'idmce.. lba, niuMJ. doc:. not Gilt in lM calC' of:any •• icIma other than the
!hm: typa of cooickna mallioned heK. Hmcc, if iI.aid 10 be restricted 10 JUSt thac
,.,.1iNIHi
dim:.l1<It;, ik:; : ... ~rir ..k"'6*N.fo.· " tritMIhU __ ...,-...uui n,
."..i.";"" ""J'd'I.
Thc compound 1.n1fiMh.";"-,hw mcaru tho: - mtric:Uon r,,;"""J of If.;,.n& ... · 10
mo.e three Iixnu of...-idrnc.c. and _ - ' the saint s.I- provided by ~i (PVT:6li7
• "'1.11-16') in his ("I)ttUQtnU on »VI.!. ru UK of "\Ie compound as a pitlft .." .....
ho.,"'f. u p«UlW 10 the specific cant.,., of~fyinl thc number of types of rdi&f"Yi-
dm.cc. a.. ...... ~. "w: compound U bcs< undmwod» a ....;
14.,.."'1& F.... """",pic. bu..
.*
f' lJl' Olin ilutrummw
radinp KUJI W poGdobic in <he aM: ofPV,." . Fino. a
fWU
(lIJ4hhit4rtt): mar is. since our rule (niydma) leUs us mal the evidence (s uch
as ,make) cannot occur without me predicate (such as fire). the evidence
musl ~ an indicator flllmllltllJ of Ihe predicale (e.g., smoke will a1w.ays
indicate fi re). In short, Dharmaldrti ugucs for a ntUSSllry "um'In belWttn
evidence and predic lte. and he calls ,his rcl:ation a wabhllvapratibam/ha, a
term that I wililcavc untranslatcd. allhough il may be rendered as a "nal-
",.,.1 "1:II;on." 11
f"oro.o.-ud (through """vrm1t/w; md(t\Oe 100ft!y thrtt reliable forms of evidmot from pVl.I,
for m. di·.... "ion wrrounding PVI.) I iud( docs not allow for even:&ll implicit mama: to
IDe rumaioA r";"_J of IllfiIlJMi"" to enly Ihlft (emu ef eYidma. Scwnd, while
Slky:abuddhi'l own comrmtlD art: ambiguous, lhey dcarfy do not fUppon a scnili~ ..t _
ptiJ'lllll radinll of mfflbhivtut;",_ II pVt.,.. His commenD (Pvr~sa6-4 sb,) raid: 1""1
ti ~,tk by hi,. • .....,,,,....,. ,.,.".,. .,vir",. tki pbJo', / 'I7" J.r", ."... ft.i
• ,. Itf" pm ~ ,.; ~ "" "", 6. "'''IIf"''' ,. bt ."j lImJ,." Sf"''' pm ~
,. .."'...., kin"..,..,." ~,. ~'i 6J.& npJ It:fi "''''-'..,uJ1Iu ms'" J.r"l '''''I J,J,;"
oi ""l'~.,.,.", 6. mnI v ",j ~"l...,..,..,. ,"'1_ We can rroo¥Cf much of m.
Sorukril of thtaoe commenD by alraain! me appropriate phruc:s (rom K;afl)akaJOmin
(".,"-,) . fo!Iows;,,- tM'!' , ..ti6.",u",..!M p...a.rv.ft w ....t I u.,.u,..~.-.I'"
,,7ti-U, 'fA; " .4.~~~, _M.i.-.lwl ...u""Y"t.r~~" ~t
~ IIJII.~ (II 1i"'~IIi";""".q, ~r",} lliM Nl Ml~ r"tn ,., p N V },
Althoup me, aaa oquinlmlofthe TiMan ",..,. P" VarI""" be indubimbly de!crmincd
r.om lW"~in'l ~i,. i,;' " .... ~, ........ clca.r <ha, Sikym.MMt,i;' 00< hc.~allPIor-
inS'" urliu gloM or .,..tM kt..,. .",....~ ttiJII-!J. Ir he wert: ro&wins ,!QI poa,
and ifW(" P~ m. po.. of .IIiIlJIIhivII. d~ lfilllli~ Nl MJNlJ, Inc liucr POI-
!ion of~'sCOfYlmcnlf shoWd read .arnnhinsaionB <he.t: lineJ: ...., ... -'''1'''''''
kin" oi """ It. "" tI.r.""'" "'" 6. "'ttl .... ".j '''''''t 6.r "tn,.]iN IN .. u,."
,..#1M". alilit' '''''' eMt"."'". ....... M.i~ I"l"'"" P" vi. Of 0lWlC, _ _ flill kfi
with the p,obkm ofinlnprW"I the r~ "l"" P" ..... but in doinll lll, we muse be QU-
riouI not 10 auu~ thai Ihis TIbetan phl"Ul< M""nruy fcprucnlf lOme form of "j-ylUl..
~mjn's own commenD roIp1o)' IOrms eM " i/<l {I T.', a .rI-lS, and apKWly a.l}),
and ifKaro;Wagomin is ~ing upon the final word!.] of Sikyabuddhi', commmu., t/w;
final "1ft ,. P" V """Y rq>rtICnl :&II orip~ orUri.". Althoup. Ka.r~i n'. hiAoric:al
poeilion (nol 10 men lion the qw.liry of hi. c:ommenwyl maltetl his own, indCf><"n.knl
ranarb 1m rdcvanl. he aplicidy adopts a umwlJM,~ ~ing. Stt, for eumpk:87.13-
~ .... hi.u,-.!l eM"""," vi,.;..., u.""tilJ~ and a .lo-ll-Nry.-1Ii t.IIII.f]m-
ulJi"" (I) .. u ~""Y"woup.rtislJll.w,... j"., (I) w- rMI.i"""''''''"'JN'tti"ikttyII I'M
*",.NiJuyo /til P""" .rlII~" "iytuu>eik~ ~, II iI this UIagt" lUI-
pted by S1ky.buddhi and confirmed by Kanpbgomin, wIlereby muMlIJII.";"'_ is
WlCIusrood.a .,, 'U"'''J'IO chat is ..... reionnl to tho: di:scu.oion or lOItI~ti~
17 This lrarulation is the chotu adopled by Ha,..,. and Gillon (1991). Al<hou&h StrinkdJllCf
m.
preICIID IlJUmmD lOr a putiruW anaIyW {~of compound _~1IIi1wNlhta,
be hi.....H" ~ (19460): ' II it pouibk 10 dedlKll: Dfurmakini"s immlioru with thc tum
_~,."ti6...JiM. i.e.. the 'comet' in~ion ofm. compound, dim;dy from any of
his nattmc:n~ Iu fatal I an _ . Dhamu.kIrti a p _ himsdr nowhcrt: in. Wly tlu.i1lKh
a dtdualon 1& pouIbIt. - T e IIIla wry Iv::I.pfuI ot.c:t .... oon I would likt 10 add that me anaIy-
,,;, of the compound doc. """ j" iru/fclarify .he phik.ophW;:al istuetI in q\lalion.
Ip . FOUNDATIONS OF DHARMAKIRTI'S PHI LOSOP HY
Dharmakini h~ SClrcs Ih2t one may use eimcr an effect (hirya) as evi-
dence for its cause, or a IlJIlbhaUil as evidence for a IIIIIbh4wt. The rdiabil-
18 PVI.l: *..,..~
Mtl""INbnI".rwJhi,.., a:
nw"'".i,,. JM;, ."d6l.hi Ur~ I ~ ,.....", M.l ...
<he tnrub<ion by Haya and Gillon (1?5I1:S). Whlk lnmrpondns
,i
Oharmakirci', wrnmmu in MY (G:}.I~J) on this wux. 1""~ IrIIUblcd il . irit ~
~Kp>mdy &om <he prow 1011 in which ic is tmbcddod in MY. ~ me ptVOIC
may h:3vr b«n inltgnUlo <he wux !"rom the:: Q\l1KI:. TM key difficuhy in rmcIoerillllhis wux
is the:: phrur nww..;, Jlr-lWr, whi<:h I>h..umWrri dearly cot\HJUeI with both the:: df'ea
(Ury.}aains as ,.idalCl: and m.. a .... ~inkrmllJom wr dfut. Dlwmaldni·, _
mabieI; him 10 apply the: ph __ _ ~i, Jia 'J.l.jr fO both the:: cffca (U".".) and tho
aux (Ur"!W) beaux the irutnllnmw e- ol ...M.lNi,,.. "unr""y pbw.i1My ~ con-
_f\N:d with ~thcr. I "'vr allmlpted 10 con"o'!J' til;. ambipiil}'.;tII tIw: Engiiah ph.-- -in
rnnu oi"; the adnnnct' of this InRSlalion i. thaI. I f wilh 1M Suuluil iud{, illpplin /!'10ft
darIy fO tho: awe than tho: dIi!a.
Ir !rambled .. tlTobcddcd. the finttw<:I,... and Jim word of tho: .hird,.w. dIould ~
rcnde,cd Ilipdy difkrmdy. H_ is Illy IranaIa.!ion of thar ponion of tho: wux induding m.,
im~ldy MlrfoundinS proM from PYSV. Worda of tho YftX an undntincd:
PVT. INt first (ar.I9-W) offnJ an a1tnnati.., _1\131 of doe Ioeali ... III Un!"'....".
mabIins him 10 in~ me illMrwnmallll im""".Mwu~{fJ(P:L}.U). Thilllntt lUI'
~;" intricWn&. but il ;"'iOIn die mation ofdle iNlnunmallO .,,;,..6:w...1 hi... abo
IlOl foIlowaI Karqakapnin', p = ton of 00ftItNi1ll thr: pbruc ,.t/t.I.rytuM";'-Mbwith
boch me awe and the dUa (K:17.1)-l.4 and l.I.1-S). all d'liI ~ unn«eonrWy ~....
19 Sued on me ~ of lhI: tnnI 6JM... and IlOl _MiN in the wale Olt<! hac (FYI.I).
Harrs (1910) Iw arpIC'Ii mal dIeK all' sicnifica.nldi.fNmitOtt btlUtu. the pwilion found in
die PtJ"_!'."mi~and dw ofDhamualru',laler rexu. Steinlcdlnu (J"o). ho .. coer. haI
~":II~ aonYino:inslr dw dUt ClI)iOCUil ;. mioplaad.1u Srftnkdlner noc~ aonI within
pv and PVSV DlwmakIrri makes it dc:u thaI both,."". (Ihe -indicalCdo 01' plopeu, 10
~ plOoth) and r-k- {tho: -inl.tica1Ot"" or~) ~ _ t v - . Set l'V:!i V "" I'VI."
(Gc&+,.) and PV,.19l with PVSV "'" til. (G:,&.IJ-9].7).
IS4 FO UNDATION S OF DHARMA](IRTI'S PHILO SOPHY
are swbh.i1lll5 ofsmoke:, and in the previous chapter, much of our ralk about
"proj>(rties'" rdlect~ th~ term IINIbIMVII in our primaI)' sources.· ""hal is
common m both of mese meanings of swbh.i1l4 is the notion that il is essen-
tWto themtity in question: on DharmalUrri's way of putting it, flO funhcr
causes arc l1!quircd for th.al entity 10 possess a Wtlbh4"., I,
Since an understanding of IlIffbhilWlpratilJlmJhll requires an undl'r-
standing of thl' term twthhdlNl in both of its senses, pan of our taSk in Ihe
fOllowing pages will ~ 10 unpack both mcaninp; of me lerm, AJ we do so,
however, (WO points must ~ kl'pl in mind, First, as Sieinkeliner notes,
Dharmakini hinudf doo nOI provide a compkte accounl of me distinction
and relationship bnwttn IlNlbhiv. as property (or "property-swrbh.it"'-)
and IWlbh.i1lll as natufC' (or "narurc-fwbh.iv.t). Indeed, he docs not even
provKt~ any dear demarcation for the usage of the term in one $Cost or thl'
olher. Henu, if we arc to anempl 50me explication of IWlbhdlltl. we will at
20 Slrinkdl~1 (19711 IOU m., lim 10 d,- a d4tincrion bet'Olccn o;wo di!fcrmr ~ of
.......".,., The fxt dulINM.lrwU naNn can only be appIiedJin&ularlyto any p-m. caUq
iI best iUURrartd by CMoS whca the ram oco:u" in i",uummrallinpdu (_MIWU) Of
wirh tbc ,.,; tufIU: (,...w.-I4J. For c:umpIa in Pvsv, _ "" PVI,nab (G:1.1.I6 ); ""
PV•...-v. (e ,)., .... ), .. rvL 7); .... rv •.,. " ,\1 (e. ,o.~). U-J in mi> ':11K. ...J ~ ' .. ;.
a oynonym of ,.../trri wbOdt it .........i ...... !uM ..... sit...- lOr N.H.""" .... "",....,. For ttIIm,
pia of mil rype of ...... of ,...Jrrri in PVSV. Itt: G'40.u "" PVI.n ; G:41'" "'" PVL7J;
G:$O.) ';PVI.,. ' 9.b: G:f7.). ';PVI.I09 (mmpan!his instanatof~wilh tbc ncatby
~ of rNM.i_l4a, G:S6.lOO; G:60.1J JPVI.IlI; G'1'O-16 .;PV,.,.... ).
Wbn! I.I$td in !be pllInI, _#b.hwdoes noc ownurily rand for !be '"•.",. ...II~y ,""'1JJb.
of a Jin&k mlitr; inMcad, dw plural ohm oa:un whca many mcitXi art beitls ~ as
when one sptoLIu of many rh.inp who.: nalwa _ di!fcrm, or noc difftrtm (in PVSV I«:
Go,6.7 .... PVUI; 67.. ';PVI.IJ7-14l; G:'JuJ ""PVI.1,,). In conttm fO ,n- CUCI. plu-
n.! ~ of -'IM_ ... P'0f"'"T include .-MJ_;~).. U b;~in N •. ).....J _ut~"..
...-,.u.,.",,,; p.,....''''''oi/I
in PVSV JI"VI" J?-141 (G:6I.u - I.J). Similar ~_
f'ound in HB (~s., 1..11: u.-,.nwu......".
~~ . In -'<filion 10 dtil ~
of plun.! ~, many if\IW>[Q of ....M.i... in dw Jinpllar _ alto in m., ICftIC of·p'~lry'
p~ Sirinkdlnct. in IIOmC coneoru compoonda cadins in -jl",--MM... and ")6_ _ '
_M4w~p:od~
211u Scrinlcdlnn (1974'J1})."... -,...MiowfJiI INI P'~hl (til•• ...;. 6bi"';) of IOIl'of-
UUfl& which ;. noc cuatd by M'>mnhinll doc. bw ;. mw
Jivtn wid. , .... 'hina iud(.- ,"'"
notion due a JNJJ.illftii inrrimic or ~IW 10 an nWry in dw dUI muty does ..... ~..,
any fiuthcr ClUIa 10 be q.ulillrd by thai _1rIoi. iI made al ~ poi"" in PVSV. includ-
ing .... PVI.7 (G='.l6); "" PVI.16 (G:l7. lo-u): .J PVI.zI (G:I,.ls- uuh ""!'VI.J'; 1
*'
(G::ao.»-'lI; 11.6-7); ""PVI.., (G:t.6 ..,...): PVI.I66 (G:a.,u); ""rvl.l" (G:99.lo-lI);
"'" PVI.161 {G : ~ . J.O-lIh "'" PVl.ln (G:14S....,): .JPVI.l7'6 (G:14S.lot-'S). AU of dwx
.QlmlCnU may be unden!ood ... grounded in at.. !wOe "OM of ntUIM.. ..
~tuMhi.. in PVI.1. and !he claim in tbt commcawy on tNl 'tttIC thai ... caliry
.......J.d no< ...h .....·~ F " I I . 'l.....u<y ,...........i_ h..~ " _ for ita "';0.1 (e .... ,-1'
lhoi,..,... ~ iJI.hw";"'/IfJ~ I Ur.~ ""'~MimilJ.
SVABHJVAPRA TI!ANDHA: TH E BASIS OF INFERENCE ISS
timcs be obligttf to denvt ugumenu and positions !h2! alT only suggested
by Dharmalcini's though!. Second, wr mw! also K'CIll a crucial point raised
in the previow chapter. n2ffidy, thiU although Dh.trmakini speaks of spa-
tially c:xttndc:d things such as a Wllter·jugs :as being emidcs (bh4t'llS) th:1' 1
h2ve effc:cu, he doc:s not mean that they arc paniculan (sINIld!"!'flS).
Instead, as we have seen, Dharmakini specifics th2t, on the Exttrnal Real-
ist view (bihyirthttl!4J4), spatially extended emitics such as Wllter-jug.s arc
actually .tggrcgations of multiple infinitesimal p.trticles wh~ proximity
en2bles them to produa: effectS that they would not produce othenvi.se.
Talk about extended entities such as Wllter.jugs and the effectS of those:
entities is simply a matter of convtnience wd conventionY In the COntext
of the term WIIbhil,,;z. this point is crucial, for sVIlbhiiva as n:ature :and
lVilhh4lN1:H prnpc:ny c m both be applied t:ither to sp,uially l'Xtendc:d enti-
ties, which :are only convcnriOn2l1y real, or to noncJ.tended p2rticul2l'S,
which art ultimatdy ml Thc potential confusion hefC is that. as YIC shall
5«, both.senses of Wtlbhiiv4 rest upon notions of causal funaKmaiity, which
is restricted to particulars in Dh.trmaldrti's ontology. Hcncc, if one YlCfC not
to k«p in mind that talk about the causal functionality of extended cnti-
(iel u simply:l. nuna or cnnvt:nil'ncr, one might believe th2( Dh.trm:l.Kirti's
use of ~hb.i1Nl .as propeny :l.nd IIIahh,;,," :u n:alllte in the contl'XI of
atended entities would imply that such emitics themselves have cau.uJ
functionality, and that Ihey are therefore particulars. We havt a1re:ady seen
mat this is not the case, and below we will .see that the causal functionality
implicit in both senses of Iwtbhilwt is 2C(ually reducible to the causal func-
tionality of paniculars, To avoid complexity, however, we will follow Dhar·
mrnni and foc:u l primmlyon th~ ~ of[~ IrrmJ :II 2ppl ir.d (0 l'X(~lkd
entities such U Wll(er-jugs.
Hen«, when the term wabhdlHl is used in the sense of "property," if sim-
ply stands for an o::clw ion. h is aucial, however. fO nole that, since a prop-
erty formulated as an exclusion is constructed in terms of only one aspect
of WI subject, the subject in question still has numerow other aspcas lOr
which o::cIwions arc possible. In other words, the subject in question may
be considered to have numerow property-wabh4"VdS."
26 5«. for rnmpk. PVSV "PVI.n (,f3rub,od in m., a~J. 5« abo I"" di ....,"io n
of ......- of dka in ,he prMoui cMpm (1I!i1J).
SVA'H1V,4PIlATIIJANDHA : THE BASIS OF INFER.ENCE IS9
27 a. Sirinkdlnn (19]1:190).
28 n... notion I~I.(""":U-C fiud wi<hin COfIVl:nu.on. if apraKd al a nlUllbcr of pbca in
PV and Pvsv, lndOOi"l Obumaklni 'l iniWl lUlnrlml ofhil ...~throry al PVI-4o;1!
All cmilia (M.n.), 1xauic lhey arc elIuobliabai in lheir own nIolUltS (_Mohwj. by
nalUft (_~ ....) F .... (lJMti"JI nd ... ioru; (~") from all bo~ and
~mOSt'_lhinp ("",MI~~ n.u.ro.c, abjccu (...m.-) arc con-
crioal u h:.vins~' IUnds of qu.alilirs ~riIJWi...i> d.... arc indic:r.riVI: {•• .pItt.¢
of d>oK obj«u' discina~ propntia: too.e qlAlilirs au NICd on this and WI from
..tUch mm, is !he a ciusion of rhaK ob;.cu.d 1lwdOrc. W I discinc:t:i~ plOp"" .,
which is copim:l by mc:&III of IOIIlC pm!ica~ (~C&II/lOII ~ known by mans
of any ocher pmficale; lima, dw ptc:Kntl.ion ("7'" +ii) [of uh imaldy idmUal
~ abanctcd !Tom Ilk _ "'bt<al is diffcrm. [r... Heft pruiia ..~J. 1-
~ ..... ~ ........ tNlMMi .."".NSthi~ 1 '''~T.M.lIMM,i", ]IIM1dtI
?hi ..iMiti...... /1 UUMJJ ",Ill ",Ill 'nItIu", "J'lrrnis ulflfjl-JJw~ I jiti~
p,w,fnau ~~~ H --WJ#"""""~ ~ -".n,...u l ... III
~,.. '''J'"'l'' _ MiJtJtJ f'J"lWSIbo·,./JJ.
a ~ {PVI',sQ,. )(;,01) p..e.1thJrj1f U ~.nd he pouc.!he
compound "J'l.,mM.ltiIf U 'pot.,...;", adusiom" (IM:,...;"'", /J...",... ..
"""!rri-"~ no! recordcci in 10. K (loI.wff) COfUlrun il U P"M,!
(P'J.L'-4' - I.4!), i.e.. as 'Jhr'titfl ~u.
b~i (pVf''i6bI.-J'' K:lIl.H) gIouajtln·MJ...", tlhtmMMJ.. and
he offtn !he awnplc of impcrmancna and JO on {rifr ~ u", ~ . . " j
160 FOUNDATI O N S O f DHARMAKlRTI 'S PH I LOSOPHY
dJ« iii ~ ;,..,~ II<' ",i ""f ,. ~ lit UfJ ,.;,; ". JlNtmuMrJi "lIi".,.
lqukfi!J"!l .,,}.
' 'The ~ion "indicative:" for .'Wfihi" it manl 10 ClpClUT rho: pc of rho:
sJou oAm.:! by Siky.abuddhi (1'VT:16b: ""fI 'grtr H) and Ka~n
{K.: II1: _IthNU/JII~tuIilL}.
dGt)II\maDa1Jy, rho: lint mnnbn ollll1l"iJwNlhll~ should Ix corutNC"d wilb
lhe "d,uM pronoun.! of, -,.• ... ~ :and I haw: {rantbltd aaconI-
inpy. Sik:yabuddhi (PVT:s6\IJ.). ho...ettl:r. unckmands t1111"iHNiJM~:II
~1IiN~ "baicd on theadl<lion," and ~in (K:lIr.,IoA)
ofi'tn FJIl",nlNJlli.,..! uiNlwJlllh.*A "based on the adudtd. the limit and
the: adusion. "
291n addition 10 the: p.....V' <:iltd in the: prrnou. wpm. _ . (or aampk. PVSV ""
rvl .n (G:,fO.U--...6; mdlMkd abo in thc: ~pPtndisJ:
~ou maintain lhal diffrmll dtinp haw: a nondifferml dTcct. whereby choN: Ihinp
an Aid {O ~ I nondifkmK:,r; in dlar dq an aU be dwaaQ'Ucd by I diff~
ffom dtlnp dial _ Other !han t ' - tIw ha..., rho: aI"oIcnlCilriontd dt"ea. Bul how can
tilit be II\( cud"
-t
."" '1111'_" (, ••qn) rhi"t' iJ l..m tIM, .Idwwp """ .,., iijfrmtl, ., /hri....hO"
(,...hJN) _ .[IIK.t It" rnlrifrn til IIw Mrf""IiJJWlnJ1 .f1M $11_ ut.. ("rth.)
•..n, <II ;...t..n"t ....._ j~. (dut,"''Y-'' _ _'''-ju,..) #r ,...IIKi,,& ,." .. _ .....
IIt1S.["" Hfrtt; 1M snfU fotljlrin,,'" ~ . .. "" nu",,u.. [PVI.n] For cumplc,
SVA6HAVAPRATf6ANOHA: T HE IlASIS O f INFERENCE 161
reason thaL even if Dharmakirti 5UCSSCS the sameness of efreer as the war-
nm for the application of Ihe same lerm to multiple entities, he is in the
end also appealing to a sameness of cause. In other words, if we can call cer-
ain entities "water-jugs" bccawc Ihey have the Arne type of dfccu, we art
al the same time s.aying that those entities have come from the same type
of causes. In short, we arc appealing to those entities' naturt-lV4bh4lN15: the
toalil)' of thor causal charxtcr1srics.
This emphasis upon cau.s:aliry leads one naturally to su p~ a clost rela-
tionship bctwccn the notion of IV4bhallltas naturt and Dharmakirti's the-
ory of causality. $ro nkdlner poinu p.1nicu1arly to the rd.ationship between
Dharmakirti's notion of narurt-lV4bhdV4and his theory of the "causal com-
pla" (hmmlmagri), the new of causes and conditioru mal arc required for
t he production of an cfJ'ect. Indc«l. lOme PllQgel in Dh:umooni', W()rk
IUgesl a rt:bdoruhip bcrwccn an enlity's naturc-lV4hhdlW and ill partici-
p.1tion in a prc:scm causal compla)< or iu uisal from a past causal complc:L
dfca in queKion l iadr aIIl'It:I f'rom lOIN' ocher aUk. Th.... lilt; JrqualCr or aWCI io
~ For. by dw:: ~ nmu>: orwa..,... and wdko;Q, ~I dUnp ~
IIIInC thinp Wt an: their ausa, .... hik other mi"8' an: IlOI. their aUla. There II no
i ..............uJ ....,u,.......1",,,,,-,",<. ma, euo •......u.:.. ohi.. 1",""'__ /,i Wtn/, 'Ii-u.!o
bin" u ~ I ~ }of... I rUsy. jDuUJ 6muU ~....tM/.I--".
~".u u Ml ~.JtV iii ",u ~ ,.., '11". ... IIiJq.&fI_Mrlwt I "'.. Mt.tU..." )i
h.1IIIriJ 1m.l~1 ~Mkij ~ ...,.r'[' 1 111 IJi W}'I ,..MI". u,-,.. I ... J.;
...".""" ....,..". "?,,"",-p'"
IJit hi. }of'!' .5.,.. ."'"Ii "".
-pir U'-J ~ .. ,..J.U", i ii I
iii I "'·... tW.rw ....,rlqi"" ,,;,,_J,hj~
" . . , Itt
I WlUI _M.#w .,. noM... ilJ WJfU' I WJip· llljjltu1fl.novwl IiIM",..,..u .trUir
-"'f'D-'!"fU# .m"d...". It; JWrJ &n.r ""~ ,.M6..u U, .~ _ 1ri1!friil
i"
~'Ow"", "' .
• Altt..:.up the (:I« cndinp of lukWandtoon arr~ . • pluni KIIIf.wn-
~ by In impliri. - Ui It" Stkpboaddhi PVT:l,abJ (dickel in PVT-
t: lbJI- KJ14,.lj).
J.4 PnfupI W molt tcpI'DCntalM: such p"PV is I~ ont mentioned ~nd Il1.Iulucd by
S,nftkdl.... (1\11"111-116). M y ""n.!a.ion (PVSV .JPV'.:r. G " ........,) rnds:
TIN . n,.J . /... 4J«o "- u i..Jmw' it -.1 ./. (4-J " -,Iex it d.~ "" 4
_Mho. I{"- ......J ..",,ta .............. {r.,..m, forI "" 1foc1 i ,~" ' - _
MtH""'." .",un", tW. (PVI.71 ThaI (cap«il)' IOrl the production of an dI'ea doco
IlOl dtpmd on an,.minll otMr man ruch I roUooliion of w cn,i,ift in the aual
mmpla. lMrd'ort;, !Nt [apacil)'l. ...t.icb if 'nnrUbly mOJrq1lC01 from W!llm: or
dw :0":>1''''''' or c:awct.;". ~... or dw Ulil)'. [namely, tN' causal compIexl. Bu,
in thiI a..e. _ """ infw only W pouibWl)' of the ariA! or W dJea from the aUA!
complo bcawe _ [only] int'cn Inc: apaciIJ ~U) or dw: ausal (Ompla to p~
duac the cfka. TIw aW LI)'. dnot tt IIlnvvbbly consoequcn' wim the mm: praer..:c
or Inc: complete (OmpM:o: or a..-. io jwt. _M.l"., and""'" is ....Iu. is infaTC'd,
16.. fO UNDATION S Of DHAlMAlClRTI'S PHILOSO PHY
)7 Vuubuldhu (AKL61 and AKBh ..J til. ; IW - )1t1) ,efers 1O.1Msic: diRinaion br~ CCIi
primary OIlK (,.~) and rupponifll coocIilkIN (,.J.Uri,,). DN:rrmki"i', mat
oonunmtllOl'J .lIude 10 mit lwic diRinction II nrious poinu. indtldin&- lOt namp&c, in
IXvo:ndn.buddhi'. wmmcnll on I'VJ.U H ,... (PVP::a67a) : ...... , . R,. lti t. &-,.
J..itI,.; "'P ..,. "" ~ w,.r", PlUM , . , k". Jiyd "" ~". M £e}lf", """",., JIwJ
,. "" .., In! ,. i "D"';II WIIf ",it '" »t' ,. IIJ' u",11 rit ~ ,. 'i """"';11 ... .... FOI'
arMKbtra:amp&c,1fe ~i'lcommmuOll PVSV ..JPVI.I7OW (PVT:l.OlI )ffi PVT-
..:1a6ff)·
38 DNrmakirti'l \lit of ~ tmn · ou..W t'Ompla: " (4muJIIUIJri) is ~ ",doer (Oft-
fusi"" Jina il in¥Olvu mlft diffnml liruarioru: ;1.11 "incompltfc ou..W compla:" M".II
1Ht-.l,....). • "complete 0 .-1 com pln." (uJtJ.l I"t>ubuf:rI). ....<I. "o...al eornplc. "
~ 1k Itml - inmmplac" (1Ii.U1J} wulty sc:and.1Or. oUAI complex mal bdu
_ ofia QUICI 01' supponi"" conditions. A teed and JOiI, for c::ampk, ..ouId be ;1.11 inoom-
p&cte causal mmp&a for !Ix produaion of I 'PfOUl- This UA«t would lead OM 10 IUIfl«' tIw
• "mmplctcOUAI oompG" &houId ~ 10 WCMr.mm, no .'pportin&tondirioruan: bdr-
in" BUI dcspilC Ihis 1UI£CIIi.-e condalion br, "ecn "incomplat" (,,;It#U) II "bckinc 10m(
OUKI and conditions.- and "completc" (uUlJ)lI "nor bcki"l;l.lly cawa and condi,iona."
a QI~ o;ompIcI wi,h aU OUKI and condi,iQru in pba: it aill nor "complnt:.- T., bot Q)m-
plett, . c:wPl compla: mill! not only h.lvc a11lhe requimi CWJCS and mndlIiona: in addi-
n-. ~.......d pooentialo ( J,..lft.) do.. " ' - .......... ~ ..... condiriotu p n do-.h ......
proximiry mUM alto ~vc undnplc dc'>dopmm, {Jtui~-.) in I tonunuwn (1II1114Mjtudl
W I lhey are __ fully c:apabk of producin& ;1.11 cff«L ThUl, in the case of the prod..aion
of I sproul, the X'td and all ia MLppOI'tinc condilioru an: nor complat wbo:n W teed is fin!
pbntrd. and they do nor become complat umil mal causal compla dcuiops to rhc pornl
of aaually prodlCins iu dfc,;t. In tbon, W complat oUAI oompla is the orw mal 0«W1
in the moonml prior to thccfka. 11 ia wonh nori", WI dUs IMKion of -complac- accoonu
for DIwmaklni'scWm!ha1 one can_ inftr w uiaI of an tffet:t: 110m a ClIlK, o:wn if WI
__ io ...;.>I>ffip-ic.I by 011 <he.--y ... ppo<u"I ".,...Jicioou.
T., mtrr:alt, ;1.11 "inc:omplcu· cauW compIa: ia one WI Ixb _ OUKI 01' ClIfWfilions.
and a • complac" c:auAI mmpla is one in ..nich all requim! cawt:lIOO condilioru an: pret-
ml .""N1Iy dc.doped, Whal lbcn do _ all. caw.al compla wilh al l call1tl:and condi-
lioN preenl W l lIO! fully J",dop.:d, .. ..tom one lin! plana I X'td1 Dharmakirti '*' w
Ilf\II.brwd renn, "causal compk:>.."
In this rtprd. DharmakJn i', WIt of lbe ItTm "calUal tompla" (~) is wdI iU....
U"lred by rhc IOtIowins PIS", (PVSV _PVI_I ~5; G:~I. I6-99.I). H ~, DtwmaIdni ~
t.ht iaut oIt.ht ousa! tompla in t.ht tonlal oIatab1ishinllNl an ml il)' cs.- nor requi",
any fimh.cr OUKI eM' tondilioru in order lOt il 10 cnsc:
166 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
For il is noc ll«CIWiIy ,Ix CUt ,....", C&IUCII havot drew Ioinu i, if pouibk dUll Ihrit
eond.itioru: be incomplrtc or thai tbero be a hirI<fnna, 110 m., dew:Iopmenl of their
r--cWoJ. .,4,../ ,. r ......
-r-_'''-.fo/J-,f''$;.lUu.I H "'- ................... do..
~ it r.--'.. ....-9, dvt.u _ Jk,rJ ~wn.r ...
",Ii,, ; ~ or"u,·
'-Ii,., ., ~ • .["- Jnm.nN.. ~... dvt tffin 1i.1., tbr non" i .m..J Jmrw..
a.../ it _~ (. ~0InW"",""'" ;lJcmunnt""""" IPVI.I91). ThU
is.n in.mnnli<uy "'""".
ThUl, an entity It.. is nor ~ 00 oomethins dtr in ordtt ICI ~ ilia. ""Iun:
[i.L. ofhti"l perishabLe) is '"'"~ (,,;,,.~ ICI han"" Ihar natUft. at is rho: CUt with
a romp/ctt QlUAI compla in rduion 10 the produaion ofita dfcod irwmucb aI ;1 can-
nor be obStrUCIeci.
~ 8 ... _ _ .,.unl.,.,....pku,o . ......", d.ouy. •...." ~ __ ~. ~ ' ~pno _
ma:: of any od>c:r Q\Ua or MlpponIns conditions. do nor .~ "'ily Iwoo: th.M ""t1.Ul:
rl.f• • of produO,,& ">ti,
di:aJ in _ CISr:t brooo.- fVft1 when ~ iI the QluaI com-
pia of ca.nh, 1C'fd and Wilm, ~ lOIna"imcs it no production of a Iprout:
Thil u nor an arpmnll apiMl our pCIIitjon, becawoe in "'" cue II -JI. rho: cawaI
rocnpkJ: ~ upon ~ do:.dopmcnl ofthar COOlpiel of QlUiet and eoodiliont in
a continuum. Bul in Of00 ICI petish.. thins doa /101 do:pelld upon i&lIy rul1.Ul: cXvd-
opmmL MOfCO'IeI", in tha,1 connnuum of QII.LIoaI ODmplaa. .Ix final ON" • ..ru.:t. iI /101
;, ..<.....,....,.j by ~ fwuoa inM..." in do< "",,,,,,,urn "" ,he ,...oJ",,""" of ,ho< cfk.ct,
definiuly Iw tha.! dfc.n. Within WI COtI.inu),llll of (:II.IQI wmpbcs, only dull final
(:IuW oompIa iI rho: sptoUI·SClIW. ~ pm'iow momml ofrbc-oontinuu.m'.doevd-
I>pmftII is CKbn lhan m., fin.aI causal compkr. tach such momml is jwt lOt Ihc pur-
l"*' of thai hnal rnornftl. in the coo linuum. And nochinS can 11',","",,1 W
oonlinUUI!>'. final moment from poooIuci.nS ill dkct.. [.Ii 1 _ ", .~ bN/1
~"'"" lNlii"t".,~.riH~",U."'" / t""" r,rrMi""'"",,,
.!wl!f ~.try.hJ<f
...wnU; 1~ .'"~I!f 4mI1U11UUnJi,..", [pVI.I,sl ;"."III..n.+-& 11IIII
.".'!' MJ.. ._~ ,.{tbtl..,!, "..n uullt/NI""..".,,. "ut",MttINl",..nHMhtw
U ... ~",.".. 1II~.u ~""""'IfI I _ _ ~ .,; ~jt " ' - ' . 1OJ""1!f
~ 6hi;",iiij. WhM~ .pi kMkiJ IlihtilrlllJllt~ I ,... utripi J.rIlWM
SVA6HAVAPIlATlBANDHA: TH E BASIS OF INFEREN CE 167
wiU actually Jnnicipllce o nly in a C20UW co mplex for which ,he nrhe!' suI'""
porting conditions arc prescnl. For example, some seeds, in addition 10
h.aving [he capaciry- to produce sprouts. may also be crushed to produce
oil. AJ with me case of.a SproUl and a subsequent moment of me seed. me
effeas-a sprout .and oil- rtquire different causes and conditions. but
unlike m.at previous case. me causal complexes in question are incompati-
ble. This incomplltibilily doer not . however. mitigate the potential for [he
seed 10 participate in eimer complex."
Wim mese poinu about the notion of causal complex in mind. it should
be evident mal one should not equate an emiry's nalU~-IVilbhdva with iu
participation any specific. prescnt causal complex. The primary re.ason for
avoiding Ihis tqU2rion is dUll. while an entiry-'s nature-svabh.iV4 stands:as a
marker fo r the rOt:Llily of thar entity'6C:lwai c:har:taeriniC$, the c:auui co m-
plex in which mat enri[)' is acrually situated will nOi necessarily include all
those possibilicies. A sesame seed has the polentiallo produce .a SproUl or
oil, so if our interprtt.nion of IWlbhiva as nature is correa, the nOlion of
thai seed's n.ature-svabhdva must subsume both of these: pmenci.als. We
h.avt: a1re.ady Sttn, however, mat both of these polenlials cannot be .actual-
ized in a ,insle Clausal co mplex, si na [hose co mplc:xa a~ ine;ompatible.
A , imilar argument can be made without appea1ing to the case of con-
uadictory causal complexes. We can note Ih.at a water-jug, for ex.ample,
has the potenri.a1 10 be Sttn (i.e., i[ can produce an image in visu.al aware-
ness), and it also hu the potential for carrying waler. These: (W() polenti.als
musl be subsumed by its nature, and if nature is equivalem ( 0 causal com-
pia, (hen those two polentials should necessarily occur in the same causal
compla . This <orould mean then when one clfect oc;o;un, the; othe r should
nccesnrily occur, but Ihis is clearly not the CISI!. When watt.r is presem in
il$ hollow imerior, a water-jug fulfilb il$ potential to contain waler, bUI if
it hap~s to be in the dark, it wiU not fulfill its polenrial for producing an
image in visual awucness, Ukt:wise, if light and the other requim:l sup-
porting conditions are presem, the water-jug fulfdls the potentia.! for pro-
ducing an image in awareness, but if Ihe~ is no water in its interior, il wiu
not fulfill its pon:ncialto contain waler.
Thus. if an entity's nantr(' were oquivakm 10 the causal complex in which
il iJ presently participating. then a Wller·jug in the d2rk is not a waler-jug.
or dse a waler-jug is not by nalure mible. Likewise, a water-jug without
water would nOI be a water-jug. or me a waler-jug does not by its natu~
have the capacity ro contlin water. So 100, a sesame s.eed being crush~ in
an oil prcu would not be a sesame Sttd, sintt it is nO! pan of the causal
complex for the production of a sproUl; or else a sesame seed planted in fer·
tile and watered ground would nOI be a sesame seed, since il is nOI siruared
in the causal complex for the produaion of oil.
II should thus be dear thai the lotality of the causal char.actttistia that
constitutes the nalU~-I"..bh.i".. of an entity such as a seed or a water-jug is
rIot oquivalenllo that entity's paniqn.rion in one or more presenl causal
complexes, primarily because the number of causal complexes in which an
emiry actually panicipalC'S could nC'VC'1 oqual the number in which it could
potentially panicipale. In other words, its actual functions will never
o:ru.wt its potC'Otial functions, bUI the norion of a Wilbhdva as nature must
accoum for all those potemia.! functions. 00
In addition to the: problematic equation of an emity's natufe-tv.bhQ"..
with its participation in some present causal complex, we should also note
the temptltion to oqu:ate an entity's natu~ with the causal compla thai
produced thaI entity. AI first gIanct, this would make some sense, sintt we
[ .ww"tIc""'~
uJ WMwr II _
~. i"
Ii ~ "'vj~ 1,~i..rntdi",!" ......iMm04,,_
tUi""", M~ "rIIfUlf",WMIM!J I m,!, "u
*i'!"riJ ....".
M.b.'!I ' " .tvU~ 1.-1.
.,i
NOIe INI tJw; cblm dw an mllty p""........ ndlhcr I single: auAJ pcxmli2l nor mu.l lipk
pcxmtiak is meanllO tdVlC: dw: rftfio:acioa of IcauAi pocmWl u an uldmacdy c:aistcnl eIllity.
"0 Yn anoI~r . ..mIni ~I apinst lhe ~uival=« of ,.,.MMNII IU.luJ"e with the
l"IOIion of I atiAI compla is dw, in the cu< of an end!)' such as I WlIltf·jU&> (nrilia lhal
_ would no! inruil:ivdy tOftWUI: as upeas 011 Wlla-juc'l lUlUft' would Ix "fir- included
in dUI IUlurr. For nampk, the causal compla lOr the prodlKlion 01 dw: i~ oil ""'""-
ius in risuaI ....... _ includes demma -.h u lill\! and.sornsory conlXC (",.,;.J. If the
na,,,,,,-nwMw_ of .....c ....i... c<[Ui..kn. to ..... co.....J <>om ....... u..c.. cn';.>eo ....a. ..
llshl and ICIIIOty cntllXC alC pm oil WilICf. ju.g·IIUIUfIt.
SVA6H,tVAPRA TI&4NDHA, T ilt BASI S OF I NFEREN CE 16,
"2 S« a~. ~ 1 (9IJ1f) and J>VSV ..JPVC.IJi'-141 (IraruJ..lIIrd in dot o.ppnocfiJ.). Com-
1MII1illl on pan of Ihe bner~. Kn~abgumjn noca I~' tbcK iI no 'in9c a.ual
po<","i..a1 (/lUtijtha. caW in me d",irw;t nllilia thac , due <0 _lUnll~bn 10 produce
~;" ~ •. *... n tLod . _ •......;..'S (_ 'h]"1.: ..... ,..._ "'1*',u idH. ~uj. Ju St,.;"Ia,tt.
ncr (1' 71;116') has lICKed, Dtwmakim (1"\',.16.4 and J>VSV Mrit.~ G:I).7-u ) aplOcitly rcjrocu
die nocion of I .inp a usal po<cndal OC" 1IlI11IK-1lWMI* !.hal. bcins dismooced 0Ya" omain
cntilia, - . . u fOr mnr apariry to produce me, sam.: df«t(o) .
.4) In addilion 10 me pr"""-ly cited prcocntlllion in F'VSV .... PVI. I)7-141 (_ 'M appen.
dix fOr I complnc lra.nWtion). 1M notion of I paniruI.u «IM:rurd q ,. f'«1"Clly•....uMW
an I:IC' fOund in me, twO IDIIowinl!. rda~ I"m".... 11 is impomnt 10 1ICU!.ha1 me IX)fIIClI
ofboth pmlV" ia 1M dnnuNoCDIKm WI, while dot urUvmaU (()1U(n>(lrd /III meobj«u of
conapa mow M rorui<.kmI dot dim:c objco:u of tho.r c:ortupu. ~ uniVft"Al. midi WI;'
mardy I:IC' constructed on dot hIuU of a wally effkicnl panicubn. The firM ~C pauagt'
ir. PVSV MPV ••171C-171 (G :14i.1i-f7.7). wIUc:h n:adr.
'f'bcm'orr II",. uniqlK individual (~aJc.M iI fI". Hjm [.[~.,,]. • ..J NINr
.IN"" {i., ..
...J if.>btI,
.,,.rnu/y..".,.jk ,.wpM] .1r.n...u, nl"tUuNIOI; • • ,,~ i.JillUl·
i, ulkJ." -~fftt1 ".r. ·c...~. " .M _ d.i". lIN, ir " . ""nin<u,r"
(1Wd.~~J. AU lINN,., .j"fnM,.,.tt WNlnt.Imf .. /u" rtn<lt ill liN M"~..u;n.. .r
.N,uIlU .fthmt fPV1.171c,,- '711. It II» already been .uccd .hal only llul which is
""p*bic of !die funttion ia a roI 1hins- It i,I"""'" 'M Sirpkhyu aUl a . . . . a UN'
wmI it iuM the ~of the.w,.1fom tall) Other thinp. For iftM univm:al wen:
SVA8H1VAPR,ATIBANDHA: THE BASIS O F INFEREN CE 171
shon. ~ can. at the same time:. speak of the - naru~" of an infinitesi mal
partide. Now. on the Extanal Realist view, an infinitesimal panide is iuelf
a paniculu, so one might swP«t that the distinet nature-sWlbhdva of tach
infinitesimal panicle could iuclfbe a J»nicuiar. This would be ,upponed
by the: plausible: notion that the: subject to which the nature:-slIIIbhlillll
applies involves no distriburion in this case: it is the: infinitl'$imal panicle:
itself. But rven if this is [rue:, a panicular's narure-JINlbJuitw is srill not ifself
a particular. The: most straightforward reason for this is simply that a
nature-lIKIbhillll is nc:cessa.rily treated :as a dhttmw-a qualifier or aurit>-
ute dinina from the: infinitl'$imal panicle: which is in dhtt""in, the: e:ntity
that pos:sc:s:sc::s or is qualified by that dhttmt4. This is best iIlwtr.ued by a gm-
itive conS1n1Clion such as, -The: naru~ of the: infinitesimal particle:,"
Dhnm,.ki"; m,. i nt::l ;n~ ,h,., in such ~pra.ions the: tOm....,., iJ :actto.:L!ly ideo-
cical to the: tiharmin iuclf. The: appare:nt sc:p:ar.ation of the: dhtt""", from the:
I real thin" rhm il -..Id IlOl ~ ~ fur ;110 ~ Unpnoep!:ibic, beaUie tM cos-
nition of diffumr thinp aI nondiffnmr.t.ouJd Ix a _ by tM p.,,«ption ofil aI
;" .. ~" ...,...t i" .'-n. , .... ~ ..... 1 ... ...... ,"- '-J'" ~ ,..., H .... ~!'
n.t "'ffi •• hM • ..J ,..",.u"jforrwn tlwt.rr i,ultmpw PI " ' - rhtt,.rr HsN ..,." rIMt
nwrtW,.,.1ftTJ (u-) tlwnthfonW", ",." ~. !-Mil - . tJ,.t!"NII MII(-
.......iJJn,. .....i" •••• ..;J ... rfforr..J.id, u,." , ..,.,.." ~ n.",,.,.,.,..,,
it ., in __ Mnorr iiJfrrnot r- rw11'hi", • . • ,,4 in 11_ _ {t<Jirh ,.".,.; ,. thi"l'
"'''- _ "-1 it tIw ,. ." _{",;., ... A ".I iJn", U _ tJistrihW "'" ."Y.thrr ruJ
1hi",~_.tl.tn..iN, ..... -..ut,........ u ..J.Jt J.", ..... .....JJ . . /..,..,;,., ..4N:J.
"- _ _ _,IiJJII. ap«W tffrrt/.[ PVr.l19-rlol To accomptith _ p;o.iJ ill~-
_ i~inlO mediffermca and nondifft.tencu ba"hC( •• obj«u......:t. ... by inqui,·
ins aboul thr WII" thu I row it dilknnl from I hone and tM ways thar il it no!"
diIkR"," Inquiri"'l in rbis 1\uhJon, all pt'1O)o1l tMn ..:r:..;m ~ 10 just me . . .
which wt" call thr '"_66 'wf of. IN... r"."'broIlMI!"'~""'" )orr'!'".."""
_ f unU'¥J1wJ.NkM"''' ~M<tJ... " 'ff...,.M H ti,,~ .lIimwd ~
"Jh!fIJi til Iil1ffiMd f - ' ' ' ' _no Nltt..nj'N~,rau,.,.1lI!t H IPVI,r77-1101
u ... ,.. V'" ~ M;"...
MI"... wri Mnt.....~'!' 1M ,.rrd1mo ....... ,.,.
IN~ ""H r.:dIiJ'll".Jhi1trrJ- p._11 f l.
171 FOUNDATIONS OF DHARMAKIRTJ ' S PH ILOSOPHY
« For the wh;cct-prtdicatt ,mUon JeC PVSV "" PVI.n (uarubttd in n. 1, and in the
appmdill). ~ also the f'oIIowin, pMnp- (PVI_S9-60 and PVSV "" til.; G:)LI)-}).J}:
I .. _ -tdN "-'1.f~'" thm .",41' ....... ..oJ tJmn,.;...~ "'P;~
-t u" fo ...., ·tIw c«/tuiMo ;, .", /hi", .". lIN arl..MtI n.riIJ ;, • .u-... 11", ,.....
41rruiMa • ..J Jnn.i ...~ "'Pi,.,,",,..r-- jout III _teN rMl'nuN.... IPV•. !,)
On lilt theory 01 ot:htr~lwkln. il;' not W CUI:' that the adwion;' om thllOK and
thc adOOed mi"l lrhal the adution qlUliha) is anothu, This~ ~ 110 btao.ue
one would lor fDn:rd. 10 rorw::l~ thai dw: adutkd thin, lauch U I cowl ill alJo
adudcd from iu adution li~, il WOlIId he adudtd!Tom thc adusion from __
COWl. In Wi cut, Illtexdudtd mi"l IAOdo .. lcaw) WOlIId br rhal llTom....t.ich il _
adudcd, sud! as. honotl. I(dw wm' til( c:uc, men thcR wou.Id be: no adusion lof
•ccw" &om -honc:"]. ~, thc adusion is just thc adudtd milll iadf. The di(-
krmc:r in the apraaioN lOr and linguistic: ~Iioru: of the eulusion and euludtd
COOK from the diiklttlOt bawUIl tho: onnanlK: oonWtuioru: applied 10 them. 1bm:
ia no aawI diff'cmICIt in mcrnKt (~
~Sina: IW iJ no diik.cna: in ,dUuk<,. diff'dU>Ol in the ..... anric ronw'lfions
make. no Ioe1Ue brcawc the cxprcuiom fo, IUbjc.oa and prtdk:m would drnotc
(.~) thc I&InC tftin&. likcwioc. a gramlNltical QJf>ordarion (ruch II thc p i-
I;W of 'thc cownatof I cow'} thai inwotw.. distinction ben. CUI the «me"u reWed
aUo makcf no IoI1lK btawc it dtpmdI OIl. diff'a'ttla t.n.UIl the «ma\lI. ~
Au'-tII "'" ~ W 11_ ,bi,." • l'W",_tK.J m.n... met tIini"l' ;,1m ill
rtr-.n i..JiuuP dK,rwIicut,;,., is u" ~ (.nJ",) • ifit fIInY Jilform, "".. tIw
,~~_" " ,..J""J~;• ..J... f ... "''''1'''''' {(~ [PVI _""! It;' "",,n.,
cut thai any UK olbn~ is mtrintd til thecsaa>tColan ob;ta brawt ont wouJd
be fOrud 10condudr thal ... himsOal (.aN""'; UAft would not OCC\Ir. n- typaol
"AlP an uxntd 10 mu 10 fW'I) ditUna thinp whether thc thin" art aaw1Iy di..
lina or noc. 8ci1ll anployal in INI WJ.y. U>cy naiJy poinl ow thc .dcl(ilI .. if ;1
wen: IWOdiffft(Dt thing. ThctJon:, e¥m lhou&h both ' (II(TW' and · w .. , =IT~ rtkr 10
the I&InC ob;m (urh.c). diffrrml orrnatIOc oonwnlioru: an conHruatd JO tNt one:
mishl kMw lbar 10ft!(: auribult is prtdicaml of _ oub;ta- \Vhcn ...m JCnWuic
"""'............ ...., ~ a ~ w+- "",,..bo. arc dUotinco "f'f>aIn in """"
nition in tuch a (uJ,ion rhal il indici.,u thc pmiic::aoc: .. if'il wen: diHcrct\1 from
SVAlIHJVAI'RA TIIlANDHA: T HE BASIS OF INFEREN CE 17 )
me ~. nom though it u [in ba) noc diff«ulI from iL T1W ocrun hm"'" _ u
habinlamd by fq)Orcdly obcrn". that the r""""tic:aI rd.atioll iI ...ro in dw way.
Thlt ~ of. gnrnnurical rtblion doc:. not iudf CONtinlle ill ~ QK VI actual
di..incrior! bra. . mere i, noohi", '0 ~I dw IUC of...dI • <dation in tQIYK CQCI
j\aR in ar:c:ooi with. ~'Ilksira and irllmlioN r,..~..J6t). For inIuntt. in
«rUin coma.. a Jingle thi"ll is ap<cacd in dw li~. whik. in orda 10 ........
rapm.. thaI AIM mins mip u ~ apUMed in the plunJ, evm!hough il UI\OI dilfu-
m, in that it io.uU. linPc dUng. [... -nI'"
alny.~' .~1IJhfrY il}.,. 1ulMW
t. "iiaJM t. i.,. ~". ... _...u".u frV,_,,) un-Jn ~ ... ¥JW"!fM .",.,.".,
..... ~ '4,n!7In' "i~ U*1tbq • .. Ifb l whtl,. "'''!fU'1'.~
I""""'",.,iN -';wrui/> ,. ..... ..".., rw/I Iuu. ,nri,.niMeJ.. til "'!"Itn.MttU, / ...
-,dh.• izjl _ ... *?t~,.".wMN. .".,.....u.,mtlt~
1 uN ttl .".,litdi~ ";w,,,w IIM"'"""
~tNJ l .n.,.r ~~ 'pi
vi!N.ftli, '7.6lirrfti,,; 1 MI"_,,, .~". i ....vni NtJt w.~u[1 trV,.60) ... IV;
u#NUu,!, ftkjJ ..,.,......~tW vrffl, iLthJl4 ",,'Jt'MJHI',.,.lit,b 1 u ,.Ih.i
IlJUiriftu -"mriku iii ".,..,..", itJtmu Ulthi ,,~ "", .rrhmr .,,.,.·..tVhnu
,,.,ft.U6Jrntri / «'1111 P "' P""'''' ;1} ~"""IIInJJM1U ft~ ~
~u.m.,!, ftrw s.~ .".,liriktbthi .;/t4dti, .nJ,J",.,."" n..u.,;",.,v;
,i
".,IiMJ" ....nhi1llltrr "pi whi p1TJtIfMnltmi~t I ... ul-Wll,.,...trw w.. ,,~ /
........ ~u uM~
1
,i,.II,••
The Ira...t..;.,., of this pnraK {"the .wwion is o n, thinS and
._1.
"'Y"n-Jpi '"~ ~ ,,.,,.,,,Jt#IhJ#JhJllit I ,.,u,.jJ.,.,!" ftw,riJ tlt.t1llN.-
,Uiltl,yJ,.1Iirthtt'!' 1whII__
the QCIu.W
~...i'r .. ... outI"",-" ."?-'7".""j, .~"!',., i'JI ;........ I"""~Ic:> .I"" llll~rp<n>'-
rion of the 0'CnC pm by DharmaIdni in pvsv, ....t.cn ~ is &k-d
aj; I1Jhrttir -,Jand -,.1IJhrrU:. .61IJIII nw ",""UlJ. H<l'"I'm'U, if ,his pnraK
45 The faa mal a IUb;.a (Jhn".i,,) iI. corut~ prope"Y. _M.hwconm"cd aj; a sub-
ita has b.tn diIcwscd abo...: (n. ll).
17<4 FOUNDATION S OF DHARMAKIRTI 'S PHILOSOPHY
The (W() basic nOlions of swbhtiva thai we have summarized above enable
us to describe the produaion-mock ci the Ivabhdwpr'luibflNihll-i.e., what
it is about cause and. effttt that enables w to inkr infoUibly rbc: former from
the laner. In brief, one can simply say that the effi:ct's narure-swbh4tIA and
the cause', rururc-SlNlbhtitIA mutually "rcsuia· each other: for the clfca to
have: the naturc-swbluiVfl thac ic has. ic must come from thar specific kind
of cause, and for the cause (0 have the naturc-sV4bh.iva thai it has, it must
have the capaciry [0 produa that specific kind of effect. Although Dhar-
mm"i clearly understands this restricrion that constitutes the production-
mode of the SlIilbhllNfprlltiiNlntiJM to be one that pertains mutually bctwccn
cause and effecl, he expresses his theory by explanations thar emphasize
either the cause or the dfca. In both cases, his theory is expressed quite
d liptically, but when ptuC:nted prinurily in terms of the effect, his State-
menu :unount to the following:
.lOme orher. non-fire entity causes smou. brause die gme unacceptable
oonclusion would apply: smoke can occur cvm when dlat other, non-fire
enrity is absellI. In response, one might attempt to hroge one's argument
by allowing for some type: of"fuzzyM production, whereby something sim-
ilar to smokc comes from something other than fi ~. But, assuming that
"similar" means "having the 5aJT\e effccu," Dharmakini responds that one
amid nor aanunr fOr the .unoke's differc:rxr from orher entities because dif-
fermces in causes account for difkrcnces in effecu. His point is that it is
precisely by appealing to dilTc:renu or nondifTucnu in their dTa.-u that we
say that some enciries are similar or dissimilar. Having abandoned any such
notion of"fuuy· production, if one limply insists that smoke is causcless,
then one could not acco Unt for the observed fact that, rather than occur-
rin g hllphu-'lnlly. ~ mokc OCCUI"l' in ~pccir.<: rime< lI n ri pb<:t!ll." Th<':$C rhrox
unacuptable conclusions from the SW"!'tti-i.e., that smoke would be
causcless. that the dilfncncc in causes accounu fo r the dif'krcnu in df'ccu.
and that spatiotemporally rC'S tncted occurrence does not apply to causeless
entities---also appear in the HrtllbilUk There, hown'tt the argument oon-
urning dilTc:rencc is replaced by a more specific argument: namdy, that if
lmok~ can N- producM bolh by lh:u wnich by it!: " :lnl", producrs tmoJu.
and that which by in narure does not produce smoke, then smoke itself
would be by iu nature both smoke and nonsmoke."
In the SWlI.rtti. (he argumems used to defend these claims thaI focus upon
Ihe cawe arc summarized in the rollowing brief pa.ssage
...-hich iI of a dlifCf'm1 kind dun rlw CI..,. of tIw kind of drea. Hena. an dIln iI
IV)I mi.,bdinS aboul iu ClUK.. llIcrd"o~ if the: rdarion of ClUK and df=; II maO-
duu smoiuo: even ona!. Nor an th.u which cornel from lOme
other caU1C' be: smoiuo:. for it has come from an entity that docs
not have the propc:rty-lVItbh4lA1tof producing smoke:. And if that
other entity actually does have the propc:rty-/VoI'bh4va of pro-
ducing smoke. then that other thing must be: fi~. Hence:. the:
re:lation be:lWttn an c:ffc:ct and a caU1C' is not misle:ading."
~
IiUlMMiu.orItby.j4_~N/I f 1MJi"J"AI
""*_
j.,..." J ,... N $II ~ __ ,i
PVSV iii PVI.n «;:1).,....)): ujj."i. In J ... I>hJ*~ tf),M.... iti f ...rhIltm.r.,;
M.hot ,... $II IiUJII ,...M,m, in f utkrrI4,i ,...
bJtJ..u Mmlt I uI:JWlNHlItZ_ C4 $II ....,.,i,
it]
-,.""*"";'
5 1 PVI.J6-)7: .",inwtM.h.w{l t.1muy4".~ M."..''''' ~ f .t.IM,...",inMMb. ,...
~_ ....diwl!< ~ H tIhtt~ 1II...m,. JNt6Un~" f~"',!to ..;
""*IIUSJ'"6hJw $II ~ ~ H.
178 FOUNDATIONS OF DHARMAKI RTI' S PHILOSOPHY
10 be: a seed , and if lOiI and such also panicipatc in the production of a
SproUl, then lOiI and such ~ a seed .1O
Dharmakini himsdf does not address this problem dirrctiy in the
S"'vrni, but he does appear 10 recogni:r.e il 10 lOme cxtenl. His recognition
of the problem is suggested by w claim that tht causal complex is what one
:iCtually infen through :.n inference," bUI if Ih .. c1:t.im were inlerpreted
nrictly, hi. ugumentll for the produccion-mode of the fWlbhJ""p,."tilNttJdh4
in terms of the cause would f2iJ: that is, tht dnermination of an enlity as
having me capacity to produtt smokt' would not be a necessary and suffi-
cient condition for applying the concept ·fire~ to th:.t entity, since rnt
entity in question could:as easily be wood. II mw sa:rTlS likdy mat in claim-
ing llut the causal compla: is inferred in an inference, he also means to pre-
KrYe lO me distinction between the primary cause r..phUtwlutu) and Ihe
supponing condilions ("","4ri,.). In shon , mt properry-twohiva "capa-
ble of producing smoke" mWI IOmehow apply to the primary Gillie in a
way lhat differentiates it from the s«ondary cause. DharmalOni's StVlvrt,i
contains some argumtnu tMt might supply this testriction, but ovenJl his
pre:senr.ltion there: iI incomplete in this regard." His Ht'tubi"J.. o£fus mon::
,So4 The probkm men Ooncd hen:: i.~., m., '1xWon of u.. ooopo: of the po¥.,,.... U.......
•~ of produci",...>Okc--it ~mibr 10 0flC' noti«d by On.kt 1199').
5' PVSV ...tPVI.J7 (G:IJ.lO).
56 The diiawion of priCTW)'QIUC andwpponinfi c:oodJtion in PVSV MpV,.19S mar pro-
vick rhc: crpt of rtltriction thac it MCCUlIy!wK, lOt DlwnuJdrti thcr~ sugpu dw m., pri.
mary awe it..hac deI:=ninccl eM type of dJ"ta Ih.:al it produud by I QUAI compko:. In
Ibon, lui VJWl'Il'nI .. A '(J thai ~ m..:...;. lOll, """IU iUld IUd! nu;r occur In IxKh eM
alUll complel. dw produc:es I batky iplOll' and eM causal compk:c; thac produces I ria
181 FOUNOATIONS OF OHAlMAKllTl' S PHILOSOPHY
Iptolli. !her ~ do not M"': lhe _b4J"" or Prodoon8 a ria ' ptOLIl whcta in I
QutaI compla duol i"d"des I. barley Ked (and 1\01 I. ria I«d). The reuon for !his is tNl
~ .,il. ~~ and od>t. condilioN in ~ c:ompla pin the ,,,,,MiN
of prodoon8 I. ritt
~I from the p~na: or !he rice seed. Tht moll tdtYanl put or ,he p»A8C. whim
raponds '0 ,ul"rnenu '"Pinot IN: nOI;on ,hal an tnli!),', dDUVClion (,,;dJ;,) I . t'¥t!)'
!IN)lntn1 is not inm.wc (W ~nriallO WI tI1riry. reads II follows (G:'J9"'-' 4):
"Bm . in wow w ........ ricttprotJI" 10 be pmtIl.ICIed. a hatlry - ' and wch nuy .....
or
be in n«d of :anythift8 .ddiDonal beaUK ,tl 1M aupponin8 oondioons rom .. ooil
or
for tbt- production the riu 'PfO"1 ""'r
JOmtIimtI be in !he proximity or ~n !he
harky -=I and ..m."
How is il l\Ol i" need of -miftl? Tho, is, in.umuch .. they (i.e.. the IN.rky K'nI
and aucbl do /IOl ha." tht wUhi"" of produc:in8 the.itt JPIO'" thai a rice r«d two
or
they an: in ncni WI ,,,,Hh,iN.
"In ,hal caK. IOtM fibricaral or QUlC..II I:fllilies nur;abo 001 haw thll J''''iJhh"
".eo..
,I.., .....a.c. .k.:.r."
A ria: oud and ouch ha~ eha,_~of produc:i"81 riu~e due CO thcir own
QUK{,J. Heoot, W I which doe.: /IOl haw thoK cawa doe.: 1\01 ha...: .hal _Mbra.
And i, is pc:rcciwd mac I CIUK [or I ri« «Cd 01 a hariq...-dl ... I rauiacd autal
poccnUal. Nor is il rt:IIOI1Ibk- 10 claim dw d>t rsriaion in J""bh.iw WI pm:ai1U 10
dUnp (.nI!.a) is random (~iU) beaClK WI which dna DO{ rcqui .... awes and
~ ;; uq..m" ..
conditioN for io MlI:nI:t antI(K be rcsoicted in span. ,irr>t and ",bllIf1«. [ ....,,"
,...... jII"" ......~ I ~ai, ~1Ii'!' Witi, w",fpi
,.,.*"
"',!"r<UilJJhJt I ".,..,. ... ~ l ,....tA III ntIijI'!"
U!iJ,ijtuJt:tJ· ,.~~ I "",!, ,.,It; ~M'"
_1>Ith. Msri,.
.,i dIn,. ,u"tfJMu!I
..,.~., "'Ii,!, N III nw
";-,,,-,.1t6I1fIlI "mw~,.,.u. ... .,;~ I Iili~rtbrt
qi III "",MJJW!J JI"i 1M iii,. ...
'S'*""';,. '",,,,,,,MiN!! ~ 1";",.;.Jn;J,. III
IomtfJ ""'rii~ , ..filii " " I ... r. "",~"""iy.r_ "niMu"" ."iWffiM
~fJ I
....~ ~",...~.ttt.
57 In noUn8 we the ~O;n H B do not "apJlQl"" 10 addre. thicl probkm.1 mnn 10
Ic:awopen ohc: pooo-iobilKr ...... ~ u-uuW'
Kopt of ~ pnxnl woriI--mi&hlllncowu
-.I,....
uI" 1-10- ...........t,...;.. ......... bc)"'l<'lo<l chc
lJOiUUon.
SVA"HAVAPRATI&A.NDHA.: THE IASIS OF INFEREN CE 183
erty iJ prM.iared tb ft. then the suhjfit' in que...ion nea:ssarily h:a.s me iden-
tity in qUClition tU llrubjm. If we claim, for example, that "being smokt~
is predicated tk n of some individual, then the identity of that individual
is "rcnurily smoke. In other words, it is not poS5ible for u.s not to treat
that individual as -smoke." The necessity here concerns the relation
between the individual and a particular property, and nOl (he rdation
~n prorerries l.hem.<l!Ives, In ront l"2~l , wht:n Wt: <:bim Ihn "h.,ing
smoke" is predicated udicto of an individual, we mean mal,
ifwe predi-
cate "is smoke" of some individual, tl1I1y thtorrwiJl we assert (hat orner prop-
erties (concomitant wilh smoke) are nect:Swily predicable of [hat
individual. For oample, on Dharmakirti ', view, if one predicates "is
smoke" of an individual, "producal from fire~ musl also be predicable of
(har individual: it is n~ril y (~ OS!' that :a.nything wh ich " is l moke" is
:also "produced from lire," In this case, the necessity is tb r/i!'U1 in that it con-
cerns the relation among properties, and not the relation of properties to
an individual. II
Now, my contention is that, at last in the contot of the lVIIbhiVtlp,If,i-
bttnJh.J, Dharmakirti'J sy'$~m does nOf allow for tk rt neccsJity, and if the
m.rularion ~essential propetty ~ for properry-lvttbhdVtl implie$ (hat type of
neat:.ity, then th::u tr.:tnsl:a.lio n l hould be :a.voided, We Cln mou e:uily J«
that tk ft necessity does not work for Db.armakini by r«a.Iling mat, before
~ 1M p'ellmllilion hm:of ill.&. and tk~lOOCbIjlicJ" modified IOXOOW'lI for !he man-
IICf in...tUcb DlwnW:ini ~ W nlnion bURuil. JiM,."j" (~ ·Mibfca") and a
" ' - - (. ' prcdiaIC"), T cdtnic:ally, in ill Jk. modaIiry n«cUiry appIicllO I whok propo-
"lion (i.e.. ";1 iii fW"C*4Irily IN< mol dvub;.aof. ddiniti", daaminllion of blue is blue;
.. OFF" .! <0 J;~fIJ ,M;ry, in....t.;d, ~'YappI'" <0 the indjYi.du.oi {i.e., "i , iii ...... u.,..
the obj«t of . ddiniu~ dnmninarion ofblUo! is DeCaAriJy bI""~J (0. Pbtuinp 1974't-I}),
~ 1_ iI, that- notioru ~ conYUtibk 10 DbrmoJdrri'. way of conoo:i"u,S"poopaoitioru·
(.u..""i"''''-- Uructurftj by daimins rha!, in ill kM mod:aIiry, dK' ncuaity penai'"
bt""",, twO JiM,.. ril is nco;.a.sarily UUoe!ha, the ~'" nun u ,be objca by W
ddinili~ danmination of'bluc' pcKKSM:Ilt\( """,-- 'bluc", wbik in tk rt modality, !he
nc'Ctaity penal'" be",«n thc ,u""""i1l snd ib '" "we.) r'l ir. UUoe ~ W Jhn".;" nIu:n
as !he objca by w dcfinitM dctnminarion of'blue' .... nrily ~ the ..""",.. 'bIue- ).
F.... u .... ~ un ,-.aei'T' _ till: uf.~<.a! (if _oai"lD """uu.cnW) P"':'''''~' to I'bn".
np (1974), S« ~Iso me bnefbutlUCid ditawion of c.ma: by Sou (I 99SJ,
184 FOUNDATION S OF D HAItMAKI II.TI ·s PHILOSOPH Y
an individual can lx: ron,n rucd :as a subject (Jhitrminj flU ~s moke: the
propmy-1V4bh4V4 ofbcing ·smoke" musllx: prcdicucd of that individual,
With this in mind. we can then nOle dur, ifbring ·smoke" wtte prcd.iatcd
' " rt of thar individual. it would not lx: possible foe dUll individual 10 not
bear thar propcny. In simple terms. one can say that that individual would
lx: in nsmusmokc, But Dhannakini's theory of nUu]tl C"dellnitiw deter-
mination" or "correa judgment") belies any such tk rt CSSlentiaiism.
Dh:umaltini claims that. in a COITCCt judgment immediardy subsequent
to a pcrupnon, the pmUcations one makes of an individual arc marWy
conditiont'd by mind-depcndcnt Faaon such as expectation. need. context,
perceptual acuity. habituation and so on, Thus. when a child who stUdies
under hil b.ther SttS him coming from afar, he will Ilnt conceive of that
person as ·father" r.nhcr man "tcacher," O r. in a morC: gruesome example.
whcn a dog. a libcnine, and a Jlltin gaze upon a dead womm's body. the
dog S«I it as food , the man sen it as a woman, and the Jlltin seer it as a
corpse," Dharmakini's point in adducing these examples is that the indi-
vidual may lx: construed in multiple Wll)'S with tilde owrIap: the mind-
dependent facton m:uunderlie the manner that a dog interpreu his
perception of a woman's body will have little to do with the facton that
unckrtic a libertinc's intClprC:t:ltion ofsuch :l. perception. In orncr word.t, the:
59 Tht paUlV i" qUCltion is PVSV .... PVI,sf (G:JI,I6-JLU). I" addition 10 IOcwincp-
nJlly upon ddin iti.., dnamin.ation (~ Obumaklrci ho:n: c.ondudcs his ~l
tnat, C¥m thoogh any ptia:pUon OW""HNy conWIIJ &II the dllla that the ob;ut can prO'l'ide
to the prrtl:i.."., the Ikwmi"uiont that the ~ mOW ITom thallblll:an: dq>ctldml
upon 1M patti",,', dUpoJirions. In rlw eumpk of tho: woman ', body. ~i
(rVt':7'Qblif .. K:I#o l~ p'tnicb lIWIyc(lhrdoeWls, Lt.., INI w~". dad_',
body. and tnallht pnmven:an:. ~11, a doc
IIMi .libntilK. The pM "rads:
"But why .. il that, C¥m though one .... hid. p"ICCpl.w ""p"Iicnoa: of"'" n.a1;UlC of
• re:aI di'''fi (_"'I'M"') thai is diRina from :aU od>cr thinp. _ c&oa ISO'! ha.., •
mnemon ic: dnerminalion ofil ar wchl"
BtcauJe 1M 1t1ppIlfti", condirioou:an: IIdtins. And IhtlJo.~ . wnt tIN.p HV IMI
."",," .J,J (>~""fI1 ~M • tIM"" ".ti" (.rJq.) ,.., it {rJti_iJJJ MMJ '"
alImpk suggests that the dog will nevrr COnMVl': of the bodY:lS hi.5 lovn.
:&J1d the libmine will never conc:tive of it :1.1 food.
Now, if these propeny-lV4bh4uas--"bdng a love,- and "being food--
wm: tk rt n«:essary of mt: individual in question, mrn that individu.aJ "'lUI
bear thost- predicatcs, regardless of any Otht:, predications mat may have
bttn mack. In Euroamerican philosophy, ont: ()'pical Wlly of dtmOlUmlt-
ing th:u a prMial {~ P iJ not'" rt' n~ry of cktermin~t~ individual x is
to <it:monstf:lte that in at lcut ont: pcwible casc:, xis not-P, which is under-
$tood to be equival~nt to <it:moll5tf:lring that it is not rru~ rhat in all cases
it is nor possible for xro be not-P." For cwnpk:, if one wished to show that
mt: coia, of my coat is not tk rt rut, one demonstrates that in at lC2St one
possible world I could own the very same coat, but the COat would not be
red. !king ud .. thU$ no t th ,,_nrialto Wt thins', identiry ~ my <:02t.
This prottdure. hoWNer, may not be feasible: in Dharmwni's philoso-
phy for variOUl rcaJOns. Some of these reasons have to do with the fact that
the aforementioned Euroame.rican :&J1aJysis resu upon some notion of a
trans-a.sc or transworld identity in which a change of prcdicates is possi-
ble. In the c.umple of my coar, me :&J1aJysis assumes :&J1 abiliry to posit :&J1
individual that .. my coat in rwo difTuent cues, which are molt onen
understood to be twO possible worlds, and then to apply the predicate -is
red- in one ca.sc and " is not rcd~ in anomer ~. But for Dharmwni, a
prcdiart: is constructed on .he. ~is of an (:fltiry's naru~, and an t:ntiry's
narurc is the IDl4li" of its causal charxtcrinia. A change in prcdica.Cf-
M;._...
~ J w.lu J~u6iu )i ~ ;,.J-i-;;u J:Jt¥~ iti~~
,. INn,.- H1*9", ~t.; _ _ 1tIJh. ""'... "''''''''UII IIIIiMfi _
,.,.,.MnIq" tlPm ~ M.iIffti I ftm(lh~lIIiI J.".JH.". t,; ,.tlJ.ir,iit.J,J-
..,.,.". ~n".,.,,,~'; l,.tIJl riJMMn.~ fi h't",.u,.;,,;~
,,;~ 1_ NUIn,.,...
,!, ,... . . .u, h~ Ir.n't"- ;" u.,. "'odJ.HJ
M. __ .~. '+rJ.H/MlI ' IqtIM 1M rtf ,...".nhlntUMJlliiWNrlll ,..., ; 'JO'li'i
""*"mann,
,,.tIJl~~ fi"f6n"~"-""n!Nfid_""d.'; ...'"
60 I am
i,; 01.
he.... 10 arp.ommu by -r of uusworid idnlliry. S« Pbnlinp (In ...,lR).
Vap""'"' (I",) oIttn a ckar md ckWIcd accIOIlnr of IUCh uswnmu.
186 FOUNDATIONS O F DHARMAKIRTI'S PHI LOSOPHY
61 Up to this point I!u.w wntd tho: qu=ioJn of ",hrdwr aU thl'tll: of tbnt <kcnminatio<u
r lovn," "food," rod "cotpKj mwt MconWemi COft«l (","')'!d). ln I private communi·
arion , Strinkdlntt Iwtugosc:cd that no ~ than twOcould be-. For r:umpk, if tI'M:' body
in quation iI tn. of a dad WOInUl. tho: Jlltilf and the doc would be comet. Burinl an
improl»bk rd".,.ma [0 n«rophil.ia. the libminr'l pctapnuI judpnmt would havr [0 be-
If~ i.~., bucd on bvth ...... t is raJ (!~ WQIILUI'. body) .oo th~ no km",r rnI (h"
[UU ...urinnoftir ,..;,." io.n). I ... ny .,..... _ """ J>«>h.hIy'V"" ,h.o, nn ~"y ..... fi&",.:,,;"n~'
Ir:u. rwo olthac judp>mu m.... be- dem>cd .. IJJJ~ (basal on rnI thinp), and hma,
the: mulripliciry of correa but II Ir:u. aathcUWIy inc:om~tibk judpnrnt still appIia. Prob-
ably put of DIwmaldni', point ~ is that ~ th_ intnpmrn----thr 00s. tho: libmilW
and tlw Mi........u ICtUI!Iy livins in differ"1 brmic workU (1M';' and IS • mult, dwir
intnpfft:ltio<u differ ~I,. Iftdftd, tho: nolion of tho: ~iona of prla ; OWl Ioattd in
diff"amr bnnic worids (II in tho: OK -...lIne water can M corrcaly prroei-J as I home by
1Uh, I Waki"'tl drink by hUmaN. or • noxiow ..,.... una by /'"'A) m.lIr haw IUnctionC'd u •
~ touI am "" P-ib&.: ......1J. in the Ewo;omc,;u" t~,iono. I alI...Jc: '0 mil ~
below in tht- Coodwion.
S VA6H)VAPRATIMNDHA: T H E SASIS O F INFEREN C E 187
propeny-svahh.iva "being food ... but it doc=s not bear the proptny-ltwbhdVd
"il a lover: And when me libertine is praent but the dog is not, Ihe: body
bean [he propeny-lIIfI'bhdt.., "is a lover" but does not bear the property-
svabhtillll .. is food ...
Now, mis may seem 10 be an inadc:quate: way to show that individual )C
does not tk rt bear P, which is me same as saying that the identiry of)C is
1'\1)1 ~urily (or ~ on l nt"('.l5.urily ind utk) P. Ooe: mighl m.:linrain. for
c:umple, mat even if me property Pis not constructed. xstill has the causal
characteristics mat make it capable of bearing P, hence. since the causal
characteristics can themsdvo be construed as amounting to a prt'dicate,
Dharmwni's position still rests upon tb rt essentialism. But such a
response:: would be ignoring the:: fact that me "causal characteristics" me::m-
~Ives do f\O'I ~Iy exist: for e:x:tm ple:. in terms of the: individU.:!l' , cap::!t;.
iry to produce entain dtects. the~ u e no separate causal potentials (11&,;)
distinct from the indiv\dual iudf, nor is ~ even one sing.le::, UlllVrutrJU'Ud
nature or undifferentiated causal potential from which w Other potentials
might be constructed.t.I Instead, a1llhcsc entioQ arc consuuaed, and in the
case of spatially extended or distributed entities such as a water-jug. the
individoal iudfisQaIUUUCted.. Of eou~, 0<1 Dlumulurti'J vi _ _ can s:ay
mat. in the ~ where an individu:al xu Jl-whl"n, for ex:ll1lple. me nu<knt
sea a pC'.non and identifies him as his f.uher- it is not possible that x does
not bear P in that case. But this docs not commit DharmaK1rti to tk u
C$SCntialism, fo r it is the same as saying that, in the possible case or wond
where my coat is blue, it is not possible mat my coat is not blue.
Since DharmakIni's system docs not allow tk rt predication, and since
$Orne notion of n~i ty is al leasl implicil in the: StMbh.i_p,..,iln"JJ,., we:
might as.sume that Dharmaldrti's syste::m involves an inchoate norion of tk
Jictll na:es.sity pertaining among propmy. sWlbhtiwa. But while:: we may
usefully make:: this dai.m for hronnie purposes. we must do so with exurme
caution. lbc first reason 10 exerclsc caution hal already been pointed OUt:
an entity docs nO( aau:ally bear a propeny· s""bh.i1-1t1 umil mat property·
nMbIM~... hall b.ec:n c:oruuvacO in n:lalioo 10 a mind, aod if minod·de:JKnd-
ent factors such as imerest arc not in place, then me propcny-slIlI'bhtivas in
question will nOt be construCted, eYen if they tk dUro stand in a n...-ncary
mation to w me other property-lIIfI'bh.i1lRS that have been constructed. In
Dharmakini's discussions of inference. this iuue is transparent, for it is
naturally assumed that one:: is indeed interested in knowing, for example,
62 See abcwe, ft . • L
188 FOUNDATIONS OF D HARMAKlRT I'S PHILOSOP HY
whether or not there is:l /itt in OIlC!'S room." H~, upon .sttingsomcthing
identified as smoke, the propeny-I",,6h4VA ~produced from fire," which is
tk Jicto n«nary [0 smoke. will be readily contnucred. But tht:~ arc orhcr
tk Jicto n«C'5sary properties, such as being momentary (~!,jk(l), that may
not Ix constructed. since they arc irrelevant (or cvm anrithrucaJ) to the
conurns of ,he pttson drawing the infe~nu. In soon. if wt wish ro claim
that DharnulUni's system involves an implicit norion of tk tikto ncussiry.
wt must remper that daim with lhe considerable emphasis on the rela-
tionship between psychologism and onrology in his system.
An additional reason to be cautious concerning any claims for an implicit
tk JinD nC'CCSSiry in Dharmakini's system conccnu his failure to fo rmulate
and provide adequate terminology for a dislinaion bctw«n nea:s.sary and
accidenw propcrtic:s, even though ~ such dislinction is dearly required
by his thcory of infcre:n«. If wt conUder, for c:umple, die case of .stting
something thai I identify as smoke billowing &om my room, my abiliry to
infer the pronmate prcscncc offire from mat perception relies upon die faa:
that · produc.cd from fire" is tk Jirt# nC'CCSSaty to smoke. But any givc:n
instance of smokc, such as thc onc I sec billowing from my room. has
numttOw propcny-lWfbhd_ that, although not nMc!pry to smoke, can
Ix and may acrually be constructed of that particular instance of,moh. I
may notice that the smoke is reflecting a blue light, or I may see that il is
swirling quiddy, or the shape of its doud may resemble a rhiflOCCTOS. BUI
these propcnic:s are nor tk Jim ncccssary 10 smoke: I cannot say mat, if
something is smoke. then it necessarily rcscmblc:s a rhinoceros.
Indeed, on a &w occasions Dharmakirti himself docs recognize a need
fo r such a disrincrion. For c:umple. in the basic formulation of inference
cited earlier .... Dharmilklni spccific:s that for elkct--cvKkn«, 0 1\(: is oon·
cerned with that number or propcrty-n.wbh.i1l45 in the cawe withOl.lI which
a some other number of propcrty-wabhall45 in the effect oouId IlOt occur. In
giving a quantilative speciflCation (with the tcrm yawul) of property-
1VIIbh.i-., Dhannakini answm the objcction dial. sin« all the propcnies
off'ire . for c:umple. panicipalC in the production of an df"ect.such as smoke.
:lnd liner:lll rlv. rmprnleE nf Imnlrr: arr: rqu:;al1y r:Fra.-u. onr: might IU~
mat any propeny*sr.wbh.ill# of me effect can be wed as evidence for any
property*W4bhdV4' of the caUSl!!~ As Sakyabuddhi suggests. this objection
points to the problemacic cue in which. by focusing upon me subscmcial*
ity of lmolre and the brightncu of some individual fire, one rttiuccs me
inference offirr: from smolre to an infe:rr:ncc of brightness from substantial*
ity. Glossing the beginning of Dharmwni's rr:spon~, ~kpbuddhi noles
mat such an infert:nce would dearly not Ix rdiable bcaUSl! the properties in
quesrion-substanci.ality and brighmess---<io not provide me kind of resuic*
cion no dJ uy to form a trustworthy in&rroce. That is, since substantiality
an be a propeny*SVllblmva of entities orner man smoke, it an occur even
when fire is absent. And sina not all fires are bright. a fire an Ix p~nt
even ifbrighme$f;' ahunr." 11..a, 2$ Silcy:.buddhi de$.c:ribc& m;.. MgUrncnr,
the basic point hac is that somt propnty*svabh4v.ars art not rdeY2nt to tht
infttma of fire from smolre: in the case of smoke. tht irrelevant properties
are those that, even ifcommon to aU cases of smoke, are also present in non*
smoke enlilies. And in tht c:ue of fire, the irrdC'V2nt propmies are those
mal, CVUI if applicable to $Omt fires. are not applicable to aU fires.v
On ~ik~buddh i " intetpreauon, m;.. argument :already refers 10 acme
inchoate nouon of aociden.ta1 and e:ssenti.a.l property*swbhJ_: specifically.
if one daims thai certain proIX'rties of individual fires must be ignored
b«aust: they are not present in all fires, one has tk ftUtD identified those
propertics as accidcntal-and nOI essential-to firr: . As for smoke, Sikya*
buddhi's intetpmation is problematic. sioce our analysis of the produc.
boMI'l(Xie of me 1INIhh.ilNlprtllibttnt/h. has already shown thai: Dhatmakirri's
formuladon of ,he au.tal rciuion il ba.scd upon a property-i.e .. the prop-
erry*WIlhh4r.w of being Mproduced from mal (caUSl!r (t.JffllnJd}-which is
common to other, nonsmou encities, such as ashes. Instead, as Dhannakirri
71 Gillon (,""st).
n ~ I..aak (1m) ups m,,1 GiIJon'. a ;liWm "miucs ill oO;ea" brnUK;1 doeo noc
:KCUrlldy IkpKI Dlwmakini', IQrIhod for Ikcumini,. aUlOllity. I am indinod 10 ~
m"t Gillon', tIUUl"lMI, especiaU y his RrJI ~ rrquira Unp!o"'''I(ftt; Lasic'. own accounc
(' m :an);'. much ..... p c:Ooe: radi ... N~ Cillon'luW,... is...dW in ill IUrlt
pracnmion of lhe: buM: pobkm. one dUI' rmWlIf ~ on 1Mic:'.~: namdy, !hal •
.....- of pt.« p!ions .nd nonpCl ccp1ioru cannoc ptoo'idc- ~n« of I causal td.Ilion
without !he Q>UI~I UIIUIIp1ion Ihai the pct«ptual judpnulu in question ar~ aJrady
"",",a:...buo.. • .be.......J ch..-...:ou;.u.:.,,( .... cno.;, iu in <[_;."".. Wi .............. aMUrnfKion.
Gillon'. donkq will mntinuc 10 be a 1lIOII WI ... dcomc au.
SVA!H.iVAPRATJ6ANOHA , THE BASIS OF INFERENCE 19}
71 SlrinkdIne:r (1971:1.09): "In ok. OntOlosie bakultl wUbi...fJ die KnIt lin Din~, :w
,".v.., :Us Prinzip ihra Sri.... ... "
7. P'VI.I~ and PVSV.Jril. (G:.... r6) : "'lbM which it~of I~ funaion uan wo·
mudy raJ t:hin& ror dail aIont ill the chuaaeriaOc WI disUngu4hes dw: raJ and dw: unraJ,
namely, WI the: former hal dw apacify for tdic: function,...t.ik the: lmer doer noc t...-..: thaI
~Iy. • [lol ,.,."..nhi"-
W .. ,. ~fl l iM1ff tN hi IItfItJWHltllur
t.~.".,....m,.I ..).,.'ll M ~r.j a .the: uuWuion by Suinkdlra (1",:1'), n..10).
n; Strinlcdlra inlendi mal z-d be ukm ill both dw JeNcI he allribulCl 10 .,,;,.. i.t"., at
both -pi" and "fUnaion..- 5« 1911:11""11), n..,.
76 StrinIcdlner MlppotDIlhiI daim wida the ~ tnsubcion /iQm PVSV (1911, n.Il):
"Abo thar alone brinp abouc Iht" fiiliiUmcm of iI function/pi only beaWC;1 poueff:ts
the corTaPOOdilllt 1NMiN.· [Auch du eint: YOIIbrin".fit" ErfbJIq de5 Z..-kes nUl. wei!
er den m~endm SnbhlVll baim; G:.,.I,ff. .... )i u".~". ,.nNM.h.,."u
"'" "',.,,1·
194 fO U/'liOATlO/'li S O f O HARMAKIRTI' S PHILOSOPUY
n Stcinkdlntt (1'71!1h- IIJ):' ... m- ~I ckI Dinp hmLhl auf dmI SvabhI....
Ob rin Din,
'riMll Zwa:Ic niWll,' tww; auuehlicllidr. datIOn 3b, ob os den ~
S¥3bbM hat..
78 Scrinltdlntt (lJlI:IIJ): ·S..,bhba rinG DiIlJ'G itc die Knfi (~). aM Wirkuns Mr·
'to<'J.Ubri"F" •
79 5« Sc.rillkdlntt', diJcuaion of PVSV .J PVI.I67a-< (Inl: IU , n .)} and )4). Set: aUo
~, n.}).
80 Son: Strinl"d.",. ('71.,,16) .aud .... u...u... uf OI .... "..Io.i,,; (rv...6.... ..t I"VSV "" ,';.q
G:',,7- II) ,
SVA8HAI'APRATIMNDHA: THE BA SIS OF IN FERENCE 19'
amditioncd mrough their own ~pKtivt: causes. produce on~ and me: sam~
effect. ... , In the case of an ~ntity such as a water-j ug, a nature-w.ofMdw is
mus -a mere, conv~ni~nt sign {Siv1lfor (h~ f..ct that th~ appa~nt 'work-
ing together' of the causes for the production of a single effect is 10 be
grounded exclusively on the bcginningkss pr«Onditioning (of those causes)
by {theirl resp«ti~ determinate causc:s.~"
Since Sicinkcltncr mainf:lin.'l thai thc~ il no ult.imateiy rc::al nature-
svahhirNlof things, he would presumably:agree Wt this ItNlbhiflll must also
be conltructed in some fashion . And in Dharm:akirti's philosophy con-
struction always involves the process of exclusion described in his a~ha
theory. Of course, this does not mean that, in constructing a thing's
narure-ltNlbhoiw, the mind of the perceiver alone simply creates the ca.wa.I
characteriniCl th:u correspond ro the th ing'." n:Ul1rc. Thin~ :lri.~ from
renncted auses and produce resuicted effects ~Icss of whether their
·capacity" to do so has been conttptualiud."
We have undcrsrood 5reinkdlner [Q mean that a n.:nure-lvabhoit'll i5 itSelf
an ultimatdy unrc:a.l entity conStructed through exclusions, and on this
basis I am largely in ~t with him on tht: discinClion bmvttn the twO
~nses of nMbhJv... One :l.$pecl . however, of SIe;nkellnc,'r prescnf:ltion
need.r funhcr expbn:uion. That U, in dc5cnbing nature:l.$ the ·ontologia.!"
mCllling of IVtlbh.i1llJ, Steinkellner could be misinterpreted as suggesting
that an entity's ·ontological" svabbiva (itt narurc-svabh.iva) is somcnow
"mote real" than its "1ogic:al" wabh411i1S (iu propcny-svabhillilS).
Now, as Steinkdlner has shown (and :as we M.~ also noted), many pas-
sages in Dharmaldni's worXs do indeed indiate that an entity's propcrty-
rwlhiM- are ontologically . ubordin,ue to iu IUture-.m<IMo.i..... The d(;lUlt
of these are passages in which Dharm:ak.ini justifies the conJtruaion of
exdwioru :as the warrantt {p,a,,!"inimitlA} for words. Th:1;[ is, even with-
out a tc:al univ~rW-such as "bl uencss" (,oriuuv.r}-iost2nriated in all the
entities in question. one can correctly and wcfulJy apply a word such ali
81 S.. inkdlrou (1'71:,16); • ... I..."n d iu n... in iibtn• • _ Si,,_ .,mc.in...in. urn J;c
T.u.adlC III bn.richnen, daB mchrac UrAdoea, die: aUf ih,en cpn UlAChc:n hcnlll
Cflupr«hn>d bcd.inll lind, rin wod d~ Wi"'",,! bmoorbri~.·
82 S,nnkdlncr ('nr:117): "(So ~ "'=II ckt 'Svabhlva des Unxhcn.kompkus· .. .abJ
bIoBa, ~ue ...... Si&d fur die T auachr:, doll ..... ochrinban: 'Zusammcn""irun' ¥OIl
Uruchm lUI" Enaog\Ing riner ~ Wirku"8 auadtlidUich _ lin anfant;lo-n Von.\U-
JCrnI"I _ Cf1tspitd"nd bacimmendcn UrsadIcn tu Iqri1ndcn itt.'
IIJ It II impoItant to r«:a1l that -cao.uc-- and - dfca- an :aln:ldylXM'lCC'pU- We ann.oI; ~
dUo cirdc.
196 FOU NDATI ONS OF DHAkMAKl k T I 'S PH ILOSOPH Y
nature-nwMIN = ~
SVAIIHAvA'RATI6AftlDHA , T HE BASIS OF I NFEREN CE 197
philosophy. But I beliNt that we can profiably eomine these ideas further
by formulating an explanation that. while based on his philosophy, high-
lightS ideas or th~ries that att only implicit in his work.
II ICCffiS that me best way to attempt an explanation is to coin a hand-
ful ofheurmic temu. Although they have no dear Sanslui6c equivalents,
these temu wiU prove useful for seeing how cm::a.in aspetts of Dharmak.im'5
"P"I-thmry:art: rt:lev2nt to interpreting the senses of IIJiIbJuJlNI in his phi-
Jo.ophy.
The first tttm is ..bstr4dioll, which we can define as concept-formation
through the exclusion process. The procc:u of :abstraction is based upon an
enol)' conmued as:a subject. Ttut is, whcn we ab5tr:act Mhaving branches"
from ,he concept "nee: we are "selecting" or :abstr:acting amLin causal
ct..:.r:ocrr.ri~io fmm wi rh in rhe ror.:aliry nf fM CloDl du~l!f'driol ",!lr d: rho-:
narure-twbh.twt of a "uee," and we construct the predicate "having
branchcs" on me lwis of those abltracted causal characteristics.
Abstraction, however, can also rest upon anomer process that Dharma-
kim dearly discusses, although he provides no single term for it. To dcscribt
this process, I will introduce a KCOnd heuristic rerm, «/nuptJuJ ctM!ncttlrt.
Through W pl"OCrU of concqnu:a.l ~~. individuals rh:u :iN' ..ceu _
ally distinct art: construed :is if they wert: aU COlUtitutive of :a single indi_
vidual. Hencc, in the case of "having branches; the formation of rh:l[
predicate is implicidy pru:eded by the conceptual coaIc:scc:na of the infin-
itesimal paniclc:s in a manner mat colUlrua; Ihem in tenns of the loality
of causal cluraattistia that is the n:arurt--JVIlbh4J11fof wn:l[ we call a "tree."
This "coaIescencc" must be conceptual, because although me physical char-
:>C'tenst>c. of rhof.e inlinitesim:&l panidet tilow us to co ruickr them,... arU-
ing from ceruin auses and conditions SO:is to have ccm.in common efli:cu.
we are not oblixnJ to consider those infinitesimal panicles in that fashion;
we can, if we choose. consickr each infinitesimal particle individually.
Hen«, when we abstract properties such as "having branches" from a
" tll'(:,~ we art: not $latting with a "tree" as a given. Instead. th~ M
tree- must
:&1.0 be concepru:tlly constructed. This means that the form :u;o n of the
predicate Mhaving branches" as :applied to a "tree- acrually has rwo aspcc;u:
fim. one c:onceprually coaIcscc:s those infinitesimal panicles by considering
them in temu of ccmLin causal characteristics (i.e.• those resuicted [0 wh:ar
we call a "tree"), and second, one abstraas the predic ue "having branches"
from that [o[:a.iity of causal charaaeristia iuelf.
It is impomun to distinguilh co nc.epru:al coale&eenc:e from :mother fun -
damental concqxual process in the "pN;w-theory. I will call mis process
t98 fOUNOATI O NS O F OHAlMAKlllT" s PHILOSOP HY
(lVIlhhillmA) is capable: of producing [he: dfc:cts that att the: basis for [he:
abstraction of th:u predicne. By conceptually reverting this proccu of
abstraction, we can sec th:u the proptrty-Jlldbh4v.r ~having branches is
~uciblc to the: narure-lVII'hh41f4 of the subject that it qualifies. Thus, we
can undertcand that propc-rty-lvabb.iva as subordinate to that naturc-
swbM/HI..
Tlut na~swbh.iva itself. however, is nothing but the totality of causal
characterinia tNt is the ntt itsdf. Thus, for an ap~ to the "Utt's narurc-
IVIlbhliva ~ to scrve as the ontological ground for applying the predicate
"having branches," we must dcmonnrate (he ontological grounds for me
"uC'C· itself. We do so by noting that the concept of a "tree" is formed by
conceptually coalescing individual infinitesimal panicles that have gained
cemUn causal porentials due ro me mannc:r in which they have been pro-
duced. In this way, by ~ucing the concept "uC'C· iudf ro the causal poten-
liab of panicularJ mal have been produced in a certain fashion , we can
ground our ontological cla.ims in paniculars-thc only u1timatdy raI enti-
ties in Dharmwnj'J system-withom hyposwiring some real naturc-
tlNZbh41Jt1 of the ~ trte."
In me end, what we have formubtcd is a principle of ontolofi~4' miw-
tum that appears 10 underlie Dharmakini'J system." Proptnies can be
reduced to the narure-swbh.iv.r of the Jubjea (dIM,.",in) that they qualify.
This amounts to a reduction of the proptnics ro (he subjea iudf, since its
rnllru~swbh.i1l4 is a marker for the totality of the cau.sal c.haracteristia mal
is tNt subject. And if Dharma1drri's ontology does noc allow mat subj«t
iuelf 10 be a particular, a consinent onrotogy would require that it be
reduccd to the particulars mat, by arising from a certain type of cau.sal com-
pia, account fo r the cau.sal potentials of that subject.
One of the advan~ of this way of interpreting Dharma1drri's we of
tIIIIbh.i1Jll is mat it also cn:abks us to actlOunt for an implicit notion of "lev-
els" or "ordcrJ" of concepts in Dharmakirri's work.· A concept at me low-
eSt kw:1. which Yo'(: might call a "first-ordcr" concept, is one that invol\U
only one of the conceptual procesSC$ mentioned above. One can argue mat
"infinitesimal particle." for example, is a firJt-Qrdcr concept. On the Exter-
nal Rnlist view, an infinitesimll panicle is itself a particular, so cooccptual
88 K.tIl.If1II (1'79 and 1991) aIJlIC' thai. hicnrchy of uni-w. is c:kariy mjuiml at leal! by
Oip>1p', vm.ion of tbe ..,..thco.y.
FO UNDATION S O F OHAIlMAKIR.T[ 'S PHILOSO PHY
coalescence is not pan of the formation of dllll concept: unlike tm- colleq)!
" tltt,- th~ conttpt "infinitesimal partid~- does not rtqui~ us to conJtru~
multipl~ parric.uJ.an as constituting a singl~ ~ntity. And if we treat an "infin-
itesimal particle" as a subject. this conupc is of course consuuaed. but nol
through absmcrion, Indeed. tilt on1ycoocqxual plOCt$S applicabI~ to "infin-
itesimal particl~ - would be )d~Juificuion , for m:u concept does present each
infinitesimal p.1rtide as the same: (rIfA) as all other infinitesimal particla.
If Dhannakirri's philosophy is 10 remain coh~rent. this norion of "levels"
or "onkn" of concepts s«ms ineviable, for it appears (0 be th~ only solu-
tion to the problem of tht subjca-shifi. d.i.scusscd earli('l". This shin is the one
thaI occun between the natut'e--nwbh.tllll' of, for a2mpl~. a "water-jug" and
th~ natut'e--sPllbhiv.r of an individual that ell" be construed as a particular
instance of a "Woller-jug." This problem becomes most obvious when we
choose to translate this usage of nwbhilNlu "e:ssencc." To be specific, we will
find it difficuh to account for why an individual identified as a "Woller-jug"
appears to have rwo naruru or "C5S('nces": on th~ on~ hand, as an instl.nCt
of a "Woller-jug." mat individual has the "essencc· of a warer-jug. which
accountS for the faa dlln we can colTlXtly abstraa propenks such as "imptt-
manence" from it. But on the other hand, that individual bnn; other prop-
erties. such as "bdng mad~ in KiSi," thaI are acddemal to a "w:l.ter-jug.· If
~ other prediCltCS art: comctlyapplkd to thal individual, then mer
must
be warnmed by itS essence: by iu csstntt (swbh4l1t1J11) that individual must
have the df"tas that enable us 10 abslf2Ct the property "made in KiSi.-
Now, sincc "made in KiSi" is accidental to being a "water-jug." we can-
nOI say thaI il is th~ naturc-JWbhilloll' or "cs.scncc" of thaI individual qUll
w:l.t"-jug to be made in KiSi, for water-jugs arc not neccuarily made in
KiSi. But it is the nature-lVtIbhtilNl or "essence" of that individual 'lUll indi-
vidual to be made in KifI. This ~double-essencc~ would lead us to con-
clude Ihat mat individual is "essentially" or "by its nature" (swblMlIt"NI)
both "made in Kali" and nOI "made in KiIi"!
By appealing to som~ implicit notion of orders or 1~1s of concepu in
Dturmakirci', thought, we can avoid this contradictory imposition of (W()
narure-swblMV4S or -es.smces- in fWD ways. First, by understanding ~n
Katsura goes on fO (XInfess trult his research on PV4 h2d ·convened" him
to the so.alled JII",lNzrulhttlNi4a. This conversion m.akes good sense, for Ka[-
sun's earlier position did not take adequate .aCCOUnt of the ontological
appeal implicit in the norion of nNlbhilltlpwibltnJh.. Ncvmheless, some of
his earlier suspicions wecc indeed well founded , for as we h.ave s«n,
tINIbhlllltl ("natucc·) in SVIIbhivapraribandhtt cannot iudf be .a panicular,
hence, it must be an exclusion, as Ka[liun argued in his earlier artide. With
the presmtation I haW' given here, we can follow Kauura's initial inruirion
white noting tlur. li~ properry-n!ltbhtiVdli arc nru:curiiy higher order (XIn-
cepts than the narure-tINIbhiva on which they are based. SVIIbhtivtl as naturC'
can still amount to an appeal fO some underlying ontological reality.
9 L Kaaura (1916:17).
92 Kaaun. (1"l:)6-J]).
SVABHAVAPRA TJ.fANDHA: T H E IlASIS Of INFE REN CE 103
931 haw ct.o.m ben: to ~ Dlwmaklm ', UIU.i cwnplr of ~ to"wp nupk:
whidt ihouJd be ~ funilia! It lease to rndm in Nonh AmcriQ.
94 ~ indwion of -is" in u.c. prcdicariorlJ (i.e., -il . uu" ramn than iI a "lfft") "",.,..,.
twO functions, I'lfR. I mean CO !how mal "lrft: nnI if "p•.,.,...cd by a ....ooallli", in
nomina,;", QIC (~). iI ~ KUIIf;:1I a prtdicatc or property (JNmu) .ppI>cd
10 • JUbjea (IN,,,,;,,) in • JU;lrmml "a~ of dw wb;ect" (IN_ ivtlritt; _ chapocr 1,
n. 11,). Second, 1M inclusion of -is" _ to mark. IUC of an abAlXI ~ (Mbty ...'1'9-J.
We o;an mw.void awkward nroIopsma. aJCh ill "produan_" (ip.unw). ·1i~1IOI·
(Ii~"""'). and 10 00. A1thoup mil ~ doe. diminalc 1M diKinaion betw rrn sm e-
mmu "r:lpreWw of IM .w,;.a" and me.."a~ of dw p«diaIC" (IN- .M6_). dw
1mW\00: dilkom« berw.cn t'-IU'~U is lOt in 00 directly rdcvant co the 1:>Iw.......
kim ', inr~lial theo.y, This iI :lmply dcrnotmn ml by.he faa thll DNnrWdn i In::IIQ the
~ _~u u !op;ally imm:hanpble itllirp!. "'~ and "'~1Iu,._,!, (_
PVSV .JPVI,1S: G"4J.II: .l'llISbml in 1M appendu),
W4 FOUNDA TI ONS OF DHHMAKIRTI 'S PHILOSOPH Y
claims that the causes that give rise to 2n entity that "is produced" art
euctly the causes Wt give risc to tnat entity's Mimpermanence." In other
words, if an entity ~is produced, ~ no further causes are needed for irs
"impermanence" because the causes that mm it "produced" also rmkc it
"imperm2Mnt." lndeed, ifboth "is produced" and "is impermanent" are to
be: propeny-mlohdvas of the entity in question. they mwt be: there from
th2t entiry's inception, for otherwise 2 thing would have to ch2nge iI'S
nature-sVll'bh.i!.l<:l, which is by definition unch:tngeable. In orner words, if me
tow ity of the causal chU2Cteristia (i.e., the n2rure-mlohdva) of mat entity
cannot provide the warnnt for the ctmstruaion of me propenies "is imper-
manent" and "is produced." then those propenies will never be: 2pplicable
to that individual.
How is this an argument for the "identity" of "is impermanent" and "is
produced" ~ It would.s«m mat, thw far. Dharmakini appeals to something
qui te similar ro the principle of ontological reduction that I luve coined
above.. llut is. the aUSC'S that give rise to an instance of a "water-jug." for
example. can be: used as the basis for correctly abstracting the COnttpl'S of
both "is impermanent" and "is produced" from that "water-jug." Thw,
rhr:se: abnrxted propttties can be reduced to the subject or propcny-bc:an:r
(JharminJ from which they have been abstnaed. The twO propcnies ~
th us "identical- in that they ~ in fact nothing but that a;,.rmin itself.
The same analysis an be applied to thl." case of " is a tree" and -is a sugar
maple." If \rt can eorrecdy say of an individual th2t it -is 2 sugar maple"
and also th2t il "is a tra:: then what \rt mean is that no funher causes are
n~ry for us to apply one of thest> predicates [ 0 it. In shon. both of
these: p~icates are ontologically reducible to the subject that they qualify.91
Now, if this ontological reduction were all th2t Dharmakin i meant by
"identity,- then we could raise numerow criticisms against this notion.
O ne obviow critici5m is that, on thi5 theory, me inferentt could be: inter-
preted as a tautology.'" If the predicate to be pro~n (uldhyttJ 2nd the evi-
To sec bow one can formu late a spurious infcren« of this kind. we
must fint rtca.Il that the subject (Jh4,..".in) of the proposition to be inferred
(prtuij U) cannot iudfbe the basis for formulating tbe twO forms of con-
comir2nc:c: we must appeal to CUC$ other than those included in the
proposition to be proven." With this in mind. Dbarmakini asks us to
consider a case where we have some nuits of a «min kind before us. and
we arc seeking those: mat bave a puticular taSte. Let us assume. as Dharma·
kini implies. that (he tasre we scdc is a p,;micularly delicious taSte that this
_ ·jllH liu the ~ I ha~ jw:! Q!tfI: III mis illkrena, aU doe: _ III' ripem.iu
of ... hich one ...ut.e. 1'0 speak han lIftn ir>dudcd ill lhe ,ubjc-n under diKuaion
~ TN. brinr; rhc cate, ~ the pm;Iicuc in quation (~;. IMX p<aml,
m.. ..,.;.0....,. .. __ .-ceiw.cl. H_ ...... could m.. inklColloCC .... mi I di-w
Some IpbiloooJli...-nauch II '~I claim, ·In mil f)'pC of litw.rion. doe: _""
in hcm¥"c(IUI1:a$CII of !he p!opt'nJ adcIooed II C"¥id.mcc iI incondwM blol.lK
onot IWpC'CII ~I i. mipi be coon.umandni by. puupilon of one of dot ITWu in
q.....ion.·
&u .hil QlIIIIOI be .he (:tK blo_ therr iI no ....m counrnmandinr; pe.-upnW
~III of III object thai baa hem akt;n II the wbtea of III inu.ua.
· Bu • •• ..,..,., lain rime one ..upl .... ot:. po:rupcU&l ~ 01 the t::Iltt .ha• ..uJ
coon,cmund ,ha. a d lUion ofthc •• ide .... in hctn"8"na>UI o:ua. .
If one ~ ...... such tn. . . . . . . ohe caw. then one .. commirou.s ohe fi...h 01
Oot:iUtmSion bloUK thm:........ud be no ncu:uuy ratriaion (,,~ WI guan&-
I~ the abKnu 01. oowumnandilll perorption in oche. lituatiool II wdl,IWithour
admitting • ....U"....,..ti~I_(:I.1I, ..;Imit only tiw. a o.IttmUy acrivt: insllu,
mental • ...-armatl C'Otu"mmrw;Is ,hil faulty aduaion of dot nidma. If_not 1ft
ac:ti¥t.wu-maI ~ 10 oountermand the~, tbm inrcm- would be unreli·
able: in aU I:a$CII (bcco'I»" IllY "iduo« mip' ~IIWIy be ooun,crmandoed).
Insuad of relying on simple nonob.crvalMwl. ,~aclusloa of cbt ~ from
<hr ....~.""' •....,... o:aoca requires the kind oicaWn.y (-.,.J~ in. thc~ .
mioution .•~ is the c:sublisIunml of rht cvidma only when the probandum is
aablUhftI.· OM n«d! .hit kind 01 anainty blowe. although ~ _y atabIw.
mrouch ooo .......... . Mion (.-"""'".JJ",)I~' I~ r.-idcna- i. abtml in IOInC M(.......
- . . CMU. i. q DOl the~ poon" thaI m., mdma, ..t.ich iI not (d-oown 10 be)
ItbmIIO the probmdum (by ..... y of _i\ ...,..,;.....Jborj, it ab.ml in all ....,......
- . . cuu. Hena. due ,odoubt. mm, iI no codwion of r=Wnokr~ in ....·
~ caws, al>d mil on&kes io: midading. ("'''01 wH6.Jw ·"..,.,t.",'\>lr ~
.,.-., ..;; IMMrHJI '""rffl!I f7krlNfMtwI.,..u;.i>; .,,,, I,,.,,,Ji'UhI rW:r rdtu
,-ullJ '''''If'l'I1Jhj "" ~i1UqM/ d,u''''',,..w-.tr1tIJ H ~ i,; I ."qi
"~1MqII,.1q1U",!" ~ ..ftIltJiiblw ;.01,-,..... I' ;'; ...,..". ",.MHd,~ I
...
,..."..,.Lf'hiM.a..,.~"'iIJ""I /~~'lrlliair / ~U.wJ ..ti
err l MN ~ .1i,~ 1 ~rnn~MJIIIh 1.,-" "..~",
4jj6d'_ 1 4"r'!4IMJJMM _".~ 1 "",tirJur ,., sWUM nw WiiM ...", iii
whiMh-..,.. ~ 1."y.J.",6hb ,., ilwriJ .iU., .j#y~ 4f7 .".ti""".
,Q,.q. uJ..... ~ ~ ..".ti~ ~~"IM}"' .
"a: Tillcmaru: (I"a:Si and n.61.
S VAIHAVAPRATIIANDHA : T HE BA SI S O F I NFEREN CE 10']
kind offruit can have whe:n ripe. Dhaflnakjrti th~n asks us to consider twO
different contats. The first contot il; me practical simation of deciding
which fruia on a particular trrc will have thai special tll5fe. We are uncon·
cemed wilh other fruia, and thus the fru its on this tree form the ~ntirety
of our induction domain. With this in mind, we taste a few fruiu from all
the branches of the nrc, and we find that the fruits we sdected from a
particular branch all have Ihal delicious tasle, while allihe others we have
lUted do not. Pointing to that one branch, we then infer, "all thc:sc fruits
ha~ that delicious tute beaUS( they come from mat same branch, like the
ones we ha~ lUted, and unlike all me others.- Now, it just 50 happens
that only that branch does have delicious tasting fruits (perhaps only mat
bra.nch received enough light), Hence, when we make our test for ncg:acive
conoomiWlCC. we will tk flU"tD f:a..i11O find any instance m.u contradiCts our
inference: any fruit that f:a..ils 10 be on that branch will aho fail to have
thaI lUte. But docs mil; mean that all the fruiu on that branch do have ,hat
WIt:?
[n the IeCOnd comar, we are anending to the color of the fruiu in front
of us. I..e!: us suppose mal they are blueberries. and that we sce:k sweet ones.
We tlSte a few of various hues, and OUt observations suggest that all blue-
berria with a panicular blue·bladc hue are SWtt( , while mhers do not.
Selecting the remaining benies of tlUt color, we claim, -all these blueber-
ries are SWttC beaUS( they have mat same blue-black shade.- Now, only
blueberries of that shade (I'" be sweet, 50 we have selected all of me possi-
ble candtdatc:s as pan of our proposition 10 be proven . For this reason, tk
flUt4 the inference passes the test for negative concomitance: in me sample
available for obstrvacion, all non-blue-black blueberries are not S'A"Oet , Thus,
in both thil; contat and me one above, we might assume our inf'c.rences (0
be wdl formed.
Any bluebeny-gourmand, however, knows that even a blue-black blue-
berry can be brusquely bitter. But if we posit an idemi£), relation on the
baJi.s of ontologic:aJ reduaion, we would have: no way of avoiding this type
off.illacious inference. The color and lUte may be "identical" in clm they
are both reducible to the subject that they qualify. But if we happen to ini-
rialIy formulate the positive: concomitance of the pervasion relation bctwccn
color and taste on the basil; or blueberri~ that are both bluc--black and
sweet, we would ha~ no reason," UtOrtjcct that formulation. One might
claim mal the sample should be broadened beforr we draw our inference,
but broa<kning the sample docs not in itself guaramcc me reliabili£), of (he
pervasion relation. We could be' unlucky (or luckf.) and encounter only
108 fO UNDATION S Of DIIAIM AKIRTI 'S PHILOSOPHY
100 1 h.ovc . 'mpLined .he dioaouion h"u. If ............ 0 ~ mo<e ..,.,......e, .... m .... ..,..
n(l(e duol . fOl' Dbannalrlni (PVI.9-IO and PVSV .J tTl. : G '7.1H.1j). Ihis type of infet·
ma acllwly employJ dl"CCI...,.,idmc.c, It.~ta. Dh'llmU;ni is obIi&cd 10 inletpr~ this
type of inJ"nm« aJ hued..pt'ln It.",.;,.m. because. according 10 the Abhidharma typal.
OIY tlAl nc adopu and Miapu 10 Erlemal Rnlism, infinilesimal panicles of color au
di"intl tTom ;nfinilctimal panicla of wle. B UI as Haya and Gillon have Ihown
(lnl :6,), Ihis I)'IK of inf"raKe (whi.eh •• ""1.9-10 is an infercDU fTom lute 10 color)
implicitly rd~ on the U$I: of nw.... ~. w-hcrcby one infen lhocapacity (J'I';O '1)
oflhe roIot-alomllO act aJ IlIpportlnl con4illoru for 11K IUIC... IO ....., 'T'he dJffi:...,nce In
11K prae... ~_I"" in(vcnec (rom color 10 lU!e--i. mal we ue no< oaly infernn,.
capacilY, INI abo the fact thai lhal Clpa(i'Y is fUUy aclival~. Thus, 10 Hilt !he praenl
aamplc mort prcciKly. _Ihoold nolt tim 11K IWO proptr'Y.,.•• Mi... in qualion lit
no< ptM";n;nl '0 the blueberry, INI r:nhrr 10 11K colot-atonu of 1M blueberry. Sp«if'i.
cally. Ihe ptrccption of color allows one It> dettnnine thaI thOM: color·alOIN an: actUally
1\>CIled in a aU$:ll compla whereby Ih.1 potmtial ;1 activated. HenO', the property.
,.,.Mi"., actinl &I rvidenu in our IirJbciow infcuna: of lurt from color illnc proptrty
of "ha.,inl,n activated potcrtliallo produo;c I d'llk-bJuc color." Ulinr; Ihil Pf0pt.'Y·
,.bbtI"", II cvkIma:, another propmy·nwMl"", is fallariouJ,ly infarm, namc:ly. <he: prop-
erty· 'tV~N. "having an xliv,ued potenlw 10 aa II the: lupporting condil;"n for the
produaion of a ddicious Wft. " 1M complai!)' of thit: infem>(:l' .t.ouId mili il c:ka, why
Dharmakirti rudy chootct: 10 ~nl his anaIysil'l an 1I0mk Irvd. It ;1 woMh nocifl&
that . unlike the ;nfercnu from color 10 WI(, 1M in~na (rom color [0 ".ipt" is more
dim;tly an in(crena by .....iN..,..rvidelKc. • inee "nptons" refen 10 • Itlrufotmltion
applicable 10 all thc .IOms corulrued as the blueberl')'.
tOJ",.. .. ,I.e p. ofdai.... ...ch_
... nodUn,othet [than an ~I or a propmy._MoIJW\ an lCfVC II cvi.oknu bcc:ow,c
!he I'CllricUon of unx.:ompanim non-ariA"I docs not pcmoin 10 IIAI w'-c: naNre iI
not rcbled 10 11K probandum. IPVSV .J""1.I0l G:I.I1-IJ.: /inUI"~ Iwt..rp....8
iti l .".ti4-Jl'wwWM_Jt;";NlW.""",ry.1IIibh4N~ .
....,... one who acuptt thai, by muon of Ihe ntpl»n of one: crtti!}' bud! II fire}.
another [1UCb 1I1fnOI<e] 11 otp.rd. mUll ..... MUpI ma. ohcn: is 0IHnc: ._~;.
HNiIM peruininl M""ccn lhote tntitiea. [PYSV PVI.14: G:IO.l)-14: .. , .......
JWO
";~IIJ'6""rffl" w,.,. ,.,. Ukit lAd ~ 'rvJrAtiI¥NihII ." q~ .
SVAlIHA"VAPRATIlIANDHA: THE BASIS OF INFEREN CE
w.bJMw-cvidence and sec dut. in :I. lubder form. the s:ame concern is evi-
dent. Thu is, the ddinition itsdfis cr:afted in a manner dut aims to move
beyond the simple onrologial identity ofevkIc:nce to predicate. Let us recall
me definicion as found in PrllmrIt;Jllvirttiltll:
I inuead argue WI the lerm "mete" does indeed have:l. "logical" func-
cion, in the th"l it restricu the evidence ro the pn:dieate by p~en t
K'nR'
1004 On me ...., of ,.._M.l... and if:!: cqlli...mn1S (ap«ialty wilhb.), Itt smnkdlna
Cl9I41 md IWllIa (UIClJ). Cua when dICK mmpoondJ afr applied as doacriprioru of the
n'idmor; Ilte 1M II"1(II( .devanl 10 (M,lf ddaulion, bul dICK (Ompollnds may alto Km' ;U
doacriprioru of dw: pmIigIR 10 "" provm (~ 1M larm ~ is puticubriy
common wirh EN: compound .. 4b1v... For. paDdip.u.:~. Itt PVSV -J1'\'1.I 7-J.8:
G:I'.''''):
WIwn tnc pmiic.&tc 10 "" p~ is in thai w;lYacabIishcd 10 ""either the I,..]Ml..
of dw: CYidmor Of EN: ClUK of the Mdmcr, rhm IlI"Ir an IftIOfI as f'oIIo-: ifimP'"-
IIWICrIOr itat..ml in attrtain thl"" then tlw thin(l. no! conwuaaf; iffin: .a!.rnl.
~. oolmOltr. Ono: can raJ(\II in this WIIylxo;.o...., (Iti) dw: pmlic:au: in qtlaOOn
is m.. evidence', ....IIIM.. or (:I""" A.nd ~ c:ouId tho: ~ IICXlIr wirtlClUl ill
ClWD p'.,t, -nwMiw Of CKIfC HCIIOe. t¥m withoola Ioxw (~). tlw Mp!;~
COftCOmitana (:In Rill br ckmorulnlrci in me contrary ewnpk . [,..IIM' ,~.n~
vAt,.... 1tm......H wllti:y.nWMl.r Itp+ ~ _ M-Iti UJ".,..Mht ell ~-; 1
ttUiJI ftj ,. "'JIII_.w.. ..",., wl l Mm.,!, ,....,. _ .........IJIO blJlO ...'""""'"' ~
it] lirwJ- IllllMt"/fllfi ~"u ,_;.o".,ti ..,.,ti'"4 iI.
1050_SlCinkdhxr (I~n).
111 FOU N DATIO NS OF DHAI.MAKIRTI ·S PHILOSO PHY
Thus, if the evidence (" is a sugar maple1 is the osence of the prcdicate to
M prov~n ("is :I I~"), we :In' actually claiming WI all treef are Jugar
maples. which leaves us with a numbcr of Stidty problenu. For cx:ampk.
rithrr maple syrup can br mw from the 53p of:an oak trtt; or oaks arc not
acrually mcs; or if oW arc trees but do not produce maple ~p, lhen the
ability to produce the sap ror maple syrup is not one of the caus:al charac-
tcristia required for an individual to bc call~ a "sugar maple." The prob-
I('m h('r(' i. :on inversion of thc relation between pe!'V2ded (vy4P7"') 2o.l'id
pervader (.".JMlt~). ,''' To usc the 11fl~ o f cxlclUionl, we cxpect the evi-
dence to bc of smaller or equal extension wim me pred.icue. such that all
insranca of the evidence arc included in the CJ:tension of the predicate: as
in the claim, "all sugar maples are trees." But in ba. by ~ying that the evi-
dence is the rsscnce of the pr~ica le, we have inven~ this relationship,
which amounu 10 me claim, "alltree5 are sugar maples."
A pocsible f'QIp<)rue to thil: criticil:m is (0 m:um:un WI it improperly
COnslNeS :an abstract predicate (such as ~tl!4, "is a ncc" or "uttnessi
as having an essr:ncc. Since such predicateS arc in faa aclusions, they can-
not be said to have an essence al all.'· Although J agree with this criticism,
one must bc careful about interpreting iq implkations. OcarIy. U " Irter\CS5"
cannot ilJdfbcaran esKna. then we mwt modify our inlcrprcution orthc
cbim Wt the evidence is chana('ri~ :;II wlSWI,/M",. relali~ 10 rhe pred-
icale. Instead, we should undcrsttnd the compound tllulMbhd". not in
terms of the evidence and predicate thcmsdves, but r.Hher in lerms of
instances that instantiate the evidence and prcdicate. Thus, in lieu of say·
1fJ7 Hays fim Iflade. point of dW ...., h 9l7;,11- JU). and lwua (&00,", has cbrifoal •
numbn of iu nmifoadono..
114 FOUNDATION S OF DHAIlMAKiIlTr5 PHILOSOPH Y
as option ).'" Iwata cites one case. ho~r. that it unambiguously t/41
amenable to option }. h is a vmc from the third dupter of the Praml~
INirttilil:
What W( havt tT2lUlatro hert as "the Iwbhillllof that thing" is aaually the
compound Iillsllilbbiva in the original Sanskrit_The grunmar hert it such
that W( cannot possibly interpm I4Jmlbhtiw to mean -has that thing as its
Wflbhi"tf (option }). And if we inttrprtt IfJiIbhilitl as narurt-lViIbh.i"" or
essence. then W( have dearly arrived at option I , "the evidence is the essence:
of the predicate. ~ But perhaps we should first considtr the verst thaI coma
ne n in the Prtl1M!'4Nmilil.·
For Dharmaldni the main point of thit verst is the assertion that the -char-
acrer- of dfcct-cvidtnce and IVilbhiJNl.n'idtnce must apply to all cases of
rtliable evidence; in other words. all forms of rdiabk evidence must oon-
fo rm to these two in some fashion. For our purposes. however what is most
inlertSting ~ it that he again usc:s me term IViIbh4va, but now fIIiIbhiw
rerm 710110 I}" nnJnrrt. bl4t to 1M prrJicillt. Thus, in PV}.70. he laYJ that
me evidma: is the: n..bhi/lll of the prediOK, but in PV}.71. he laYJ that the
1091W::1I.1~ -osniu:l duol in mol! CUD op!ion I (I.e. "'lI~ is no!: ~Iy
RqIIimI by m., ~ of m., J>UAF in '1uaOOn. Commomuillo to. exunpk. on an OClaIr-
rcnot of ",""",Miw in H 8 (,.'-4), IWilla IIOlCS, -of"""""" i! .. abo 1*Jibk: «> inu:.pfCf
till_INN U IIrr.r.Ino-mhicompound [i.e, option }I."
I ] 0 PV)'1'" ,.uJ.> u.u.;, tiMIIM .". f'Wtitir ->M!t bwit f "'" II.c,.'!' III,. jlJ4 _
w.. fi .....WI If. Colt<! br I""a (100):1), ft.}j). I ba"", lnIUbud this and dw follow-
i", ...... in ra:otd..;do ..... ;..~ of£lev.endrabooddtoi (,f<)blffl .
tU M.a.
] I I PVJ.71' ~"'i_,!, -MJ_'!' _ ,,;U ~ _Mttow.!o f"... ~ ~ 1117" tInI"'!'
1c"'!"'M II.
SVA 'HA'VA PRATJ.tANDHA.: THE BASIS OF INF ERENCE 117
11211111 atlUmcm II made dKwbm: In I'V and I'VSV. Xc, for aatnpk, I'VSV U
PVI.I7-lI; G:lI.f..., (tnIutucd abo-oc, n.I04/.
118 fOU NDATIONS O f DH"aMAdRT I 'S PH I LOS OPH Y
that, while wt ha~ numerous textual rtaSOns to Stt rwo difftttnt mn-nings
in (he term svabhiiwt, wt should not forget mat Dharmilini never clearly
stares that he is using the term in two distinct ways. Often the context of a
usage forcefully indicates me preferred meaning of swbhllllll (as · propnty"
or -nature' . but in many crucial cases, including rhe verses that we ha~
just considered, Dharmakini's words offtt no unequivocal choice. We nced
to k«p this ambiguity in mind. fur it may suggest thar Dhanna1cini has
deliberatdy hidden some specific pu~ in this equivocation-a purpose
that modem inrerprttm ofDharmaldni have yet 10 detecr. BUI it nuy also
indicate that Dharmakini's theory of inference through nwbh4~dencc:
s..imply was not very well worked out, With this in mind, kr: us examine
$Orne potential problems.
11411 is worth llOOIi"l dw,. ~ thoup DlwmUiru doe. IlOOI ap/icitly lOrmulate thU pro-
«dun: in rvsv,:an inchoale f'onn ofil is IUund al PVI.191:and PVSV M rit.
no FOU NDATION S O F D HAlMAKlRTI 'S PH ILOSO PHY
_u.di" (f.ikh.uJinulttvtlj. On thi ~ In~it. 'Nt would verify the pervasion rd.uion
in O Uf inlf:rt:ncc by me nonpelccpnon of Khaving branches and such· as a
correct predicate of non-caa. We know that the range and vicissimdcs of
our obscrv:l.tioru should have no impact On the outcome of this tCSt. so we
JUSt c:u.mine thl'tt non-caa that happen to be at hand: a dog. a car, and a
hooligan. We find that the causal characteristics necessary for predicating
Kb""n".hes "nd such" " .... nOf per«ivt:d in "ny nf fh~ CL~ . _'" _ Ih L L~
conclude that the sugar maple in front of us is a cat. Obviously, something
is flOt quite right here.
The problem, of course, is that we have failed to cstablish that "having
branchcs and such- is an -wenda!" or ptl'Vllding proptny (VJlipalta-
Jhamus), flOt only of a sugar maple (the evidence) , but also of a cat (the
predicate). In other word.. before resling for the nonpcrception of th:tt per_
vading propt'ftY, ~ first need to know whether "having branches and such"
is a pervading property of cats such that its prcdicability is relevant to decid-
ing whether the items we tCSt are indeed contradictorics of a cat.
But if Steinkdlner's account is COIT«t (and I think it is), then the pro-
cedure offered by Dh:umakini docs not give us any means of determining
how to sdeo;t tho: pcJv.lding property for whkh 10 test. This is not terribly
surprising. for the only way we could determine whether, for example,
Khaving branchcs" is a pervading property of a cat is by using svabh4va<vi-
dencc, as in: "This hasJdocs not have branches and such because il is a cat."
In shon, we would faU into an infinite regress. Perhaps it is for this reason
that Dh:umwni docs not include I way to determine which ptl'Vllding
property is rclCVl1nt to our tCSt, but in not doing so, the tCSt obviously
rt:rnains incondwiyo:.
Thus, it appors that Dhannakini has once again f.a.iled in the second pan
of hiJ ask. That is, although his formulation of the identity relation that
underlies the usc of IViIbh.iv4-cvidencc may provide a theoretical guanmC'C
for the accuracy of such inferences, he has not provided us with a rt:liable
procedure for determining when thlt relation is in placc. This failure is
nOt, howcvc:r, panicularly damning. fo r it rc::so on an intracu.ble problem:
namdy, the distinction between the a.scntial and the accidental.m
One possible response to this probkm is to fall bade upon an appca1 to
worldly convt:ntion (IIJIIvahtlra, IoItaprasUJdha, etc.) as the basis for the iden-
tity rdation. In cff«t, an inference by Iwbh.iva-evidencc becomcs an exer-
cise in ltarning the proper definition of the terms involved. While this
FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
prolnbly is /HIr1 of the identity relation. 1II it does nOi entirely solve our
problem. Fint, we must determiot: which properties of any given individ·
uaI arc: germane 10 the predicale in question; is hdght, for o:ample, relevant
to being a me, and if so, how shon mighl a [rc:c be? This ilITlounl'S to the;
rhorny problem of describing oactly what constitutes a worldly conven-
rion. &cond, if we N.sc: idrnlity wholJy on ronvefuion, we lose the OntO-
logK:al appc:a1 implicit in Dharmaldni's nOlion of wabhivapr.lilNtIUiha. ln
shon. the oerci5e of reasoning ITom the filet of "existence" to "momen-
G1rindS" would be reduced to a mere word-game ofleaming the accepted
rules for applying these terms. and it would ignore the question of whether
,hose con~ntions are based upon the causal charaaerlsda of u1timardy rcaJ
paniculars. On Dharmwni's view, to ignore (hat que:uion would be dis-
astrous, for if WIlbhJllflpr.,ilNurtlJM is. as we shall $«. to act u the: warrant
for the trustworthiness of inference. it must enable inference to lead us to
things that we can use to accomplish our goals. It must, in other words, lead
us to particuWs th:n arc capable of the telie functions that we seck. Thil
brings us to the question of the instrumentality (prifPllilJJtl) of inference
and ('\'en of JXrcqxion; why should we believe that either is a ltwrwonhy
means of knowled;e? In the nexl chapler we will examine lU(;h quenio(U.
118 Ser:, !'or t:nmpk, IN: p"FIfII' fTom D1unnocora'l ~..pUcif.td by Sctinkdlntt
!l991;n,sI), H~. DtwmotI2l1 dearly ~ IIw: ~ UAF of ganVlml>oN 10 be:
,.ur;i.aI
'v;,~ by .......~iookoo;c. n..: I". 5'" (NI)T:"" r~») mob (d. ..... ,nu...
lMion by S.rinlrdlntr, ' ; N .);
In ,hil rtprel, mppow ,Iu"
in a pba abwIdanl in 1iItU,..1OfMOI'Iot..now. a nil
1i".u,.10 I dimwit who is I.Inhmilia "';,h m., lam (.,."whmt) "Ii'!fltl"t" and ma.
Idh him, ",his ill trtt." When lIu, happmI. ,be- dim"';l, bccalJlC of hil aupidil)',
bd~!Iu, me: IiIflUll'I tallnaa iuboa ICnWllie ICWK (" i",itUjb lhoe fUm "11ft."
Thil dull iI ma. ;rllrod1.iUd by ,1w: 0!h0 fdLow 10 tIw: &a dut!hac o1I li~i ~
10', 'r - ....... oerruon.tcc:o..... 1Or ~ ............ "1.-" .0 il . ~...t.e. kl.1ow
"1"htn- ~ no od\tf ~R\aIllic 12:aeJ, IUCh II mc: indica,oed f,''!fItI'" 011_ and
.u.yt::
ouch, for 1M appIic:arion ofllw: tam ' \Ift' 10 d'" indiridlUll""cad" iu mctt /" w~
nell u,be- .muncie CIUK. ' In ocher oronia. eM ImWldc CMIK ulhoe faa of pcMWM"
incbrancNs and IUch wNo:h;' irw:luded in WOOiMXpI ."..,.." [",,,..,,•• ,Ji~
tJm iNiiM1i'!fltl~~'*'ltnwriiwnl",Ii~"..,M dJW) t • • '"
S
O FA. WI! HAVE EXAM INED some issues in Dhannakini's ontology by
way of his views on instru~nw objttts (pr(l"'~); ~ have also dis-
cussed flNfbhlvaprmibantihll as the basis for his theory of infttena:. In
this chaplet, we will inquire imo the fou.mbrioru o f Dh:umaldni', episte-
mology. More specifically, w.! shall examine what it is mat justifies the claim
that perception and inference arc each 2n instrurncm of knowtrogc
(p1'ltmJity). In the language of PrlIImlQa Theory, this is Ibe question of
~i nstrumcntalii}'· {pr4m4!fJ4}, an issue that Pr.tmil).a Theorists generally
tmu as a tOp ic: of discw.sion in and ofirsclf. It a mc:crru flOl the workings of
an a1legN instrument of knowl«l~ it a pt."lCq)tion. an in~nce. the
smsc faculties, a sacred tal, or ~me other candidate-Oul r.athcr what it is
mal ;wrifies the d.ai.m WI one or another of th~ is ind~ an itutrurncnt
of knowledge. To introduce ,he ccmnl iSliues al st3kc in discourse on
irut.nunmtaliry {pri'm4tryrt}. YooT will first discuss some critical problems of
rransl.:uion. I.n doing 50, we will darify thc manner in which Dharmakini's
theory of innrumcnrality is tied to his conception of the uhinutc goa.! of
Buddhist pracrio:. 1bc remncc ofDharnu.lcirti'l ulrimatc pi wililcad Ul
to thc problem of a tttming circularity in Dhannakirtj'l theory of imlN-
men tality, which in tum will prompr Ul IO consider thc implicitly axiolog-
ical contat o rOharmakini'1 thought. With aU thac ill$ua; in h;md, we will
then assay a dOSt: analysis ofDharnu.lcirti's theory of ilUtrumcnraiit)' it5c1f.
U)
11.4 FOUNDAT IONS OF DHARMAKIRTI 'S PHILOSOPHY
grammatical concerns that frame thili discourse across all South Asian tra·
didons. Specifically. Pramil)a Theorists across all mditions recognize thai
the gnmmatK:al fo rm of the word prllmi!", rinstrumcnl of knowledWt)
ili indicati~ of an inslrument (lAr.!U) in a verbal action. For aample. in
the sentence. '" cut Ihe tree with an axe," the rerm "axe" would be in the
instrumental case in Sanskrit. Liktwisc. Prami!:,a TheoriSts universally
unlkrst:lnd the issue of prtlm4!'y. as mated to the claim that the thing in
question- the sense faculties. scriprurc, an act of inferell«. or some other
candidate-is the instrument (prtlm41J11) in an act of knowing (,rlltd or
pl'ltmili). Thus. when arguing for tht prtlfNi!'J4 of one or another candKbte,
these philosophers ineviably rum to the quesbon of whether thai candidate
manifests the propenies of a grammatK:al instrument. In extended discw:·
sions on prtlm4!fYt1. (hili concern Ic:a<b inevitably to the ubiquirotu citation
of s.iJJHUfllllmtlh114 the widdy accepted defining characteristic of a gram·
nutical instrument as the "most prominent causal factor" in an action!
While the starus of a P'flnu/!'" (an "instrument ofk.now\edgej as a gram.
nutical ilUtrument dearly frames all ate:nded discussions of what coruu·
tUtes a pl'ltmJi!fll, a Pr.uni !:,a Theorist is not thereby (lIIlit,. to argue that
grammatial irutrumcnlality-i.e.• the fact of bcing the "most prominent
ausal factor" (wJh.1t414m.) in an action-is our best or only way to cs~
luh thai the candidate in question is a prtlm41J11. As we will K'C, Dharmaldni
is a thinker who l'CIiisu thili standHd approadl . and he proposes the qual.
ity of "immediacy- (.1I1inlflhil4tv4) as an alternative: mark ofinsuumcntal·
iry. Ncverthdess. the manner in which Dharmakini makes this argument
presumes that the grammatical issue of being the "most prominent causal
factor- cannot simply be ignored: in olhu words, the innovative move
toward immediacy is only made: possible by the need to COntest what is
meant by the grammatial inStrumentaliry of a prllmA~ Our interpretive
lesson here is thus that, if ~ arc to appreciate fully what it means For
Dhumakirti to claim thai cem.in kinds of cognitions arc instruments of
knowledge, we musl keep track of Ihe grammatical concenu that frame
that dixussion.
~ ....
IN ST RUMENTALITY: J UST IFYIN G THE SOU RC ES O f KNOWLEDGE 11.5
yogic percq nion (yofiprlltytl*",), the content of some yogic cognitions aTC'
false in that their contents aTC' no mort Mtrue" th:tn a hallucination. NeveT-
Iheles.s, such cognido ns ti4 havC' prlimli1JJIl bec:all5C' they pl"C'dictably lead to
desirC'd outcomes.' In anorhC'Tcoma l, Dharmakirti discusses IhC' well
known position mal, in IhC' case of SC'Ilsory percrption, Ihe sensory organ
is the instrument of knowlC"d~ {prmnfi'JA}, and as such, it has prlimll,!J4.
Dharmoorti rC'jecu this pmilion. but hC' is obligC"d to takC' it seriously
bec2use it is a plausible approach 10 the issue of prlimll'.'JIl. Can We". how-
ever, plausibly maimain that a sensory organ has Mtruth -? How can a ph)'$-
leal object rhat does nOI haVe" anything like a propositional or senll:nrial
u ructUTC' be" considcrC"d eitheT"true" or "faIse-? Clearly. then. We" run the
risk of considerable confusion when We" translatC' pnl1nl1'!J1l as "(fUm.'"
When v.'C' tum to "validity: anomer common translation for pramli!lJll.
we encounter many of the same problems. On this translation. a pramR'JA
is a "valid cognition" or "means of valid cognition," and although the C'X:llct
meaning of "vaHd" here is nol of'ttn cl.ariflC"d. in most cases "validity" in pan
means "vcridic:aJity." Again. in the case of a percq>lual judgment. we 'WOuld
think thai "validity" should apply: if I sec Ihe color blue, and on the basis
of that pcTCq)tu:li CO ntCn! I form the judgment. "that is blue: should we
not consider this judgment "valid"? Even though it SC'tms ve:ridical in itself,
fo r Dharmaldrti that judgment is nelt "valid " bec2use il is n OI a prllm4'JA'
O n the other hand. an adept fJotin) can induce a mental perception in
which he dirKtly intuits Ihe conc~pl of impermanence. bur since the
Idnlitrn. M I>(Ja pl:K.n is "mJe: in dw Ihty arc all aff~ by the "internal distortion"
(."ury.,tn.). ~ the coriti.-.: ~ in puapl:ual """ilion appall moneousIy 10
~ Q tcmal h~ PVJ,lS~>,11. For a lre~unen l of ,uc.h iuua in rd~>on etp«ially 10
Kamalatila', appropNtion ofDturmUini', Ihc-o:wy, _ Funayun.-a (Im1 .
.( This up«I oiDtwnWdni'J Ihrofy of yosk per«pl'ion (ill found al PVJ.lII-lI6) Iw ra'
10 be adeqlUld y diKwKd. TIw: Uy _ ilNJ.l1S: -rhnd'orr. mUlo whid!. On( rnrdia.
ti..-dy conditioN onacIf, wbnhc. i, be real or ",nreal. will ~I in a c:ku. I'IOn<oouI"'u.al
"""ilion whO! that nwdiali.-.: oondilionins rncha iu allmination" (,-.wI M.iu",
.Wtiu'!''''~~ft"'Mi""'J 'I f iN.."..nllif/WIWMflU ,phMfJiev-8iJIMI.,w1). One
misfn asily miKolUI,," the tenn Un;,.1-Io.n to lMatI thaI 11M: ,..,,1i1ll of 11M: CIOJf'ifion, is
",rue:: bul Dhannakirti rn.aka;1 quilC dear d\;.t in ~>on to d.t:i. phenomnW col'llmu
alone. yocic pcrorpcionI <lrc indiSlinsuiJhabk 110m dlol: ha.lI ",cinalioru of a lovesick peqon.
IrtJlead. Obannaki rti $pfcifict (PVJ.1.U) thai roPe paal"'iom;uc ,,.,...... breal.lle d.cy
:He "1I'\UlWOnhy· (,.".Mi.',,), and _will_ that ,his Itrm rd"m to the resuiuobtained (01
obWnable1 lhrousf! tIM: oosnil>on.. M(>IC _ dy. Woo {zooJ1 loucha on thc:K i _
,Acrordlns 10~. the nodon lhal lhe ImJ( facultkI:on: th=ueivu ,.._~
is raiJed by Dhannakirti II PV:t..J (PVP:Jb7ff, I « bdow. n.7s).
INSTRUMENTALITY, JUST l fY1NC THE SOU ftCES O F KNO WLEDGE 117
6 Of (:ou.w. tht ~ ' vaIid- doa oftU KlInt tmlpQ", (and vny b.>d) puN when _ con·
Jidc:r dx nocioo mat t~ ICNC orpna Ihctmd_ can pbu.ibly ~ CDNItufti as In,u>:W' if
I haft ClwactI and .... um., hai.., u",~, my cyc:s ...., invalid bn:::t.u. I am an innlid..
If I COfpIC IMdICfUC orpm dx)' ~ would no!: be: valid. Whr. lkaUK they would
Ju,¥I: paMCd theil expiration date.
7 Sec Porttr (,,--.,,.6). wIxte 1x .........1u:
".,u /i.e:., an insuncr oi, copUOOn for wt.ich I ~ ill tht inlllUnx'lI) may
• •• I
noo: c:ormpoad...;m tIw llatlln of thinp {it ill -CIDftIp"ubk.n1h nn:w," as MolwlI),
~ il). "WIxdw:r or noc ,,.,u!'J'6 is c:onfincd in ics appIiallion 10 l~fOQKI
wt.ich mribuu: 10 theil colllmt propmia which an ob;«t COI lcspolldin& 10 mat con·
mil amWIy Iw II noo: I mat'CI"ofthtddinition of ~ {u il ;,Wcm 10 bc: I mal ·
ler of !he dcliniUon of ·,.... th· in WQlmI eo<Itmlponl}' d>(I\~II}. bul corutirulft
ram.:r fiuthn dxoty aboul .,J,fd, rwarmesK:I Alisf'y 1M putpolCl motiYlri"l'Ixm.
N~ <hi........... [d>cJ lim,," lhcoty .....kh r<q .. itu __ oupondcncc ;" .... UK·
~ OM: othn 1)'Nmu.1UCh u BuddhiJm and Advaiu, do not.
118 FOUNDAT IONS OF DHAItM .... "'IRTI'S PHILOSO PHY
8 Ponei' ill ~ primarily with the question of....nnw u\lm ill. r.. ' pry <»ndirion
for a copi.ion [() I,c, • ,,..,,,., i.~.. the reruh of the application of I , ......'!"'- H~ noIQ
(' ~)1I1.
To d«i.c:k ..Mthn uum ill • condirion lOr an a.....-.rmeM [() I,c, a, ...... OM
nealAl)'
m.... ardUIIy IWHI the rdation whid! a fN"" mwr beu 10 iu OXItcn •. W~.nnt a
.cU.ioo. R.. whid! holds bd"cu. any ,rlI1Nland iu CDntcnl,' rdcion which can Iw
admiutd by C'<'n'f Indian phi~ to Idd bduuo, a".,uand iu «>nlfti. ~
kfII of the pouDc:ubr theory M 01" any od>cr Indian philotopher prop<iKI aboul the
na.u~ of 01" propa- anal,.,..
of R.
UU","C 10 repracnr «>nl<:flt and or
10 "'PI_III the cosniOOn {i.~.. j /U"'" iudi. POIIH
"'mcm!hal Rbe Khmwized UUII (I~}n):
R iff J apprehe...... (Ii•• " _ ..to. 0... 1 C in o.cc<>rdaac.
J -i.n ,h. P"o-poH 'M'
mor;"ltd J.
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 229
;j;UIlmatical is~ue~ that are the typical poi~t of ~epar:ure for extended dis-
lttssions on prama1}ya. Nevertheless, Potter s baslc pomt should not be lost:
i~1fpose does indeed playa significant role in all theories of priimii1}ya.
a ence , with both the notion of purpose and the importance of grammat-
~ concerns in mind, we will translate priimii1}ya as "instrumentality"-
m~t which warrants the claim that perception, inference, or some other
~didate is an instrument of knowledge or pramii1}a.
;9 In Dharmakirti's work, rhis notion is best illustrated by rhe metaphor of rhe eunuch (see
:!lVI.2IO-2II and PVSV ad cit.; translated below, 3IO). See also the discussion of purpose in
,fhapter I (45ff).
;to The highest goal attained through such spiritual practices is variously described by terms
§uch as mok!a ("liberation"), naifreyasa ("rhe highest good"), svarga ("heaven") and so on. See
the fine study by Chaltravarrhi Ram-Prasad (200I) for details.
l}O FOU NDATIONS OF D HARMAKUtTn PHILOSOPHY
must be -good for something,~ one pre$umably will show a sp«ial concern
for verifying the instrumenta.l ity of an alleged instrument of knowledge
that is meant to be good fOr the obt:linmcnt of ooc's highest purp<l6C. The
coma! of a highe$t purp<l6C, however, confron ts Pmn~a Theorists with
a basic problem: each rndition conaiVC$ its ultimate purpose-and me
path mat leads to it-in sum a fashion ma, to some significant extent some
clements of that pUrpoK and its palh ~ nOI amenable to cmpirica.l exam-
ination." Oharmakini , fo r example. posits buddhahood u his ultimate
goal . and whik he argues tenaciously !lUt the possibility and genera.! char-
acteristics ofbuddhahood can be inferred prior to ia att2inment. he admia
that there are aspects of ,hat mile which we Qnnot know or understand
prior to obtaining it." Likewise, in terrru of the path thai is meant to lead
to mal ~ . our spiritual exercises may rest in some cases on belicfi; about
the lr.ulSempirica.l. O ne ROClbie example is the doctrine ofbrma, accord·
ing 10 which the activilies of this life willlC3d to specific resulu in the nen
life. Again. Dharmakini wiU nu.imain that much that is at stake in me doc-
trine of karma-including the cxUtence of former and future lives-is
indc:N. amenable to aamination mrough an ordinary person's senses or
inferences based 0 0 what is available to the~. N~nhdeu, the tk,"ib
of karm.a·s workings arc tntirdy beyond such empirical aamination,"
In this way, Buddhists such as Dharmakini-and ind«d all PmnlJ)a
Theorists-are confronted with the nc:N. to suppon dainu about trans-
Naplomi and Vmrr. it koo rdcnnllOr our purpoMS. Smnkdl"" (I~l) laler mlen~ and
arJII"d for Veun's petition. bul lit: Iw N.naoplcd (Slonkdlner 100) for a IUbdcr vnsion
01 circularilY I~I we wiD dilaua bdow.
16 I'S ( ••OJ d . H.....n a,), , _ ~' Ao';"".1'-~.~':W ,-':W~ u.- nop"".1'- ~.., I
, ..,,"~.,',/,J¥ j ~dt ~.~ luNtJt- ",,..,uJ ooluw/! /.
INSTRUMENTALITY: JUSTIFYING n-tE SOURCES O F KN OWLEDG E 1}S
I.n relation 10 Ihe notion mat me Buddha is Monr who has become instru-
ment of knowIMW"" tpram4!",hhii14). rhl': comml':nr.ubl rl":ldirion u:nl':f-
:ally interprets these lines as simply rescuing Ihe long argument Ihrough
the four epithcu thai precedes this VCt5C. That is, (I) our of his compassion
(44,.), rhr Buddha "sedcs the wc:aI of the world,- and (1) because ofhu
knowledge {jfi4IUl)conccrning what i.s lrue, he is "One who has wdl under*
srood," Since (J) he rcaches nOI only that truth, but also a means (wihalUl)
10 realiu il. he is the -reacher. and sinu (4) hr i~ rffortfuJ (ahhiyofIlIN!lITt)
M
in that taSk, he is me MProtector" (Iayin;' And thw, as one who has these
17 Set- T.tlenw1s and Irwni (1916) and FmICO (1m) lOr diKl ·qioos of 1M IClUCTUre or ,he
Pr.tfIJ/WI;,uJ,i duPlu.
18 PV"L"'a-"')C ~my.""""'rw- ~'!' M . M _ ... u _~~ .... -t"":,,
JIfUII4IlIItI,,"'~" / M~"~.muis ";,,,' ,(*"" f~ ..
236 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
21 Fnnm (Im:6sj arpa; apinst the ima ptt......... of the ¥aX • roottd in • drwbr ~
As a lnllU't offXt, the immed.i.alt mmco!! of Iht vnx does ncx Mlppon Vmer', and
N.-ni', inwpetcac>oo [~. Iha, doc~. 1>0:,., it; rimJ1uJ ......,,'C OharmaI<irti
arpa htK and in doc foUowirc twO ¥mQ mal doc Buddha U"Cd ptlUpc;o.l and ink.-·
~. ncx tha • ....,.. an.-ilid bea..... ofhim. Nor doa thecommmwOf$' in.~
cion Juppor1 Vetter', or Nap-Iomi', h)'JlOthcxs, (Of none: of OIwnulr..ini',
commml1lon dnccu hat I~ ciK,,]..uil)' ptrccivN by Vmcr or !he: rmproO,y
daiJMd by Nap.omi.
Whik K mly be lfIK dIa,.he vnx (which it; aawIJy a ¥ftX and a IWf) is not twapribk w
doc vrniono of rin:uWi1)' "'M MN by Nac;o.omi and Vmcr. it iI diffi..."".o ckfmd >ipinn
doc rircului<y _ baY.: akctdw:d~, if _lppcak ooly 10 !his vax and iu i~ c:on-
IUL 'iUhotn. in the _ tN. foUo,.-, DtwrrWUni ma!he 8<iddha', Wot of ~,a:pJon and
wac.." . _ omainly"Cml juaUfwd in cb.imiilfi tNt we "-'Id acupI doc irumunmt:lll'1
of pmxpJoo and in(",u... bcuusc:!he 8tiddJg comidrmI thmo iiUlNmCntal. ~, it, lhe
I"act!hal ~ and inr.--1U"l" iftMl'tIft\CiIlJI", dCfIW:>nsm,eed by ~ flC! that.he: Bud·
diu. who if; • Cftdible ",",.non.,. 0 . . . uch .......- . .....d ........ (cf. Fnnco '?99:U). And of
couno:, we t.."" alrady UIIed paupeion.nd infcmKo"'o dcmonan.~ that.he: Buddha iI .
CRdibk ",,!hori'1 on web m.lRcrL
FitWly, while Iht tonuncntaton do not mention any such rircubril)', docU "lena on the
maLler rnit;:hl Jimply Ix an 1.1Cftlp! 10 avoid an intnaabk dauc. Hmot, 1douIx Wolwe CIIl
diam;' d..- amninsIY circu1u Irpmcn. Jtl!F'Ie.t by the _ illS! on the '-is of iu QOr\.
ICIII alone. InckaL if anyt/llilfi. iNsmum IS ....,.. oocur " the end of an argLimml for o:ht Bod·
dha', e'Nibilil)' at "._~u.,., the ¥ftX'. immcdi.atr to,lIa' lUpp<'lru .ha. cirrulu
..p ........ loMcMl, .. f.1Ul<.U..Lou ..tru ...... it io u",b.... ·I;'.'"~", uf .... iptlAni io,{CfnlU: u...
allow. UI 1O.qca mon: ddinimdy lhe: ..:cua.1oo of ci1rulari1)'. .. we will...., below.
z}i f OU NDAT IONS O f DlIARMAKIRTI"S PH Il.OSO PHY
22 F~ (1999'66),
2.J franc.o (1997'19). ~ at... McOinrodt ( IOO~J~I.
INSTRUMENTALITY, JUSTIFYIN G TH[ SO URCES O F KNOWLEDGE 1} 9
tnding of the S(2temenu ciled above (with their anendam ECfE circu-
larity) art superseded by a clearly amcuLned position in Dharmakirti's work
that sea the fauh in an appc:tl (0 credibility! '
~ it rums out, we can point to a dearly articulated position of thu
kind, :tnd to do so, we mwt turn to Dhannaltirti's comments on another
of Digniga's S(2lemena , mu time concerning credibility and the innru-
mentality of scripture,
'}n other words, the Acarya [Dignagal did not say that knowledge from scripture is a
type of instrumental inference by claiming that it is actually or truly (bhiivika) instru-
~mental. Rather, it is instrumental with regard to the way in which a person should
proceed.
242 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
1:.··
SakrabUddhl (247a) does not identify who these others might be; K(96) glosses apareas
;~ainafJ, while Manorathanandin provides
~.We accept the point that has just been stated-we do not reject that kind ofidea. But
e.it is accepted only if one is able to know that superiority, defined as the experience of
~.·things as they truly are, etc., as being a defmite aspect of that person. But one is not
!~able to know that. [bshail ma thagpa'i don gangyin pa de nyiil ni kho bo cag 'aodpa kho
~•.na yin te / kho bo cag ni de Ita bur gyur pa spong bar byed pa ni mi byed do / 'on kyang
~gal te don)i Ita ba bzhin du mthong ba fa sogs pa'i mtshan nyiil can gyi phul du byung ba
.de skyes bu'i [D.: busJ nges par shes pa nus na de yang shes par mi nus so 1].
6) preserves this statement almost entirely:
'yam anantarokto 'rtha sa it!O Smiikam / ki1J'l tu fakyeta )fziitu1J'l pU7Ufanaiyamyena yo
.ayo yathiiilarfaniidifak!atJo na tu fakyafJ [S~ayana reads: yathii ilarfanafak!atzasYa
fakyafJ; this has been emended ex con). to accord with PVT: fak!atJo na tu '" in
ord with SakyabuddhiJ .
. he portion corresponding to Sakyabuddhi (2483.2) ... de Ita bur gyur pa'i don)i Ita ba
FOUNDATION S OF DHARMAKIRTI 'S PHILOSOPHY
/dilll "III ..w-, .. IiI,.",. i JM ,.,. ""I """ ,. j MJn .. '" Jw. 1M ,.i "'" ". ..,u It,;
#.hould (l((IIt afio:r iiI!' ~I!' ~ WJ- ,.~ in KC,j)6.J)-u;).
.)2 The pm...... if onm nnWtcd simply as "difficWt,' ~ this Enpoo word if roo wak
fOr .ht-lmKtN. dx prda mI"c)l.ln ncarty~c:&tC, ~cxprencI ~ins_than
Eitpish "difficult" but nof ~ qw~,.. RTidmI" 'im~ibk:
Noc( mac, aocotdI"Ito ~i (Z4ia)"K(m), tlJ,rIwIhJ(lIJ if l'anini.w nornina.iw
sinsubr wbm ~ with "w.,ui but muculinc nominative plural wbm ~
with .T, u,Yt...
JJ Iu UI Cl<2IIIpk, ~i (J..48b).K(J97)) _ -nut it, penoN who have dctirc
may makr .bo:rmdvct appa.i' .. ir they ~ dcsirdca. and dairdas ~ rn.y rnab
Uwmwm. appear .. if they had dnOra.. • IMIN iii ~ . . .,,~ mu-,. ~",;
I "Illrip/ {. Ml7'ip_".
34 PVSV "" rVI.u8-'J" (G :H)9.1)-UO.I.4): ""~I!' ,.,hJn/M", .,.,., ~ I
,.,~uJip~,.,. qus w,u~'" niIIIIfI..w. i1'7.~ f w. .,.,.
",,~ ~ jUn.IfI» '~J*Ij lIulll ,.,..,." nofp_,. •...,,,""',,, H"'~
',,~ ~,...,.-.nu f .';*-I"H!!''''''' _ ~ ".~". mr"H" IJ'#'"
iti f_~~...,.",.~IdJ'i,_ f ~lH1~ 'Lb~,.
f,.~ ftllN"fTIh .,,.~'.,... f ~ ",thi~j~",...,.1Hl ~
f~ me•.•
:1\ ' ''IIM f uthi It; f .,.. ,...", ... ...". .TJ .1'1' ,unlq,.,. M f "'rlIlMtt,
",., prnU!li1fti'Jf •• ,
u"",",,,.,.,.,.,,.~ "
r.'i~
.,.... •• t.M N,.J) { .. U_~
~"" f "'" .,.",IiIfpu~,."
.
iIH".,.,., ~ (PVu191
_,,.Iolw
(G...,.Mr J.i l"'?'"{ J UJ4 ~
..rA.,..>1.,.,, ... hri,,.....,,.,b
...,..J,J,.. "'""!" £'1.:1-" ,..,...""w'l"i_
u.'" ."ikiIf".,. ".tI~
~ f
~ <VI
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 245
The above passage leaves little doubt that, for Dharmaklrti, one cannot
demonstrate that scriptural statements are a reliable source of knowledge on
the basis of the claim that their author has the kind of extraordinary qual-
ities that constitute a credible person. For Dharmaklrti, one cannot use this
procedure for the simple reason that one cannot reliably determine whether
the person in question has the requisite qualities or not.
On my view, the above passage constitutes the kind of clearly articu-
lated rejection of an appeal to credibility that allows us to discount any
apparent ECTE circularity in Dharmaklrti's thought. On this interpreta-
tion, the seeming ECTE circularity of the statements' that end Dhar-
maklrti's discussion of the Buddha's extraordinary qualities (i.e., the
qualities that make him pramii1}abhiita) must reflect a calculated excess of
apologetic rhetoric along the lines suggested by Franco. Whatever else
Dharmaklrti may be up to, his affirmation of the metaphorical instrumen-
tality of the Buddha as "become an instrument of knowledge" (pramii1}a-
bhiita) is not meant to suggest that the Buddha's words are what literally
prove the instrumentality of perception and inference.
We might think that the passage cited above, which has also been dis-
cussed most notably by Tillemans, would put an end to all talk of circu-
larity, but it has not. Instead, it has prompted Steinkellner to reassess the
issue of circularity in terms of a far subtler issue, namely, the relation
between Dharmaklrti's conception of ultimate goal and the need to define
the instrumentality of perception and inference in a manner that suits that
ultimate goal. This subtler approach creates an important interpretive con-
text for our study of instrumentality, and it especially helps to clarifY the
curious ambivalence-the oscillation between the empirical and the
transempirical-in Dharmaklrti's discussion of scripture.
35 Steinkellner (200J:323).
46 FOUNDATIONS OF QHAI.MAKIRTI ·S PHILOSOPHY
TilklNlU. Steinkdl~r, howevn, does not mcrdy sum:nder his earlier posi.
tion; inst(ad, be aho s«ks to explain t~ un<krlying qualm dut prompted
him to interpret Dhannakini's argument abour inStrumenulity as circular,
and ,his rd'ormuiated circularity is of great hermeneutical interest. Ind«d.
for our purposes Stcinkdlntt certainly is not - making n~ mistakes."
Describing this underlying circularity as "conctpfual," Steinkdlner sum·
marizes it schematically:)i
I. Our ordinary valid cognitioru (prllmA'."') Clablish the aumority of
the" Buddha's t(aching (bsuJdhtt-lNN:lll14),
1.. the" validity of our cognitions (prJm4!']il) is undentood as the"ir rdi·
ability (Iluiu",VIlIiilWl).
J. tC'liability dqlC'nds on succmful activity (pNn¢r1Jw·,iJJhi).
-<t. ..JI hwrum guoili alc dcu:rmillCl.l by lilC ullin ... u: gv;d (,.inJrl!"l),
S. the" - ultimate: goal" is indicatC'd by the" Buddha's [(aching (bwJJhIl·
INIClltul).
As we shall 1«, Dharmakini's notion of a cognition's insuume"ntality
(which Steinkdlne"r calls "validity") resu on me cognition 's tfUSrwonhi·
net$ or reliability, lind Ihal IfUStwOrmine.s is largely connilun~d by one',
accomplishment of a goal through me knowledge supplied by that cogni·
tion. Stdnkdlne"r'. point hel'(' is mat, if we do not have a notion of what
constitutes our goal. how CUl we appeal to the accomplishment of a goal as
a cental criterion of a cognition's instrumentality? And if. as a good Bud·
dhist, one UrivCl toward the ultimate goal of "imi!'ll. will not all omC'f
goals be defined in terms of it? Likewise. if all OUf goals at(' defined in terms
of that ultimate goal. then since only the Buddha (or some other IIrhllnt)
has Icnowlcdgcof mat goal. we must IPpeal 10 his words (i.e., the Buddhist
scriptures) in oukr to posit it as the u1timatC' tournuonC' for a cognition's
innrumentality.R Finally. Steinkellner draws a hetmeneuticalleuon hefC"-
onC' th;lt is dearest whC'n he refers 10 the parallel Ca.5C' of Varpg:lI~ya 's
s",!iulntrll, an early S2'!lkhya ([earise that discu.uc:s the instruments of
J6 Sicinkdlntt (1OO):J~).
J7 See Strinkdlncr'1 filnhn dOOcbtion (UIO}:U9):
'J'hcft _ 10bf a pp of IW;\m>C$l on !be: part of Dlwmaldrti,.."...I] .. the whole
tr.odilion Iv: rrprelftlU c:onarning 1M analysil of,~ il doa; 001 Kml to Iu""
'-n Clptc:Md in any oontC:1I that I''""'' (UJJhi) in auaini", (..,uJuJ or aYOidi",
(w-). ""' ........ O«ti.....,. p (-m.J i. dact-rNnai in £.eo, if only indirudy, in ...,.
opinion by the ckIlnition of the "wtimate pi: for atWnmml of..-hid. all ~
INST.UMENT ALlTY: JUS TIF YI NG THE SOURCES OF KNOWLEDGE 1..47
~toward nirvtirta: once one has established the real possibility of nirvtirta, one
!has already set out on the path. At the same time, Dharmakirti clearly means
~'that except for their objects and results, there is nothing distinctive about the
~Qrdinary perceptions and empirical inferences that lead us, however incre-
fihentally, toward nirvtirta. In other words, in their nature they are the same
,lis every other perception and empirical inference. Hence, when he presents
ffhe instrumentality of ordinary perception and inference, his discuSsion must
!~pply to all such cognitions, not just those that lead directly or indirectly to
l:nirviirta. Dharmakirti thus has two concerns: that some instrumental cogni-
~ons of ordinary persons be capable of knowing the truth of nirviirta, and
~at his account of instrumentality construes these cognitions to be the same
in their basic nature as every other ordinary instrumental cognition.
When these two concerns are combined, we arrive at a clear constraint
~!bh the way that Dharmakirti must approach instrumentality. That is, since
isome ordinary instrumental cognitions move one toward nirviirta (by giv-
liing one access to the important part of the Buddha's teaching), those cog-
lhitions cannot be inimical or contradictory to the obtainment of nirvtirta;
l!"ltherwise, they would be moving one away from it, not toward it. At the
~~ame time, his general account of instrumentality must apply not only to
~fhose cognitions, but to all ordinary perceptions and inferences; otherwise,
ll)rdinary
t;
instrumental cognitions could not do the work that he claims they
~$ll do. In other words, if the ordinary instrumental cognitions that lead us
~toward nirvtirta are not identical in nature to all other ordinary instrumen-
t~al cognitions, then Dharmakirti would in effect be asserting that those
~erceptions and inferences are not ordinary. Hence, since the account of
!tnstrumentality must apply to all ordinary cognitions, and since that gen-
!eraI account must also apply to the specific cognitions that lead to nirvtirta,
!l)harmakirti's general account of the instrumentality of perception and
:tnference must avoid any features that would be inimical to the obtain-
iment of nirvtirta. In short, the nature of all ordinary instrumental cognitions
~ust be compatible with nirviirta, but those cognitions must also be capa-
pie of establishing the important part of the Buddha's teaching without
:I..:· ;.
reference to scripture or transempirical knowledge.
From an interpretive standpoint, this "compatibility-capability" con-
~traint may help to identifY and explain a number of features of Dhar-
_,~1nakirti's thought. He must give an account of inference, for example, that
l~nables inference to be a powerful analytic tool extending far beyond the
I.'f.ange of our perceptions, since inference is our .primary means of estab-
Ilishing the truth of the teaching's "important part." Dharmakirti must,
~f
lSO FO UNDATIONS O f DUAII.MAKJRTI 'S PH JLOSOP HY
040 Villi H uvry (1m) Fa a liM sumnwy« 1M "hcrmmnllia« iWpicion" in the ImIC
a1lwkd 10 hcn:.
<I I My lhinki"l aboul d>r nioiopol nalu~« JUdI_raiI!u Iw b«n lnAumud by KIp-
Ilrin (WOI),
<12 s... Ed<d (1991) and Dunne (1996>.
INSTRUMENTALITY, JUSTIFY ING THE SOU RC ES OF KNOWLEDGE lSi
1(; In lhc dim..;"" thI.. tollowl, only anain ~ will be chosen OUr of. IMp diJcw-
oiuu. Tu o-l. II,", C Il Un: ................ (PV .... -<I ..i ... "'" <;:uIQIn<:n .. vi ~ and
~. tc'C lhc tnnloIM~ in lhc oppcndilI (n4ffl,
254 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
47 PV2.Ia-b l : pramii1Jam avisa7]'lviidi jfiiinam. Note here that jfiiina has been translated as
"awareness." The general question of translating the term jfiiina is of greater importance than
it might first appear. Potter has remarked on this issue in a somewhat different context.
Questioning whether "knowledge" as "justified true belief" is applicable to Pramar,a Theory,
Porter has noted:
A jfiiina is ... an act of awareness. It does not name a disposition (say, to respond in a
certain way when meeting a certain sort of thing). A jfiiina is ... an occurrent. If it
involves belief, it does so only in the sense of a believing as a fleeting act of awareness.
A jfiiina is not a belief in a dispositional sense .... Any act of awareness which has
intentionality constitutes ajfiiina. (1984:309-310)
This aspect of Potter's argument supports the point I am making here: namely, that a
jfiiina need not involve the determinacy implicit in the term "knowledge." This is especially
the case ifwe understand the "beliefs" that constitute dispositions to be necessarily determi-
nate, or even propositional, in nature.
48 PV2.5c: ajfiiitiirthaprakiifo vii.
49 Both Dreyfus (1997:291) and Franco (1997:47) maintain that Devendrabuddhi under-
stands these two statements to be taken separately, but Devendrabuddhi explicidy construes
them together at the end of this discussion when he remarks: "[A cognition] which by nature
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 255
leOn the most straightforward reading, Defl means that, if perception and
[ir:nference are instrumental, then they must be cognitions of that which
~accomp1ishes the desired goal. Devendrabuddhi notes that as such, percep-
i'tion and inference are trustworthy, for they are what "cause one to obtain
lI(prapaka) the desired object (artha). "51 As we shall see, on Devendrabuddhi's
tview, the obtainment of the desired object consists in the activation of a
%ognition in which there appears the accomplishment (kriya) of one's aim
1~r tdos (artha). Within the context of this interpretation, Devendrabuddhi
irefers to a statement from Prama1Javinifcaya, where Dharmakirti remarks:
'1
I*"
I'
~~'
ste khams kyi don ma tshogs pa nyid kyi phyir ro].
PVP:2aI-2, translated below, 280.
1;:'
tt
lS6 FOUNDAT IONS OF DHAII.MAK IRT I' S PH ILOSOP HY
'2 PVin,., , .......'!" ,.'!'J'#tiU-' ~'" .,.........'!" «Ii "" ., • ..,... .n6.'!",..,..
~" ..., ...1JIh# ,*If~ ~M.
5J Thil foIJowa dw ilMnpml Uon of~i, c:sp«iaIIy in Innu of1"'".,mIM. s..e
bdow, (161).
}t: EnmpIosof a111O.u marunpabow1d.. bw in DhannaIc1ni'l mmmmwyon!he fine YaK
cI ,., .... Nt............. Alone. one can find . . . _d ck&rl, in thlU mean. aim (Col" ' C"-,,
etb4V,.u".,;__-,.,._. . umy-,., rM,i,,.,,,,-,.rm 1Id~_~. . . ... ), meaninl
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 257
to some further aim: the distant fire that is the object (artha) of one's per-
ception may also be one's aim or purpose (artha) in that one wishes to obtain
the warmth that is one of its property-svabhavas. This overlap between "aim"
and "object" appears to be based upon the degree to which causal efficiency
is linked to purpose: considered just in terms of the causal efficiency of pro-
ducing perceptions and (indirectly) inferences, an artha is an object. And
considered primarily in terms of purpose, an artha is an aim. 55
Although artha as "aim" and "object" may thus be distinguished, even
'when the term artha is best rendered as "object" it retains some connection
to the notion of a purpose or goal. Such is the opinion of Devendrabuddhi,
:is is apparent when he comments upon Dharmakirti's statement at the
Outset ofPV3 that the illusory "hairs" or "flies" perceived by a person with
¢ataracts cannot be considered "objects" (arthas).56 As noted earlier,57 the
trux of the problem here is that Dharmakirti might be obliged to posit not
~nly particulars and universals as objects of knowledge (jneya), but also a
third type, namely, illusory objects. Commenting on this problem, Deven-
dtabuddhi raises an objection and offers his reply:
7"JoJ.. .... 'J.. M; "hPT tk pJ rJ-" 1M' p"", - I pJ . -.. IMl'7'> ",. 1M - Ji,. _ •
.... I ",j rip,. _ ",. ~ I iii Ju, •• shU J.IIIf1 Mit ",ilf /W I rii ""', Jw NI l Mit J.._
,. - " # ,.1. " ""' • ..",. ,..t",tk I. ... J.. _,. -',.i,., .. I...".
,.,,; iii"It ", I pi ~ ..j rij WI I. ,." ,.. tb.itr ,. j __. , _ I. ,." ,. J,..J"If,... :ow
.. #II;'*,.If ""
.I... ,.; ....
"If •
I ri #II ..'" ti ",,... ..,.." ......", tP- ""*,,., ~ ..." .. NI l iii IsIlT
1","1 ti ,.,., .."" ito tIM ~ ""',.,. tk .., Ii J.n ,. P"I I. .......,.
'Ji,.,. ""
. 1.
.... J.,.",. ..../If ... i ,;".,. tk; pJ J,n tk Ut J. - , . .... ]1,. _ a: P'VV-n
(m. n. I ), ._, ' I,,} r~ _ J . . Hr~ ~"·.,,.,.....,....,,·trII~'!''''
iti".,.,.,_,.l/JtmI
3 PVP: plM.. -Jilt. bul PVV·n (III , ft. *- ~'"""""'" ~
"""""''''''''
60 S«chapur I (U.4-1,) &nd Dhumaluni'. commmo ift PVSV u'PVI.IJ7- 14J. tn~
in dw appmdix tlWfJ. II it worth nocinc tN.! dw indcpt-1IIknor of an ob;M:'. (:IUJaI fUn,;-
tioNJilJ' from dw;!dic ~ ofdw; !'CIeri", rnaka il d.i.ffic~h 10 a«IIflItdy ~rxKritc:
DIwmaIdni', phiIotopby ... anyl'orm of pncr!YliJm. dopil( «nain 'pragmatic OOtllOl'G'
ift hU philotophy. a: ~ (1m).
61 Napromi (' 967-61:H-~4).
260 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
62 See Nagatomi (1967-68:55-57). Among the passages cited by Nagatomi, it is not at all clear
that in PVp-3 arthakriyii is used primatily in the sense of causal efficiency.
63 Among the passages cited by Nagatomi (1967-68:56), one is PVI.98-99ab {}fiiiniidyartha-
kriyii1[l tii1[l tii1[l dntvii bhede 'pi kurvatap / arthii1[ls tadanyavifle!avi!ayair dhvanibhip saha II
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 161
ant, ror rhcsc two differmt emphases mrrcspond to two different ways of
undc:rsWlding Dhannwn:i', norion ofinmumentaliry, csp:cially as int~r·
pm~ by lkvendrabuddhi. To Itt how this is dK ase, ~ need to tim
rcca.Il me definition of an insuumcnl ofknowledgc within tbe /d'4kA I)'S'"
tem as the:: -most prominent ooor'" (s41ih4Jr4l11lff4) in me production of
me resulting activiry (ItriJl}. specifically, thai rcsulwlI :teriviry is the insnu·
menml dkc:r r,m"u!'Ap".IA), which most Pf':lmiJ;l~ Th~rimI oekfi~:u thoe
resultant act of knowing (pn.",;ti) .... On ~endrabuddhi 's vi~, Dhu·
makini proposes two al{~rnaliv~ inlerprc:tations of wbat constitutes the
instrumental dfc:ct. H~ claiffil dllll -... there :ll'C two kinds of inslrumc:n·
tal effccu: one call~ a -human aim (pJl11if4nha), - which is a medialed
(dNJJ,. • II]IllNlhilll) ~ffcct, and a distinctive on~ (lthytuJ,.r w, - viI;'.III)
that is not mediated. ~
When Ocvendn.buddhi speW of a human aim as the mediated
(.".Vllhilll) instrumental cffw: resulting from d~ application of an inslru·
menl of knowiedgc. he docs nOt explicitly define what is meant by "medi·
ated- (1I]Il/)"";III). Neverthdess, his point is dear: me effect is medialed
(IIJIIVllhilll) in mal there is an -intem.r (.".wu/h.iM; betw«n the func.-
tioning of thoe iNiu'Um~m (the: ItIflTIfl!fA. i.e.. the P"IflMli':'lfl) and the dfc:ct. In
o th~r Vo'Ords, th~ ~ff«:t is separated (vpllllhil4) from the: insuum~m by
vinuc of other intervening causes and conditions that mwt be in place for
the: cffw: to occur. The: cffw: is tbw -remo(~" (."..,.u,illl) rdativ~ to the
instrument. and since: some obstruction can therefore occur bctwttn tbe
functioning of me instrument and th~ production of the effect, me dfca
may not neceuari1y occur. even if the innrummt functions corru:dy. Thw.
although the Engluh term - meciiued" bat conveys th~ notion that the
prodUCl»on of me cffw: rcqu.irc:s other causes and conditions subsequent to
me functioning of the instrument, ~ could as easily refer to thit dfttt as
-remote," "separated" and so on.
In contrast [0 the mtdiated dfca, the unmedialed effect is "distinctive" in
.... 1J'fiJ- ""'jU,..'!I ..".. ." "",..,rUM). Evm in Ihia cue• .-hm: ""_ h I}oI is
axucnaed mady with rhr pn.xiu<:Don all ftauI awarmat, rhr dUau.ion of iUwion bin
inthr:po B ....... (d,citindim:dyj doat_norianolplorpwpooe " ... iric""
limply duifyins rhr KaNl or 00(' . ~ia.iU mevanl.
64 oS« chapm I, 1.1.
6) Onn.dnbuddhi (t'YP:Jb4ff): .. Iur~_'i ..... '"' IN
din .". .. dI.J,. -.." ~,., (". d.t/,. ~,. ..
mao,.,.,.. ti l .."..,",;"
161 FOUNDATI ON S O F DHARMAKI RT r S PHILOSO PHY
70 Ie iI-.onh notinsdw Uddyou.Ic:an (NY f1.ill) mW:s I JirniJ.u ~ whm dc:a.li", with
a probkm in Vliuylyana'l ~wion 01' i~. To Ix prKiK. 1M problem here 1I1h.11•
... inf...",,,,,, i> • m.an. <0 11a~ Ju..:",...Ic..~. men .in"" i"fr"""." ;. ;todf a ~e"" of &II
ob;ea. i, c:ouId Iu~ not Iu..,. "",,;,..,., r,r.",i ti) .. iu rQU/t fI,..~bta,*;1I
obi«t Iw alrady brcn """iud. After lim propoti", apmlNlkalIlMUI""'. Udd)'OQhn
t\lI'flt to. pIUIotoophiaI......, WI haI_ reonanc.c,.;m DharmaJdrti'. approKh:
.~ ... MlppG'<' chae the IUlcmHll in dw tat tw t1w 1OIIowi", man.ing: inf...mu is
• ropIirion thaI involva thaI wlUch is marka! by an infnmliallicn (J.itipi). Bul in
chal QK, dIuc would Ix eM huh olh.1vi1l& no IUIIllanc aa olJcno.,..inS-"
W~..,.dw ~"I' 1'1ti. io.-~ ~ ~~ ...... oftlwoh;eco .. w+u..
_ .bouId J'/'Oid, employ, or ig>or~ ;oK tho: 'T"Uh. MO«'O\'TI'. in ITpfd 10 iu own
object. any iruu ummtal """ilion runaioru II dw wabaJ aaMty, JUCh due "inscru-
mn\1 ol~· mnru ·,ftUlti. . . .tion fln ...,'ti)." With ~ 10 aMdIn
ob;ea, all ,..._~ fVnaioo II t1w inmumcm, luch thaI ""..'!"i' lnQRI "thal by
rMIIII of which eM obj«t ("..~ iI known r,r.",/ytI,,).-
" If dw; won! ',~.~ funnioftl II eM ..mw xtivity. chen wNol would M eM
mulcanl ..:tmcr.1'hm: could be no ..-Ir bta....., the: ob;ea it; alrndy "'I7'i....t.-
[In Viuyiyana] huoUd ...... dw: C05"i....... of dwoOjcct ....... whOch one
rcopona<'
ohouId .¥Oid, employ or isnorr lIT m.
raull boaill<' ....ben eM ob;oa it;1mown. onr
tw me- copUDona. nut is. ....ben t1w oO;ca is known • ......, hal. duu IcindJ of me-
N lioll' 0"" copiza dw: ob;.ct II mae which .. 10 tw 1MIidod, ""'piorrd "';~
[Mw.nt NJ- __ t...+J.ti ,,.n,.m~ ....".._
iii f ...... (}I ~6htIw..,. ~
!f ...... " ' ' ' #.4~".", ~mh f ~ u
,...Ii~"., " ...iti!l , ...w-,. iii I ~.,.",.. ".,. ....~_ I
,,..,u!f6'!l_"'''
'''''';,.'' '' ' ' ; ,r••
!",,,, I ,.A Mbw'USw..~ "."..~u"" J.i".
",""F) *"I-Y rtNU ! Iiku,!, ,,., ~'!' 1M....·.. ,ur.n,.
iti I pt." .,<Mw..., 1frw"
,-1.",.
1tJN,J",.,. triJJ,i.wUhir M.nwti ~ ",,! i t.,. ~/I'" _ 1,
264 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
71 PVP (3b5ft): shes pa blang bar bya [bal 'i don shes nas de sgrub par byed pa nan tan du byed
pas mngon sum du byed pa 'i mtshan nyid can gyi Jug pa skyes bus thob par gyur ro / de bzhin
du dor bar bya ba shes nas de sgrub par byed pa nan tan du mi byed pas / de mngon sum du mi
byed pa 'i mtshan nyid can gyi Jug pa thob par gyur te / blo ni tshad ma nyid / bfang dang dor
bya'i dngos po yi / Jug fa de gtso nyid phyir dang / bfang bar bya ba dang dor bar bya ba'i dngos
po 'i yul can dang / skyes bu'i don zhes bya ba can gyi Jug pa de fa / de gtso bo nyid kyi phyir te
shes pa gtso bo nyid yin pa 'i rgyu 'i phyir / blo ni tshad ma nyid yin no.
n Devendrabuddhi first uses the term arthakriyiinirbhiisa explicitly at PVP:2b7 (ji ltar 'dod
pa'i don byed par snang ba can = yathiibhimatiirthakriyiinirbhiisa), but this notion is dearly
what he has in mind when he refers repeatedly (starting at 2a3) to cognitions such as "a per-
ception whose object is the sensation of warmth of a fire" (de'i dro ba'i regpa fa sogs pa'i yul
can gyi mngon sum), "a subsequent cognition whose object is burning, cooking, etc." (4<13: sreg
pa dang 'tshed pa fa sogs pa'i yul can gyi phyis kyi shes pal, and so on.
73 Perhaps the best known example is the negative statement in the context of nonpercep-
tion (PVI.3a apravrttib pramiiIJiiniim apravrttiphaliisatz).
INSTRUMENTALITY, JUSTIFYING THE SOU RCES OF KNOWLE DGE 16S
74 ~ an:: twO /'CQOnS dvol prompc me to ..wn~ that dU. _ is~. Fim ,
Drvcndr-MNddhi'. din.-a po.. of ,._ptA; in tho: yux rd'-.n only ro p..,m
in tIN: _ of
the "aaMtion" of:an in&uwDM1II copiOon (J'VP:)bs-4). S«ond, in • •btc:qolt:l'll poillll in
1m oommmu on I'VLI-ti, h .. dw nrlkr InUO: mar .. apln mlCl'llttd, and n(IC thc ~
of ".",m ... penon'. aaion IOW:tfd :an ob;ca.
FOUNDATIONS OF DHARMAKUlTrS PH ILO SOPHY
:aocnmplimm.-:nr nf nn.-:'s aim: :md (2) in a .K!COndary se:ns.-:, th.-: action is rh.-:
initial action of Sttting out to accomplish thai aim.
Before moving on ro the notion of an unmroiated effect, we nttd to
make one final obscrv:arion: namely, that on DcvendlOlbuddhi's imerprel1l-
tion, Dharmwni accrpu the notion that, from at least the perspective of
attaining a human aim, an iruuument ofknowlcdgc and its effect may be
conuru~ H diuinct. ThiJ i~ indiotM morl 5f::"kly hy ~ndr.abudclh; 'J
claim that contrD)' Mon may imef'YC:ne bctwttn the instrumental cogni-
tion and the effect, such that the eff«t is not necesArily attained. in each
case. At first glance, this separation of instrument and instrumenra! effect
(pmM'.I4ph414) might not s«m problematic, inasmuch as both Dignaga
and Dharmwni t'ltplicidy note that such a distinction may be made in
conventional ~rms .'" But Digniga =d Dlunn21cini', ~plicil n:uemcnu
about such an imputed orconvenrional sepat:l.lion rest upon the claim that
the inruumcnt and its cffi:ct arc u1rimatdy a single cntiry-: the cognition that
is the "instrument is iuclf the "effCCt," It is difficult to s« how this
W
approach [() the singularity of instrument and effect can apply to Devt-n-
dnbuddhi's notion of a mediated instrumental effect, since the sensation
of warmlh , for example, that is obtained wr approa.ching a h~ dearly
cannot be considered identieal in:my f.u.hion with the initial perception of
that hre from a distance.
It would thus appear that Dcvendrabuddhi's notion of the mediated.
effect, if it is not to contradict Digniga and Dharmwrri's claims about
the ultimate identity of instrument and effect, must be construed as con-
ventional in a difWenr f.ashion . Specifically, rather than :a conceptual dif-
fe n :nti:l.lion of the l ingul:l.J" "nliry thaI is awaren,," inlO a cliitinci
instrumenl and effect, for Devt-ndral>uddhi an emirely differem comext
must be aslumed whereby entities dlllt arc u1tim:ncly distinct arc related
through ausaliry-: the dfect (whctheraction as initial motivation or action
as subsequent perception) is auscd, even ifindircctly, by till: instrurnenw
c.ogrmlon.
How useful i. this :aspect of Dcv.-:ndr.abucldhi'. interprct:&tion of
Dlwmakirti's theory ofinstrumcnraliry? For many sub5cqUCnt commen-
t:UOB, the answc:r might wdI be "nOi at all,· inasmuch as the subsequent
tradition of Buddhist PramiJ)lI Thcorisu, which often relies quile squarely
upon Dcvendrabuddhi's commentary, has largely abandoned this portion
FOUNDATIONS O F DHARMAK IRTI 'S PH ILOSO PHY
Dcv... ndrabuddhi notcs that Dharmakirti does not argo... for the instru-
m..-nwity of aWllR:ncss $Oldy on the basis of its primacy in aa;on, in wbich
COntext 1M norion of a mediated instrumental df«.t is panicularly rele-
vant, Instead. imm<diatdy following PV1.jb-d, Dharmaldrti prtsents an
entirely different argumenl for tll<: instrurnentalil)' of aw.nencss:
78 The f.,ult in OIurmaklrri 'l rheoty is specil1cally dull, in PV}. he oonsO!krs.., inItrumm-
W wsnilion only in fa" .... of an u.. mediaRd irutrummw dim: with link rden,," 10
human ~ms ~ bon his preICfIwioa in PVl..-6, wt.id! dearlyeon.rrucl an iJuuu...
menial WJPition in _ ofhuman ai..... 3ppcars.o rntui~ ~'I notion of a
medialM illlmUl1tllai dI"ect. In "-t, DlwmM!rti'l atmsm unrmml ofirumunmw
dfm (,'.",.!W~ in pvJ doIS IMII cormpond wdl 10 hil accou.nc of inwummeaI cos-
nilion in PVl.l-6.
19 PVl.4a-<: ~Itm~ u ~ ~_MrJ.t4J I i++W',..,. wi" .....
I NSTRUM ENTALITY: JUS TIFYI NG T HE SOU RCES OF KNOWLEDGE 169
tM II? N .... ~ p", I.u m.J,. __ ,.r1'I#fI ,. j tIM """" thH ,. .... tU ~ .... Jill ...
I .... m.J,. tU ,.", Jfd ~ 'Jn .. t.s ,wll IA JWI ,. _ "II "" I In ...... IIi ...,.,.., III
~ N 'i u., ",.. J1. .....~ - ' J1. .;. ..,.", tU ;.,. ,. ,. IA tit. . . PM ~,.""
... 1tU Iur ... Jfd.-... [PVP_D: .-.ul uII .. i'" """""',,, 1M,., r4M Jilt,.i';";, 1
M. ~,.. ... J.,/,. " .. ,.." 'f!tI tU; JI7i''' / tU JWI .... ur pJ fPVP-D: IA",]"'. .......",
"Ully.J ... inI h it; y.J""". IM. iii tsIMJ".,. tt]U "It .../p",JWI,..IIJi4Ji"'" P"K
Dw N tU ..i tU, P,;1t ,., P ,., ~,. Jill'" "" _ I tU ,-J .... ,.,., p'un, ti Mt '"""
h - " ~~.. _ iii I lui pM ..,. N pbttlt iii "')1M t41hn ~ ~ Dow,.. / tU j ' Mt vrr•
.... tU .pJ /1M,.,.,. 1M r'- J" tNtI ,.. i ""', ,;,.J tU. .".",N UII oi "".,. .pJ tI..
"btl.., ,. ,." ,..
",i D-,. I tU ...... ""..,. tt]U J. -:1" ..... 1
__,.". " 1",,",,, m ... ,..
~
~ 'i phJir,. / 'Jn N .pJJWI- ..
,."" l tU;tU iii ;""'P.i;lor",tMJ,. ...." ..... 1
i,... "';
y.tI ,. ,. "'''K .,. N
JwJ,.,i..,. .. ..,..".,.i
&f N, '"
#r'-
IA lui,. -",. UII of h h "."Iff ~, "...., ,.. i ..... "" .... Rad Jba ,.'; tNt for r.htJ
~ 'i tb.. Ct-9'
83 S« 1hIM:. 1.1.
M For DisnJp'. pl'ClCnucion., I « ~ PSI.).)O-<f and PSV . . til.
INST RUMENTALI T Y: JUSTIFYIN G 111£ SOURC ES O F KNOWLEDGE 2.71
8S T aIU"fj ..... pooi ................. "..,., . ..... wlt.id. io ... I.e -.bIioheci (~by ..... i .......-
mmt (Iu,.!y a Wihtuu), Dlwm:alcJni intnpml lht tyiJMltlu", u m( niyi beaUK il is
~isbcd in IftmI: of the ~ whid. is thus the ;rutrumml. 'fhQ, is d", basil lOr the
claim !luI ~ :and ~ a:and in the ..dation of ""'~"7"_t/:.qd"M.hwo.
Set. lOt aunpk. PV'.JI4- )l S:
W( do IlOI acap! Ow thnc ill diR'uma benocu. Uw objea of the acnon and Uw
objm ofUw iNmIment. Iftlx:y hac: dx:$UlCoO;m (nrJ,.). dono !hi: notion !luI thty
an: aaually rwo ditkmll cnliries is mo:aninpe... M..,w.(t. if lhey 1u.¥'C ~ me:
ob;ca. chq Jhould IlOI ocau ~lially. l( tMy OCCUr III~ Aim lime, !hen IlItre
is no rdatiorl oftlx:", u tbc cttabIilhinl mnN (uJtD.NI) and the cstatMiIhcd (nliry
(~ rSur on our new]. tho: .... bjutm:: and vbj«tin: ~ of tho:.OOo ~
ddined u tIx: em.biishinl mearu and ew.b1Wx:d mriry braUM: me:
establishment
(~ IIW) of me:
subjcaM Mpea iI bued upon tIx: obj«ti.., up«t. r ~.;,."..
MnU .,.~.~ Jdbrb.M....,.". ."."..". ow ,.. f1ir !r._1Ihhi1J
II 'w, 'd1H... rM.w.~ ~ Jh?- ~ J WIIJi'_IMir.,..- ~
MAl
___ _L ,-,L N,
"'Im"!' I. .
Nocc "- aI"-'&/I ~.b."Uhi here (rYl':l1i1b~) ....d cb>ewtx:n:..den "" tho: __ "PMU·
(.".u)of awata'IftI (Jbi.ftC.) u tIx: -objea..imubcrwn" (~)and tho: "copilion
,......- ......."".
of tIx: objca" (~). I u.ndcnu.nd tbcK (WI) 10 ~ Iynonyma for ,..~ and
To suppon the claim mat innrument and instrumental dfect are actually
identica.l, Dhartllak.ini effectively proposes an enrirdy n~ way of un~r
sl1Inding what constitutes an instrument: rather than being the aUK of
some t('$uitant activity (1m,.,). it il instead that which is unmediated
(IIvyaWlhitJl) with regard to me activity. Al Devendrabuddhi explains in
the comments cited a~. when the activity (kriy4) in question is an indu-
bit.lble cognition (prllmiti), the instrument is thus -that mrough which,
when all other causes are in place. the convention of 'knowing' is Atisfied
without runher mediation. ~ In shon, it is that which requir('$ nothing fur-
mer in order for one to be cunently having a cognition of the objta.
In emblishing the innrumc:ntality of an in.mumenw cognition in these:
temu , Dharmalcini (as interpreted by Dcvendnlbuddhi) rccogni-z.es that
one must be able to distinguish between cogrurioru. That is, if an irutru-
ment of knowledge is mat which enabl('$ one: to claim that a cognition of
an objta is occurring, one must be able to distinguish the case: where a
cognition of that objta is occurring from a CIS(' where such a cognition is
not occurring. With this in mind. Dharmalcini also claims that the irutru-
rnefl[ of knowledge is the "final diffe:l'C'ntiator~ (a"tyabhui4It1t) of cogni-
doru. Thus. nex only ~ it provide [he basil fo r claiming that a cognition
is occurring. bUi it also accounts for the differences between the contents
of cognitions."' As Devmdrabuddhi points our, on Dharmakini's vi~ the
only Dcer of knowing that can meet thc:sc: critena of an instrument of
knowledge is the "objtaive image- WJ/ry.HJ'II) or "object-5i mulacrum ~
("~nya~i . e .• the appearance: (pratibh4.sa, pnnibimbll) in a cogni-
tion. Funhermore, since the image is actually an aspect of me mind arising
in the form of an image:. and since the mind is ultimately undifferentiated.
the instrument is uhim:uely nothing but the mind (i.e., the cognition)
iudE. Thus. for an entirdy different reason than the one proposed in the
COntext of a human aim. an instrument of knowledge is once again shown
to be nothing but the awareness itself, i.e.• the instrumental cognidon.
90 Although Devendrabuddhi does not discuss it specifically, Sakyabuddhi notices that the
problem of infinite regress must be addressed (PVT, nye74a5-6 and 74bz).
91 This certainly becomes the opinion oflate commentators. See, for example, Dharmottara's
comments to that effect (Steinkellner and Krasser 1989:13.3ff).
INSTRUMENTALI TY, JUST ifYIN G TI-I£ SOU RCES Of KN OWLEDGE 17S
h would thU1 appear that 50me notion of aim or td05 is applicable to the
interprcution of IInhttlrriyJ as the crilerion for trwtworthincu even in this
particular type of cognition with unmcd.iated effect, namely, a cognition in
which the accomplishment of one's aim already appcan. But even if this is
the case, ~buddhi's discussion of unmcdiated inmumenw dfCCl-
and Dharmakirti's initial devdopmenl of the idea-is probably also
intended to account for in.strumenralilY in another fa.shion. Specifically,
the norion of an unmediared effect also enables one 10 understand a cog-
nition's insuumentality without any concern for human aims at all. O n
this a1tcrnalivc inttrprctation, one can claim that III/cognitions arc instru-
mental in a minimal sense. Although neither Dharmaldni nor Dcvcn-
drabuddhi is explicit on this i.uuc, il would appear that an allemarive
interpretation of IIrtlMltriy4 must also be applied on this imerprct:u ion,
since the entil>t poim here is to cvalWlC a cognition without reference to
goals. Following Nagaromi, the a1ternalivc interprcation suggcsted-bUt
nCVff ckarly statcd-by Dharmooni or dlC: earliest commentators is that
of IInhUriy4 as mere causal functional icy: an emiry has IIrthttlrriy4 in the
simple sefUC that il has effect!. On this interpretation of IInhllkriyJ. an
""'~1!1$ would ~ tnllrwnnhy in th~ minim,.1 v:n~ th.:u it i~ ,. rr:al m~ _
tal ~nt: it has ~iti in thot ~ seruc thu it is ~r.tbliJhed (,u,jlll)
as a causally dFiciem moment of coruciousncu. This minimal trustwor-
thincu amounu to the claim that, reg.ardkss of the determinate interpre-
tation of a cognition', contem, one can always reliably know that one is
cognizing. Since mis minimal truStworthiness is applicable to III/ aware-
ncues, all awarcncs.scs can be coruidcrcd trUStworthy.
Although ;1 m:ay U\"m odd to m,unmin t"'", ~'Y ;nst:llla of ,.W:lten~n
is trwtwonhy in the minimallCl1SC discussed abovr, this interpretation
helps 10 aplain £WO other nOlions proposed by Dharrnak.ini. The first is
one we have discussed previously, namriy. mat illusory objras such as the
hairs ICCn by a person with cataracts are not "objecu· (IInhlls) bccaU5C they
arc not considered such by persoru engaged in practical action (lIJIIwthtJrll);
in other WONk from the ~pcct:ivc of ~lc.in8" hum,.n ,um flU.,},.,) . th,.,
visual cognition of "hairs" in the mind of a person with cataractS is not
inslrumental. BUI for thai penon, those "hairs" IIl'tobjcct!l when consKictcd
simply as mental cvenu.'1 If we undel'5tand Dcvendrabuddhi', twOfold
explanation of instrumentality in terms of thr twO afon!mentioned sensc5
of IIrthllltriJ4. the perception of those hl.irs is not instrumental on the dcf-
276 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
93 That reflexive awareness is noncausal follows from its simultaneity with its object, namely,
the awareness that is reflexively perceived itself. Indeed, what can be most confusing about
reflexive awareness is the notion that it is a cognition distinct from its object. This distinc-
tion is clearly the case for all forms of perception, including mental perception (manasa-
pratyak!a), for in all cases the object (grahya) of perception is its cause (see, for example,
PV3.224). The same is true of inference, since it too arises indirectly from its object (see the
succinct statement in PVSV adPVI.I; G:3.8: avyabhicaraf canyasya ko 'nyas tadutpattep, where
tadutpatti is meant to apply to both effect- and svabhava-evidence). In contrast, what Dignaga
first identifies as the three aspects of an awareness-namely, reflexive awareness, the objec-
tive aspect (grahyakara), and the subjective aspect (grahakakara)-are all ultimately identi-
cal and hence simultaneous. The notion that reflexive awareness is cognizing the subjective-
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 277
;4ualification to the way we have characterized it so far. That is, even if such
:"icognition combines a telic meaning of arthakriya with the notion of an
!~nmediated instrumental effect, it also places greater emphasis upon
;;rihakriya as ca~sal efficiency, in comparison to instrumental cognitions
:that are considered trustworthy in terms of mediated effects. In the case of
~ese latter cognitions, trustworthiness means that they lead to some sub-
~:ind objective-aspects is merely a way of conceptualizing the process of knowing (see the locus
;classicus in PSI.I.IO: yadabhiisa1J1 prameya1J1 tatpramti/}aphalate punap I griihakiikiirasa1J1vitti
;triiya1J1 niitap p.rthak k.rtam). Although Dignaga maintains that External Realism can propose
:i:similar way of analyzing cognitions (PSI.I.9), this way of explaining cognitions assumes the
Epistemic Idealist position. Dharmakirti accepts and elaborates upon Dignaga's opinion in
Hie following verses [PV3-333-339]:
"If we maintain that an external object is experienced, what would be wrong with that
claim?"
There is nothing wrong with it, but what is the point of saying this: "An external
object is experienced"? [PV3.333] If awareness has the image of the object, then it must
have something that distinguishes [each] image [for each awareness]. [PV3.334] It
would be wise to look into whether that differentiation must come from something
external, or whether it might just as easily come from something else.
[I] There is no apprehension of an object devoid of qualification by the experience
of it; and [2] when that experience itself is apprehended, the object is apprehended.
Therefore for these two reasons, the cognitive appearance of blue is the experience
(darfana) of blue. There is no independent (kevalap; cf. PVP:223a7) external object.
Instead, something activates the internal imprint for some experience. It is due to that
awakening of an imprint that there is the restriction [of a particular image] to a [par-
ticular] cognition; that restriction is not dependent on an external object
[PV3-335-336].
Therefore that one awareness which is experienced and remembered in that fashion
has two aspects (dviriipa); the instrumental result is the reflexive awareness of both
aspects [PV3-337].
When the object is considered to be other than the mind and established with a
nature that is desired or not desired, then the object is the cause of the representation
and the effect is the experience of that representation in that way, i.e., as desired or not
desired [PV3-338].
If awareness includes the object (yadii savifaya1J1 jftiina1J1) due to the positing
(vyavasthiti) of the object as an aspect (a1J1fa) of awareness [and not as actuallyexter-
nal], then the determination (vinifcaya) of the object is just the awareness' experience
ofitself. [PV3.339] [yadi biihyo 'nubhiiyeta ko dOfO naiva kafcana I idam eva kim ukta1J1
syiit sa biihyo 'rtho 'nubhiiyate II yadi buddhis tadakiirii siisty iikiiravifefi1}i I sii biihyiid
anyato veti viciiram idam arhati II darfanopiidhirahitasyiigrahiit tadgrahe grahiit I
darfana1J1 nilanirbhiisa1J1 niirtho biihyo 'sti kevalap II kasyacit ki1J1cid eViintarviisaniiyiip
prabodhakam I tato dhiyii1J1 viniyamo na biihyiirthavyapekfayii II tasmad dviriipam asty
eka1J1 yad evam anubhiiyate I smaryate cobhayiikiirasyiisya sa1J1vedana1J1 phalam II yada
nifpannatadbhiiva iHo 'niHo 'pi vii parap I vijftaptihetur vifayas tasyiif ciinubhavas tathii
II yadii savifaya1J1 jftiina1J1 jftiinii1J1fe 'rthavyavasthitep I tada ya iitmiinubhavap sa
eViirthavinifcayap II].
2.78 FOUNOATI ONS OF DHARMAKIRT I'S PHILOSOP HY
,., N&T I, .,ft". FOf a nady Idmllalarpimnu In rVinT, _ SlrinkdLna and JCn.ua
(1989:16-1.9)·
280 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
prccisdy what makes that initial cognition instrumental. In other words. the
subsequent cognition in which appears me desired tel k function (i.e., me
accomplishment of one's goal) confirms the trunwonhincss of the initial
awarencu. A3 such, it confirms the fact Ula[ me instrumenal cognition is
-Wt cawes one to obtain" (;r4jH1J:IIJ an object that acco mplishes one's
goal.'"
Thc oonfirmnion_modcl and !'ducd claim) :loom trus(WOrthineu l~
Devendrabuddhi inw a .series of problems and attempted solutions. We
will now consider a Few of these, first bycxamining [W() general WUC5, and
then by turning to problems that arise in the specific COntcm of ~rc:cprion
and inference. Although many of these problcnu and solutions resurface in
the works of subsequent Buddhist commentators, I wiU restrict m~f to
occuion:.i Icft"mcc o nly to the lub.<omment:u y of~iky:abuddhi.
Dcvt:ndrabuddhi poina to [W() issues that cause trouble fo r the defini·
tion of instrumentality in relation to the attainment of a human aim. The
first of these is a disparity in time, that is. a problematic lag in time; the $Ceo
ond concerns the possibility that activity initiated by an insTrumental cog-
nition might be obstructed.
instrumental cognition is that it makes one obtain the (die function of the
expected or required paniculars. The problem, ho~, is that in the case of
pclttplion, sina: meobjca (r;rihJtl) of pcla:pcion is me panicuJ.u{s) that
caused it, the object of peiCxpcion has a:ascd at the time mat the perception
occurs, The same applies ro inference, inasmuch as me indirect objccu of
inkrencc:s art: abo partku!us,IM Thus, the crux of the problem is th2t th~
instrumental cognition.t----pt:rc.eption and inference---aIf: meant to be instru-
mental for 2 hunun :aim beausc: they take:as their direct or indirect objeru
panieulars that art: capable of the n!'!Cnnry tdie functions, But precisdy
b«J,ug th~ p:articulus CUI function in 2 causally efficient ITl2Ilner, they
haY!! ncccsnrily cc:ascd to aist: by the time one has acted upon them.
For Devcndrabuddhi, this problem of.1. rime-I:ag is most acute in the con-
tat of the confirm.1.tion-modd: when one acts upon some initial cognition,
the instrumentality of mat oognition consists of its raulting in 2 cognition
in whkh appears the required tdie function (,1rthtthiyt1"irbh4Sllj, i,e., one in
which appears the accomplishment of one's :aim. As noted prevWusly, the
subsequent, confirming cognition would be the sensation of warmth, for
a:ample, that one has when, having Ittn 2 fitt from 2 dinana, one draws
near to il. On Devendrabuddhi's view, this !)'pC of subsoequent cognilion is
panicularly important, for he d:aims that it establishes [he trunwonhincss
of the firsl, Although we wiU rttWtl [0 this notion of oonfinruuion below,
the foUowing poISS2gt' offers a dear summary ofDevendrabuddhi's position:
For one who acts through being prompted to act by a f.lulry cog-
nition that apprehends something that is not fitt as fitt, a sub-
sequent awareness that has as its object burning and cooking
does not arise. That awareness docs not arise beausc: an aware--
ocss in which the expected tdie function .1.ppears is b:ased on a
real thing. If Uur subsequent awvencss docs arise, Ihen the for-
mer can only be Itwrwonhy because: I) one obtains a tdie func-
tion in accord with one's OIpcaatioN and because 1.) the cause
of just th.1.1 kind of .1.wareness of tdie function is a rea.I thing.
Therefore, if the latter has a re:aI thing as iu object, rnen the for-
mer is ttwtwonhy with rtgard to it, '"'
104 Thu. u. Iht 1On:c of Dhannaldrti'l dain Ihal Ihcr-c il only OM iruuumcnw object
(J>V",~-d): ..,.",,.,q'!' 1. .w.,.!fM'L
lOS PVPbb)R): _ -'1"~ _ iW"1" UPIqi 1..,1"'''''1'' w / k1"imt1",ullf
'nMr/". J. -rr 1" i yoJ .,." ai ~ ItJiIiwf ~ ~ ... ,.J1" - Ji" '" / ... III ;",.. ,.; ""'
.,." """:fl" ,. j #,. /"" .... JDff"'; liM .. iJ. .... Ji" r / ji Iu, "'''P'' 1'" 'MIl,. kJ,;"
." ." .".,. . .,.,JI1JIr " " ' . " III tk ...'" oi 'JJ'" ",up" ,. i fhIi, ,. / tk lui,
". JIhp' _ ;",.. ,. i pJ .,." ",oJ:fl" ". / sttp _ '" u ,..; "" ... r.",." _
INSTRUMENTALITY: JUSTIF YING THE SOURCES OF KNOWLEDGE 18,
Dcvendrabuddhi answers:
106 PVP11h4If}: pi ~ InF" mil!' ...j .... ' " ' " 9i "'-' _ ""I:J'" ..... yi.. _ , u ji ltv-
_ tin 1"'l' 1M wu1" i J,n, ~ , . i 7fd UJf P"I:U IV f u j _ ~1" u ,.. ".; ,u, ... i
fbfo '" '" nNJ _ Ji" YiN ....
1f11 Slkyabuddhi (PVT. "J'"'71b1H) comtnmU that tht formtr and bner ~ an: ddin-
dfterm~
itivdy .. ~ umc IxaIlK "lhq- ClIIK a OJSIIilion with a lillJk imagt;"
(~ti~ em the buR of tbiJ deunnination tht""'lltkl posiu- thtm .. the tanV
ob;ca (ttrrIM). and thil happetu Ixa..uc tht po.itin@:ofan iruuumenw rosnition iI con-
_Donal [.-...1",n, ""'1"; 'U" W#Ji"1'" pJtJi, p, ....... ..,-,.,.Jllt7itrit ttl..,..,
I"""1a' n"fu·j 1Jftnl ~ _u (1m....",,., 111"]1" ~ / 1IhU _ 'i mPlI"" y.,1"
,,; ,;", I1t)'t'i (111 W# )'i.. ,. i ,;,,;, .. tlNJ bJu.. 1"" Dw,,.j·
284 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
Therefore, since the real thing toward which one acted was estab-
lished prior to the cognition in which its telic function appeared,
that initial cognition is instrumental because through it the latter
cognition engages with the telic function (arthakriya}.108
~hBSTRUCTED ACTION
~ addition to the aforementioned difficulty involving time, a second gen-
~ra1problem persists with the notion of instrumentality in terms of a
human aim, and it concerns the identification of an instrumental cognition
"that which makes one obtain" that aim {artha} by virtue of obtaining an
gbject {artha} that accomplishes it. This issue is broached when Devendra-
buddhi notes, "Since an instrumental cognition is this or that cognition
~hose trustworthiness has been ascertained, doubtful cognitions and such
~e not instrumental."109 In response, one may ask:
;;i.p9 PVP (3aIff): mi sLu ba de dang de nges pa gang yin pa de ni tshad ma yin pa 'i phyir / the
~tshom za bat shes pa La sogs pa ni tshad ma ma yin no. Note the typical subject/predicate inver-
:sion in the Tibetan translation.
~UO PVP (Ja2): gaL te skyes bu'i jugpa La yanggagsyodpas / tshad ma yang sLu basridpa'i phyir
tshad ma ma yin no zhe na.
~lll Sakyabuddhi notes (PVT, nyq6bI-2), "By implication one supplies (adhyiihiira): 'In all
1B6 FOUNDATIONS OF DHARMAKIRTr5 PH ILOSOPH Y
aKI when Of>( is I><K obMruacd..• •• [..... r.J tI. PI' I>]tJ ". 'j '17"' - " ... rhn.,. N
..... - a,."" .....
112 PVP {)a.I.fO: M If; nlNJ _ "t" f'D MJ,. j " If ""I pImuI,. If; _ Jilf IW 1 ; ... n
p" ,;" ... 1M £U "'''P'' 1M 'J.JJI4 wpJ4."lfhuJ,. Ji,. ... I M,.,.J ... ohM - fh-I
I'D I>]tJ JI4 ..,;JJilf ,. M Wr ... ~ I'D - JI4 ohM - ..,uIJi. 0' fhuJ ,. tt]iJ ,.i -
yilf ... I ."',. M",.. iii;"'; do. Np,.,.. lwjM,.i pJtyi, sJr,.,. ~,. _ yi,. _
II J 5« PVf, "J""7'a7ft
INSTRUM ENTALITY: JUSTIFYING T H E SOURCES OF KN OWLEDGE 1.117
in orner ca.so: I will find only th~ hO( sand of a mirage. In short. what dis-
tinguishes an instrummtal cognilion is not that it n('('N S1 ri1y leads one to
me result, but ramer trun it has me ~"parilJ to lad one to mat rault if all
omer conditions are: in platt.
An. important outcome of the definition of insftum~n taliry in terms of
capacity is dlllt it amounts to a rejection of Devendrabuddhi's earlier ·con-
firmalion-model- of trusrworthiness in mt'! m nrar nf a human aim. Tn
reiterate, me confirmation-model is: with some goal in mind, one h:as an
instrumental cognition of an object thai has the capacity to fulfill mal goal;
one implements the means ro obtain that object, and one Ihm has a cog-
nition of the fulfillment of one's goal. On this mood, the trustwOrthiness
of that initial inn rumenw cognition consists of Ihe fact thaI it leads to
anothr:r innrum~t:al cognitinn whmr m ntetlf i. rht'! ~i roi fdie runction.
i.e.• the ach i~menl of one's goal BUI if trustwOrthiness is redefined in
terms of capacity, then the production of thal rubsequt:nt, confirming cog-
nition is irrdevanl to nustworthiness.
Why does Devendrabuddhi find himself in such a muddle? Cenainly,
part of wh.at aplains this tension is limply mal Dcvendrabuddhi h:as
b:adct'!tl hinudfinm rhecoOW'!". ~dr:1lbuddhi 's ~nition of an instru-
menw cognition in lerms JUSt of the ~1l"'L.'ilJ to result in the aehievC'm~",
of one's goal comes after his initial presenation of the confirmation-modd,
and it IIrikcs me :as an attempt to avoid a claim made in Ihat conlat:
namdy, thai a trustworthy cognition acru.aJJy thn lead one 10 accomplish
one's aim.'" BUI although the move to mctt capacity is a aaKal rettcat in
hiJ argumenl, for our purposes it serves 10 highlight another, implicil :aspect
ofDev.endnbuddhi'. accou nr. n2lTlCi y • the bipanit" norio n of extrilUic and
intrinsic instrumentality. To see how mis is the case. let w now rum 10
Devcndrabuddhi's discussion of me confirmation of the inslrum~tality
of perception.
For example, for the person who. having cognized a fire through
PClUprion, then actS (j~r ptt) on fire',1II capaciry to bum. rook
and 50 on. there is the activation of a pclccprion whose ob;ca is
the sensation of w:ltmth and such. Or, for aampk. on ccnain
oc:x:asions there might be a cognitive error due 10 something which
has a form similar.o fire and so on; at that rime, there is for tha.
person the aaivarion of an inference through smoke which defin-
itively determines the firc..(Dcpmding on the motal, one of thcst:
twO-{hc engagement of a subsc'Quent pcn:cprion or infttenoe-
confirmsl the tn&$lWOnhincss of a pciOepcton.n. lkawe various
causes of error ate possible in the case only of pcrcc:ptton, it is
known to be instrumenral through the activation of a 5UbscqUCOI
if\$trulllCnral cognition thai h.a.s as iu objca that thing's (.rtIM)
· 1: __ •
IeIIC runctlon ....
'It
In thiJ puugt!. n.:v.. ndr::ahuddhi .q-.r:alu nf rwn C25l'!l involving :action thu
is motiv:atcd by an initW petceptton tim requires confirm:ation. In the fin;(
c:ase, :an initial perttption of an object capable of the desired tdie functi on
motiv:ates one to :act tow:ard th:at object. For enmple, one perceives wh:at
one believes to be:a fire , :and intent upon obtaining he:u , one walks toward
th:at fire. In this first c:ase, th:at :action results in the :acrir.ltion (prlllJ!fti) of
:I poerC\'f'tion wh~ conr<ml ;1 t~ de<i~ l..Ji.: h,nt:fion, ont' l':Wf'Hi/':n~
heat. In me second c:ase, one is also prompted to act by a pera ption, but
one's action results in:a perception ch:at is not definitive: per~ps one only
draws close enough 10 perceive a color th:at might be II fi re, but might also
be a brilliant dump of flowers. On the view of DharmaIoni :and Deven-
drabuddhi, in this second case the perception following upon the :action is
urable to produce:an immediately subsequent definitiw determin:ltion or
correct perceptual judgment (prllty4/q11P."l.thlllAbJhanikllJll) of the object as
havi ng the desired telie function. UI In this latter case, the initial perception
mWl be confirmed. by m inferellCC.
The key here is that, regardless of whether the initial, motiv:ning per-
oeprion is confirmed. by a subsequent perception or:a subsequent inferena ,
that initial pc:rcoeprion cannot in and ofiuelf guar::amee th,u one will att:l;n
one's goal. In shon , that initw pe:rctption does not enable one to deter-
mine whether the content of that perceplion is something (such as fi re)
capable of the: desired tdic function (such as heat). Ncvenhdess, it is cru-
cial to note that Dcvendrabuddhi does not wish (0 claim m:at such a per-
ception cannOt be instrumental. Instead, accord ing (0 Dcvendrabuddhi
and the sub-commenra[or Sikyabuddhi, when one acts on the basis of
,ue:h a pe:rce:ption and one: then auai n$ an o bject with the de$ired or
expcctro [die function , m..t initial perception was insuumental; one was
simply unable [0 determine the instrumentality of dut perception :at (he
time of the perception. III
To llCCOUnt for these cases and preserve Ihe inslrumentality of such pe:r-
In Noce llul I un noc deaJin& hen: wilb dw:qlUllion <lit- ~n inflrmnoflm frum......,
could be. qnilion "in which a.p~m Ihe accomplUhtlM'nt of OM', aim" (",.,;,.m~.
IIirMJ.).
1 23~; ............. 0-..0 • • ~ p<>in.. ;n ....,d~ _ ....Ii............ (_, .... '
PVr. II;"'.7)~ff).
INSTRUMENTALITY, JUSTIFYING HIE SO URCES OF KNOWLEDGE 19 1
Siky2buddhi's point here is rhar not all actions butd upon perttprions are
tentative, for in some cases one's action is ~habitual ~ ("bhjNis4lN1ti), in me
sense thar the one has undergone Mhabituation M or · conditioning M
(.bhytIu). In such cases, one's perctption is able to Mdet'ermine its image,M
which means that the perctption itself is capable of directly producing a
subsequent correct judgment that can guide one's actionY1 Somrone
might, for ewnple, become sufficiently familiar with the fire-like flower
that she can easily distinguish fires from a d ump of such flowers, and she
is thus able to determine without hC$itation that the enti ty on the far side
of the field is ind«d a fi re.
Sakyabuddhi later goes on to specify that this distinction amounts to one
between petuptions that have intrinsic instrumentality (lWub pr4m1i!'JAm)
and amnsic instrumentality (pttrtlltll; prlmJ!'J'Zm), He comments;
fu,,..,
i .. ...'" N ..... ".., ]i..... 1tk k ~ iJtJ,. '" ... '" ~ ..... N .."' ....
",..,.,,..,uJ
"'",...,., '17" ".."',. p]iI Ji- ... J1Mr,. JUt """ twJ ,. '" u" nh.", U N
J "-..N ~ 1tft1,.pyis ...... ]i.. ,.. , . 1t]i 'l)W pyU ftrU k,."
"1tf
Dw' N p1JfJi.. ,.....,,.,., MM,. pfo 9U 1Mf,., ;po, N p1Jf';' J¥ '" ,.; ".,
, . utpII i .. -r N .... Ji.. 1M' j~ ".." ... "'iM M Jpi ... "'" ,. ,.,w 1t]i 'P - k
,., iJtJ,., Dw' N "t' ..oJNo]i" 1U J;, .....i zhi1tfJ¥ "',.,.,,. zhi", "'''''''' '"
kIM iJtJ,. p1Nr p, .... ..,"'".., ."",.;)'" '" ""',. p,. '"
~,.; nNJ ".., ~,. ".., ]i.. ,.,.j,
"tn'"
127 ~ (PVT, 9"-7-tbUfl nocct:
I .. • ,M.-..-.-tn [och«) cua of pcrccpion--ladU", habilUJ.OOn, w capmmu
(",dJ.,., N • J.rlMw) U noc cku. 10 IIx panicIIIu idenutY (~,., • .w,.) of me
obj«l is IlOI dcfinilivtly cklmniMd; Ibil apWru the aK wncn one: il initially
prompied loaa [uoppoKd 10 habit\latnl aaion.o IoI:c ~J, AI tNlUmc, OM aIao
.:u OUI of doub!. By implicaDon mis IWCI W.I if, d~ 10 iI.:IbilUOllion, OM ckhnitiwly
appehmdJ (.,,.,,P"f ~!he panicuIu identity. e-om paotption docs nor Ikpcnd
-'-t
I,lpoD !he n\~1 of a sub.cq\ln\1 awumca, 1pooJ,. - ',.,
j..,,.,
N pi N
JlbJir ItI.,.J,., "l'" ,. twJ,.".. ... 1 'Jis ..i ..,.,,. ..,..1t]i
, . __, . j' iJtJ
,.;0 pM _ " MIIII M I J, 'i nJ...* ....oil A::-t kIM'''''' M JP"'I" Ie A::-t
~ "'"~,....,~ .. ~ .... ,.,... J,yJ ..J-I_ i M'" J.. ""'r" _,..., J_
oU
,. "..,]i.. ,. ziN1 ..... ~,., ~, Dw'"
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 293
~B PVT (nye:y5a6ff): de 'i phyir don byedpar nus pa 'i yul can gyi mngon sum ni 'khrul pa'i rgyu
'fitshan medpa'i phyir tshad ma nyid kyi bdag nyid du gyur pas rang rigpas yongs su beadpa yin
,'ii}lji Itar yongs su bead pa bzhin du de la nges par skyed par byed pa de Itar na rang nyid kyis
~d ma yin pa de 'i phyir thug pa med pa ma yin no.
:~~, .
;U9 See HB (2*.13-14): tatra pak!adharmasya sadhyadharmiIJi pratyak!ato 'numanato va prasid-
'i!r~r nifcayab, yatha pradefe dhumasya ....
~~~ Sakyabuddhi addresses this problem earlier in the argument (PVT, nye:y3b6ff) by rais-
~ti1m objection in the context of inferring fire from smoke:
'~s it not the case that, if smoke is apprehended through perception, then since there
are causes for error, the evidence will not be definitively determined? That is, there are
:#Ie causes for error that prompt one to wonder, "Is this a thing that has the appear-
ance of smoke but is actually magically created by a yogin? Or is it coming from a ter-
:mite tower?" [du ba mngon sumgyis gzung ba na de la yang 'khrul pa'i rgyu yodpa'i phyir
!tags nges pa nyid medpa ma yin nam I de Itar na ci 'di ni rnal 'byor pas sprulpa'i du ba'i
t;tam pa 'i dngos po 'am lei brgya byin gyi spyi bo las byung ba yin zhes 'khrul pa 'i rgyu yod
710 zhe na IJ.
buddhi goes on (nyt,:? 4aIff) to argue that this is not a problem since the evidence can
lImed by a perception that is "fully habituated" (shin tu goms pa), i.e., one that is
... sicaIly instrumental.
19-i fOUNDATIO NS OF D HARMAKIRTI 'S PHILOSOPHY
133 PVP (5a5): ... bzung ba'i rnam pa gangyin pa dag fa yang nges pa skyed par byed pa de fa
/ Jugpar byedpa 'i phyir tshad ma nyid du 'dod kyi / gzhan du ni ma yin no / mthong ba las khyad
par med na yang / rnam pa gang don du gnyer ba dang / goms pa dang skabs fa sogs pa nges pa 'i
rgyu yod na de nges pa yin gyi / gzhan chod pa ni ma yin no.
134 See chapter 3 (n. 59).
296 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
memaiity-as intrinsic and (Xtri nsK:- rclates to the "muddle- th.:1.( closed
the previous section. There, we noted that rHvendrabuddhi begins by
claiming mat, in the como:t of a mediated instrumental eff«t. an instru-
mental cognition is "what makes one obtain" (prlipUllJ an object with th~
desired tdie function, and that the fact of obtaining such an object conni-
rutes the instrumental cognition 's tNSt"N'Orthiness. But toward the ~nd of
his arpment he must account for the possibility of obsnuat'd :action. and
he is thus obligM to reddlne instrumentality simply in tttnu of an inuru-
m~ntal cognition's ClfJ'iUi1J to make one obtain one's goal (i.e., an object
with th~ desired tdic function).
We can ICC this tension Ixrween actual and potencial obtainment of one's
aim as paralld to the division ~n extrinsic and imnnsic instrum~n
tality. Specifically. for an extrinsially instrumental perception (0 be known
as insuumental, it musl 3Ctually lead to the accomplishment of one's aim ,
whe:rc this is defint'd as a cognition in which appears the desired telie func-
tion. This is 50 txau.sc: when a cognition's instrumentality is exninsic, it
rc:quires confirmuion by a subsequent instrumental cognition whose con·
tent is the desired (elK: function , which is the same as saying thai the con-
firmation requires mat one actually obllli n one', aim.
In contrasl to the CISe of cxtriruic illSTnimenuliry, a cognit}on who$("
instrumentality is intrinsic requires no substquent confirmation, and mis
amouna to the claim that the cognition is inmumcm:a1 even if one's aim is
not: aaually obnincd. This notion oorresponds to the charact:eriz.u ion of
instrumentality as the cognition's capacity 00 make one: obain one's aim
(i.e., ia capacity to lead one to a cognition of an object with the desired ~Iie
limo;","). If Om' urv:ierl'l':IodJ lM rerm ·C1 pacity~ (/A},';) to include cua
where the goa1 is actually raliz.ed, one can inttfpm such a cognition as me
trivial case of a cognition that contains the appearance of the desired telic
function (IlrtJmhi]dllirbh4sJlJ, since th:tt ap pearance is c:quinlent to the
obtainment of one's aim. But mo~ inte:restingly. such a cognition can also
be understood as one that, even though it does not dir«t:ly contain an
"Pf>e1r:m~ of the deircd tdie fi.mCtion. noevcnhelen is indubiubk :about ito:
object', capacity for that [elic function. Although ~buddhi poina O UI
that 5Om(: ascs of ~tion fall into mu category, [)(:vendrabuddhi places
particular SU'CS$ upon inference as intrinsically instrumental. For as we have
ICCn, a we'll-formed inference must be bucd upon a JlIIlbh4~'ilpr4tibandJu,
pmaining bctwttn cvU:knce and prtd.icate. Sincc mat relation guarantees
rru.t me encity inft'llcd h.n:a CftUin type of IU.ture, it Iilcewise guv:mc«c ch:i.r
that entity is capable: of sptt:ific telic funaioru; this 50 because tdie funct}on
298 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
138 There is no doubt that on Dharmakirti's view, what he sometimes refers to as selfless
(e.g., at PV2.135-13G) and sometimes as emptiness (e.g., at PV2.214d-215ab) is to serve as the
"antidote" (vipak!a) to satkiiyadnti, and hence to all suffering. Dharmakirti does not, how-
ever, provide any extended refutation of the self in either PV or PVSV. But as with the
Abhidharmic analysis of sa1f1vrti- and paramiirtha-satya (see chapter I, 41), it appears that his
refutations of distributed entities-especially universals-are meant to serve as the argu-
ments to be employed for refuting the self.
139 For more on Dharmakirti's soteriology, see chapter 2 (Goff).
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 299
issue here is the notion that part of what constitutes the instrumentality of
a cognition in the context of human aims is that it is a "motivator" (pravar-
taka) of action. As noted earlier, the notion of "action" (pravrtti) here is the
secondary sense applied to that term by Devendrabuddhi when he addresses
Dharmaklrti's claim that "awareness is instrumental because it is the pri-
mary factor in one's action toward an entity that one wishes to obtain or
avoid" (= PV2.3b-d). It is clear that Dharmaklrti himself is concerned with
the argument that perception, as indeterminate, could not be instrumen-
tal because it could not motivate action. In his Pramii1}avinifcaya, he raises
the problem with an objector's voice:
This poses no problem, since just when one sees the object, there
occurs a mnemonic awareness that arises due to that perceptual
awareness. Due to that mnemonic awareness, practical action occurs
because ofone's desire [to avoid or obtain some aim]. 141 [PVin:I.I8]
Although direct awareness is the bare perception (Ita ba tsam =
drHimiitra) of an object, just when one sees the object, there occurs
a mnemonic awarenesJ'-i.e., one that is immediately following
that experience-and due to that awareness, purposeful action
occurs with regard to what one has seen as being what is desired
or what is not desired. 142
140 PVin ad 1.18 (58.I2ff): '0 na da ni mngon sum ma nges pa'i bdag nyid las ji ltar tha snyad
du gyur I 'di'o zhes bya bar nges na ni bde ba dang ,-dug bsngal gyi sgrub par byed pa dag thob
pa dang spong ba'i don du Jug pa 'i phyir ro zhe na.
141 PVinr.I8: taddnfiiv eva dmefU sa1?Zvitsiimarthyabhiivina~ I smara?Jiid abhiliifepa
vyavahiira~ pravartate / /.
142 PVin adLI8 (58.I5ft): I skyon 'di med de I ganggi phyirl don mthongba nyid mthongrnam
300 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
la / myong ba'i mthu las byung ba yi / dran las mngon par 'dod pa yis / tha snyad rab tu Jugpa
yin / mngon sum don la Ita ba tsam yin yang nyams su myong ba'i mthu las byung ba myong ba
mtshams sbyor ba'i dran pa las / de mthong ba nyid na mthong ba rnams la mngon par 'dod pa
dang cig shos dag gis tha snyad du 'gyur ba yin no.
143 See Dreyfus (I996) for an account of these issues; for the title, Dreyfus draws on Tibetan
authors, one of whom compares the problem to "the fool leading the blind." The question
here is how the "fool" (perception, which cannot think thoughts) can lead the "blind" (con-
ceptual thought, which cannot directly see any objects).
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 301
The major problem here is definitive determination itself. That is, fol-
lowing upon a habituated perception of a water-jug, in the proper context
one will have the definitive determination, "This is a water-jug." Since that
perception leads to such a determination, it is considered instrumental.
That is, in the case where one seeks the function(s) that a water-jug per-
'forms, that perception can (indirectly) prompt one to act in such a fashion
that one will definitely obtain what one seeks, provided that nothing hin-
ders one's actions. But clearly, the same is true of that definitive determi-
hation itself: it can effectively prompt a person who seeks a water-jug to act
in such a fashion that, provided there are no obstructions, she will defi-
nitely obtain the water-jug. 146 In short, it is both "what makes one obtain
one's aim" (prapaka) and "what prompts (or guides) action" (pravartaka).
We have already seen that Dharmakirti is unwilling to admit such cog-
nitions as instrumental, in part because that admission would lead him
down a slippery slope to determinate perception and, hence, the ultimate
Jeality of universals. Thus, he must add some further specification that will
disallow the instrumentality of definitive determinations. In a reference to
Dignaga's work,147 he does so by maintaining that such cognitions are "con-
'144 PVP (3a7ff): gal te gang gis shes nas Jug par 'gyur ba'i don byed par mi slu ba de ni de fa
:tshad ma nyid yin pa de'i tshe I bum pa 'i mam par rtog pa 'i shes pa las I bum pa fa Jug pa de'i
:.aon byed par mi slu bar 'gyur ba de ltar na I de yang tshad mar gyur na mi 'dod pa de bas na
mtshan nyid khrugs [ex con). for PVP-D: zhugsl pa yin no I zhe na.
,145 PVZ.5d-6a: svariipiidhigate{l para1'(l II priipta1'(l siimiinyavijfiiinam.
146 It is worth noting that, in terms ofDharmaldrti's apoha-theory, a definitive determina-
tion must therefore be a concept based on real things (bhiiviifraya), at least in relation to the
:!Spects (iikiira) of the object relevant to the desired telic function. For the notion of expres-
sions and concepts based on real things, see PVL205-206 with PVSV ad cit. (G:I05.24ff).
147 I understand Dharmaldrti's use of sii1'(lv.rtaat PVZ.3a-br (grhitagraha1}iin neffa1'(l sii1'(lv.rta1'(l)
(0 be an allusion to Dignaga's use of sa1'(lvrtisatin the context of perception (PSLI.6cd-7ab:
JOl FOUNDATIONS Of DHAlMAtr;;IRT I'S PHILO SO PHY
ventionar in that they art' -apprdlending whar has already been appre-
hended ~ (lfhi14trW!'ll); as such, they art' not instrumenru. ,.. Devendra-
buddhi, citing Dharmakini's response, offers these commenrs:
151 I'VP{)blft): '*'" M,..,. _ "1...."' ..fot iW" pbJi, .....",...,.; / ".; 'J.JlfJ 'Ji
IsM h",,. "",.,,..,.1fyN/
Jhn,. If; rm.J _1fYN/ J.. "'; 'J.J M J til #.M _ I P"",,, Ui",..
""", I ,...."lf ""'" iitrr,. '" /J#ff,.i,.J r.oJlrti ...... "'"'?i
'V"'"' pbJi, ... 1
M '" """, ~ "'tA...t N ff]iJ r:JhtJ ...... """ '" i"'l,.,."..,.]i" IW 1M i ",tho. t., ",",,,,
... roo_ rti ""'''' ,., ....,,. i ,..., ,. ""...... ji Iu NI>Jn" J.. ~"fI ,.. ko..t ... j Mil oi ~
r.o" IJu" IdUt ,u, ~ ... M Ju, /U in ... jyd. .,.,."..1M i J,.,.. ,. (II,., uJ .i;Jw"l
sMI ,. -",. M ...... P"I ~ sMlNU k,., .",.,., ;0..'"
152 This iI. in shan. an upuntnlapiM admining~ (..riufpt~pauption. See,
INSTR UMENTALITY, J UST IFYING T HE SO URCES O F KN O WLEDGE 303
I SS 5ft pvsv tUiPVI.})ab (G:1I.6/l) and the doIcIy ,cblrd dixlWion I I PV). IOI-I 01.
156 PV). Ia,d.
I S7 S« PVr h7Sa6): .1."., cbm,.. chn",.. .. ";......,,. W ...,,.;~,, ",,;,.,,, .. it.
IS8 PVJ. IQ7Cd: "",~"d"'~"" ,.,..,w,.." rNtbtUh~.
l S95ft D1wmWm'l loooum of"';~ {PV}. 181 47)' np«i.lly.,jlh dw; oontmc'nu
ofOcvaw:lr:lbuddhi (PVP:l lobJffl.
160 NOIC mal alihou&h 1M oonlCl.l IJ,en, iI the appIicIIion of the propcny. ,..}JM.. JdfIe:M.
new to iIO enti!y on the b.is of 1M ~ of tNl mliJ)'. Ihil doa; noc rman dw KIf·
anaa ' " '" iI pthxplible. ~ibtr. the di$rinaion hen, iI bth"UO the lbili!y 10 h~ I
<kf'in','..,dacrminl'ion ofodllcaonal dircctlr from the jX,aption and 1M fIftd 10 raon to
CYickncx to apply thai p.open,........""..,.
)06 FOU NDATI O NS O F OHAilM .... ICIIlTI'S PHI LOSO PHY
......,..,.,..nn un
IIw meditation is ~rfm:nL' [tc""" MitlfM ...-",,,, .. ,.. ,.J tfIfIi/,iMtlo,." I
",",.tU..t,.JM"",u". /~.
productd: and (4) as an ordinary person, she can nC'Vtt detennine jWt from
~rcrption that me water-jug is momentary.
The impornm point to recall here is that, on Dharmakirti's view, a per-
ccptton app~hends (lDthe a5p«tS of its object, which is simply to say mat
iU\ obj«t contributes in its emimy to the produaion of a pctcqnion.·"
This means mat alllhc: propcny-Jl4i'bhd_ of an object have been appre-
hended (ghiu), ("Ven if one has no determinate cognition of memo Hence.
it iJ clear that if a definitive determination immediately following a per-
ccptton is not iU\ insuumental cognition b«ausc ilapprdlends the already
apprehended, men the instrumentalil)' of mis type ofinferencc should also
be rejected. The reason for mis is mat me water-jug has already been ~r
oreiVC'd: henore, c:vt:n if the ordinal}' penon did not d("(ermine from that ini-
cial pertq)lion that the water-jug is momentaI}'. that propeny-1WIbhdV4
has tk fortD been app~hendcd by thal initial percqnion. Thus, as wim a
defini(i~ d("(ermination immediatdy following a pelception, this type of
i nfe~nce should not be instrumental because it apprehends what has
already been apprmcnded.
We have 5CCfl. however, tha, on our interpretation ofDhannakirti's sote-
riology. he is obliged to preserve: the instrumemaliry of such inferences.
Hence. ht must ~pccify some additional mlerion ,har will ~how how maw:
inferences do nOI involve tht app~hcnsion of what has already been appre-
hended (KfhirRf/'Ilh/lJ;Ul). Dcvendrabuddhi maintains that Dharmwni
establishes this additional criterion by claiming that. since the property-
IINIbhdva about which one is concerned is perceptually indeterminable in
that case. irutrumentaliry an still apply to thai inference. N Oharmakirti
put!; it. such inftrtn~ "illumin:!lt whn hu nnr yt'l ~n Icn(lw(l
flfjfiillil,.u",prRktJJ.) (. Defl .2r in a case "when me particular has not been
disc::trned" (tllljjfiilu '''''~!'t) wim respect to a propcrty-IfIftbh.i"" that is
M'" "'ry for the fulfiUmem of one's aim.'" In shon. it is what has been
163 PVI ,,~ ""I'hc:rdotr• ..ncn a thins (lrbMJ iu:xpniroud th<OtJ&h 1"'.Ct:prion. all oliQ
qualilia an: apniMad ....·IWJJW/~ ~-,. ~ ~J.,..!fI'!t. Thill ¥UK MiIn-
II"IaIira the UJUIMII I olPV.....r44;md PVS\' .J riI. (G:-l.6. I-17.l)I.
16<4 Sec PV Lsc--'. •mulJ.fti in the appmdiL
JoB FOUNDAT IONS OF DHAl.\tAlrWlTl 'S PHILOSO PHY
mental.'"
165 PVP (6aS): pi /Jt ria"""''''' oiMJ _""'<111"" "'M l p''''1ff hhillloi ",i
""''''' ",w ...", .. *' "" i J#7i,,. ........ ,." "tIM .. r.s IMf ""; .,..; ""'r",w
.. ,..
ali
""',. ..", _ Ji,. "" k.u.t ... ..;" ~ 1- "", 1"'" " ..,.,.,. ... tin TfS "" " . " ""
"'IIfli Jr.. ... ,.,. ~ t,tJ "" *,.,¥ tIM.,,-J J.. "'", .. blff" F"f1i ttM" .... ,.,.
"" M Iur ....... h..y" wyiJ" i-t""i" Ij4 "'.,,"",.; IIIMJ .... Ji" IW• •••
166 For $iJry.buddhi'. pooibon, Xl' PVT, "Jf'!]6u-J. For an aampk 011 l.ua oommm·
1.OIOI...no doa.- cIi.c- ---uty in d;';ncUon fn>m .nwo o •• h.i...,.,....,
Kn..et (,", .....
'''JJ on DharmcloU:aB'. thmcy ofi...uwnmulity.
INSTRUMENTALITY: JUSTIFYING THE SOURCES OF KNOWLEDGE 309
li167 See, for example, PVT, nye:79a5ff, where Sakyabuddhi, by reducing the discussion of a
l~econd characteristic to a matter of worldly convention, seems to suggest that an explicit
[~tement of novelty is not essential to the definition of an insrtumental cognition.
t1t68 Franco (1991 and 1997) has argued that Dharmakirti does not give a genetal definition
Ipf an instrument of knowledge (pramii7Jll) anywhere in his works. This interpretation is
~argely based upon the argument that the particle vii in PV2.5 can only properly be con-
~i~l:rued as a disjunction. Franco's work, which seeks a historical reading that at points must
!:resist the commentaries, has elicited considerable comment (see for example, the exchange
~~etween Franco and Oetke in Katsura 1999). Beyond Franco's insightful suggestions con-
ferning the relation ofDharmakirti's work to the wider philosophical context of his time, one
k
310, FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
salient outcome of this" vii controversy" may be the finding that, if we do not resort to com-
mentarial interpretations, our conclusions on the matter are necessarily underdetermined by
the available evidence, namely, DharmakIrti's laconic statements. This is not a problem in
itself, but it is one that must be tolerated when we assay the difficult exploration of Dhar-
maklrti's immediate historical context in isolation from the commentarial tradition. When,
however, we seek to situate DharmakIrti within the commentarial tradition that follows him,
our task is less difficult. Indeed, we find unanimity on the notion that DharmakIrti does pro-
vide a general definition ofpramii1Jd- and also that both trustworthiness and novelty are com-
ponents of that definition (the only question being the issue of whether both must be
explicitly stated). Since I am thankfully concerned with this latter, less assiduous task, I have
forgone any extensive discussion of the "vii controversy."
169 See chapter 2 (84ffl.
INSTRUMENTALITY: JUSTifYING THE SOU RCES OF KNOWLEDGE , II
..'_1..1...,...
b i l L " ..
~
H .~_nb.sy. ....",ijo iii". UiUnJn ... ' ,.~ ,.,._
",;~ 1Ii~,.,.~ H.
17 1 ~ (14a,)oIfm:,Jlishdydifkmll imttpfnllion:
Ir iI«J _ -t? ..w-i, doo ~ aiH- i>o Iiw ...., Ih.u;. it ;-pun"", ""'"-
..;~ha, .... a ~ imapnalivdy drcmtlincl the npc -ion·, .... 10 be ~
bk of ao:omplWU"l dor [<birrd[ Mfa And dorapreasion',"';"is POe aublishc>d
fI- - mnM) i .. tiN II"'J u,., illImMIIy m., u .. in dor OQy Ihli dor partia.lu it
...bIiWd tl- - n40'...v.
K{JII) tQ(IowI ~yabuddhl firM p/u"aIf. bul 00 ttlt: I«'O!Id phruc he rmurb:
00 lhe
8«_ it it _ ilflnfMt/ i.. ,., _ ......,.1iNu It.,...u, nri,UI bec:wx dor ".,,;,. U noc
akm u VI ob;cct by dor c:ap<asOon in dor IYWIMf thaI oM .nA.panirubr or _
tal ~io.tbr ;, a ub&hcd 10 alsc. lJouloI~ ......;h;IiIIIIIll' ,.M>4s-t.~!W".
~!M'" fill ubiu" u'" M~ktutfllltl .
172 ~u..n. wbieh ~ ((l.4u}.KClU)) c\oua <&II ~k.tM.
17) PVSV .J PVI.llo--lIl (C:I07. I....,l: ... hi t.UJ. i:.",." fIII"".m ,.~.
312 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
Although all conC(pnl,-1 cogni rionJ such :IS inferen« ,-re con·
fused, wt still define some as instrumental and some as spurious.
We do so because we '-gree on the ifllended capacity or lack
thereof for tdic function . We do so until wt arc fou ndationally
transformed. fThc cognitions of sclAessness and such] arc: f.Use
(bcausc they proceed from the imput2tion of unral universals].
N~rthM,. (My lI«' inffrum~r:oI <»gnirionJ bot!.ea~ they ue
conducive to the pacification (of ncgarive ment2i StatC!l] .'1II An
Despite the fundamental error in conceptuality, one can still speak of some
conceptual cognitions-specifically, inferences-as instrumental because
they are trustworthy in that they reliably enable us to achieve our goals,
such as the goal of pacifying one's mind and attaining spiritual liberation.
But one admits their instrumentality until one attains "foundational trans-
formation," which refers to the elimination of the form of ignorance
(avidya) that underlies the errors in conceptual cognitions. 178 This implies
that after foundational transformation, inferences are no longer instrumen-
tal for that person. In short, Dharmaki:rti appears to maintain that, from at
least some rarefied perspective, inferences are not ultimately instrumental,
for once a person has obtained the human aims for which such cognitions
are useful, why would one continue to admit the instrumentality of such
cognitions? Why would one continue to carry one's raft on one's back?
sUg&csts that. while the Buddha's leachings arc true to his followers, they
:are not true to him. In this way, the met:l.phor appart:ndy pointS to the
pragmatism of the Buddha', teachings: they are only to be :KUpted in rcb~
tion to a goal, :and oncc thaI go:aI has reached, the pragmatic truth of the
teachings mould be abandoned, since they wac only (0 be considered true
for the purposes of reaching that go:aI. This appeal 10 pr2gl'1t:l.tism, h~r,
is in some Wll)'S mi5lc:1ding, especially when we consider Dharmaldni',
notion of·conventional- (Ul",lIJilvahirilu.} :and · ultimale" {pirllmtinhiluj
instruments ofknowtcdge,
Dharmwrti', notion of convenlional and uhimale inllrumena of
knowledge: falls within dtc EpiStemic Idc:alisl CriliqUC of pel<xption. AI mal
level of analysis, not only inferencc, but also percqnion is distorted. or en o-
ncous (bhr4Ifu) bcoUK perception involves an · internal distortion"
(IInlllfllJHIpiaWl), This distortion. a form of ignorance. ntakes the cognitive
image: in perception seem as if the objects of pcrcepcion :are external, evc:n
though, according (Q Epistcmic Idealism, no such objccts exiSt ouuide Ihe
mind. On :an tvcn stronger rtlding. the distortion :abo causes the ·varie~
ga.tKm'" (citrlltJi} of the object, such that it appears 10 have various attributes
(Iuch H colnrs) :and dimen.!:inn~ (~ uda :Ill heighT :a nd widrh). Whether
u nd.erllrnod in io Sirong or weal.: form , rhe error is nonconccpn.:al: rhe cog-
nitive: im. iudf is distorted.'- If, hOWC'V(:t, evc:n perception is somc:how
distorted, what distinguishes a rdiable or trustwOrthy cognition from one
thai is nol? If all our ordinary cognitions are distonoed such that none actu-
ally !dim or rt:presttlt things the Wlly!hey uoly:uc, then would not all cog-
nitions become equally unreliable~ In his PWM!"'w"ikitp, Dnarmakirti
OOlUidel'f. (hi.!: U:IUe-: (OI
Although all perceptions are distorted by the internal distortion, one can
still distinguish between perceptions that are instrumental and those that
are not. In effect, one can do so because the internal distortion is always
present in an ordinary being's mind: for as long as a being is in saYj1sara, the
defect that causes the distortion remains. Since all beings in saYj1sara have
the imprint, they are all equally confused in this regard, and they thus can
interact consistently within that mutual confusion. In a similar context, a
later Buddhist thinker speaks of two persons who, because they have the
same ocular defect, can hold an intelligent conversation about the two
moons that they see. 183 Other, additional, errors are not "stable" (drdha) and
all pervasive, and they therefore can be contravened by other perceptions
182 PVin adI.59 (roO.12ft): de yang shes pa thams cad kyi yul bzlogpa na bsfadpa dangcigshos
su ji ltar smra ste / khyad par med pa'i phyir ro zhe na / nye bar bsfad pa'i bag chags mtshams
sbyor ba medpa'i skyon gyis mi mkhas pas kyang tha snyad fa yid brtan du med par mthong bas
gcig ni tshad ma ma yin par brjod fa / gzhan ni bag chags brtan pa 'i phyir 'khor ba ji srid bar mi
'bral bar rjes su 'breI pa ni tha snyad fa mi slu ba fa ltos nas 'dir tshad ma yin no /'di ni kun tu
tha snyad pa 'i tshad ma'i rang bzhin brjod pa yin te / 'di fa yang pha rol rmongs pas Jig rten slu
bar byedpa'i phyir ro / bsam pa las byung ba nyid kyi shes rab goms par byas pas rnam par 'khrul
pas dben zhing dri ma med fa log pa medpa don dam pa 'i tshad ma mngon sum du byed do. Cf.
PVin as collated against the fragments collected by Birgit Kellner: upapfavavasaniiniibhi-
sandhido!iid aprabuddhasyiipy aniifvasikaY(l vyavahiiram utpafyann ekam apramiiIJam iicak!ita,
aparam ii sa?psiiram avifli!fiinubandhad.r¢havasaniitviid iha vyavahiiriivisaY(lviiditpek!ayii pra-
miiIJam, sa1pvyavahiirikasya caitat pramiiIJasya riipam uktam atriipi pare vimii¢hii visaY(l-
viidayanti lokam iti cintiimayim eva prajfiiim anufifayanto vibhramavivekanirmafam anapiiyi
piiramiirthikaY(l pramiiIJam abhimukhikurvanti.
183 Such is my liberal interpretation ofrhe metaphor as used by KamalaiJla (TSP adTS 908).
The actual context here is rhe "unconscious error" required by rhe apoha-rheory. In order for
us to apprehend language as related to rhe external world, we must misconstrue rhe mental
INSTRUMENTALITY, JUST I fYING THE SOURCES OF KN O WLEDGE JI7
conltnQ of couap!uaI and IinsuOOc CGpIino... .....:h ~ ..-.:: e&kc dMlIC mmmtllO aomebow
~ idmOOd 10 dw; real, t:m::rnaI thinp chat they rcpiuu.t. lr anyone ~ DOl fa make dw error,
!ben _ would ~ wubIc: IOspcak in~ with wr penon, but Una: all I&KQ ol~
do maU dw~, ..-.:: an 5PCaIr intdJiajblywidl achodlcr.bow rbc...,.id, Stt:abowor: (rl,).
1M Stt PVT .sJPV).U l.-u J, uarubtcd in the appendix.
}III FOUNDAT IONS OF DHAIMAKIRTI"S PH ILOSOPH Y
T
HROUCHOUT THIS BOOK
ing of the g:n u;u i.»uo in Dhilrlllakini'l philwophy KI lUi tu pn..•-
vide an imerprtution of his work as it is prestnl~ by tht: earlieJI
known commenalors. In doing 50, 1 hope [0 have contributed 10 the his-
tory ofidc::as embodi~ by a line ofimagincd Dharmakirtis found in the var·
ious commenW'ies to Dharmakini's worb. BUII:also hope: thai an acc:oum
of Dhannwni's thouglu from the perspective of a single commennri.al
SlralQm will prompl questions that arc otherwl.K more difficuh to ask. For
example. in.lOffiC comaa one nuy find it bcsl to examine a specific issue--
such as the notion of insuumenlality--al various commentarial stnf:l. To
do so, one would uace the imerprmcion of that issue through many cen-
turies of commc.narics, perhaps including nOl only commentaton writing
in Sanskrit, bUI a150 Ihose who compo5C their works in Tibetan. To be
worthwhilt:, such a srudy would have ( 0 provide a history of the interprc-
talion ofinslrumentality, and one wculd therefo re hi! obIign!:also (0 notice
and interpret the distinctions among various commcntarial interpretations.
As a result, we will ccnainly learn a grt:al deal about the issue under exam-
inarion, but we will also find it difficult to gain an undmtanding of the way
in which that issue relata to other alpttlS of DhannakJrti's thought. Our
difficulty in gaining such an undentanding siems largcly from the enonnity
of the task in question: on the one hand, we wish to discuss the way in
which various aspccu ofDharmakini's thought are intcnwin~ , but on the
other, we have committ~ ountlvCli to aamining several centuries of com·
mentary. To accomplish both mks, we would bt: compelled to give an
accounllhat:addresses ICYerai aspccu ofDhannakini's thought in terms of
ICYerai distina commentarial strata. The sheer quantity of m:ucrial and the
diw~nt tendencies of the difft:rcnt commenr.uial sinra involYfti render
that app~ impossible. both for iU2udicncc and iu 2uthor(S), AJ a rc:suit,
'"
JI0 FOU N DAT IO N S O F D HARMAKIRT I 'S PHILOSO PHY
judgment about that a5p«t of the object in question which has the expcaed
tdic efficacy rabhimilurth.Itri]t1). In the cut of SOtnt ~r~tions . instru-
mentality may depend on a subsequent irutrumental cognition. For eum-
pic, when I see a bright color on the othtt sidt of the fidd, 1 may think that
1am seeing a fire, but that ~rcrpt:ion's lack of acuity will r~uirt that I 1M
some other instrumental cognition to come to a dtfinirivc determil12Uon
(nika;lll) conoernina this isslK': in such a cut, I will only know mat I am
mng fire when, for example, I inkr its presence from mng the smoke ru-
ing above that spot. The instrutntntality of this type of ~iCt:ption is con-
sidered lO be -o:mnsic- rpaTllt4f1). Undu other conditions, a ~rception
produces such a determination direaly in the form of a correa ~lUprual
judgment. A perception of this latter kind is considered to have -intrinsic
insuumenrality" (SVtIIil/1 pribMlfYII). Sum pt:.ceptioru are instnuncntaJ only
in that they direaly produce a definitive determination of (rhe capacity
for) the expected tdic efficacy; hence, even though any such pe:rcrpt:ion is
nCCt1suily MMDnaphull its iruuumentaJity rests on a correct amaphull
cognition, namdy, the correct judgmenr (a definitive dtlerminalion or nil-
clIJ4) mal it immediatdy produces. Thai judgment, moreover, is correa in
rMr if JlIII·Y'.,..crully ~ rn lm Im!iry fMr ruu the 6pc:crM. rdic dJic:aq. II
;1 in rd:nion especially 10 these bntt £lCl ccl"ions ,nd rhe judgmenrs [hey
mUll produce mat Dharmakini's notion of I12turt (Wilbhlllll.) plays an a~
cialIy aucw role.
The nOlion of nature ClplUre5 our attention when we note how the
apoN-rheory accounts for a correa judgment'SsUCCC'Ssful reference 10 such
an entity. Let us suppose lhal we h2ve a perception lha, immediately
inGlIcu the judgment. -nil i. :I. w::I.tCf'_jllg. • R.eetlling our dU.curllon of me
apoh.theory, we know that this determination cannot refer by way of its
rdation to some real universals because Dlwmaldrti denies the ultimate
eristencc of universab. Instead. Dtwmakini musl aca)unt for reference
simply on the basis of particulars alone. He does so by appealing to the
norian that each entity that ~ call a &water-jug" is the same as every other
emiry wr -e call :I. w:ltcr_j ug in wr all those enrities I"ve the J:lI1l.e eff'ea.
One may thus differentiare them from other entilies thar do not have the
dfttr in qucstion. Hence, even though every Imtity is entirely unique, one
effectively ignores the uniqueness of "water-jugs" in reluKm ( 0 each orner
and focuses on their distinction from tho« entities that do nOt have the
effect in qucstion. One thus construcu a universal that, in the final analy-
.~. eGmin. o f th~ cxdwion of ~ entities that do not lu._ the (:;1.1.1.01
chatacteristia expected of what we aU II &water-jug."
}1l. FOU NDATIO NS O F OHAIMAKJRTrs PHI LOSOPHY
;II hand. BUI 10 be insnumenw in mis fashion, the perttpdon need nOI
produa ;I correct judgmenl aboul twry asp«!: of:lll objro:. Hena. my
paotprion of a patch ofbluc: may lad directly ro lhe determirnuion, lnat
is blue. - And if that judgment is rebant ro my purpose, then that per-
aption is intrinsically irutrumenw, Suppose, however, that I am con-
cerned with me question of whether ,hal pen::cplual objro: endures for
more than one iruwu. In my current scue. my perceptual d:atll. alone will
not enable me (0 directly determine. -'T'hal is momentary-; hence, dUI
perception is ".1 instrumental in rq:ud ro mome:ntarine:u. Likewise, if my
purpose: rtquircs that I know whether that perctptual object is in faa extra-
mental. then my perception wiU again not be instrumental, sina it is con·
taminated by the: internal distortion that makes all of my sensory obje:cu
srem extemal.
Despite such dinon;olU and inadequacies. :lIl ordinary penon 's pttce:p-
twIU have whal we mighl call a -quotidian instrumentality-; that is, they
;Itt insuummtal about numerous ordinary iuue:s. Indeed, most of our quo-
tMiian conarns (Is mil teacup hOf: ~ Is this w:&ter cold?) are matters aboul
which OUI pen:qKioru::lIe irutrument.al, despite:lllY internal distortion:llld
",her $1Id.. probll!mc. 111;$ 'l'U)fidi:ln in$f",menl'2liry i$ OO!rT:I. inly (':I" of
wM.t Dhumakirti moIU when he tpeW of such perceptions :u rtJnvnl-
hOM/ in5trumenu of knowledge. Moreover, Ihe Sanskrit term 1I11!'1IJll1l4-
hlrill., the ·conventional,· suggesu that a perception iJ conventionally
ac::c.epa.ble in thai it somehow cohercs with the distoned li~rld of ordi-
nary beinp.' One pbyfu1 metaphor that applies in this amtext is the coher-
enl (if limited) conversation that two drunbrds may have about the (W()
moon. dUl both 1ft. Both beine; eqiWly soused. dtey both see [WI) moon • .
and they can each remm to the omer about mis remarkable even[. O r per-
haps even more i propos is the example of the -pus-rive:rs" seen by IfthlS
r hungry ghosuj . ThaI is, when humans and 1ft" gue at the ume Aow-
ing substaoa:. humans see :II river of w:&ler, but 1ft" see a river of pw.
The f.&a mal their perc:eptiOIlS cohe~ diffcrencly-humans aglU that it is
w.uer, hungry ghom:agree th:u i, is pUlo--is a ray.h of the karma ,hal h:u
shaped the likworld that they inhabit. From the perspective ofhumaru. the
hungry ghostS' perttptwllS att dUlorted. :lIld from me point of view of :II
hungry ghOSt. the humans see it wrong, Nevenhdess, the humans aglU
among themselves that what they are seeing is water, just as the hungry
ghosts agree that they are seeing pus. 6 This cosmological example is useful
for understanding Dharmakirti's notion of conventional perception. That
is, even though the lifeworld experienced by ordinary persons is contami-
nated by the internal distortion and other such problems, they can usually
agree on what they are seeing. In other words, their perceptions cohere in
regard to most daily issues.?
We might think, then, that we have returned to some admixture of
internalist foundationalism, confirmed by an appeal to coherence. The
ultimate arbiter of a habituated perception's intrinsic instrumentality
amounts to the fact that our perceptual content appears a certain way to
us, and in this sense the theory resembles an internalist foundationalism.
But when we examine our perceptual process rationally, we uncover cer-
tain contradictions that point to fundamental distortions (such as the inter-
nal error) in that process. Concerned that our sense data may therefore be
somehow compromised, we confirm our interpretations of our perceptions
by appealing to what others report, and we find that everyone from whom
we can receive a verifiable report is in agreement, at least to a degree that
6 Along the lines of Asa.ti.ga' s Mahiiyiinasal?lgraha (I5b-I6b), this case is raised by Vasubandhu
in his Virrtfatikii, a text that Dharmakirti probably considered important. Sakyabuddhi and
Devendrabuddhi refer quite clearly to the Vil?lsatikii (see the translation in Appendix 7, espe-
cially notes 14 and 18). The point of the example is that the same locus is seen by different
communities of beings in divergent ways. Vinltadeva (I77a) limits his interpretation to dif-
ferent communities of hungry ghosts; this is less striking (and less well known) than the
notion that beings in entirely different karmic states are having radically different experi-
ences of the same spatiotemporallocus. In Dharmakirti's time, Candrakirti refers to such a
case (Madhyamakiivatiira 6.71), but a particularly clear account is offered much later by Asva-
bhava in his commentary (82b-83a) on a verse from Kampala's Alokamiilii:
Due to the variety in their karma, when pretas look at one river, it is filled with pus,
and it likewise has the scum of urine and excrement; that is, it has a scum that is a mix-
ture of feces and urine. But when humans look at that same river, they see it as clean;
and they drink it. ['las sna tsogs pas chu bo gcig fa yang rnag tu 'brub cing de bzhin du
bshang gcin rnyog pa dang /dan na / phyis dang gcin 'dres pa'i rnyog pa dang !dan par yi
dags kyis mthong ngo / de bzhin du chu bo de nyid fa mi yis chu bo dri ma medpar mthong
ba dang ni 'thung ba yang yin no I]
7 I will not discuss here the mechanisms of that coherence, but in addition to an obvious
appeal to karma, Dharmakirti's theory also likely rests on some notion of intersubjectivity as
briefly described by Vasubandhu (Vil?lsatikii I8ab and Vrtti): "Cognitive representations are
mutually restricted due to mutual influence. That is, due to the mutual influence of their cog-
nitive representations, all beings' cognitive representations mutually restrict each other, as is
appropriate" [anyonyiidhipatitvena vijfiaptiniyamo mitha~ /r8ab/ sarve!iil?l hi sattviiniim any-
onyavijfiaptyiidhipatyena mitho vijfiiipter niyamo bhavati yathayogam}
CONCLUSION
~
;o. not fully correspond to ontological realities, and to the extent that they
" er, they are located within different conventions which are sustained by
l~e beings within each world. Hence, inasmuch as some aspects of our per-
tption~ are err~neous, the ~y we perceive the :world does not r~flect an
pntologlCal realIty. In part thIS means that, even If we feel and belIeve that
I
~'~r perceptions correspond to the way things really are, the lifeworld with
.l.·•hi
..·•.. ch those perceptions cohere is not ontologically given. Leaving that life-
'prld behind therefore does not require some kind of radical ontological
j,
328 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
8 I am alluding again to the verse cited earlier from Santideva (BCA 9:3-4ab): "In this regard,
it is observed that there are two kinds of persons, the spiritual adept (yogin) and the ordinary
person. Among these, the ordinary person is refuted by the adept. And adepts are refuted by
successively more advanced yogins through a distinctive quality of their understanding" [tatra
loko dvidha dmo yogi prak.rtakas tatha I tatra prakrtako loko yogilokena badhyate II badhyante
dhivife!era yogino 'py uttarottarai{ll}. Prajiiakaramati provides the gloss of loka as "person(s)"
(jana), but it is tempting here to translate it as a "lifeworld."
CONCLUSION
'"
If lOme such modd underlies Dharmakini', notion of rdinement. then
reasoning p\;ays a crucial rok in the trmsition from an etTOntow lifcworld
10 a more rdined one. NcvcnhelCQ, tcaIOning is dearly not sufficient in
irsdf: it mUSt be guided by certain interesu and questions, without which
one could not conslrue an object's nature in the required fashion.' That is,
10 begin an inquiry into the ICCming pcrd urance of entities, one must be
motiv:ued by some desire ro know (jij/ils;J) directed toward some p~
{JmtJojIlNlJ. Here, one central purpose is to eliminate the suffering associ-
ated with loss: when my favorite water-jug brcab, I am disturbed: as my
body deays and dies, I fttl unhappy. But to:addm.s the problem oHm, I
must ac:ccp1lhat I am suffering in that way. Liitcwise, I must believe thaI
it is possible to eliminate thaI sufferi ng, and I musl agree that rational
inquiry into my belief in pcrdurancc is It least pan: of the mcaru to do so.
Numetow other Buddhist claims soon become crucial : contemplation
enables reasoning to refine me mind: mora.! rectitude is a prerequisite for
effective contemplation; a morality rooted in compassio n is the most eff«-
tive COnla 1 for such praaices. Dharmakini would surel y maintain that ,
without these and numerous other conditions, refi nement through rca-
IOning cannot ~ begin. In shon. the procr:u of rdiOC'ml':flf reql1i r~ nor
jwr Buddhist rc2JOning. hur a Buddhist w.ay of lik
Oharmmrri's recommendarion that we refine: O UrsdVCli through rca-
IOning and conte:mplation can thus be read as an indirect argument for the
importance of Buddhist institutions. To PUt it baJdly, in order to become
more refined one must learn to seC' me: world like a Buddhist. And $«ing
the world in mat fashion requires me pcrvuive conta l provided by the
Buddhi.u community. In other words, 10 see the wo rld :as a Budd.hUr, one
must aJlow Buddhist irutitutiollJ [ 0 mold one accordingly. And we can
thus swpcct that, wnatCVCf cbc they may do, Dharmakini's atgUmenlS arise
from-and scrve to defend-thosc institutions, NcvenhelCQ, s.ctting aside
our suspicion, we can draw orner lcuollJ from Dharmmn:i'l notion of
refinement: to be rdined, one mUSt acknowledge the contingency of one's
lifeworld. including iu institution•. Rcfinemenl also requires mat one
accepl me possibiliry mal orner equally contingent lifcworids and inn.itu-
9 To pul 11 anochn way, r:hc pi'obkm of aJUbjea-Wli in DhamWWti', norion of Nrul'C' <-
1OOffl, aIonl";d! hit nocion of r:hc conditioninc of pcrorpruaI judgnent (_ II4ffl .tquiru
dul, ~ any rcuonin&an M:applied. one must fi,.. ~ r:hc ~b;m of one', inkrcna in
1M approptble fashion. If, for exampIt;. one annor I « that 01' .rtf II • JU.tbinl penon..
inlttenca concHn;ng r:hc QUIlt of dut $Wkring ~ M applied.
330 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
tions, whose otherness places them at the edge of our ken, may have
resolved some of what we find unsatisfactory in our own. On this charita-
ble reading, Dharmaklrti counsels us to seek out those institutions that
accept contingency and the possibiliry of refinement, and he urges us to
inquire into those other lifeworlds that, while not yet understood, may one
day aid us in the goals that we seek.
App~ndix of Translations
T
HE FOLLOWING TRAN SLATI ONS provide me larger comot of the
poRioN cited in me body of this vrork. Where possible, me trotHS-
lations have been prepared from critical aiitions (I« the bibliog-
raphy). In me case of some Tibetan tou-most noably the commentaries
of Devcndrabuddhi and Sakyabuddhi-<ritic:al editiolU are not anilable.
for these texts, I employro me Sde-dge edition of m e Tibttan canon. On
some ocx:asioN, a passage in one of these tOU would prt:Knl, a.! il sQnds,
a meaning that seemed either grammalic:ally incoherenl or irreconcilable
wim me contOI of the argumenl. Only such passages were chod,ed for
variations in the Peking edition of the T ibetan canon ,
Some of the p~CI u;uularcd heft have been previously tnnslatcd,
and the chan below lists those previous versions. In my a~ricna, the best
translations build on the innluable work of their predecessors, and I am
dearly indebted 10 all who have worked on this material. I generally
avoided, hO'NCVCf. any consultation of previous translations umil the point
of preparing final drafts of the: pasnges presented below. It is hoped that this
proadun: preserved a certain nohnw ofinterprtt.uion.
Trarulalionl inlO German :and J:ap:anese wete consulud only on qUef-
lionable readings. TranslatioN iOlo English or French were checked mort
thoroughly, In some: cases, an inirial Muadon of a translation prompted
me to avoid coruulting it further, One or mort of mree basic reasons under-
lie such decisions: (I) the translarion follows primarily an indigenous
Tibetan imerpret3tion of an Indian tat whert Ihe afortmentioncd inler-
precuion cannot be reconciled with Dharmakirti's earliest commenQlon;
(1) the translation is unintdligible; andlor u) the translation consistently
interprcu passages in a manner that is 50 grammatically or semantically
'"
}31 FOUNDATIONS OF DHARMAKIRTl'S PHILOSOPHY
~
~
•
1 '3:J ~
1
:J
'3
~
1 1Ji
:J
R
j
•~ j
! !13 ....:;j ]
E
:J
=I
Ji j ....:; 1
j
:J
-§
•
~ !
~
~
~
! j i..:; j
" "" "- <
g• ~
$ z
"'" 1 "'~" ~
J• $ -! ~
~
5
~
~ J .
,z• z~ , ~ ,
. •
>
•
S i, f ~
~
~
~ i, ~
c5
!
1
,
•
if
"
•Z~ ~
~ ~ ~
-
~ ~
" ~- ~ N
•
r
Z
;
~
;
if. ~
~
~
it ;
~ "1 it
~
T
-~ ,i - •
•T ,
if.
it
~ l l
~l
"1
~ if. E •• if. E
•
~
l ~
~ li r E-
~
i ~ ~ l
J J~
~ l '1
~ "l if. ~ ~ ~ J~ l ;; ~ 1
"
~
~ N
" ~
"
I A lnIUbclOn of "1dea1OOl" mdlkkl up 10 _ dmi aIm. pam" A I'ftIdmn& of "_ "
Uoduda ~ lhan _1hW, buc lipific:anl ponjoru luoo: '-n dJded. A InnSblion of"alI" iI
of dw;mti~pr . .
2 A "NIl' COI' ... harioo IDanI that I ....00: read alI"!he lnNbUon and CIOIllfM"'d. ~ili-
lillnmn. (199) and 19t5I) AU ,..u
PVSV .... PVI.lI ..... llj
Eo"""
.Ht.!
.a: f ·1.
Yalta (1917 and. 1911)
Franco (1"7)
PVT .... NSV ../)7VI.lI4-11) Non;e
All
Sdutioou ......
En,,"" 'w'
,..u
'i'
K MlPVSV ..... PVI.U4-UJ N_
"
• d.·
.. qt
I
PVw", ...... .......
"~
J~(99)l All
c
, ~l~ Vaa Bijkn All Enslisb
'w' z
brl:
r
Kauun (1910)
NlpComi (1911)
Vmnh944l
All
.......
All
~
......
"""
c..~
...
,..u
,..u
Limited o
z
~
ti~
rll!
i ~a.
PVP ..... PV1.1-6
PVT .... rvr .... f'V'.I-6
Van 8ijlcn
N_
T_
M~ Enslish ,..u
•o
~
o
~
AlI J."...,. ~
JY!
~~,
.. E- ..
PV,.I- IO
SiT
v.,., c.s. M~
M~
English ... """
Umiatd
,HI
Iw.u h99l ) Mm, c.~" Limited •z
• 5 • pvp uPV,.I'4- 114 IW1II31199I) ....... ~ ~ Umilcd
<
~~l
I &: i.,
PVf u'PVP .... PV,.•".- ll.4 l'lQta 11991) ....... ~ ~ Limiled
;"\Vell, then, if observation and nonobservation are not the basis for know-
Ing the positive and negative concomitance, then how does one know that
smoke is invariable as an indicator of fire?"
One knows it since
1 Unless otherwise noted, all translations from PVSV are based upon the interpretation of
;Siikyabuddhi.
32 Siikyabuddhi (PVT:50b) supplies:
One might wonder, "One has seen that one instance of smoke came from a particu-
lar causal complex of fire and so on. But if that which one sees later is something dif-
ferent, how can one establish that it is the effect of fire by observing that smoke came
from fire just once?"
335
J)6 FOUN DATIO NS O F D HA I.MAKIRTI 'S PH i l OSO PHY
If smoke were causdess, then, since it would not depend on anything for
its existence, it would never be nonexistent because there would be no
incompleteness in the conditiON for its existence, JUSt as there is no incom-
pleteness in the c;onditions (or its aiuence at the time that it is accepted to
have been produced. O r dse it would not even occur at the time when one
accepts that it has been produced because that time would not be distinct
from the time when it is not aistmt. For it is by virtUe of their dependence
Ion a completc causal complex at a specific time and placeJ that thingt are
intermittcnt (ItllJiritlu.).
3 I.. ock. _..is. Oft< could ..... ..abIiah d.a. " ..... Y "'" d'n:..-. by ....,....t,,,& .0 u.. r..c.
w. x and , haft dilfncclI a .......
APPENDIX OF TRANSLATIONS 337
Entities must be dependent in this fashion because the time [and place]
en and where an entity does occur have the capacity required for that
".;. tity to occur, but the time [and place] when and where an entity does not
!~cur do not have the capacity required for that entity to occur. This must
\: the case because otherwise the time [and place] that are a locus of that
~1tity's occurrence and the time [and place] that are not such a loc~ would
I
\t~ equal in terms of whether or not they had that capacity. As such, the fact
being ~ locus for the occurrence of that e~tity would not be restricte~ [to
:.;.if.·
...
"!pst the time and place that had that capacIty] and the fact of not be10g a
OC~cus for the occurrence of that entity would not be restricted [to just the
Itme and place that do not have that capacity]. And what else but the pres-
i.!1ce of the cause could constitute that capacity?
Therefore, a thing that is occurring in one time and place in distinction
Ip another time and place where it is not occurring is dependent upon the
~bnditions in the aforementioned time and place. In other words, its occur-
'ence in that fashion [i ..e:, in .a speci~c time and place] consti~utes its
Jependence on the conditions 10 that tIme and place because a th10g that
~d not depend on the causal support provided by those conditions in that
Ime and place could not be restricted in its occurrence to that time and
Ilace. Hence, since smok~ is restricted to a s~ecific tim~ and place, ~e
f,),ature-svabhiiva of smoke IS the product of that 10 whose tIme and location
Loke is observed [at least] once and in whose absence it is then not
1&1;,:
~bserved. This must be the case because otherwise, smoke would not occur
J.en once. And being restricted to that time and place where its cause is
~resent, how could smoke occur elsewhere [i.e., where its cause is absent]?
mf it were to occur elsewhere, it would not be smoke because a particular
~\,,!
" A urmhc: wwu (Ukr...,;nJM..j .....y ..mctima em;' omoIu: due ,u ,he ........yinS nu.Lle'
wed 1<'1 wum !he "1!1 du.mbas.
2. PVSV,d PVl.6IJ-75
JJ9
HO FOUNDATIONS OF O HAIMAKIRTI 'S PH ILOS O PHY
the difference among those: things; even though they arc themselves differ-
em , they appc=ar nondiffemlt in that they appear with some single form, In
conformi ty with those things. appeanncc. we c:a.IJ Idle mentally con.structed
image) the ~ univas:al~ of those mentally occurrent real things, which appear
[Q be atcmal' due [Q being apprehended in ,he form of a specific kind of
.. .
cogn U iVC 1m• •
- But how is it that a universal is an other-uclwio n?~
That lconcqxually oonsuuCled i~1 is itsdf an other..er;dwion.1 Con-
crptuaI awareness apprehends that. and since conttpu are by nature erro-
neow, the aw.neness appart as ifi, ~re apprehending a raJ thing. That
conceptual cognition occun only with regard [Q things that are excluded
from othen Imat do not have the same ClUSC5 or effects) ; as such, that
awareness is undernood to have an adusion as iu obiea.'
"But the distinct things in quc:stion are a ternal. and concepu do not
7 The objeaor'. paim ~ is !hal. on Dtwnukin i'IIh«wr. (OI>Olptl dilUdy rmr only 10
• c:onaptuaIIy CUI'UtJUCIcd ~ Hcnu. bowc:anOM aaouI\l eo. rdltl'Ula 10Klual. atft'-
naI thinp!
8 A«.ording 10 ~ (PVT:aIl1 - K:171.19ff). · and to on" indicaus lWeI oeM rn-
IxaIU( _pea do 1\0( _10 aiIl.mm WKm: noobto..bk [pm'iaIbn)
JOnI: •• ••
and bea..... lpanic:ulanl 11m: ,riM;: funaion. but loou"pul do noc:
9 ~hi (PVT:I~ • 1<.:1]).1) &Ioua "individual." r..,u,.,.;) ... "putinllan·
}41 FOUNDATIONS OF DHARMAK IRTI'S PH ILOSOP HY
Individuals arc not distributed over each other. They annot be distrio.
utcd because if they ~ there would be no distina individuals; henc:e, o~
would be forced to condude th:1l there would aho be no universal." More-
over, it is not the caK that something that is other- i.e., dininct from those
individuals-----appears that way in awar~ness. And how c:an that which has
no cognitive ap~ct: induce one to apprehend or designate somtthing
dse as itsdP." Nor is it the caK that something is universal just because it
is a ungle thing whim is connected to those [pmkulanl because one would
incur an overextension. I have already explained this.It
"A single univrnal is the wan-am for expressions and cognitions that con-
strue many individuals as nondiffacnt. It is not the caK that aU [entitW::s
that inhere in things, such as numbus, SJm'~ as wamnrs in that fashion! .-
But how c:an one have: a cognition of one thing from anoth~r thing~
"Because the on~ thing has a rdation to the other.~
Then one would be forced to conclude that number. effect-substance.
and so on arc also univcn:al5.u
"They arc not universals beause they do not have the nature of uni-
versah."
Well, we should investigare what ,he nature of a universal might be. In
this regard. SOnlC philosophen say that when a a:ruin kind of singular. real
thing has a rdation fO multiple individuals, one has a cognition of those
individuals as the same; thus, they c:a11 that rca.I thing a univrnal. But here,
we say the: following. One could have a cognition of some individuals u the:
same from effect-substance and so on. which arc also conncctcd to many
II '" Sakyabuddhi mUa deu (pVf:19b4If- 1(;17).ull). the noOon Iwrc;. dw in I «In •
.,...,.ual ... li......iuic cop;lion. indmduaklNI pnMrM'" il'UWltial~ a ccruin uniwrul an:
contmICd in ro:rmf 01 mal uniwnal For ~, aD indiYiduaiJ dw in.tanlia~ "wall:r-jUl-
_" an:appt~!daI:u · wam.jup" by~irt\IC of~ POQ(DCIC of"mal UlIi¥a$l.l in cad! indi-
.......w. 1M upunml ~ ;. lhal 10 do 10, W univuul mUlt aaually Ippear in _·1
....'-
12 Stt PVSV tlliPVI--40'-41 (G:l1..ffl.
I) An ·drax..JUbKana:" (i~",c) .11IIb.cana Wtl'Clll!tI from me «Injunction of
om.c. ...' - - .. ~1'01 )db IM",.-CO (,0., - 1<;'1+'1) oM Ietm often -..do ............nok.
(-:1"ltitt). Sft, for aunpk, iu IWIE II PDS (I l l).
APPENDIX OF TRANSLATIONS 343
And how can that which is not different from awareness occur
in another object? [pVI.7Icd]
Objects are not the same in terms of any essence present in all of them,
whether it be distinct or not distinct from each thing that instantiates it.
Therefore the apprehension of those things in that way is just a false con-
ceptual cognition.
from (his and that oth~r object; one ~ngagcs in the formatwn of
linguisric conventions (SIIf!tjfil1)" for th~ purpoK of knowing that
difference, (PVl.]lCC1J
For cxampk, the sensory faculty. the ob;cct. ligl'\l, and mentation (I'M"-
llIilc4ra}-or [according to orner philosophiesl. the self, the mulry. mind,
the object, and contact-produce a lingle effect. namdy, an awareness of
a visible form . Thq produce this one, same effect even though thq do not
instantiate a univerWthal is restricted to ha:ving Ih~ nature: of producing
(hat effect. Likewise. distinct instances of tKCS such as iif!I/Rp4s and so on
by their very naNre produce the same cffi:ct. namdy, a rccognitional aware-
ness that has an image of each instance as the same aJ the others, i.e., as a
~ Utt." All tree-instances have thosc same effects even though they are not
distributed the one over the other. Or thq accomplish some other [die
function Ihat is [0 be done by wood, such as combustion. housing, and so
I" ~lIr.r"'..Idh;
(PVT:9.b. _ 1(,.",_...,) p.._ ~,.. "fon.",.;",., ,,(1;"8";..... """'_
lion" (aw ~ ,. ~ ,.".m.~.
APPEN D IX O F TRANSLATION S
'"
on. in accord wiili me oonditions. But even iliouth Wo:uer and such au abo
not distinct from tl't'CS in that any entity is difftrent from all omen. water
and so on nevcnhdCSli do not perfonn the afo~mencioned tdic funaions,
JUSt as the ear and so on cannot produtt an awarenCSli of visible form.
Also, the universal is not what pafOmu functions i1«lIug, since the uni-
versal iJ trlNtAnJ, it (linn!), (1l1lSll/Jy Jupport IllfJthilfl. (PYI.7Sdl
[Sioclt it is unmanging,] a unMrsal a nno t bit augmltntw, so if it .....-c~
to provide causa! suppon , it would produtt all Onts dr«u at once. Oth-
erwise, it would no t have the n~ lUre-llIIlbh4l11l of producing that effect.
But individuals occur in diJ[incrivc waY' due to the time, place. and mod-
ifying conditions of their occurrentt. The~ is no contradiction if one
says that those individuals produce distinctive effcas. likewise, even
though they are by natu~ different, ttrtain things accomplish the g me
telic function. sum as producing the same recognirional awareness
(prillJllbhijfi4lfll). As such, they arc diltt~nt from other things that do
not IX'rform that function. and they are therefore said to be nondifferenl.
O r [hose variow things that arc produced by a single cause arc said to be
nondiffe~nt beca~ they are different flom [hose Ihings that arc nOI
produced by that cause.
"Well, through the universal---which you have dcfinw as diffen:nce
(from that which does not have the intended. telic function}-does one
cognize the particular as the same as other particulars, or does one cognize
something other than a paniculu in that falhion ? What problem comes
from th is~ Well, if the particular is what one cognizes as the same, then
how can it be an object of co ncq)(uality~ And if one cognizes something
cIsc. then how could there be cdic fun ction fro m something other than a
particu1ar~ And sino: one would not cognize impermanence and such in a
particular, the particular would not have the nature of being impermanent
and so on. Funhermo~, sintt they arc not cognized in paItKulan, those
univenab such as impermanence would nOl be qualitic:s of real thinp;."
This Dult does nOt apply lxau$C it is in reLuion to the :lppear.mcc in
conceptual cognition that wt form conventions for univerWs, oo-referm-
ti;&iity, and the subject-predicate mation. This type of aw:lf'enc:u arises in
dependence on imprinu tru.t have been left by pclCCprual experience, which
apprehends the natufCS of real things. The awareneu thar arisc:s in this fash-
ion is conceprual; as sum. even though it docs not have those real, extr:a-
menw particulU$ U its objca, ronttptual cognition s«ms to have them
u iu object. In other words, being conceprual, that cognition bas a nature
such that iu obj«t is imagin«i (1lIiJryilINllil4) 10 have tru.t natun: [i.e., the
nature ofbcing an extra-menw puticularl. Conoeprual cognition operateS
in that fuhion bccau.sc it is by nature produced by imprints thar have been
placed in the mind by e:lpcriences of those particulars [i.e.. the onc:s that
prompt the concept in qucstiOrll. And sintt conceptual cognition is lindi-
APPEN DIX OF TRAN SLAT IONS
tewcl when ont" s«s the glimmer of the jewel. Other cognitioN are not
trwrwonhy beca~, ~ though thq also arise from a disrinct;on of the
real thing. these other cognitions Fail to determine" me distinctive qualities
of me thing in accord with the way in which it waJ aperienccd through me
serues; having Faikd to nuke that d«uminat;on. they impute some other
distinaion onto the thing by apprehmding some slight (Iti~f) similarity,-
An example is the cognition of a jcwcl when one S('CS lamplight.
The above response 10 me objectioN raised above shows that Ime capK·
it)' forI performing (die functions docs nOI apply to things that are con·
ttptual objKU. II also shows that it is nollhe ca.sc Ihat impermanence and
so on do nOI pcnain 10 particulars. This is 50 because there is no imper·
manCntt othtt than me' Ruauating lhing (II"J'U ~1t/4J INIIhl1UIf,) irself. Thai
is, one apprehends a minI which is of mat kind Ii.e., impermanent] 10 ~
something that perdurcs for only an wan .. henet'. one has cognitions such
as: ..,-nis is impcrmanem" or - Lmpttmanentt pertains 10 this," Focusing on
the nalure (tihllmwt4) of the particubn in quotion, such conttpu prcscm
various qualitio [conceptualized as aruibules of a single ming], single qual-
ities iconcq)(ualiud as common to a group of thingsl . and qualities which
are eonceptualiud as distinct [ho m what they qualify). ThOS( conceptual
cognitions art nOt ground.l~ beca~ they are based on the experience of
an asptcr (bhd4) of the rca.! thing.lI Nor are conttptualiud qualities such
as impetmane'ntt nOI qualities of real things bttause: that which has thaI
nature (uUVllbhilltl} appan in that way.
However, it is still errontous to apprehend a real thing as having various
distinct anribules, or as the same as Olhn' reallhings, or as distinct from me
qualities abstracted from iL It is erroneous because: an indivKiual that per-
formS:l function that is the $:ImC U lhe function perfonned by omer indi-
viduals or th;u pC'rforms many functions is defined in that way jwt for me
purpose of making thOS( who wish 10 know aboUI such things aware that
it has such a nature. One does nOI define it in that way bttause of me' real
nce of distinct aspects of real things because: [I] a single thing cannot
anifold; [2] many things cannot be singular; and [3] we have already
fed the distinction of an abstracted predicate from the subject from
iJch it is abstracted. Since property-svabhavas are conceptualized as being
~}fpature di~erent from each other and from t~e subject of which the~ are
p'i~icated, If one wer~ to acc:pt tha: t~e objects (art~as) of expreSSIOns
'~1"esponded to real thmgs, co-mstantlatIOn (samanyadhtkara!lya) would be
~possible.
"'''~omeone objects, "Expressions for both quality-universals and dass-uni-
~~fsals denote the one substance that possesses those qualities and univer-
~~sas its delimiting qualities. Hence, there is no such problem."
i""%~If the substance does not causally support a universal or any other alleged
aelimiting quality, then that universal or other such entity is not that sub-
~;i;ihce's delimiting quality because that universal or other entity cannot
~i~end on that substance. If that universal or other such entity were
~4:jj>endent on that substance, then substance and universal, etc. would
~d in the relation of producer and produced. Hence, those two could not
,l!Jt'cur together [i.e., simultaneously], so one could not express both of them
",¥<~
possesses it are the causally suppon C'd and the causal suppo" ; :IS such. they
occur together. TherefOre". me fiult you have raised does not hold true."
No. mis response is not correct. Something which is already created is
not de~ndc: nt on anything, so it could not be a delimiting quali[),. And
something mat is not yet created cannot be a delimiting qualiry because its
narure (IVIIriipa) is not esf1lblished (such that it could, by that nature", be a
delimiting quali[),J. Thus. whC't her the delimiting quali[), be already aeated
or not yet created, mere" is no d~ndena: whatsOC'Ver. H en ~, the conven-
tion fo r me qualiry/qualified m ation must be Qablished after one has con-
ceptually imputed the existcna: of these emities. If (hat is me ClSC, why
should conceptual and linguistic cognitions not occur in conformi[), to
conventions mat have been consrructcd through conceptual imputn ion in
all cases~)oI
It has been prC'V iousiy stated that if a single expression or instrumental
cognition rakes some quali[), of a thing as an object by the force of rnI
things memsC'l vcs. men other expressions or instrumental cognitions would
be useless because all omer qualities of the object would already be implied.
But if such cognit)ons confo rm to convt:ntions in the aforementioned man-
ner, this would not be the ClSC. Thai is, since the appearance in cognition
is unreal. the problems that arisC' due 10 {the claim that me objects of such
cognitions] are real do nOI follow.
Therefore (uuJ),n a universal is not contradictory in accord with the way
it is cogniud because me cognitive image appears in cognition to be non-
different from some orner images. The relationship of qualifier and quali-
fied (as in the expression -blue lotus"] is also not contradictory in Ihe way
that it is cogniud for the following reason: even mough one aspect (1Iuir.)
{such as ~ bl ue"J has b«n taken as an object, the other (such as ~ Iolus"l has
not been determined; th:1f other aspect is what is apprehended by a cogni-
tion wh ich involveli lOme doubt or apcctation about mat Other aspect. [as
~" ..
~~(; Glossed by Sakyabuddhi (PVT:99a = K:r87) as pratipiidana.
~7 Sakyabuddhi (PVT:99b = K:r88) suggests: tad eva buddhipratibhiisabhiita1J! vastu.
~I]akagomin offers a second interpretation in which the vastu is understood to be biihya.
,-hat is, instead of saying "The distinction of subject and predicate pertaining to a cognitive
~ppearance is also not contradictory in the way that it is cognized," he says "The distinction
~~ubject and predicate pertaining to a real thing is also not contradictory in the way that it
]i.cognized." He then makes the apptopriate changes in his commentary:
yad vii dharmadharmibhedo 'py asya vastuno na virudhyata iti sambandhab / kuta ityiiha
I anekasmiid arthiid biihyasya bhedasambhave sati tasyaikasmiid yo bhedas tasya vidhi-
prati!edhajijfiiisiiyii1J! kim anityab fabdo bhavati ciik!u!o na bhavatiti tad eva biihyam
vastu pradarfyata iti sambandhab.
~'rhe complicated statement from dharmadharmibhedo to pradarfyate is difficult to render
iD~intelligible English, and I have therefore chosen to break the statement into more man-
[~~ble sentences.
~Sakyabuddhi (PVT:99b6ff = K:r89.II) claims that this complex passage treats two sepa-
~e cases of the subject/predicate relation: a predicate-expression (dharmafabda), such as
"'IJp.bdasyiinityatvam, "and a subject-expression, such as "anityab fabdab." This interpretation
35 2 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
is weakened by the absence of a conjunction, and I have thus not followed it. One might sug-
gest, however, that the current translation problematically suggests that a single statement can
be both predicate- and subject-expressive. Since this objection is plausible, I will note here a
translation based upon Sakyabuddhi's interpretation:
The distinction of subject and predicate pertaining to a cognitive appearance is also not
contradictory in the way that it is cognized. A cognitive appearance can be distinct
from various objects (arthtts). That being the case, when someone wishes to know
whether that cognitive appearance is established or rejected as distinct from one of
those objects (arthtts), the respondent presents that mentally occurrent real thing hav-
ing expressed it by means of a predicate-expression that precludes other distinctions
that could be predicates. Thus expressed, it seems as if a separate predicate is expressed
because it appears that way ro cognition. But he may also present it having established
another properry-svabhava of that mental entiry as the subject without the distinction
of precluding other predicates. To this extent, subject and predicate are slightly dif-
ferent; hence, a conceptual cognition of this rype appears in such away that it seems
to be differentiated. Such a cognition is not, however, differentiated due to some dif-
ferentiation in the real thing because this would entail the above discussed problems.
3. PVSV,d PV1.137-142
Ixcause: III rhey would have no similariry; 11J since they have no similar-
iry, thar expression would refer to only one individual; therefore, that
expression would not produce the cognition of anomer individual of that
type: IjJ if:u1 awattness of one thing as the same: as another were to ariK
without any semantic proximiry (praty4JAlti), then an overextension would
occur; 14J the application of a single expression to them would be useless
because one would not be applying me expression to a singlt thing (artha)
(since they do not share a universal]: :and 1,1 if one were ro apply one apro--
sion separately [0 each individual for the sake of cognizing' those individ·
uah whose essential natures are diffirem from each other, one's listener
would not know that those distinct things that have been expressed in that
fashion (i.e., as distinct instances of the samt type of emiry1 are diff'trent: J
(In this regard, there are two possibilitiC'Sj . Thinlcing, "that individual
also has that single universal," peoplt apply an apression to it. Or a single
univusal manifesu the expression just by iu capacity as a real thing (i.e.,
without any human agency] . But ntither of th~ possibilities is the case.
Instead, a pt:t50n applies expressioN to something with some pu~ in
mind. That is, if differenr things are usdUI for one telic function, persons
concerned with that funetion ddinitdy (llwUyam) should express that effi-
cacy of those: things with regard to that function. If one were to express that
efficacy individually (i .e., with an expression for ach panicuJar] , it would
be extremely difficult to communicatt. And in :u1y case, it is not possible
[0 express the unique essence of a panicular; mt antmpt TO do so would also
.1 Utc,aUy, ".,..., would ...... KCOV'W: ,he di.un'ty (~) of ,he .h in", ...... had h-ecn
a pre.ccl .. NCb. "
APPE N D IX O F TRAN SLAT ION S lSI
tion docs fix meaning. then who could pre·...ent this a p~ion from refer-
ring to many things?
-But there wo uld be no purpose to wing aprmio ns lif there is nothing
to which they rc2lly rckrl; therefore. one would nO[ usc expressions.-
I have al ready Aid th:u the purpose: is to know from one exprmion that
diuinct pan Ktilars arc different from that wh ich docs not fu lfill the
[expccrcdJ purpose. It is not the c:a.se d\a t one applics expressions because
a p ro~ny-SW'bh4U4 of the things in qucstion is the A rne. And I have
already asUd. '1ne propeny-W4bhi_ of rc2l things :ue SituatM rupee-
tivdy in their rupttt i~ real things, and as such. they are not disn ibutcd
~r each other; hentt. how could an expression for those distinct things
have as iu sc:mamic ClU$C (" imift4) a single propeny-svabhdv.r occurring in
all those th i n gs ~·· But evrn though things arr different. it is nO[ contr:adic-
tory' for them to have an aelusion from thar which docs nO[ fulfill the
[expected) purpose. Hence. let US consider this no ndiR'erencc of the things
(,m h.t) to be the cause for the nondiR'Cf'COCe of the cxprcssions applied to
them. Therefore. in saying, "'These arc whar fulfill that purpose. - those
things arc Stated to be distinct from others that do nO[ ful611 thOlt purpose:.
RU! rlv.: F....::r o ( tuvi ng 1M! t:ffin i~ nor U)m O!: qll!lliry !hu ;$ n rh e r rh:an rhO!:
p:lnicul:ars them$d v~ : otherwise. they would nor be different fro m the
other things which do not have that dfect.'
For aamplt. [accordi ng to Buddhists} tht qt, color, light and menta-
tion all ha~ [undc:r cma.in conditionsl tht wnc tffm: ocular conscious-
nw. (According to others] the ~lf. tht f.aculty, tht objter and rhtir
conjunaion have that ont df«t. WhatleVu theory ont proposc:s, it il pos-
sibtt 10 o:prc:u their sameness (s4~) in ttmU of having that dfca. as
when someone asks., "From what docs coruciousnw of visible color arise"
Thil ~ing the case.' in order to facilitale practical aaion (lIJilvttM,..uf./M-
~'finIMm). somtont applies a con~ncion-esttbli.shing (SiI~;) statement:
"1l1e causes a~ such-and-such and $O-and-so." One uses these exprcssions
such that the listener somehow knows all of the causes of ocular cognition
and $0 on u once. HenCC', the~ il 00 essetlU that is distributtd over all of
them. Rather, their nondiff'trenCC' iJ this diff'tmlce--namdy. that since
they accomplish that leiOS (IInIM), m06e
things are differenl from [hose
that do nOI accomplish that telos.
To be specific, they, being all of such-and-such a kind. are expressed by
expressions lhat inmate :I amin conglomu.uion (IIlmw), continuum
(wnrlina), or State (1lIlIISthij . Those particulan thai whttl conglome:rated
This inmprmllion ma.krs lTIOft JnIit; dw! GnoIi's radin&. lOr illmtll 00r WI Dtw-
maklni is rdan"110 1M iuua rWed II FYI:l', "ThII M is raisi"l It- iuua is Iimhu Np-
portal t". hiJdccWon IOdiKwa 1M pcoblun of dw: pm..: rd.llion, kw Ii...". !.alto. lOr W.
is dw: AIm iswt- tha ht- rakes up II FYI:60. To Rippon IlKh an inmprewm. the aWII
SansIuU. in bodI K and Cnoli can be limply amr:nded by dw: addirion of In
Oftbosn.phlc dmce W1 is ohm didcd in INIllllO'ipa, The m.anwaipi ofK. ro.- campk.
In-W-
is pcppcn:d wid! nUmmKII JUCh disions. 'The amended Sansluil would mIlS read ... ,..-r
qbI ~~ u~ / .~"~",._ . ~.
Unfonunarcly. ~:obddhl (PVT:ot:c») "'- ..... pt""'ick .... y additional darlncooion,
a1thour;h he aL.o oIJm the IlfualislxtOfJ n:adins ~ u..Jb r,.;.." P- m.
J. J.
liM,.i 1't1i~ r ) wKhoullM ....,._ Reader'I of dw: SdMsr ~ have 1M addi-
tiofqJ and quile .. ~ p.obIem of :0. lup: portion of 1= WI it _I of onkr. 1"hiJ irnsu-
lui.,. ;. ' - d ... lD:o. buh ill tnnWtion.l"or ....tom eN _ "..." order ia Itst(ItN. Iht u:o.noIao:ion
I"oIIows rucdy W Satukrii of K. Hcnoe, il tmnI w. in the K¥nII cmturia ber.tt .. lhe
u:o.nWOon of riM: leO and !hot Cll'Yin& of dw: ScIe-df;e ....""..,"10011 ~ TdJewa .robe iNd·
from
tlwtIM,.
.!iob. -II'-.t.r,.,.,,.
¥maid)' copi.d I folio QUI of order. The order is. follow.:
afu:r
--',..u.. ...... Rae! 10 16li1l. F/Om [6li1l mer ·M..,..,..,• .,. ...
tJ,. ... -tkip forwIrdlO llilb. and tqin at. -M ~
J. '¥
,. _ ... "Wp b.dt.o .6,bl and tq;.n at. "ritr"'u-.,.." Rad 10 16,b,. From 16.b. mer
oft tu.- ....P:ip 10 16uII and tq;.n at.
0111of order,
0......,,..., J.." From thia poinl eN fCXI is 110Ionca
OM"'"
,."." -
""'..,..., J.n.,.
.
...
In ocher wonh.lM scaion of tal from [6tbl UN tIM,. -',."up to [6li1,'M
J. '¥,..- ... ·needs 10 be iMmN I' 16Gb1 after "phtm "" 1M Ut
perform a sing.l~ cA«t ha~ no dinincrion from och other with regard to
that cA«t_Th~reforc. it would be poindCS5 to express any such distinction.
For this reason, in order (0 mcr (niJDjtmlf) to all of them at once. peoplc
apply on~ expression to them. such as ·water-jug.·
ThOK (i.e., th~ particulars that form a wat~r-j ugl arc all equally diffcr·
Cnt from their respective homologues and hererologues. but since they con-
tribute to th~ aa:ompl ishm~nt of that single purpoK !such as containing
wat~rl. they a r~ distinguished from othm that do not do 50. H~nce, they
arc cognized as non-distina due to that nondiffetence.
fin regard to th~ point just raised, thc ClIpression · the color and SO on of
a wat~r-jug· mcam that th~ color and such that arc the propt:rty-Iu.rbh.i~...
of th~ water-jug arc capablc of an effect such as a specific:" way of contain-
ing watcr.Y Through thc exprcuion · color and so on- is indicated (,m'
II Sikpbuddhi (PVT:l60b6-7" 1<:171 .•1) P<-=I specific (.u.,.) 11 · tha, which caftOOl be
~ by non-w.un. jutp· rg,."u •.." t44yrw'!'J. II thUJ i!ppGtI '0 be I.IKd in
dili CORca. II • synonym for rU]cu.
!I ~i (PVT' I6cb7-.611l " 1<:17 •.1,) rctnarlu.:
Thr::Ibovf: mans dw fOIlowi"" 'Tbcaprmicnl "cokN ' and 10 on IndUtt lhc ..ndn-
fUndi,. dw color and 10 on an: nondiA'ea-., in dw they ate Cl!p'hk of ..... ~
dfm lhal ill '0 be aoc:om~ bycolor. But tMwonI"W::II..-iu!· GJriUICS!hac 1M
coIor-pmida: in quarion an: specifiai (~by their dill'a-ma f'roI!Irhat color and
10 on dw an: P'OfKny• .-M.I- of. doch; lbot watn-ius', color and 10 on an: Ifl'«'
i6ed in duol r..hion ~.....,. an: a.pahk d a IpOCiroc ~ (~). H rnu.
dw: ltatnnmt"the walcr-iu!', c.oIor and 10 on' pl'oduca thrOUJh the: optnrion of
..... twO words an lwatmeM w1lh. COIIIil;"" i..... of ... ni ......... ("'-'1f)'I(J (i.e .. dUll
of albl and I specif7inI diltinaion (1Ii/q4) (iA•• that of c:ontrib.tlins 10 ~ dfttu
6UCh u hoIdin,..atftl. 11x16oce. I smirm rmmaDal rdarion (",M.h;) in whid:J
dw: two manbcn an: diKinct ill appIiaIl0 aptta dw !d,tion of • Wli~ [i.e., color
and...dl in ~ I and. sp«ik (oub-'t' "'"" i.e.. rhc: mIoc and RICh of I wuCl'-j"ll
(tJtM~.
tiJJJN,) their naNn: that is wdl known (prasiJJh4) to be the means for pro-
ducing a gwm effect (uimi"J"kiry4 ,tThcy an: then specified (viIq!4) by
the specification (vilqa) (i.e., "water.jug"j that indicatcs that they also have
ttnain specific dfcct.s (vi1;"!Jlki1Jll) [such as conr.a.i.ning water]. Being so
specified. they an: called such [i.e., "the color and such of a water-jug"j. But
~r than dUll color:and such, there docs nOI o:ist here any subs-WItt thai
has charactcrUlics in tht manner dC5Cri~ by those o:prcssions. It docs
nOI exist b«ausc one docs not pcrctive Ihal kind of substance."
The expression "warer.jug" is used. in the singular 10 indicate that those
paniculars [i_c.• me many pan iculars that compose it] have mc causal
potcntial to together produa- the wne cff«t [cvcn though then: is no such
real causal potential aClually iosrantiated in all those particularsj." Or dsc,
the singular is dependent on linguistic convention (Ill",,",,,).
For irut:lOCC, ccnain kinds of causes and cffccts (i.e. , those that form a
continuum] either produtt some single effttr or art produced by a single
cawe. In order to cognize those causes and dfcct.s all at once, one expl'CS5CS
them with an expression such as "rice" for which the linguistic convention
has been formed . This should be explained as in the above expl:an.ation of
o.prcssionl fOr conglomeratCi such as watcr.jugs. lJkewise, those particu.
lars which arc dkctivc for somc single function cithcr separ.ucly or in com-
bination an: expressed by expressions mat indicate a p;tnicular state; thcy
arc exprascd in that way so that one might make thcm known aJi at once.
Enmplcs ofsuch exprt:s.sioru art ·visible" or "obmuctivc." The paniculars
in question can be expressed in mat way due fO me QIDCflCS$ of their dif·
fercnce (bhdMimi"J") from other particulaT$ that an: not occurring in
that kind of sau'.
In some cases, where some particulars have the same effect, in order to
indicate that they have that effect, they are expressed by expressions such
as "water-jug" through their difference from what is other than them; they
.are so expressed provided that a suitable convention has been formed
(krtasamaya). In the same way, in terms of having the same cause,one can
express what is non-singular with a single expression in order to facilitate
practical action. Examples are "Hereford," "Jersey cow," or "a sound arises
immediately after effort," or "sound is causally produced." One can also
express multiple things with a single expression as a negation of their abil-
ity to have the effect in question. Examples include: "sound is not visually
perceptible"; or "impermanent"; or "essenceless {aniitma)." One can also
use expressions in this fashion as negations of the notion that certain things
have the cause in question. Instances include the expressions "unowned"
(asviimikajI3 and "empty." One may also state other ways in which expres-
sions are formed in accord with the theory presented above.
In the case of expressions such as "empty" and "impermanent," expressions
perform their semantic function (vyapadefa) by [first] inducing in cognition
an image that is intended in accord with the interlocutor's concepts and then
excluding that image. Expressions work this way because all the objects (artha)
gf expressions have a distinctive aspect that is projected by cognitive intent. 14
Poor thinkers'15 bombasts which raise problems such as the assertion that
there is no [real] contrary (pratipakfa) for expressions such as "essenceless"
should be ignored.
I 5« PS 1.p.
2 Sikpbuddhi (PVf:1,fw" 1(;,,0) ~II ilII ob.i«for'$ pG'irion;
'Iran apt , ' m is noc inununmw for kllo.Ie" of .,..nm1
tal thmp. Wn wN.!
about [dw IUlemctll mMk by Diplp; namdy. lhad i1w IUlanalU of. cmIibk
~ an: dw buU for • ordl-lOrmed in(crmtt: bcaUK me,. an: pually UW(WOr-
tby' I""~'!"~ ........~ 1 ThaI ...No:h ill a ,,1<:mCft1 <Jtf. (mi.
ibIe ~ ia inmnn..-niblr. an n2II'Ipk iI dot lUI.tmftV 'AlI .......nn...-:I min&<
aft mommwy.· l1!m: aft abo ltaretnmll oi. cMibk penon COl ...... ..;'" a:.~mdy
rmMMC objent; th.crd"oR. they aft aho inc:ontrovutibk. In dUi; IIWIncr. Iuoowkdsc
dcriwoed from 1M .c:aimony oi. CftdibIc ~ ill • form oi inm.tntt beaoUK JUCh
MI'~U aft ~u, ~. ThU$, .i.d.ry. Oi~ Wd th:u I~
dnnoed &om1~ ii, (1Orm of] inkmsa coaamin& atcm.1l ob;tc....-
To sate th:al mil conrndicu wha. hal ~ IIXI:pIcd {by DhamWdni. lbe ~
_,~..,." I -Btu ..... r.. tIJiI k ~-e-
Noec th:u the ""nose: ."......c• ..,... iI /'101 .0 M ClONtNC'd with the aantpk ,.tIM
_ ...• IN. radon II the HI.tmm. oi.be,.,., in mil inlttmu.
3 Sikyabuddhi (PVT:~1h5 • K:J,o) offm the &in oiDhannaldn.i·II~
In othn words. the Ad.ya [Diglipl did /'101 Q'f duo. ~ &om Kripnm: iI an
inlU ....... by daimi. duo. it iI ~y (oe Indy] (UhiUj inRl\lftWfti:al. R>.d.cr. i. iI
instlumerual with reprd 10 the ""'y in which I ptrIOIIlhouId pt'ClCX'ed.
.. for G(lof) .~,. fad ..,,,.,. wim ~i (rVT:~)a) . whidl K'CIN . he much
pt'c~,.bk readinl-
,6.
}61 FOUN DAT IOt<S O F DHAII.M AKllI.T r S PHILOSOPII Y
Not being contradicted by pcrctption means rhu Ihose things Ihat arc:
considered to be perceptible according to me UCl(isc acrual.Iy are percepti-
ble:. Exampln include: the: color blue:, plc:asure and displeasure. the: recog-
nition of a characte:ristic, mental St1ltCS such :l5 desire, and awareneu. Also,
things fhlu are nOl considered 10 hc perttptible should not be perceptible':.
Exarnplell include: th~ claim that on~ can ~rceivt: plc:asure: and so on as a
' Sikyabuddhi (PVT:14)a) IIiI1)'1 thai ",.».mwu chat lht IU1I~mmu "poinl OWl" (-1M
~- ~.rir fL U ) that IOPic.. KOOI) oIrm no r;ka
6 ~i (PVT:14)b . K:",) pnMdcI tIw po.- tOr 1M IWU leNS oi...-m..
APPENDIX OF TRAN SLATI O NS j.j
oonjunClion ofform panicks such as sound:' and the nQlion Ih.:u ruhsr:mce.
morion, univers:W, CC)nmcl, and SO on ate real, perc;:qxible things.- Likev.rise,
tho.sc objectS mat au oonsidered to ~ infer.lble wilhout relying o n scrip-
ture mo uld be IO-lhe Nobles' Four Trudu are an example. And IhO$(:
things that are considered no t to ~ inferable mu.$( not be---aamples
include the self, This is :also the case for inftrentt based on scripture.' For
elCI.mple. hllving ~ccqnM mn rht: a"",nr;lIl nnll.t: (rii/"l) of nt:g:u;~ ml!'n _
tal states such as desire is dlihamut, and having accepted that the origin of
tho.sc negative memal H2les is tUihArmA [in the sense: of .h!'l'ma], the sug-
gestion th:at o ne perform :ablutions and fire sacrifitt in o rder [Q d imin:a!e
IltiJu,I7ftIl is not sound advitt. The scripture's purity [i.e., in lack of con-
tradictions] in regard to all those objecu which can be determined in the
lIbov.. mllnnt:r conu;rurl!':l its lruH'W'Ortltinl!':ll.
.,.._.
7 ~i (PVI':144h . K:)9}) I>O'ICI du.llw: d\C'01)'aiticQ.cd httt i, from IMSirpkhya
8 ~uddhi (rvr:~ _ K:J,,) " :"IaIL. ,I"" , ,I..: .icw 1~j«'W h.;,C u .h., uf'M
-Vailqilw and.o on,"
9 ~i (PVT:l-fsal:
0"4' migtu invaripte . Kripru..., d"" to a conlradiction between earlia and laler r--
Ages conumi", tho: rwo kind. of obj«a mal U( pu"" (in th.t they '""' 1>0'1 comn·
dieted by ~n:qxual aWU'UlClii or in~!ICU !wed 0)(1 raI thinp] and QUl!'md,.
rrmoIeobjecu. In Uw cut.;~ MsH.,. tni".." ';"""'."1>0'1 contradict dw
KripNrC" _upplkd by conU:::II. [fIi~ ~""7I ''''''''''I'''~ (.l.-~
~,,;~ JWrIi.. rill~ '-1IfI~'; _ ill dp~"'''' """"'n.".
qi 1.bUlv.w", itj ,P,.q,.", ".
10 1nc: compound "11"" 'ua.".~1t]ii an be rad in one of rwo w:t)'I. 1'he lmn
16.4 FOUNDATION S OF DHAlMAKIRTI 'S PHILOSO PHY
lhtbtJt' may to. inlcrprftCd in dw.a- 0I"..mmc:ss.."In thil caK. dw ~I would to.
WI dK INRWOnhinat f..r-,.fl44) of a Cft'diblc: pnlO'I'$ IUlcmcnll aboul ~bIc:
ohjecn is rIN 161M U W ~ncu of die. Jalnnenll with rq;ud to unob.nvabIc
obj«lJ. In otbn _1'Ch. me KJ~tJ I n tnlItWOrthy ptcc$::ly boaUK!hey _ me RlIIIe-
IMIItJ or. cmiiblc: penon. Thil 1ppc:al1 10 to. tho: imnprnllion Wen by Slkyabuddhi
(PVT. p:~"",. K:m):
codibkpaMn·..p•...cl. io .... _
... BoM' ......... i.n l.. ~ (_ ...."" H • .
Il~ Thai is.;ua u me
cmiiblc pmon', IpCUh illrusrwonhy wilh rtprd 10
UI objm thai can to. dncnnincd [throuF f'C1"p'ion at>d ordinary ;nkrmoe}. Iike-
witoe il i! mlK'II'Oftby wilh lepro to an extmndy rtmOIt abita at- pR'riJdy bceo'llC
i l it: the ~ or I cmiibIc: penon. [.,..,,. UtI,.'i 1Ihit.,,; ...i lift..., ....""
,. i ~, _ I ri fur """".. ,.... J-r .;r, ...... ,.. .,; w.. Hi .." ~"f' III pJ"'"
"III ,.IA It]tf , . uti,. 'i d.r ".; w.. H tit fur ,J,i.. no '"' no 0'",.,.'",.... r I .,..,
I
,. - ',.; lib;' ..,u,... ,.; """ .. K: , . " ....' ~1fI~~
w".,.,. ...., 'rdJ.o .,111# ]hilozt!flMllM ~..~ ;;"".~ .... 4·
I,.,...
fOSlrp~ ....... ~ ~~ __ ~~ 'T'kio" cJ-+y conup<
and Iw b«n COIlecled t¥twtj. ro conform to PVT].
.,those statements, just like the other cognitions based on those statements
~that can be verified by perception and empirical inference.
Hence, even though that cognition comes from language, it does not
make known just the speaker's intention like a cognition coming from
~[ordinary] language because in this case the cognition is also an inference
'.of the statement's objects, since it is trustworthy with regard to those objects
~(artha). 11
Alternatively, we state in another fashion the fact that, due to its trust-
;Worthiness, the speech of a credible person is the basis for an instrumental
Inference:
That credible person taught what to avoid, what to do, and the methods
for each; what he taught in regard to these issues is not erroneous, and
;hence, those teachings are trustworthy. An example is the Nobles' Four
Truths, in the way that will be explained. 13 Since those statements are trust-
worthy, the assumption (upagama) that other statements-which are use-
ful for accomplishing a human aim and suitable to be practiced-are also
trustworthy with regard to other issues will not lead to one's deception
because [1] there is no instrumental knowledge that contradicts that
assumption, and [2] it is pointless for that speaker to make false statements
without a purpose.
!4Thus, in the two ways [described in these two verses]; it is said that a
cognition coming from scripture is an inference since there is no other way
for one to proceed, as is illustrated by the thought, "If one is to act in aCcor-
dance with scripture, it is best to do it thus." However, this kind of infer-
ence is not without problems (anapaya), since expressions are not invariably
concomitant with objects, as has been already pointed out.
:I.Ct :l.t all beaU51'. one cannot know- whtttKr or not Wt person has those
kind of exuaordin:l.ry qualides. It is not the case thu penons such :I.J us
would not :let bccaUS(' of nm :a.ccepring that th~re :a.rc som~ pef5(ms with
~ qualities. since that kind of ~rson docs make true (.."illlthtl) SllIte-
ments." In other words.
"""'. *.
,. ~ _ Ji" (If I ~ .. <Wf II; M Iu .., D'" ,./fftI. NT 6yd ,. iii ,.,i 6yd .. I ...
9r"!"x.J (If"" ji Iu .. ftbi" J. ,.,,;,.,. .. t. .... ,.7..uJ.""p4,.." rJt,mJ J.
""I
h 'i (D.: "fI"I,.,.,,.
1111/" M]tUtf Ibn,., ... ""''' fl.
The tnlthfuJ (SIlmytltJ and dcttitful anions of pmons arc dut: to their good
mental qualities and their mental fla...,. Those mental attributes are super-
.sensible, and they would have to be infe' lcd from the physical and voc:aI
behavior (flJlllI4hl,.) that arises from them. And most behavior can also be
performed ddibcratdy (bwitihipiimz1lt) in a way other than the mental state
they iCem to reflect bec:au$(' those behaviors occur as one desires and
because those bduviors may be inrrnded for vWow aims." Thus. there is
an overlap of the: evidence: for f2ulu md f2uldcssnes.s. Thcrd'ore, not hav-
ing made a ddiniti~ determination, now is one to c:mblish that the: author
of the scripture is flawle:u?
-men is there no such a person who is f2uldcu?"
l8The pma J.!t-. ohtn tnNbud ~ u "difficub," bu. rbit EnsIish word it 100 wak
for Ihe ImK IhM Ihe prUu mnooryL In ncarlyroar cut, .6.!J. apnae JOmCthi,. matt !han
Ens\ish "diffiaab' bu. fIOI ~=:t u ' imj)C*ibk:
Nocc mal. KICOfdilllco . (PVr:14Io. K:ml, ",,~). fan,nine nom-
inW"" IinFu when <:OnIlI1Kd wi<h ,.i. (i .. buI muo:ullnc nonWJatj"" plwa when wn-
.. rued ....w. .>OJ .,.,
19 ~ an Q&lD.pk. ~j (PVT:l.4Ib - K:J97) ncxa: "Thu is, fIUIGN who ha~
d$n- ....y make thuNdW'Qi appear • if !her wen: ~, and daiWcM penoN may
make tbantrIYa appc:u II if Ihq Iud daira:" IloIil4oI '" " '''- .yi
""""'",; l ..uNpI U Ui"PN'.1.
'!' '"""is"""" .,.,.....
20 Slky.buddhi (PVT:1<I9U; ab6aI. in K) ammmu;;
I, it "., ..• jw.._ " " . , MrrjIIi", " -
rIJm cr _ ~ .,jill .J- .,...u;rift. lill«
"-li""malt{"Ix). - Ma ..... "..,"*"""""'
ihowul aaxpa ~ claim mal
.. Throuch
~ an _
<hil IOIlonall (otwm.Idni
prnons who aft f.IIIUdaa. Hlv-
i,,! acupud dUo daim, olna: he dnira ID ... .t.I;q, il by dnnonstnO",!he inwu-
mmral evpIilion .tulihows aKh IIW( ID bt poIIibk, he pYu bwani all oppoai"l
_ " ' _,., ; .... __ J, ~,.... 'MJ,."IP.D: 'i] ~ ' -r _ .. _,,-. __ .,..
.. ,. "" ~ 'pJ!4T ir!,wIJ. /j.
Sikpbuddhi (PVI":L4,M) ahrr .... duou&fa IIx _ , alto n-marIu:
In !he IIC"ClOOd cbaptu [DharmWni l shOWI tha. !he knowkdp of ~ mil.
__ aU Ib_ Tk,,:§....<. it is nQC mal _ do nQC attqK m. dw:n: iI any .....:h prr-
lOCI. W( aft jus( ....ri.. ~ it is c:urandrdifficWllO know...-htthcr or noc ~ ptnon
..t.a iI _ '1 ~ is dJ:, o«t of&uJu.. [..,... 1M --. eM Itp' ,..-J,. UI/I.a J u.,
--',.. i"-,..,.·" _ .... kio,...,..,..
, . j . . . . .,.; [PVI"-D and -p: ...1
l.o._,. D-Jt,n.. 1*........_,.,...,,.
MJ,. _';. ~ J
_*11. hru-r
1M
APPEND IX O F TRAN SU,T IONS ,..
All Raws. being susceptible to decrease and incra5C. have coun·
tcragcnu (vi,.ltpt); hence. due to having inculcucd the coun·
[Cfagcnts through habitw.cing onc:sdf to them. at some point the
ncgativities (IJrlIIltl) should be eliminated. (PVI .UO)
countcragcnlJ becomes inte:rue. [bose fuws have the: quality of being dim·
inaled without a mu. as is the: ~ with Are and such. Tlltrcfore:. some-
one: may ind«d be: flawless.
~Bul bow could someone: be Rawlc:u? No one could be for Ihe following
reason: c:vc:n when me fuws' coumeragcm (vip"Jt,.) is infused (t4tnuIN) in
a ~non's mind. £awes wowd Ulu ag:a.in in aecord wim the eoncUtiolU wr
becom.t: availabl.t:. jtat u that counre:ragent arisa in a ~non whose mind
is infused with the: Raws." >I
This is not a problem. sillCe
It is not possible to alter the nature of the mind without making an effort, JUSt
as a Brahman scholar (frotriya) who later becomes a kapalika cannot stop his
disgust (ghnza) without making an effort. 24 And it is by seeing the good qual-
ities in the nature-svabhava of what one seeks to obtain and faults in the
nature of what aims to eliminate that one makes an effort to obtain or elim-
inate something. But it is not possible for one who has inculcated the coun-
teragents of the flaws to see good qualities in the flaws because their opposites
[i.e., the states that counteract the flaws] are free of negativity. They are free
of negativity because [I] in them all flaws are eliminated; [2] they are devoid
of the suffering that comes from negative orientations (paryavasthana) and
birth; and [3] they never turn away from the taste of the bliss of peace. 25
A mt:nral state that has an unrttl obt«1 (abhiiJ4nhll) arises due to a pri-
'"
mary aUK (ufUitUNI) (i.e.• me: conceptual imprint for a fWe, conceprual
cognition] . AJ a result of appropriating (updJdut) what opposes the con-
rinuum of that mental nare. mat menu! srate should not continue to exist.
HO'NeV(f, a menwscue [such as an awareness of sdRcssncul trun has a
real object cannot bot halted by cultivating its opposite b«au.sc: it arues due
[0 real things themse:lves.· The Raws have: unreal objc:as. and as such. they
cannot oppose a mt:nw sate in which their counteragcnt has bC'cn infus«!..
Therdo~, the Dults do not arise: apn. This is the case: bcc:au.sc: even if one
were to strive to re-create (he flaWl in ,he mind. ,inee (he mind tends
toward positi~ qualities (XW!l4). a judicious penon [who ,till has some
Raws] will make an dfon only for wlu.t counteracu dum DullS. How much
mo~ so i. £hi. the case: for a penon who is unRawcd.
"But what i. the source of thoc naws such that by ioculcating that
source', countcragcnt th()' can bot diminatc:d?~
All rypes of R.aws are born from the belief that the evanescent
components of body and mind a~ the locus of an Clsc:ntial sdf
(utlilyiuUrt.1IJI). ' gnon.nce is thai belief. When ,hal hc:lief
ntturs. nne aperien~ clinging In that alleged ~Ir. and from
lhat clinging comes an~r and so on. IPVJ.lu]
A person who has neither the- nocion of "' - nor the- notton of "mine" is
without grasping. and as such, he- docs not ding to anybody or anything.
And a dispwiolUtc penon cannot haw hatred for anybody or anything.
27 DhannUini', point M.: ill limply tel bria& losechu slisbtly difl'uml ways of ~n,
abou. d.o h'-.I· _ ..... ca.- (· ;J,-.)olbulu. On u... one MM. __ ~.,..u ol
~O~., eM bdid" or "view" mal eM ~1IeICen1 oomponenll ofbody;l.tld mind aft
APPEND IX OF TRANS LATIONS l7l
Thus, it is possible that one might eliminate the Raws, which arise from
the bdil!'f that the evanac.c:nt psychophysical componl!'nts atl!' thl!' locw of
an essential self. One can d iminate the Raws by eliminating that belid; and
onl!' eliminates mat belief by means of its COUntcragcnt , which is ml!' real·
ization of I!'mprineu. But it is difficult to know (durllnll4J") whether thl!'
person in :aa:ord with whose teachings onl!' might practice lw indeed dim·
irnllti!'d thl!' Raws.
• Ioo:w of J.n cam.ia! KIf). This bdid". ho" hU. m.y 1M: alkd -iponna" (.m.Iyi), and
i£norana may Ix alkd "coofUsion- (-'-1. ~. rvorn thouJh both -0. and
~ an: Wei !(11M: rhc aUK of the: fb_ no o;ontndKtion is involvni. sin« dwy may
Ix Uft~ as synonymow: In dtlJ COOlo:D. DhlrmHirti d _ 10 anptwia: "'U,,J;lIi
bccaUIC it ~ readily KaMlnu for his way of tpakinl about ibws and their dimm.uion.
5. PV2. /-6 with &l«tions from PVP lind PVT '
As for rhou trustwOrthincss, - h~ving determined the objttt. when one then
acu upon il .") thai thins'. c:ou$al copaciry iI ftlabli.hcdi hen ce. [in one
$C:n$C:Jthe trusrwonhiness i.s that th:l' thins has the lUnd of nature which it
I S« Knuotr (1001) £ora numbottol usoeNI Sarukri, ~tJ, lTWIyofwbidl wnt' induded
in. prniow -uo.. of til;.. l.....ul>un (Dunne 1999:4)6A).
2 Commmtinl on ~i'J SUtcn>ent, ~j (PVf, 1IJ""7lb]ff), ~ the
following ~jtaionl and resporues:
"lnfm:nce and pacc:ption an! ilUuummw; thaI bring ,he cue, how an tbc BicsKd
One. ..t.o docs no( ha~ the IWUtt oflinfm:r.a: and pcrcc:jlOOnJ be: called such (Ihal
iJ., 'irmrummw' W
H~ iJ c:alkd an "","rumenl of~· beaUIt' he is $imihr 10 ~ ("'0 kinds
ofinJ"'In-cnc. ofk..IO~. In other 'A'Clnb. he (an be IU~ in 1M mcu.pbo.,
"iruc:nuntnw."
' Su, the B'-ni One Iw; the .... run: of the ~ nona>neq>rual, .mminl ..;..
dom thai atdots by fotu of kir medlUlicn. Hena. 'M BIeucd One ;.. actually by
naNn: I perception, 10 why doa one nttd In n:ly on. mcu.pho.~"
Sikyabuddhi (PVT:rutH") mpnndl:
Some lily rhiI ;.. no( I probkm beawc the qnthn ,"lIfh:MltlJifil is inrmded to
[ hal
n:kr 10 • ~ SUIt. But hcn:. the 0Dtta:I n:spotuc iI I i follows: Evm if the-
8ksKd OM Iw; tho: ....1Wl:' of lhe afomnmlioncd inlfrumml oIluoowkdst", he iI
~ DOl (Om..-dy kMwn II ouch. ~. he if metaphorically (Ompand
to ........., .. , •• a-a. ..... """...., ofknowl<.l!,,-
17.
APPENDIX O F TRANSLATIONS
·Sin« infertna does nor dcrermi nc the objccr in iudf, the definition of
'trustwOrthiness' does not includt it; hence. your definition is incorrccr."
We respo nd that a cognition is trustwo rthy !>«aU$(' it does not dtctivt
people; this indudes' pcra:ption and inferucc, which have the cha~er
istic of causing one to obwn the intended obj('Ct! To comment on that,
(Dharmwni) says: fnUnvo"hi"m is II ctIf1'itiD" 6/ ulic fimctiD".7 T'hi$
means that one hu a cognition of the accomplishing of the aim that is to
be accomplished by the obj('Ct that one lw delermincd through the insuu-
mental cognition (tsINuJ 1IW JD"fI1I4 "trf JNlI" ,'m
ois bwuh JNlT bJII btl I" Jo"
b]uJP'I' 1'UJt1,. ;).' For example, for the penon who, having oognizcd a flI'C
through perception, then acu (jMI/M) on fire's' capacity to bum, coole. and
so on, there is the activation of a perception whoK: obj('Ct i$ the sensation
of warmth and such. Or, for example, on certain ocasions there might be
a cognitive error due to something which lw a form similar to fire and so
on; at that time, there is for that person the activation of an inference
through smoke which dc6nitively determines the fire . (Depending on the
comat, one of these ~he engagement of a subsequent pcrccpc:ion or
infcrence-<on6mu) the tNS(W()rthiness of a perception.'"
fkc:ause various awes of error in the case only of perception arc possi-
ble, it is sometimes known (0 be: instrumental through the activation of a
subsequent instrumental cognition mal
has as its obj('Ct thai thing's tdic
function ; this is not the CiUC' with inference. That is, a propertY-s"""''''",,,
used as an inferential sign and a cause used as an inferential sign arc
restricted to being the property-Wflhhiwr of the rcaI thing in question and
the dfttt of the real thing in question, rcspectivdy. And only they (i.e., a
property-llIIrI'bh.iVtl and an effect) are the causes for me respective sign-
aware ness. H ence, if mat kind of thing {-i.e., a property-J1I«bh.i1l1l or
efl"ea-l is absent, mere is no inference. Therefore, inference does not rdy
on the enp.gement of a subsequent instrumental cognition.ll
,.->
5 pv. D: 1tb]M/,.,11:JtJ ~ ,.,~,. {a tflljJ
6 See paWId Ski. p"'l &ts in Smnkdlnn and Krasacr (1989:)1).
7 PV.1.. IbfF, lI~tilJ l ..u.,.lIrIbu1ft.
/I ~~hi (PVr. ~JLtfJ) p..... ,hi< IrnTU.:
An ttnIM is bwnillS and.., on. 1lw ·ac.c:ompliulln~m· of dUI meaN dw arWI of it.
9i fh1i',,).
9 _ -> _HatMj.).
10 Sikyabuddhi (PVT, ""..71h7) makai il dcardw!his puA&C cono;crm the inwvmcnta1ity
of ptllXp!ton.
11 AaoniinSIO~ CPVr!7-4htl. die poinl hen is thai inkrma involva no infO.
APPENDIX OF T RAN SLATIONS m
In other cases," one: may nOI be: ttrtain of the: difrcrrnce between a per-
ception and a spurious perception when they occur, in IUch c:a.ses the actual
perceptual aware:ness is known to be: trustwOrthy through the: e:ngagc:me:fl(
of a subsequent instrumental oognition.
"'Ine: latter and forme:r perceptions arc not distiner, so, since one might
doubt that one: is not acting on a real thing. one: could not be: cert:a.in that
it is not erroneous. -')
This is not the: CI.S( lxcaW( both would not occur (JUt /'4) in the ab5c:nC(
of a real thing. That is, for one: who actS through bring prompted. 10 act by
a f.&ulty or dubiow cognition thai apprehends JOmcthing that is not fi~ as
fi~, a subsequent awarmcss that has :IS its objc:ct burning and cooking docs
not ariK:. It docs not arise: beaUS( du[ [awareness in which the c:xpcaed. £die
f"unaion appe2lSl is bras«!. on a rea.I thing. If that substoquent awareness docs
ariK:, men the formercan only be: trwtworthy lxcawe: [I] one obtains a [die
function in accord with one:'$ expectations; and 1xcaW( III me: callS( ofjwl
obj«t i.o nO{ ckfinidvdy oon-mincd; dw aplli ... lbe 12M' w~ one iI initia1ly
prompted 10 I(t [ . . oppovd to ~ituJ;miKtioN--tft ~J. AI. tNt rlnw. one alto
aaa 0lIl of doubt. By implicllion thil laid dw if. due: 10 habiuwion. one <khnitM:Iy
:apprdlcnds ("fI!I'" n.."t H j thc panicWu idc:nlity, C"Vf:n pt"tuption doa no!
depend upon thc cnp&rma'II of • .w-quml aWUft"ICIS.
13 ~ I"IOfCI that dw; former and lauer I'W&rUICSICI an IlOl diffrl1:nl in W.I the: lal'
ter nrvn>eIII ofbumi,. and IXIt"lI<ins mishl .., be dubioul, • when one appem fO Itt hl1:
in a drHm and then IIJ>PC2" 10 Itt its he'llli~ and c:ooking. Thil dubiety or ~ dw; con·
firmins iI'II'U"CTIrII wooId ~ il 10 be conhrrnoed by ya .....,u..". c:onflrrning - . and
ON: mlYllIiYCI.t an infinite ~ I( on th. odwr hand. ON: admiCi w.t dw; latter ~
naI iI ...tr-conJirmi"" thm why nol admit tim mr hnt OM iI u well1 Stiryabuddhi
(PVT",,.b7-1?ll thm q _ rwo _ he onrihuta 10 KI.lmlrila, and which may wdJ be
from dw; fJrIM.tfiH. Thry _ riled H 1'5 't¥. tJS~H:
JustIII'Ihc hfSl:.~ drpmdI upon that ~ [Rlppt;ecI by mr IIeCOftd
OMj, in tbr -r ~ W3y that UUICWOfIIti_ {Rlpplird by mr 1IeCOftd] wooId:oL.o
rcquil1: ya anomer uwtwOnhinc:u [wppIicd by JOfTlC thirdJ. lf, ho .. r..:r, onU'efr 10
.unil duo IOIfIC j.wuaxa in this ctwn] uillluumcnw inuUuiaUy [i.c.., on iu own].
I" Sikyabuddhi Rm.uio:a on !he tcaion, lUlling with ~i 'l lalemenr: TIN;'
_ UN (_11«._
eo~u (nalfl)!
Nih K-lJ _ fItrIU (j"f JMJ i,. w .~ ". -' Sikyabuddhi un,.,.
a.u. !he I'tO'O ~ 11K a.ppem.ntt of fiR in the fOrm« aJ&IIiOOn md
IIIQnS
w appannce of bumin& and such in flK Ianet. Although a. tapUtivc appc:amICC of
fiR may apply 10 C"I'm unreal tbinp. wappcarana
ofbwningand wch wi.lInor. nu.
il apresxd by !o.-ndnbuddhj·I] lOtemrtlr: for ..... """" «11 "'""tit
ki",1""',w
,. lin i-y. fo .. hJ •• J..~ "'Pi';',. u,., ."",JwVt »-un,., thttt iJ _ fi" 10firr.•
11«._._mom
"'lwfW1lt
.IJN
~hils ill Hjm "'"'"","'"'
10 "riM. 't"-,..,
.-.Mi", " - _
dJu.---.namdy. dK 1 _ tMI IlUeIw"h an ob,ea that Iw 1M !die
funccion ofbumi"" cooking and to on-iJ""'",,, -' thi",. I/i, "- DiM. thnI
UN~ ...,. ... ~ WtnlJ~tN.1 is, if IlUbtequenl ____ oflMob;ea (...m.)
thaI accord. wilh OM', Clp«tatioru ....... in a paIOn who aeu through bcinS
pron1p!"ro by dw kind of former lwaleral, Inm tN.1 ii Of conmluler W formt!"
IwaKl'IQI' !nUlWOI"thi.....,.. ... TIWo is 10 M._ IN t.-.Jput tJ",t ~"" .r._NWfI
-/"lir fo1lttiH it . "'" thi,.,. Thai it. iioor: il acmmplisha: .. Irlo. (ttrth.) such as
""minI and ~ fiR is tM wtoidl aaompWher buminr.. cookinK and 10 on;
likewisr, ~Irr it tN.1 whidI it wrd 10 acoomplWl washing and drinking. This is so
b«aUK a rr.u dunS i. by ddinluon dllt willch II ~blt of Irhe funel>on
(."b.~).
" Sui done is WI kind ollruRWOnhinou with ~ 10 a lno. (ttrth.) ~n in a
d~ '
It is prd't1"IbIr mal theR be 1 reU thinS in tru.1 aoc; whm a-n. lhett is no other
INUs (rrttd for pocirinS" reU thins.
llIm dram· lwata'ICIaICI woukI be irunumrnal."
Sin<:c _ admil wt this ;, the aso., mil is nor a flzw in ow ~nli.
"How is ;llhal.".. art thaI an lwauneaa is nGI ;lUtlumcnlal?"
BroaUK OM tbinlu, "'i, is trfOnc'OW.. • Hma, a prrupr>on whoK obj«t ;, ~
bit- ofllCCOmpWhin& an aim (.nIM), Mn«;1 is dt-¥Oid of any C;Il11e1 of error.;' UCft·
lain«! 0."1'''' ..... ,. - ,.rird,i,...) 1tr rdiClivc aw::lR~ "" !xins by n.a'UR
irmtumnlal. ]1 produca I dcfin~ drrmnirulion of mat objKI in aocord wilh the
_ y ~t;' ...... ucmail\Cd. HPI(t:, il is intriruiaUy irmrummal. and thn-d"oR, IIKrr
;, no infinilr rqras.. O~ mishl nor . I........ cl. be habiru.nro 10 an iniriallw::llmCSll
...;.h ....... w-n ....... oJ r.~ '" _fff: in ..... " "'. ,h., ................. ~ .-- "-",,, .....
CII~ry 1(1 produu I dcfinitr.c dn .... mirwioon becawc, e-trn dw!u&h WI fin. or ""'t....
has bern appreMnded by an inckpmdm. (...,., 'fP'I/- _,..,,.,..; irurrumcnm cog-
nilton, done I f ( a.IUCIIW. induce ~ !hat pmoenl MIC.h I ddinilm: dncrmirulion.
In Wt caK. WI iniliallwaKl\Cll is mablmed 10 be iruuumcnal bylhtmptr;anml
of 1 subtequem illlltUtnCf\m aJ&IIilion. Hena. il iI afriMia.lJy ;1lSII\UI\Cntal. How-
tvrr. if OM Iw an Iw::IRnC:II WI involou llabitWlion and wriry. thm ill irmlU-
mmaliry is oktcrminc.t fiom irwlf (...,., W .. _~), "" _ apb.i1\Cd aboo.ot.. In!hil
....1, ;• •~ •.w.... oJ", ...... ,1_ JKlo.cptiu" is inK.W'OCI ..... ill .... 'IC c.oa ;'",i•...oc.Jj1 ,""t!
in _ CUCI ClninsiallJ'. lnfddOCt ;, in.rumcnlal in lriruically.
APPEND IX OF TRANSLATIONS
not cogniz.e me object. lIut w:lS apprehended by the former. If mat is the
'"
ColiC, how can it Mve as iu objea the ,elie function of an object. determined
by mat former awareness sueh that Ihe former awareness is instrumental
because it docs not deceive one about that object's telie function ~·
This is flOt a problem. Beings engaged in pr:roctic:al action (1IJII1IQ}",,,,j act
on those twO objeca: without differentiating them. Hence, in KO) rd with
such practic:al action, we S2y that, beings act on objectS that occur in tem-
potal stquencc as if those objccu wete a single thing. In reality, the former
and latter objectS arc distinct. However, the real thing that is the object of
the latter ios((umental cognition would not aist if the object of the former
instfUfflC:nw cognition had not been existent. Hence, we metaphorically say
that che latter awareness has as its object ju.u thai object of the former
awareness. Therefore. since the real thing toward which one aeted w:lS
established prior [to the cognition in which iu telie function appeared!.
that initial cognition is instrumental because through it the lalter awareness
engages with the ,elie functwn . '~
According fO those for whom c:r.:rcmal ob}ects do not exisI:. W obi«t which
is dctamincd by the former iJl$[nunenral oognition is not iocontrovcrtibly
the cause or me laner ilUuumenw cognition: howcvc:r, it is the cause or an
aWllrcncss that has me appearance or the desired tdic function. Whatever
docs not have mat appearance is not inmumenral. hence. there is no con-
tradiction. Even ir there ~ no external objcas. [he aWllrmc:s5 that arises
with mat kind or appearance is a human aim (JItyn 'j Jo" • PllnqArtM). "*
Thus, with regard to meir theories of tfUSlWOrmincss, mere is u1timatdy no
difference between those who maintain the existence of external objcas
and mosc who do not.
SonxoIM: obj«u.. "Bul . ina: 1M formtt insuummtal copilion is whal cauKI dIO'
actiYation (k,. """,. - ! '' ''taU) of dIO' uller inttrumcntal copiticon. tht: 1Or-
ma illilrwnmtal mplilion is illltnllnmw only when dIO' urm inKnunmtal c:opi-
don " _,,,,,.td;
p.~. ,. .... 1\01 QtutMfltN. But u..t bfter ,,,"nrmental
aJ&nition docs IIOf lui...: the apacif)' 10 lUke mal t"onner IWVmC'# whkh docs not
ha...: [dK naNrc ofbri", inouwncntal ] into an~..........ao thai doco ha...: dIO' rwurcof
bci,. ilUlrwnmw." AI Kumirit. hu aid: -Ln i. ~ known ~I all insuumm" of
kno~ aft intrinsically ilUl~na.1. 1Ot lhal which doa 001 ha¥e the apacil)' W
~ al an ilUlf\/.ml:na.1 copliOOn OP i" own (anJ)O\; ~ made: Q pi' bk: 01 doinlllO by
.mochtr IRlb.tquc,"1qnilion " I ~ ~~ ~ iii p...,m-
I .... ""'_ Wi ~ ",-_.""... ~ f1 (~V: c..u,.., 47)1.
for Ih.~. [ ~dd.hiJ AJ'E Tbw. m.n.... ln othe-wonk. withoul dIO'
obic'a of <he fOl"llM:r ilUlf\lllM:nw QOKIlilion. d>c ob;ctt of <he bner imlnuno:nw (0«.
nilion would ~ impoaibk: htn«. the b tta" _ ~tN;r. the ob;m ofdw: 1Or-
ma;,.",J/ "'"'IF d.clcfun dlt: ftnt~ 0lIl: thar is ~ tht 2~
whidl hal as in 00;..;.: aa:omplishmml 011'" [ap««'d] ,do. {~;, ;"IInI1Ifnf-
uJ. In OIhtr words. 1M firM ~ OIIIISCI d>c Iura-.~ whoK obj«1 is dw:
aa:omplisluMnl 01 1M [apeaed] Idol (.nJJ.). as MaCh tht f i n t . _ is iNtN'
mental b«awe il also Iw. ~ thinllas iu object. 176a11rthal W01: 001 dw: ax, d.m
lM fine twaftm. would Iu...: an IIIlreal!hi", u ia ob;ca: as -n. ;.
would 001 bt:
..... au", oi ...... kiM '" ... ~..,.,., _ _ _ Ii ..... ...... wiv- ~ .. ,+... ___
plishrrooenl of d>c ap«Ied .doi]. This;'10 si~ .....p;t. (,rqr._".. • .,.,.,
wid, lIN ~t If'" {".,.-J} w... TIImIp;, tnanI "ml1Mll!h me, fint
awarmc:n." In other words. this I!W:IlN' -since me, fUll ~ is the auK" of d>c
lifter ....... tcIICSf whoec ob;«t is the aaomplidlllM:flI of tht [apmed] ldoc."
TIu.1ItpfIIftI1 dmiorutnles d>c ~ When it ariICI from in own ClUICS, d>c
I"ocmu twaft_ NiKI u lui""" only d>c IWUn: 01 an iruaumcow ODpIition bca......
• raJ minI {. , .- -'"'- ~ is pmkM. Howco~ •• if tbrn aft =-lQrftTO<•
.... rely me JdU, •.;..., ck ..... ui ....ion of;. ao ;~ ...,.J io CZfH rh.roug...-l-.cr
f
[subKqumt ] cosnilMxl.
APPENDIX OF TRANSLATIONS
~l7 5akyabuddhi notes (PY[, nyq6b1-2), "By implication one supplies (adhyahara); 'In all
'cases when one is not obstructed' ... " [thams cad du gags byed pa 'i rKYu med na zhes bya ba
'khong nas dbyung ngoJ.
18 Cf. Van Bijlert (1989:130-131).
j81 fOUNDATIONS OF DHA1MAKIRTI'S PHI LOSOPHY
[TIN ot,," issw thar Ltnf'Ullt rllist:1 is ~nt thilt Ilpplits U1 CfJn«ptlUl/ ropiriorJI
in p"IlL lJnJmJr.lnulJhi aprmn rlN probkm in tht foJ/.owinl obj«rion
(,.,}oJ
•An instrumental cognition is that which is uustwonhy wirh regard (0 the
tdic function when one actS having become awa~ of the instrumental
object through that insuumem. If mat is so, coruider the case whe~ one
acts upon a water-jug through the conceptual awareness of a water-jug;
when one does so that conceptual awarenm is also rrusrwonhy with regard
to the relic function of acting in that fashion. Hence, that conccprual awar~
nm would also be instrumental, but you do nO( acccpl thai it is.. Therefore.
the definition ofinsuumcntaliry is faulry.-
[To clarifY the point being made here, Devendrabuddhi (3b4./JJ makes an
important innovation: he introduces the notion of "mediated" (vyavahita) and
"unmediated"(avyavahita) instrumental effects (pramar:taphala). He begins by
answering the question he posed just above-''Why cannot the sense faculties
themselves be considered instrumental?,]
;ulacrum that instances of knowing are distinguished from each other, even
"though they are indistinguishable in terms of their nature of being experi-
'ences. Hence, due to the differentiation of the objective image-i.e., due to
rrhat quality ofthe cognition 22-the awareness, i.e., the instance ofknowing,
.is differentiated. And since this effect exists when that is present-i.e., when
irhe object-image is present-awareness is therefore instrumental. If when
:"y"is present, "x"comes into existence, it makes sense that "y"is the most
~fficient cause (shin tu sgrubs par byedpa = sadhakatama) of "X."23 But if at
some point there were no such effect [i.e., "x'1 when "y"was present, then
one would realize that "y" depends upon some other mediating causal fac-
ior. That being the case, since that former cause ''y''is mediated by some-
thing else on which it depends, it would not be the most prominent causal
factor. Therefore, it would not be the instrumental cause. Even when the
sense faculties and so on are present, they do not [necessarily] have the
causal function of producing an awareness because they are mediated by the
object-simulacrum. But if the simulacrum is present, it is necessarily known
.because it is not mediated by anything else for that knowing to occur. Since
it is of the nature of awareness, it is the basis for positing an effect that does
:not depend on anything further for its occurrence. As such, the object-sim-
ulacrum is the cause of both kinds of effects. And since it is of the nature
of awareness, awareness alone is instrumental .
[To answer this question, Devendrabuddhi cites the first portion of Dharma-
kirti snext verse.}
{At this PO;III, INwrubabtlddhi (l4imJ tJutt Dh,mrukin; has finiJMJ his
tlntriptio" ofthe first J4i"j"ldMrlllVrism (I"a) of." ;nstnlmmJaI cor-
nilion. Ht t laimJ thlll Dhttmutltirt; "OW prtSttrtt. -s«oN4 Mfininf m.~
mVti(-:r
/This j,1II1J I(). ftw problmu. StI"" htlw tD th wim ,IN possibk ;1Utnlmmtai·
ity ojlNtUlKilUltWnJ. ilUlS",uch lIS thry Also "iUlimilUlu whttl hAS not b«n Jis·
mnd "DrwnJrilbudJhi (JM) t/4irm IIMlmis notWn is tlvmN", tlx ~ til
thr urm ../lniNt. .. • Drwnt/rtllnuJ;/hi Ju d4irm tlNr: tht tmn "/lrth4 .. nulbln
tlnr I() rollnt ntlnprrrrptWn (/lnllp4l4bJhi) lIS inJtnlmtnlAl (Jb71U). BWI thr
mtlJI imfX1rt1lnl tlbjrditln raisrJ hrrr is Ihr Imt t1JnlAinu/ in DJutmulti""; i
nat SlAlmrmr:)
27 In a kIott- _m., IftrII I . , 'r hm.lUnnioN .. a synonym lOr n"L.q.!U- In Il1O« pre-
ax 1unu,IO apprdlmd m., ~of an objm is toapplmc.Kl it in its KtWlity. ThUJ,orUy
po«ption "f'PrdKndo m., -.w,. of thinp. irwm..m »c:onapnW msniUona apptebcnd
objo:cu i... ,.._iJ'·/Uftion...........dy. ~ the mcdi.otion of. uni ........... Sec 1'\')-,,..,,
and u.,,~ .. ol~ (•..,.hl.
APPENDIX OF TRANSLATIONS
'fpharmakirti answers:}
Only the particular accomplishes a telic function. Hence, for the sake of
that telos, a person seeks an instrumental cognition with regard to an object
that is a particular, which is asserted to be the basis for [the desired] telic
function. He does not seek an instrumental cognition with regard to
another object that does not accomplish that telic function, because by
accomplishing that kind of thing, the telic function is not accomplished.
6. PV3. / - /0 with 5<kctions from PVP and PIT
Innrumcm:a.l cogni[ions are of two kindJ because mere arc two kinds of
obj«tS. There are rwo kinds of objKU because some objccu :arc apablc of
[die funaion while: omen;u-c nOI.[ IIIIUiun l JUl;h a",llhc: ha.in [dlill "ppc:ou-
in ,he visual percqxions of a person with cu:uacuJ arc: nor objecu (IIn"-)
bttau.sc they ate! !lOr consida«l to be objccu,' [PVJ.II
I ~ndrabuddhi (PVP:u,P4ffi:
·WdI. in IM......am1CII 0( a pcrMNl with ddi:aM Yiaion dxr,. ate &lK appc:annc:a
....J. iM 1....... 1\ia., , ...................J .... "'.. "..., r..t.c ~ ... _ ... .- _.pI ....
any P'WP'* (.,.",.) II all; bomo., tMy U't' no! induoXd in [die alqpy 01] panicubn.
Sui nm thou&h ,bey ate dc-toid of any fIlCh ,elie function. they all' noc inchKlcd
........ UD~ """UKthqappcu dearly and ba:awc thqan: noc diaribuu:d O¥'a'
anydtins. n. .... ~ntt tMyate ROC..bswnetI wwkr [dwalcpiel oE) parricubnand
u~ they anilllOlhcr kind of ot;m.. HclKlt, h is noc WI'«' tNllMrT arc [juMJ
rwo kinds of ob;«u. •
No, it is no!. dK ""* tNl OW' vir- .. not <OIWO. To be ~ bain and «> on ate
noc objeta (If" '). Whr. &WIM".""" t'PIWmI. H ~ 1I'V}.Idj. 1~1
;., pa...... aoppl in pncUcaI ;oc..... (.,....;.",) 00 1Io)C~,.hem 10 be obj«t..
1lw; inlenOon oidtit IUtMWllt is u IOlkrws. If tIw Aia and 10 on thai an ptla:i.ed
by • ~ wid> QI~ and II> on --= 10 be objoas. dim one ......Jd invntipte 1M
Jirualion. ukin& -11;1 a puUcuLv. Of it it • v.nivustlr Bu, Wy U't' not: ob;rcu in thai
Whim ben· ...., wid> rcprd 10 an 2WU'CftCJI in whlch !haT is dlor appc:anna: offlics
and 10 01'\, pmoru cnpgd in prxticaI aaiGn do noc Iu"" dw inlmrion, Ibis is 1M
ob;ea of mal lwaI'UKa. •
ThUl, II in the cue oflhe IUbjeai~ »pea oflW2l'mCll. cvm lhoop IMy an:
cktnminai. Ihq do nor hl~ II'Ic WIlli of bailS obJects 0( pncucal action. Hu t
ahhoup • 1UIiva'AJ is nor IlIlirrwdy 1ft oijea. ~ mppl in pnaicaI action
imaplllri-tdy ~o:rmint il 10 be I rnI thiDIS> and they a(t aeoordinsir. hrnu. il ..
ptIIiK'd u 1ft ob;co. Thia will be t:lpbincd laIn.
"The prooilUlm1m1 is as fOllows. 1luo: wt.icb is DOC 0DIUickred by pmona 10 be dw;
ob;ea o f _ l-...an:nno docs ..... glisly ~ mn....,tion of brins dw; ob;ea of duol
~ An cumpk is the: .... bj«:ri.c a5fKCt of INt lwan:nctI. ~ ~ in
pnaicalllCtion cIt> ..... cocu.lder the baln uwI oooa ""«;'00 hf PU_II wnh QtaraCU
and 10011 IObeOOjo " "TheC.idellla: here is the: ... ~pUon of. ptopc,tr_Miwl tlutl
oonlndicu a propel ry·""'Mbw tNl ~ W Mpndwn.
"The Mlb;m~ aspcu ., d~ objM of ~ I~ dxrd"orc. lhc o:umpk it
..... eII.blishaL •
This. ftOIlO, MaUlt 1M a.b;a:ti~ aspea: ismpliKd (rit,. . wt/iyJ ]n dw it iudl"
an- 21 mIaiw _ •
• ~ (PVT• ..".IJWU):
This InQl\I me fOIIowins- r~l lwam-oc:M is ftOI JOmnhins odwr tN.n lhe Mib-
jcai~ IIJpta AJCh tNI ;a objta would be ~ ~~ upea. ~lher. I~ IUb;c:cli~
J9.
391 FOUNDATIONS OF DHAlMAKIRTrs PHILOSOPH Y
The~ arc rwo obj«u txausc some uc similar across irutanccs and omers
are not similar; bcca~ SO rm' are the objttu of words and others are not mr
objccu of words; and beea.wr the cognition of some occurs when mcre: an:
causes othcr man Ihe object, and the: cognition of othcrs does nOl occur
when me~ are causes orner Ihan (he object. [PV}.l)
In this conlal, (hal which is capable of telie function is said 10 be: u1li-
Imuely real. Thc other one is said to be: conventionally real. Th~ arc,
respectively, me particular and the univenal.' [PV}.})
Wc o~ truat rhinp such as seeds ha~ a capacity for relie function in tbe
~ of sprouts, and so on.
Let it be so in the way as you have said {1fJ~ l"thtl tIlthtlj.' IPV) ...]
;\IflC'Ct uoopimI ill WI ir uiJawim rhe 1U1U~ ~ .cfltxiw awarmea. I( rbeyarc IlOl
distina, how ~ lho:rc Ix rhe rdalioll oI .... bjca and obim sud! !hat the: eumpIc
wooId IlOl be atabIished?
2 ~i (I"VP:I ~ ,,"fa the ruuo..i"l &Jc-: "- iii lrar hyi." Mi 1f:1tr/,..i
"'ItJtrI,..(.nIM/I4i",. WIJtI A,?-t1 ,&_). a . Napromi 11"7-61).
) Dc¥md"b"ddhi (PVP:l14hl maka 1M foIlcrwi"C ",nwU:
An objector Aflo 'BIU __", U ~ " " IN jiufmM. 'This ..-na, 1hc capaciv
b- [die: hilKlion don not m. u1tirnatdy ...m1llX'm".' 'J1gt it. I rca! mms IMI U
_ned w aile u1lil1Vtdydon not /u<w aJlrdwaam.cia (J.ip~wtu.IIiX .a . 'Thoe
oarpmm! j,; 'J1g, wtudo. don I'lOl e:aist in .omahin& dx Unotl dwaaeritUe of duo!
thine; ... cumpk is the: dwxtm.ic, °homN" ..ben applied 10 I borx. 'Thoe apae-
i", for tdio; funaion doa IlOl
I.
en..: in dw whkh io _cd.o Ix rhe partioabr. This
is III qumm. by fQIOII of aM nonpaaption of dw pcrvadi", quali(),.
IDtwmaldni fCIJIOI'Wbl: W, fA m. rm"l' __ ... -Js '-r. rqM"iIJjiw uW
fo..m... iff tiN UN I{If'"UI. .,.. ,. "'" Wkn he: Ars ~ the apKi", is obto:rmI.. be
maN due ilII dr.xt iii ob.a ,ed. ThiI meanr aM f"oIIowi"" When one rNllI COITICI inco
.,.iKmc:o: whc:n dw: otha- ul'fUmt. tMllann thi", .... a .-:ap;oci()' 10 pcvdu« the for·
.................. 10 "" ....... bk of ..... ~ ............ ;u.. th... Ici....t of ",-..,..,,;.. 0.:.
That (lI~'" U wI\" -.:U and .0 011 lui....: in rda.ion IO...-.U and M) _ . ,.,...., rhe
APPENDIX OF TRANSLATIONS 19'
It is not found in univmais. which art not o~rvro to have othe:t positive:
concomitance (in which a universal necessarily aiJu when the:rc i.t a cog-
nition of a univasal] or ne:garive concomiuncc [in which such a cognition
also in the: prcsmcc only and merely of a universal] with the cognidon of
a univc:rsa.l. An example in which these relations do occur is that of the qc:
faculty and the: form perceived in relation ( 0 the cognition of that form .
[PV.l.S]
tuIOII fOr tM objector'. argument is not esablW>ed. ~ ob~ III)": "The pfO>
duaion of. spro.!t and 10 0<1 when. xed and 10 on aisu is. QK of c.onvmOonal
produaion; ;t is DOl. ultinule prodllCtion. Tha&'",c. if tM ~'Y fOr Idie funcrion
is mted in lemu of uhi/I'UU prodocrion iU w ddinilion of. rcaI thing. thrn ;1 is not
iQ ddinition bcc::au« il does not COYer all ilUWlCa of w ddinimdwn. And if il is
IClled in tnms: of c.onvmOonal production. uni¥Cf'Ah would alto M~ it; tM ddlni-
tion 1fII'OUld dllll be innlid by reuon of O'¥I:IUlms>on."
1Dharm,.klrti raponds:) Ut it bt ~,... ...... ",iJ. We Illy thallMClpacity for INc
Iimcrion ill not qualificcP- lu rimer ulaltUU: or c.onOftlcion.alJ. Bw: ~ kind of~.
iry lOr ....i.e IUnmnn ...... ;. "'"1...,.;, ....,1 (-Il-t ~ _ Jm...) [by t...;"..! ;, i...:!i.,.. •...hIo>.
and what cIwactaUa a particular is rhertfore thai kind of indisputable: ClpJdty for
....
cdic Nnaion. We do not haft tM ka$I obj«tion 10 you caIIill! iI whaleva' you mip
MOreDvu. what is ;1 mat you aIlnt te be c.onvcntiorqJly cxiJtmt such chat )'00
111m QX\\'auionai produaion IUSl'1bull101 production in me otbtt..,."y (i.e.• ula-
Imfdy]! If tM c.onvul1iona1 is 111.11 which is utterly nonaiscmt in ;tiI ma<ny. thrn
whm ipCilinl of 1hc "c.onvultion.a!." _1'I'I)U\d mean that 1hc thill! hu no~
''Y' But in speak;nl of " pnlduc;tion~ .,..., would mean chal the dUnl hu [ldie1~
iry. And irwmum as dv: capaci.,. rro. rdil runcion] and .....1-..... "' .... , "'r-i.,.
~ mutually =w.ift f!--s,..,."mhtlrll1lln·ti). bow an a linBle chi", ha..., bod. dul
capacity and itl ab.mc.c?
Or ebe. if )'011 IJKIt that the tcml "OOIIvcntional" means -produr:cd" thr:n IJina:
-ulamaK" muru "oonproduction1)'OO Uft lhereby:wcru-d dI.11 ~ is noapro-
duaion in ultimate tcnnf, i.e.. by eM non;woduccd. (a . PVT. "J""ISSal] 11w bri",
tM cue. othcn (i.e•• dMIK who hold our oplniMl do not ~ 111.11 lhcrr is the pro-
duaion of 1M nonp<Odl.OC>l:d ....t ...ct.. Hence. In it waujMed ;,. tb.n II.Wy n....., ,.. you.
hs- ....", ;. ]. ,, ;. ...... .... • "' . _ ~ ... _y u...., "the p<"O<I.-d P"~1J ' <O the:
produr:cd" bca....., then:: iI [obvioualyl noodwr fOrm of production.
3pyp (r1.4b6) y"..,..,J,._>~,. .,.,. Cf. PVT. ~.1s-.b6.
~'I objeaor mpondi dw m.lCllcmeru"1,..c( it be II you haft Aid," "IOUId
mean tN! ~ univuAk ha~ !die: f"unaiDn.1ouc ~i mpondi dw they do not
bra...., they ~ not invariably cor:w:omiQn! with thoe OCCUrTellOe of In lwarmcsa of thnn.
He ~ ~ AI. tM -r Ievr. the !ciK: IUnaion of rnI rhinp ;'10 pnldua: In awa.rmc. dw
thina ... i.. 00;....; unlvt:nm do rMX rvcn M..e tM ~ w produce awarmcw
I.... U..-
[tNliuvc u tbcir ob;ecu dv: 11Ilivt:nili:1."
39.. f O UNOATIO NS Of OHAItMAKIATrs PHILOSOPHY
The fact that a uni . . ersaJ is not invariably concomitaru with the cognition
of a univerul explains cognitions of suppos«lly exua-mental entities, such
as substaruial wholes-i.e., a watcr-jug-projcctions. universals. numbers.
etc. 1bey ate' also not invariably concomitant with their cognitions because:.
like univmals, the cognition of them follows from the PJ'CS(:fn of other &c-
tors, such as signs and mental effort. [PV3.6]
'fhinF;!' such as me hairs [that appear to a person with cataractS] are not
universals because they are not considered to be objects (by persons who act
upon mem]. This fauh does not ensue fo r absences because: they are appre-
hended as knowable.1 (PV].7]
The fault also does not ensue for thO$C hair-like appearances when they are
apprehended in mat fashion (i.e.• as objects by some other awarmess]. This
is so bccawc there is no reason to deny that they are apprehended as know-
able objects. The clarity of the appearance of hairs in cognition is due to the
" o...-..!...b 'dd'" (PVP,,"'b) In,,od ...... tN. -.... with .10.. obj.ectloo>,
"'You daim !ho,. WI;ven.J .. ma, oiwhid. one: ...... ""'5"itiun ..ba. then: ale od>c.
faaon .ha. do no!: ckpend on an object (ndM). Bul ~ awamlCSi of mint;< wen ..
~ halrs lhal a penon with caWICII paw •.,. OC£\Irs without dx prClella of JWn and
such mal art apabW ofldic: funcrion; rau.u. JUCh awarmeacs 0fXIU ...t.m ~ are
odIC' causes....m .. eoptiti..e dfon r.M.p). "ThaI bci"l lhc cue. u.o.r halr-<i...,.,p!
mIlS( be univasals. •
5 Onendrabuddhi (pvp:n6b):
Tbi"" JMdI. INti" p t _
"i_1I. Whf. &r..u, ~." _t i"";lIN. ,., HjNts.
..
Hown'tt, thoughts such as "ibcsc art ha..in" have univmals as their obj«ts;
but the appeannct ofha.irs does not have any object. (PY).8-9ab)
"If a univrnal is also a (reall object in terms of having the nature of aware-
neu, then you would have to conclude mat it is a particular:
6 De¥mdrabuddh i (PVP:U7a).
7 ~ (PVP:UN'
su- the ....i....w .J.o ... by AMUN ~ '<MIr. _....,.,. ....... to ;. .~.
Hence, dlCft iI 00 mntDdiction IrWrd by ow vppotlm. '. objtuio.. l. bh.. ,.j lIP N
*
If]it/" .. ,. 'i fb7ir ¥':Iiy..." ,..'" ti ...arJ." If]it/:fl""" ....,. Un IV i.J N ...,J J.I.
8 SumrnarUinl chc point~, ~buddhi (PYP:l17b1-l) ~
SJrw it. "-, ..,,.,. _ .... ' . ... it is chc ..............d is, it ia u.., _ bco;:r.... u.., roc-
n;lion 'ppeall in thai Whion by rim.e of ~nl upon chc adwion of ochn
objecu.l. _OO.mc..;1 ia prod\KlC'd throtJ8:h OM'ltxptric,1U (......, ... .. ~
vi <.I.., c- d .. 1cd ,...a <.I............... ;''11_ MlCb, i. io rn..-....uYalid oflha. ubjca:. In mil
war il ia defined u bod! , paniaIbt and I uniYUGl.
7. PV3.194-224 with &i«rionsftom PVPand p'fi
D~ to :II rdadon with other things ,i.e., ocher particles). infinitesimal par-
tida thot arc different thAn their own p~iou.. momcnu: arUo:: (from their
own previoU$ mommu luch thatthq can produtt an aw.uencuJ. In mal
sense, they are S1id to be •aggregated: and as such, they art said to be :II con-
dition for the production of awareness.) IPV). 19SI
Moreover, the distinctive quality thai particles obtain docs nOI OCOlr with-
Qut the Q, hc r panicle. ""j,h _hid> .h~ arc in proximity. H e nce. , ince
:lwarcnCSII docs not rn.vc any n~ rduio n 10 a single panicle. aware-
,.-,.n~ thW _ ..
in 11K phiIoIophical poNI>on, The corpw of ~ _r~'I" ha.-e 4ilw1el (."",
. Haw iI il apptIlpI'Wc. lhcn. roCOlKlCi •• oftt.o. *tpcs
.. ~ """t!.k thinF F.,.- <he .... <><~.bk V.......t.....dh ........ ·Ie .. II I,d.n... okCO<J"'" of
fivoe a~ h:.n .. din. objecu _ -panicubn auc:h as ~, which iI appre-
hcn6rd by me ocular faculCY;'1 ia noo:;wmcd tim dq!aU ~partinobn u
.heir objKU.' So ho-w iI ;. duo. their ob;.cu arc paniaI1an [in !he _ mQII[ by
DipIf;l and DharmUirril ~ n..:u objtctl an:;we uN,...1, ' lei. PSI.I-4. quoc'ns
AKBh on AKI_loJ_
2 ~ (1'VP:119b1I'l0l' dw dwr an: disrina from prmow: panic:ko in dw dq
1.r in prm:im.iry ro tad. othn: U I rnuh. thq pin I pUIf'lC1f)'- _"'1N f_ •• , ...Jwhich is
their c:apacity 10 prodUCJr ....... rmc.. Pn:.umabIy. pro:ocimiry to a patti •• r'. IItIIfC ~ry if;
abo intmded heK_1M ,dtnn. portion froon ~rabuddhi rndJ:
Ow • • ,..,.... tWJ,.u-~ d... to dK 1',...,,4 ofcoadiliona which ac:u. W
poopm,........""." which is me.ao.,..my TO prodUCJr an ~ ;"foUuJi- '
,.mnn ,hal haO't d... capacity 10 produa an ilWU'Cnca arUc from their out..Wllial
aol,U(l, namdr. pohiow inl'in,ltSimal panicles ('n die IO\II'IOe QlIlI;n\l\lllli m.. do 11<0(
ba.-e WI capaciry. 'The wonI "iIQ9cpted" ezpr_1hoM: panides dial bavoe dW
"'P«I'''' aopKitia Ina. arr .tt-inn! wbrn !My Ire in proximity with thil and """
od...r panidc.
,,.
APPENDIX OF TRA NSLATIONS 197
ness is said to luve a universal rin the sense of a group of aggregated parti-
des)J as iu obj«t.'IPV3.1961
"Even though they occur in thc same perceptual field, if they do not form
a new, diu inct substance. then those various panicles arc not observed
simultancously.-
, The word lOt univual it uh!fI""" bul thlI word an also man I "wholc· or I • compot-
ilc ctlQIY." lnaunuch II chc reWiorI of. L1llMnaI II) i~ panicuhn PI*' many of the J;Une
robk....... dw: rdul.on of. wt.o&c fO i.. pano. Dharnukinl coUap.co ~ of Of><' infO
me diKu.ion of anodla. Hue. DharmalUni RCOpIz rlw I ~lN ctluIY !NI
ai&o ICfIU1'te from ill COftS(irucnu wouJd mIOUlII 10 an admiWon of. kind of raJ unmr_
..I; Mnot. he UICS tfu, word "\IJ\iwm.a!" [i.e., "rompc:w.itectlo!)'"] 10 IWnC I ronporncmcof
paniaIIan withoul mala", any ontolopal c:onun;lmctll to lhc aiacnct of thai compc:w.itc
entity in diRincrion from the panicubn of...tUch ;1 iI rornpo;.cl ThU: point it clarified by
~i (PVP:II9b-I\lC»).
The obteaion th:u awareness OCCUI'5 quickly and hence one mistakenly
apprehends them as one entity has already been refuted [at PVJ.IH]. And
why would sesam... S«ds and so on that are falli ng down Kqu ... ntially not
b..- apprehended simuit2l1rousJy? Moreover. all cognirjolU art equal in dura-
tion. so why would some hav..- Kquential conceptual appeannca whik
others art simultal1<Ous? On... would b..- forced 10 conclud..- that th ... appre-
h... ruion of any obtea is non-sequential. rpV3.1')8-1991
And how could one 5cc: a variegated form such as a multicolor (ri".) but-
terfly?
Then that multicolor is even more psychedelic than that muiticolom! but-
terfly!" (PVpoo]
- In tho$c c:::a..scJ where one sea a single color and IlOt the multicolor. one is
jusl seeing the color that is a part lof me whole).·
If aft~ eliminating thr conuirul'. nr colors such as blue. you ca.n Ifill set'
$OEm: multicolor that is other dun thoK constitueru colol1, thf."n what you
6 ~ Dhumaklrti motb mit poRIa with • pby on wonb. The word rrw- mcaru bod-.
"muhicoloted" and, by awuioft. wondrow Of amuinS- To all(mpc 10 c:on~ a link of tho:
lIavor oldU:s p.yGn -w, 1ha..... ~ tho: laner~ .. "pIJdv:ddie." in!h.. K!IaC
~w 10 Amman wltuK.
For w~ posir thar Ihings are variow because cognirions are various; [and
when one sees a variegated or multicolored (eiml) object, the vuiegation
remarked in cognition must reAect , wricry of things that produa: that
cognition]'
What men would establish thai thi ngs are different? [PYp04]
1luc ;" dw objector cbims Wt 1M ani!)' whic.h it tho: whok 01 tl-w bunm\y and dw
di$rinct entity thac it dw pvc lll 01. mulricolorocl {dw.} carper and such ~ dw mui-
tic>oloml coLor 01 bor:h; u 1Udt. thq aft dw R:UOn f'« tho: metaphor fha, lIor con·
junaion is multicolored.
Onmdrabuddhi (PVP:I,u) continua:
Thil canOOl be dwcue bcawc thcR;. no muitirolomincM li.c., \'II'ilpriool in tl-w
color 011 multicolomi bunafly. and ixn'lM' thcR iI no multicolored endry in the
po.;nlinS and wch.
~i (M', "]W'wob) dari~
n,;, U1/Mlt N tIN ret _ _ rhrrr if .....~ j .. tIN rWw.j'• ..JM.I.mI
~ Thai iI, thcR iI no mukic>olomlnal [ i~, .,.,q.tionl in the aonc.qMualIy
.........,'-"""011 ~~ en';.,. Ilwl .. the;..t.ok (~....,.. .....;. " . . i.ta. ............ ;, _ ,.,OJ.i.-
..J.nJ t1IIi" j .. lIN ,.n"'"'r "",, ___ Thai io, .u- dwtt iI Iir$c of all no wftok, dM:
oppoilCllf doc:. 001 .,,;q>C I~, mulciobecl enti", thai iI tIw wboIe. RatIwr, only
tho: ooIonoi dw puul~ . 1lwy individ&W1ydo noc ~any muldw1otM·.! ...... and
...n.;.... diltina cokJn can/lOl. be dw bo.W ()ou .w, " .tip) lOr I meuphorical we 01
·m~_· bra" ... I"" opponmt doa noc aox:qx that on< an.appn::hcnd
many minp Rmul~y.
c..m.....n,i", nn ,.,~...rNirr.Nill.. o...-t:ndrabuddhi {PVP:19W nat ...)'I:
MOI Khtr , _ _ mip;hl daim mat pantinp and P.toeh...., mru.ph.oricilly calkd
"mukic:oloml" bcawc tht paru inbeu ill I si,.kconjunaioa, but dtil aJ"o doe. DOC
maltt' KIUt'. T'lu;1 is, t~ iI no conjunaion thai ia I multic:oloml mu", in th.e indio
.-idaW paru.1UCh as tIw coIorbllK, in ~of whidl rMn: itt said 10 be I conjuncDon.
lrucnd, thoK paru ha." thcir own e-ncc orlwina blue: and 10 on: u fIlCh, they....,
individually /101. multiooMN ed: !hc~~c, Met dwy aft DOC connccwd to lsinpoo;m:
(.,.,Jw). Ih<yc:an.nor be: dw obj«t {~of a metaphor.
10 Dn-cndrmuddhi (PVP: I,th):
APPENDIX O f TRAN SLATION S
.'"
"If sincularity is nor possibk in the case of objecu lsuch as a ooncrfly',
wing] mat have varitg:utd appeannct:S, thcn how ca.n ~ be a linglc cog-
nition whose cognitive appean.na is variegation?"" WVpoSJ
''Ibost who analyze reality makc a statemcnt that is cnt.liltd by real things
tharudvcs-namdy, th.:at thc way in which mey think of obj«ts is thc way
in which those objecu disappear.
n.u iJ. jwc N mmc:ol ~ Itnds 10 apprcbMd many ...0;..:1:>. ~ wbaI con·
uadicrion is that: if sauory awarmeA aIao appc:hcnda many ob;rcu [simulwocoudyll
Hma:. il is rIO( aI all ~ !OJ tmJOfy ~ 10 do to. T1UI is. ~ thouch
tbn.:- .. no Mnp: muiricoioled or nmpud ~ in dw ~ bluc and 1O DC! of ~
painti"" dw . - . . - char arita Mm a multieolooed or 'f"Irirped ~ u tin-
par. '....fo... " Ii"p . , btu ""rir", dw iJ. • sinp COSD;Uon chal has many
~~ t/· ' +Ii,.",lOoa:ur. Hma:.!hey aft' pn..d..,....I by maar nbjc:ca, on<'
c:on .." .... ......, onurwy ......... n ni, awlpwn.
Iu- I ·un"","",] · [in ..... ~
"'*' ••
UIm chal mpirionl occur wid! cop!iri"l' inuf;a. Henu, I ~ed irnast' iI OKII·
lioIIy d>c c.opicion ifJd{ (M. i ~ l1"''';' m..1 bri.. lbe (SR, you IODeJ'I
..... ""'1"1.......... I.....q,..ed .....: ...... ....,.........." [And ..nc.. ~
(,,)U) oommmaJ. "n..../ ..., it;, ~.~ M I1ICaIU dw d>c oo;.no.lhinla d>c
~ Ihr IftIDn .lIIed ito. the bm. 'dw whid. '" ....ncpnd don loot t.- • tin-
rJc _rial nature' '" irKondwi~ bcuusc: ~n rhoush CIDpI;lkIn .. of I "nplu
Ral\lft. il can ha"l' I ..,..;q.aud copiIM: apptaraIIOe."
ThiI cririciam appIiG 10 OM who maintains mal d>c ~ (_ ,.J" u1limau, bul
II doD nor apply 10 1M b«:auK 1 do nor ICIOrpI thai 1M I. . . . aiIu in !hat faahion.
Tn okmoruuau IhiI poinl, [DtwmaldniJ AJ'I ..•.
¢l FOUNDATIO NS OF DHAllMAKIRTI 'S PHILOS OPHV
I] ~ndnbuddhi (PVP",)&} maka it dear lhal the ones !loins the aru.lyr.illl hm: afI: the
Buddh.. H~ abo rnnatb:
... * . . .,
j ,. wIncI! *'Ibi... f{Hj«rr. dial ia, wbr;n on.: ntiolWly anaIflCI than as
cithn-lin&ular" multiple, thq tIisq~ Itt ckvoid of ainilll-in INI way,
i.e., in INI faahion lu either anpbr or multipkJ. In oWr wotdJ, thty annoc M
aabliohood in ,«.... of any CSKmw narw~ ~.
~buddhj (PVT, ~.) makes lhac oommma:
... 1M ....., u. tuIHd. "-1 Ibid f{NjNrs tmn fO awnaJ bilK. rdlow and JO .... H~
AYI "ob;ca~ (.m,.)to rdill~ the nocion!Nt ;1 ia disrlna fi-om ~ (lIijflhtt4)
iudf. M is noc rdilrinc the nocion rNl ~ ;. by !\aNU,.,..tnIr& This will
abo Mcxplaincd w«. T1w ....,i,...M "~1i"rmeans thq art noc osabIiohood
u rithtr .. ncuW" multiple. [Whm ~ndl'lobuddhi JaY'-1i,. _ f{.", mnoriM
_ " ....1JM¥r; M means tNl thc oO;rn ann(lf bot esablUhed as aluna!. nor a n
il bot _blUhed as !.m", thc ruNt<! of coruciousneM. In ochn- words. thq 3M noc
acablisMd as .pptarin, Iq)"Inldy (~. Mftp!). In the immfdi~I~ly fOllowin,
phrucl hr do:monHnla 1M <Q)' ill.....t.id. i. is noc rsnIJi·ha!..
Ikomwi.m..ddhi (PVP:I')I) oominua:
AcccpIilll mal dharmu haYf b«n .no...n 10 be ~ il is from dIe~.,."
of oWn thai hr IWa the opi...... of an ob;mor, "I, ~ "..~ #r""';'"
,.n- (ri~ u.. -1I(k t9pitiMr
~ (PVT, 11)0:10" ) ~
AJ for the statarw:nl tha.lqina. .0-.... IM_ Nt. sJ.- ,. N mtrttWn.. (oW nona·
iscma (.1IIIcM _ , .J is noc the: CSKO-:r'n,II."of dharnw ........ ,... ono: would bot
focud '11. c:ondu<k tha. theft" iI Mlm afIIinion nor purii'olio.. and bco:allK thc ton·
_.iona! .. ...tw.l. in ....... an ...... apIai.-l culm. t....u. tM ... u.- ~i ... odf ....U
czpLain mil tdA-, • ..m..to the mnt rdI~ awan:naf mal is dnoid of COOl'
ccplually ~cd objea and IUbjea.
~i (PYP:I,)I) ~"inuc£
"If dw:tt tOmdtow ~ ullimal~ varicption or mu1UoolorcdncM in I • aw--.
then in ulrimarc ICfIN dIU nriep.tcd 01" II>ultkolortd 1-.,." .... would aublish thaI
the raJ mi", was Wo vvicplcd or mul.icoWnd. Likewix. il would be ""'.- This ill
"1" ~"..J.,.. -n.-
...1... oJ", '1uoUoo'" ;. ,;nu,. ~,. 1.. '<>f>U'-' oJ"' ....a-
~n- .,.".-.' ..u_;,. .............. Si~:and YIritptiooImul~_
oonmwiinOf}'. If ~ <:IOJI1ition it 1lnguIu. it ultimaldy doa noI ha.,." l.-ariq;;ucd 01"
mullicolomi ftalW C; inl ncwnhcIcM appean with • r.uiepttd or mullicolomi cot"
nul...: ~. thm mal OOJIliu.,." ~ doca . - ultimaldy aiR in !hoM
c:opitioM. One mllll aa:qK this ptIIilMln bcaUIC otherwiJc" the .. ntuluitr of the
oosnition would bot CIOUnlnmandcd. Otbcr than a di~rmt or nor>dilfcnm oosniri...:
~. thm: it ulrimatdy roo od>« - . foc the aabiiahtMrll of lOtI"Iftbins as or.e or
llWly. In Ihis Rpfd. if ono: ~ to mainwn thaI CD&"'UOII " wnma.dy both one: llId
APPENDIX OF TRANSLATIONS
~;4-:------
variegated, then the whole variegated world'would be one single substance, and that
being the case, that position would incur flaws such as the fact that everything would
be produced simultaneously. Therefore, if it is singular, then it cannot have a plural
image.
tSakyabuddhi (PVT, nye:20Ia-b) comments:
[Devendrabuddhi says,] then in ultimate terms that variegated or multicolored awareness
would establish that the real thing was also variegated or multicolored. This means the fol-
lowing. {20Ib} An ultimately singular cognition arises with a variegated/multicolored
cognitive image due to that kind of object (artha). The external real thing that deter-
mines the cognition which apprehends that image is thereby established. Likewise--
as is the case with the singular cognition that has a variegated/multicolored cognitive
image-it would be real.
!He (20rb) also comments on Devendrabuddhi's phrase, !f (the cognition) is singular, it ulti-
imately does not have a variegated nature:
That is, it does not have a variegated essential nature because if its essence were varie-
gated, then it would be contradictory for it to be singular. Instead, the cognition
appears with a variegated cognitive image due to cognitive error.
On Devendrabuddhi's comment, therefore, ifit is singular, it cannot have a plural image, he
says (20Ib-202a):
If cognition is singular, then it cannot have a plural cognitive image; rather, cognition
{202a} just has the nature of mere reflexive awareness which is devoid of dualiry.
Devendrabuddhi (PVP:r93b) also comments on the last part of the verse:
!fone is content to have this be the objects' essence-that is, even though they do not have
that nature [of being external], they become apparent (gsal ba '" vyakta) in terms of that
nature; if ultimately abiding in that manner is their suchness, why should we bother
refuting it? The idea is that one should just let it be so.
Sakyabuddhi (PVT, nye:202a) comments:
"If there is no variegated external real thing, and if there is no singular cognition with
a variegated image, then how does cognition appear with the color of the external
object in a manner that is restricted to -a particular time and place?" In response to such
a qualm, [Deveridrabuddhi] says, !fone claims.... One speaks of an "object" due to the
imaginative apprehension of that which is by nature the cognizer's (sgrub pa po = prati-
pattr) cognitive error as being an object. Those appearances are not [actually] objects
because the constructed nature does not exist at all.
What is being asserted? [Devendrabuddhi] says, even though they do not have that
nature, they become apparent in terms of that nature. The phrase even though they do
not have that nature means "even though they do not have the nature of being exter-
nal." They become apparent-they appear-in terms of that nature-in terms of
being external. The phrase, ifultimately abiding in that manner is their suchness, means
the following. Appearing as having a nature that they do not have is the way that they
ultimately-really-abide because there is an instrumental cognition that establishes
that appearance. Therefore, that appearance is not unmistaken suchness. The idea
here is that since the cognition of them as external objects is contradicted by an instru-
mental cognition, that appearance of them as external is not suchness.
FOUNDATIONS OF DUAilMAKlIlTl 'S PHILO SOPHY
It i. o:oma that apel jellU of t:lUeiM objccu iI moncow; it if no! ..... Itd IG
m~
claim tha, m,,1 apujeha: ill no! M'OMOl& Btl! what iI coma ("iP,. - ~ li.~,
thr: bet dut the atemal ~ iI an ~l is noc pn::ocnud in dIM [1CIIIOf)'1waJe-
MIl iaelfl; M¥mhtbs. we rd'ul~ wtw is ~ted by f:.IK cona:pruaI COIftirion mal
is inMn.unmWly cootradicud. Hm«, as in the- caR ~ nonaislttl' thinp ~ as
hairs and rna ~ to 1 penon whotc C)"$ ~ impalmi by CIWKU, an atcmal
ob;«t. ~ rhousb nooc::tdttnt. appcan 10 t!loK w!loK qa all' ((I,,'ut by the
ClWXlII of iponna. Siner if is "PP'''P' inc 10 pretml this noOon in thU aIfltat, the
.... thor of tho: Ircati$e A)'I, "J_In;. H ...
I.notI.u ir.'UjlituDon (-,.tri- '" _.., ...volmil-oon is. foIIo.wI. ~
pl\nw if oJti-':! Ifbiii", i" "-r _Ii ...... ;s thm ...m- mans dw; bet dw thinp
whid! do no! ha~ chat IW.Uft appear 10 have Out naN" c:ome!I about throur;b an
inMrurnmw COI1Iil:ion. '" tum. iI is no! confuxd; inRe1d. it iI ..Jtitn.t~ ThaI is, the
bet tNotlhoK Ihings lICt~y aift at~ illhrir nonoolliOO_ MKhnaa. Hm«,
what would Ill' IIll' point in my rd'utilll I ralionally rupponed obtea by thl. W1'U<'&
idc.a.> Hena:, hu Wk.o. is: how an one.n- m,,1 whal is rc:uoNbk iI f:.IK [i.~. non·
..Jrim..JuJ?
I. DcYmdrabuddhi [PVP:I , )b) c:ommmu on _ 111:
BtIl .i1Klt thc ~ of the ob;«t., r.ocnitivc inuF is noc ttllUdtnca, ,hmt...,
...mbn tIN ,~"a .. -IUCh ar~ daimai 10 be the ob;m-_.....--which iI
daimod 10 Ill' the JUbj«t---N~ any Sf»Ody atmdai ~ in other word.. tIw
which Ippatl ..-ith 1 'fWiaDy at~ ~itiw i""", does noc aUt either aterNIJy
or inlunllly. ThaI is, dlal which appnn 10 be "",tiaDyUl~ndc-d mlLll be ~ithtt lin-
par .... plwaI. A ...ltfJ<IMmtiry c:anooc i!ppQI' 10 be 'f'1Iially _ndc-d boalUC spa-
tially almlkd Mngu1ariry ",.,.m-, ...... ~ lihar aIrady bem aplaincd that
u.. _.ity oN.. .........L:I Iuvoo .po.i.! _.......... nuneil'" _ .;nsul........ 1, ~"' .....
aill. n.., ~"" .f,.",..tr-U... • ..... tially atftldcd imap-irtd. OOC aiaunl
;/1 tIN ~" ..."., ir -II}'-namcly. color, .-hid! C(IlUi.su of inlinilClUoal partida.
T1u1 ~ die: .ti~ imagt; that apport ill awamICII 10 be 'I*;'lIy afCI!dcd does OOC
aiM in the: indmdo.oal infinir.simal pmidel. And if an awcpiled enury mac if ~
• an ob;ea doe. OOC M~ Iny Ii~~, with whit could awarrIMA Iw similar
[tueh ,hal it would alto han I linpalu ~l ~
.......
Sll:yalx.ddhi (PVT.
E~
,.:,ow), diKuainlit the nln1f1' of ~tni mliriel, malus theJe
ir ir.finil6i.maI purida: arc mixed with OOMr inlinitesimal panM:la, !My do ooc
10K that naum ofball! infmitaimal panidca. E¥Crt in ~tion, thty would by
iWUR bot infinitesimal panidtr. if auch if the caR. dw:n In ~ ..-hoK COSIIiUwor
inur;or it "P"tia1ly at~ caniKiI haw: iMnitailMl "","ides. its obj«t. 1lw bcift8
.he aaoe. il c:anooc be do:n:o-rninni by. noln--c",Meoua awatmeII bealUC. COIftiUon
(,..",...J IMI hal orw:oopU<iw: Unag::annol apprehaIdJOOWOOMr mine [with a dif-
laro. COSIIiliYi: ...... I. ~. one would 'neII' an ow:rn~
APPENDIX Of TRANSLATIONS
Some umkilkd pmom A,. tIx fOIlowins: IlxR ;. no ~Iwl,. almckd ilmF
~ ~~,. or ;n.~,., and it .. ....,. t1x..a;.a of ~ Ralhcr, lhe: objm:
of paa:plion it ilII infinitai.tNI puriek _toI.lIII.kd br other infinit$..w panidel
IUd! dIM !hey _ atm"lldy dole wi~ bans InlurupmI by an objm:. n- infin-
jlcarnal ~ ho .. (>'tI. arc nO! mind tw."H' mirin& iI ....,. p""ible in tilhcr of
tIx [WO ........hk arfWlCUM'llS: if they...., cunjoi.ncd II j ..... _ poUtl , tIxn tIx: pam..
del would llaw JWU and hma: almlion; and if thcy an: conjoined in their mtimy,
dwn tIx: ..... prd dump would i_ be !he IDe 01 an inlinimirnal panicX. ~
~. [tix obj«u of ptia:pUon an: infinilesU!W ~ WI arc] am:md,. dow 10
each other RKh tIw they an: noc inlUnlplCd by an,. hner..- p;wticX6 but dg
haw &It intmla bavoun than. Moreowr, IIx apparar1O£ of. ~Iully a:tmdrd
oopUlM: Urusc w. it I pndurilll conpommr.te .. o:nono'OUi tw..LIH' il ooWd nO!
" - nm I minute inlmUor [bc""cu,1hc aroma w. cotIIpotrd il; il ..... bern "'-n
-'--- -----'- . ..
~ "'""" &It InIClJClltt .. "'"
I"
'7 .
n- who a:prs dIiI qumcn. an: lOoIisb: I!1CdiIll their own words with their
_ _ _ , th.y ... ......fU.od. [fic ic .................. ~.,( _ 'F ri_'Iy_ended
~ irnIF in an ewarmc:N..ned fO appdlClod infin~ putidelit all _ ,
then how can me IwvmtII dw ..... W.I lind of ipltialIy extmded imat;c ukc inlini-
IUinu.l panicks • m ob;ccu.' An ~ mal .,. one thin& .. ill oopitiw imasc
ClnnOI boa." _hi",d.o: .. its ob;ta bea... othnwiw ...... would inau ilII own::I!_
1mIion. 1'his Iw already bern explained. No aosniri¥d,.
mropic (rJt- ... . , Iw) per'
_could ~a:osnlzc 6-r"""~f"Jan _ _ whoK~ imacI:-
an ~ ~ (~n.. Wt, infinitesimal ~an:nO!tlxub;«lIof
paapUun. ...... _...,. .... ~ utiofaal<C. c::.u..-,........ ~......., ut """""",,-
inlWieoi..w jMItideo. __ ......" in~1C it loy...,.. of *" i~ .cigI. whid> io Ihc
abKnot of tIx,wumat!Nt is tIx dMlloflhc aDqrd obja:lI]. Otha pmom main-
Din dwe¥a\ dw; n::sultinsCOl N• ..,.... iI wprodua o,.";oIan int.enW impriDl;
even ... the ~ would noc IU'f'C at: an inkrmtiaJ lip for tho.c imprinu •
-n.. AJthout;h one an claim thai ; 1 docs occur. il K euy 10 I « thai: the sparialJy
exwodrd oopUti." imat;c ...tudt ;.., i eN 10 be ...t..I one ClfA'iaoca doo:t ....,. ulti-
nwdy CIlia Iwa"., when one anaI,.- ~ il • iin«uJar or pIunI. il doe ....,. wilh-
lUnd tuc:h an anaIyaiI.. 1lv cttmWI Ntuft: ("I"' . . . . .,. I Mh.) of infinitaimal
poonido ia ClUJalod, rnncJI.~ io me ~iouat, dauibnl DliUlllD" . Bu. io <alII ..... j,.,
dmXd • .-:h, and thnc is no irwu1.uDcnQloopition thai ~it bcoo_....Lau,
I JfNIW/y extended inuF il ia linplar by virruc: ofbri"l panIca.
On !he other hand. one michl think. "Thm: it abo no inll.NmCnW c:opilion thai
pi"O"a it. 1lw;lCtO.C. it .. doubtful Thai brine the~. it hat not b«ro indubiably
dcmoowraK'd tha. doinp It(" in all _ xtResa. •
To dIiI_ respond as rouo-. E.-m po-...be m..ab: tueh daimt mUll admil
thai infinilclimal po.rtida ..... matmal tN"" (J.u nlf • ".;m.,) beaux if ;. ....,. poI-
Iibk lOr an imtnamW tJuns 10 ~ an ,tthNiesurW parodc. ~. wh,......,.u.t one
noc o:opUu m10d and mental fi.u>nioott. rnueriillif oornnhins".
il n«ftWiI,. Iw distina Ada; and KCton beaux thai which ia ~ 01 discina
maco:rial thi...
Iidcs and ttaOn can....,. be • nwo:rial ihi. . . . . tho: cut.;u, _ And ifiOIIK-
tN"I ..... diRina Iidcs and ttCtOn. thm il ..... par1$. n,UCfOi .. ill iI dw cut .nth spa-
tial exwuion, ..t.m OM analyxt: wt.nhtr infipitaimal J*tida an:.inp1ar or plunI,
they dg 001 withRand audt an anaIysiJ. 'fhncforc, infiniluimal partida I f ( alto con·
u.diatci by an irutrummaJ copiDoo. Thia Iw alto bem dart,. prewntrd by pow-
crfW thlnkcntuc:h at: Mallin Vlp ..... todhll; hmDt. CX'IcmaI . 1 1 do....,. aiR.
,,06 fO UNDATION S Of O HA1MAKIRTI'S I'HILOSOPHY
IS lX¥mdnbuddhi (1'VP:19)b):
· E..... thoup lhut is nodlUM object. 1M awwmetlf dul haf mal duafutic appear-
ana IIhinu.tdy pisu, •
This abo doa: tKII maU IftlK for ~ fdlowilll raION. 1W ;.~ _ _/..uiM
ma.tIJ the dcfinitivdy dctmnil>C'd PpCrimce or the Mibfea-ifN&'t which is infm131
(......,,, u., ttJ¥ .. ~and del:mnlnN (O Ill' a l l . rndt)'. This Is thr one
Iiom.....nich Ihr odIU upcct istlijJirnu; doac odoer upca is ~ oar di.Jt is_MiIJW
$WJ. Ib.I in ~ coruidrntiolu of childish bcinp i. _ mn-J. Whrther or noc
dlrlNi ob;«u em.:, rfIIl'ilion haa I duallUlWr. but i. doa: noc IIItim.trly makr
ImK kw I ain&k copition to haw two oopitive i~ beclwr the copiOOn would
no Ioosel M Jinpbr.
On ~othu hand. ir~appann« oliu objca:-imlFwen: noclcosniti¥Capprat-
ana. !hen il ............ no 1onF" br dw eo&nitiw ilnafl' of tbr COJ"ition'l ob;ctt. thaI
being the c:aw. OIK:mWd tKII Sly. " . . is ~ ~ oithat' just by dw men: bet
of ~ Hmoe, OM mIlS! aro:pt mal lM i....., '"oimilar 10 tbr c:or;nirion'l
ob;«t (.nh4j. MIJI'C'O"ft", Ihrre is no similu im.asc ocher than what is inl~ to cot!:-
nibon. to copition iud( ill wha, appcan {O copition.
Sikyabuddhi (PVT• .-"r.v:>)a) <:Oo\mrIIU on thc pIuaIc 'childish bcinp':
He A.,.' d!iJ,IiJ, M.". bcut-Ithc duali", of ob;«t and .ubjra) apvc:an in WI fuh-
ion only {O IhoK who are conf'llstd. 1M bodhisattvN who haw tft!iud lnal " - -
uc rdfIeN only ~ (. . . . vj4#mcre tdlaiw 'wam>cU (NOIf rit,. u", ..
_~~.....nich is dn>oid of dll&!ity.
~IX objecu: "If thcobjm: and IUbjea do IlOl aiK, then..nar would ~ left bur
rhe IUchI1C$l of lwarmat iudP. CopUliwdy rnyopi( bcinp do IlOl 0l"'. iutc:c any-
doin8 bu! doc: objc«iw and subj«ti~ cosni Lin ilNgCl. If they ....:n: to apericntt
IOII>nhinS dR, doty would..., .uchne.. l104I l llv.1 bcinS doc: cue. beinp would ~
cfFonJe.Iy libmutd. llv.r aochMII QIIIlOf Ix ddinirivdy da"erminnl doroup. infer-
ence. Fim of all. an infnmcc by way of an GSCfIlW plopu f)' KlYinS u mdcncc is
IlOI pouiblc in thar cue beg ..... its _nnal property hao: ~ 10 ~ proven. An infer-
mot abo could IlOI rome from an dI"oct teMnc Q cvMtrn(C bcao ..... thc pacxptions
WIlC)ft-pcrttpUons mal ..-ouId aablish thc norKiu:a1iry of IwumcA and doc c:aUAl
rebtion n my lOr an infduoa: from an dfu:I an: no< ....,.bI.iohed. Nor QII one han
the kind of proof of a c:aw.al relarion mal is ddinnl doroup. the .bscncx of doc dfcu
.....twn doc c:awe is ab.enr becl\~ rhe dfea is IlOl c:nablilobai {in 101M paralkl cue
tuch mal one could nexia iu abtmcx in thc cue in qualionl. Th.1 beill8 the c:aK,
if doc non-duaJ JUChnca of doc Yopcira ill an d'fKt. il c:annex ~ -..bHshed insllu-
mmWly. The dlUlinic appearance it no< an dfC'CC .1 all bec:> ..... il is Iikc the horru
of ........i •.•
To dW _ rapond Q foUows. n.- IWmlCIIlI b;a~ b«n made by rno.c who.
d\IC 10 docir scupidif)". do no< undrntand dlUlif)". Tha, "
with the word "oubjcct. we
do fOOl mean 10 cJIm' rdIa~ awaun~ imcmal cosnilion thai ariaa in '''It.
iow fOmu JUCh Q thc pkaAnl and the Wlplc:awu-RKh thai by apratins il with
Ibe term .tub;m:" we -..ld Ix ~yi", that il doa n(K cisI. btMr. we mean the /'01-
Iowin& Cosnirlv-r appeataIlCft JUCh Q bill( l«1l'i10 ~ alllfn;oiw IWlLmIGf, bul when
one anaIyza whnhtt IhoK appc:aranca an: tingubt or plwaI. doty aft wubIe w wido-
$<3Dd du., anaI)"b; Iocna. <hey aft IlOI fUChn .... Tlo.crdQn,. thac iI uhima,dy no
obi«< tha, is dictina from , , _ i~r, and ainot thai 00;- doa 100' ""iso, _ ..y
· theaub;ca <loa IlOl cziA·; in u yins this .... mean doc · lUbjca· mal occu... in 0.,..,..
sioouor concepu d\a( 1Uf~ (IV.... ~,. . p~ in dcpnlda>ccon
dw lappamHly t:lllemal ob;eaJ. II in "This is dw Ial enlif)" ("" N . ny.,hhu is the
1Ubj«f which apprdlmds mal object. which is doc rnl ena ry dou if COVOiza.· Since
an arml and ill pariml aft c::onmuctcd in dcpmdcna upon each other. mac IWO (i.r..
... b;ca Inc! objC'CC] aft poIi~cd in dependence on e.c:h orhcr. Thf apreaioo . ...b-
;ea. does IlOI apmlln<'tf ~ "warenas. which iI ,he _nlw narun: of qni-
...... iudt. n", .:.oKm;.) "".un:.,( "'5"-i, ;.,.. it ......in mung] ~~
on lI:IIIWthin8 dR bcao01K il aNa u su.c:h from ill own a wa. 1M CMUUW n.m'l"f
of COV'ilion is oab&haI in rncrc rdIczm: ,,-=>elL Since il iI devoid ofthc ~
dacribtd obj«r and subjra.;1 u Aid 10 be non-dw!.
AJ il Iw been Aid:
"Thr appcarana: of bI"", rdlow and JUdo .. if doty WftC c:ncnullO ~
ia no! rnI: 11041>1tMrd"orc,;, ia Wd tU. the"".emaI doa I0OI cziA ulli ......dy.
""J1... ..... if)' ...... i •• Iu- ' r ' i ..........! ... I-... ,,,-, "'S"'...... ;ft ...,! . ...... ..,. d....
[apparuody a!cnul bur aaw.lly unreal obj«r); ulUd!, if it Wo nex ultimatr.
Hmcx, il ia dd'inilM/y damTLined lIw. apc:lia~ is nono4.....• (. U"f '"
MfI . -.. JJn t.J J# ..J hJr;" liM .,..", g.~ .. Jill iJ". _ ';" ill #
fhl.. Jill..,.. liM NJ' _at'" "'1"1"
,.'" ,. Jill 1""1 ill ",u ".;,. ill ""'~ f
Jill u./rytJ,. ,. 'i 1ft' N J rit f'U
at ""*',
PJU "."J "It' ;o..~ f J.
. P"I
is apprehended as part!ess (eha medpa nyid kyi phyir= ana1[lfatvena) by reflexive aware-
ness, as a result of its connection with the seeds of error, that reflexive awareness does
not produce a subsequent definitive determination of the nature of cognition as non-
dual in the way that it has been perceived (rtogs pa ; pratipatti). Therefore, even though
reflexive awareness has already apprehended the non-dual nature of cognition, it is as
if it has not been apprehended.
But this in and of itself does not establish that the cognition is not a general [i.e.,
unqualified] subject of predication. That is, the qualities (khyad par; vise!a) that one
might wish to predicate of the subject "cognition" would be duality or non-duality.
Even though these might be in dispute, no reasonable person would be able to say that
since the reality of the distinct qualities (bye brag; bheda) under dispute are not yet
determined, the mere subject of predication itself is also not determined. That is, when
"sound" and such is established as a general subject of predication, one might argue
about whether or not it is momentary; in that case, one would then use an inference
to determine whether it is momentary.
"If one cannot say that the mere subject of predication is not established just because
the quality of being momentary and such is not established prior to the inference,
then how could any evidence be unteliable by way of not having an established locus
of predication? One would have to accept that a general subject is established in all
cases of inference."
If it were the case that one could say that, then when one had definitively deter-
mined the subject, one would also definitively determine the predicate that is under
dispute. That being the case, it would be pointless to seek out evidence that had the
three requisite characteristics. Therefore, in this context one has established through
perception that cognition (rtogs ; pratipatti) is a predicate that has the quality of
being pleasant and such. Nevertheless, due to one's cognitive error, one has not defin-
itively determined its non-dual nature. Thus, in order to establish that, one employs
an inference.
The proof statement is as follows. That entity that is contradictory to some prop-
erty-svabhiiva is devoid of that property-svabhiiva. For example, heat, which con-
tradicts the property-svabhiiva of being cold, is devoid by nature of being cold. The
essential nature of cognition contradicts the properry-svabhiiva of being dualistic in
the manner discussed above. {205a} The evidence used here is the perception of a
contradictory antecedent. Therefore, it remains the case that mere reflexive aware-
ness is what constitutes selfless things (dharma) that are devoid of the aforemen-
tioned dualiry; it is not, however, devoid in all respects. When the Prajiiiipiiramitii
and so on cites the refutation of things being essential singular or plural, and when
it refutes production and so on, it does so in terms of the constructed nature; it does
not do so in terms of the non-dual dependent nature. Hence, only awareness that
appears dualistically presents distinct qualities such as production; mere reflexive
awareness does not do so.
APPENDIX OF TRANSLATIONS
~ is me cue with persons who have cuaraco. m05e who art by narurc
confused by ignorance have cognitive presentations (IIijliAp,i) wim false
images that arise in dependence on their fcsp«tive conditions. fPV).l.17) '"
16 ~ (PVP:I',.a):
TIw Mjilfi';'" (1II'f'l.mti) .["""I' III t/;/for'ml iJ NwJ." d. Jifo'rttn ~ " -.
'TlIa1 is, it has. iu (2ute Iht ditfdalClt bUiiCQ. objca and AIbjca. The ditfatiltC and
noodifr",uo« of me copoilr...: ~ io the Q .... lOr d" Wonilion oi dunp at
difftrml and nondilfmm" If theft Wtte nDthins II aU, Ibm on wIuol buU woWd ruI
!hinp 1M: dilfMnl and not diffi:rmtll..iUwi.o:, d>e ~ ~I appta to !hiIII"
&ris&l and 10 on. That it, mrWdtt d>e caKwMrt- 1 ptiupUon mac ..... pmwr IM)t
praml Wti OCQUI; thai puapcioon io ..... dw ;1 apprd><nd.o d>e ClC>!l"iu.,,;....., oi
III objea {"'*J. AI dw tUne, d>e dojcai." cosnilion of tNI objca arixI; wbcn ;1
bteoob"dnooid of d>e appeanntt oi d>e ptiupcion oithal c:opiri¥C ilNf1. il;' aid
to Iuo." "a::ucd: Bul if bod! that which pcraioea ancI tNI wIUch is pticalo'f(\ do IM)t
cUt, lbm theft is no IX,apcioft 11111 ba:awe dom: II no CIOpil;'" ~ of obj«t
ancI aab;ea. 1lw IM:ins d>e C»e. if ia mo.u detomincd tNl aNal docs not Iuon per-
orption • ;rs - . .
17 Ikvmdtabuddhi (PVP:.94hl rrmaJb:
£Ym ifor>< ~ ~mUI ob;ecu. aU thinp (tit. .. J1M1'JIW) HiDhI"" no ddiniti<wt.
For iNDntt. !be Jmi1lniwJqilfin-.[.,.,'l us;' rhac thq_ "hape:I up": "'1M,
"1I'CpIC· " - d>e mranlns of ·~ heaped up: Abo d>e Imll " CXlaWoo" {1it]tJ
IItdW _ ~Iw me_ ol"thu.hlch is cnmded": "fundammcal dancnl"
(-"-' _ JIMN) .... the ICfIK of"-.ra:- of prodlKlion. • AU audI definitioN _ q.......
;fwd byaaMl)' {~ in m.lhtyare JIIi..mJ .,.mw'1- Aamry is abo. M1pU'
impoAUon; lintt III dUnp ~ devoid oflCli";ty. il doe. IM)t ulrimudy aiR.
18 Bach Dew:ndrabuddhi {PVP:I~ and M~tharw!din (P¥V: ,J ri,) /\Ott thai the
ipIontw:t- clled here it me
IlnaJHiaai iplontu. Slkyabuddhi (PVT. "J"' wsb) dd'inc!: it u
the "the imaplAu." domrmwlion of objca: and p,bjcn" ("1-""" 1M -. - ]i" 1M j ,.;
. . 1M ,.; P""K .. """ iQ:j" JMU ""'P" JMU . . . 1M .),
lkvcndnbuddhi {PVP:I~ nioea I ~u.I probkm:
4JO FOUNDATIONS OF OHARMAklllTI 'S PHILOSO PHY
11K: ultimate nature of the cognitive content (in perception) is not known
by any [ordinM)' beings) whose vision is not supreme; they do nO( know
that ultimate fiaNce b«:ause it is impos.sible fo r them to experiencc that
content without the elror (WplrrllA) of subject and object. (PVJ.118)
"If. o;osnitivc P"'*'nUlUon does ROC dcpmd on In amnal ob;ca:, then..-by doc. il
oaur wnh mpca fO. rw:d ~ioa...d 10 on ' "
11IiJ iJ .110 the problem raiK'd by vu ... bandh .... , the ~nnillJ of IW Vi~,;U (V.I;
u-r.}.8A). ~i o&n an anrwtt
Even ~ it doa noc drpo:nd on amnal obj«u. I mplirivf: pracntlrion doc:. not
oc.cur in<kpc»<kndr 101' cauKO ...... CIOCIoditionoJ, Rother , the p«>d...ctlvc _ oS. cos'
n;.iw: pnvnr.rinn it..",.,. ~ in dw, mmw conrin ... um. That IlnIlJ the aK. il
..... in..",.,. f>p«il\cl form in 1ClCXWdwith me ~ofrbe ....u.bkc:ond'rion& nw
is wtu. (DharmUJrti] india!CI lwith the phrut i. """"" "I.. 1M /Mit "".,mw_
JitNoul. h iI ROC the Oi( dw thi!: &h i.e.,~. pceKIllJltion Iw ftOlftd..
all; it iI jwt dw one canROC point it oo.n br SI)'i"" "11oit ill iu ..m' One cannot If'«-
ify the I«d in this fuhion Ma.\IJt w i«d Iw dw ~ <If the appa=Il .mibriry
(in me "., ':'101 <If ""_",,nU oil m.. mind.nd nwnw Kala.
S« ~i wvr, ''.I'C106b) for the bl3Chud phrata ruppliocd ben. ~
(.,pJ o;onunua:
"Irthe na~ oft:hc...nI cannoc br pointed OOt, lhen how doa. it m.tr'
If it ~ nonaistcnt, then sinor then: lIT no diffio,u>cu amon, tbt nonc&iJum,
all tbt con ...... tioftJ of birth, darh Ind JO on tNl arc pcreci¥cd of f'U'OII' wbo _ per-
otivcd in drums would ¥.o br ~ of!hoM: dream penonl when one _ - - . . .
jwr Ii thtr- pncrivcd ro apply ro doc pcrJOII..no .. pua:i+in, tboeedtntn brinp.
WIw difk,ulOX """,Id there br aIIlOrI8 thex nonaittmt pcr1'Onf! E+'UI utterly non-
aisl:UII "Jllilkl oudIli • burm woman 'l MIn wooId be born and dif, tuM: Ii.,..;m.
penon wbwc bini. and death an: .c:aprcd. Othnwisc, the penon ""'- birth and
death arc acupad would aI.o not cUe. Het>(C, if thm: ~ ftO...nI fOr m'Of, then
<lODtU..i copirivf: appc:aranca would ROC oa;lU.
S« alto ~i'l atmaM commma on du. Kerion wvr, ..,.,106I-b).
19 N Manonthanandin explains (.J riI.). an elephant', ryes;oK on the sides <If ill head:
.... n.:c, il an m..o... not to look If wtu.r iI occu.rrin, on one fide by .....pIy ibullinl one rye.
20 o.e..mdnlMddhi (PVP:,')60.-M) p.....~11 ."..."._jJJ....~
APPENDIX O F TRAN SLAT IO NS
..'
not be K'Cn las distiOC1 from the varitgarionl bccawe when analyzing it las
disciOC1) . one is focusing on the object (lin},.) [that producai the awarenc:u,
not the awareness iudO. [PVpl0J"
If the colors of a d oth and such also formed a simple or single enti[}" then
thq should not be analytically dininguishable from each other. And when
the analyzed parts arc diminattd, a remaining unanalyzabk whole is nO(
observed. (PVpnJI:!
And what is the contradiction if many [panicles) that have the special char-
acteristic (of producing awareness) when aggregated are not the awe of
aw;irencss individually, as is the case with the senses and such ~ IPVj.11JJ
And except fo r something being a cause, there is nothing else that could
coru:lifUie ,hal thi ng's being Ihe apprehended objra. Thai is. the apprr_
h~nded objta of "n awa r... nen i5 u id 10 be um in the image of which
awarrocss arises. (pVPl.4J
Concerning what has jWt been said. me "definitive determination-I [of the
evidence-subj«:t relation} is the common establishment (pNJitJdhi) through
either perception or inkrence of a qwliry of the subject serving as evidence
in the ... bjea ro be p........,n. Exa.nple. include the ckt-nmination thaI ..... 0'"'
is prescnc in a loau: or thac coruuuctedncss applies to sound. That is, a
person lim o:pc:rienccs through perception a smoke-pos5CSSing place whose:
nature-swbh.nw is distinct from everything else as be:ing unique.' In that
person who has perceived that place there $ubsequendy arisa a cognition
of the evidence (liti,a";jfl41J11.); that awareness is a mnemonic cognition
(smJIrtllm) whose: object is the difference [that distinguishes smotu: from
non-smoke) in KCOrd with wh.tr has bcc:n pc;rccivcd.
In this rqard. only the initial o:periencc, whose: object is a unique thing.
iJ an irulrumental cognition.J Wllen II thing of th2f kind has been aperi-
I Nap (HST::lO.,) dai..... ""'1 the Icnn ,,;x.,. is only bot:i", wed meapborically heK.
2 nor, pasl aai~ paf1icipk ~OIII/JrU!lIw bren inlnprttcd in an ~ _ to tXilialc
brnki"l IhllIonJ Sarukril JmIC1lDe InfO fWO .nona eng!WlImtaKa. Thb putidc: II wed
in ia pas! KIIK in the I"oUowins 1m(Q1(e•
.) Arcafa mnarks (HBT:I J.17):
nac is. the dcfin ilion of an iruuummt:ilOO&Ricion is lb.' ....... objta was IlOl ~
viowIy known and WI ..hole objta ill mearu!Of u:lie: N.naion (....kri]J). This
pmairu only co paa..... ua1 upcncn(.r. In Ibis ,.ro. the qualilicalion "the inilW"
Upl"","" lbot: fan mal thai apcrima: Iw an ob;«t lbat baa not bcocn prcwiowJy
known. And "..-hoIor; objta it Wliq\IC" ~ the I'a.:I char dial experiotnce Iw an
objoct mal io • meaN lOr • relie IUncrion .~ only • ~ an bot: • mc&nI lOr relie:
."
APPENDIX OF TRANSLATIONS ..,
eneed, there occurs a mnemonic cognilion thai represt:nts that thing's
uniquefldl with regard [0 this and den other thing in terms of whkh il is
unique. 11u1 mnemonic cognition represent! thai uniqueneu as a difftr*
enee from those other things; as such, mat mnemonic cognition has as its
Obj«ltM thing's adusion from thaI which it is not. This mnemonic cog*
nilion arises by the force of that perception. Since it apprehends a cogni*
ti~ im:l&C of the observed thing in the manner in which it was observed.
that mnemonic cognition is niX ilUtrumenw. This is so because, one hav-
ing first SCCJI a unique thing, the subsequent mnemonic awarencu repre-
5tnts the image gc~r.llled in poception :as Munique M ; I't:p~cing the object
in that fashion, dut mnemonic awareness has no cognition of an object that
was not prcviowly cognized. Also that mnemonic cognition is not inmu*
mental because. since the previous perception o~ the ~rity Ihn iJ
the means for {the ~peoedJ tdie funct ion, there is no cognition through
thai c:onceprual awarenas of II prcvlowJy unseen particular that is by nature
the means for m.u (die funaion.:as is the case with inference} u "1
The reason for this is:as foUows. Setking some relie funaion . a judicious
person invenigatcs what is :and is not :an instrumental cognition. And a
univ~1 iJ nm c:a[Uhloe nf 2oI'XI)mrliJhing :ony IdOl :01 :011. for it iJ wMr iJ
apprehendc.-d by a conceptual aw;uencsi thaI. after the pcruption of:a par_
ticular, arises by the force of that perttption. An cxatnple is the concept,
"blue," thaI appears in the conapcual cognition of a person immediately
after he h:as.s«n blue. For il is only thai blue-particular thaI pcrfonns that
kind of tdic function. and it is seen with that very naruf( by perception.
However, the object of the conccprual c.opiition, ~ blue, ~ that:arises after the
apprehemOon or thae putic uhr docs noc poertonn the rd.k fWKtOon which
mould be accomplished by the color blue.' Ther~fore, ~n though it is
said, ~ an instrumental cognition has as irs obj«1 a thing that has not yet
been cognized: one should add the qualification, -in casc the particular has
not been cogniud,· But if the pan icular has been cognized, the conccprual
cognition that arises by Force of that previous cognition in correspondence
to the image lof the obj«1 in perception] is only a mnemonic cognition
because that [perceived thing) is irs object only in tenns of that conceptual
cognition's df«1 [which is to induce action directed toward the particubr].•
!u such, that mnemonic cognition is not instrumental because: [II there is
no cognition of a real thing that has nOI yet been cognized; and (1) the
determination of a cognition as being instrumental is based upon rcaI
things. This latter reason is the casc becawe the activity of persons who are
intent on the [desired] rd ic function has as irs object a thing that is capa·
ble of that tdie function. and a real thing is dchned as that which is capa·
ble of lelic function. A further reaJOn why this mnemonic, conceptual
cognition is not instrumental is aOO mat one acrs upon a real thing through
that conttprual cognition by imagining irs obj«1 lo be the panicular, (The
conceprual cognition uill guides one to the particular) because when one
acu, Ithe obj«1ion of the conceprua1 cognition] has a coherence and con-
tinuity that is not different from Iwt previous object of] perception!
"7
418 FOUNDATIONS OF DHARMAKIRTI'S PHILOSOPHY
IDh.armoIt2l'll]. (19912) Rip pll'; Ih;p JUI'; "D" eMr fre/ pll f· NJlJII-
bitulM!iIt4j. Tr2IlsLated by Jftinagarbh2 and Dh.armilob.. Edited by
Sum.atUd" i 2nd Bio ld.an shes r.ab. Bst4n (rarr, Sde d~ edition. Tsh..J
mil, Vol. Wt, }6b--9u. Reproduced in Buber, Vol. 49.
(DhumonanJ . ('99Ib) T,hAJ mil nllZm p.r n!n JUlI in/ b,btuJ f-
PrIlm4!U'villikilJ'l.tiltil). Tr2IlJLated by SumacikIrti and Blo Idan shes
rab. Bstltn Dur. Sde ~ edition. Tsh4J ntA, Vol. TSM, 178a-195a. 2nd
Vol. Du. Reproduced in Barber, Vol. .s.
Dignaga. ( 1 9SS~ lb) T,1HuJ mil /tun Lv jnu JHI 'j 'rrd JHI ,. Pnmul!fIIUmur-
uJIIIJfffi/. Translarcd by Kanwvarman and D.ad pa'j shes rab. &un
pr, Peking edition. T,hAJ ntA, Vol. Ct. Reproduced in Suzuki
(r9ss-6r).
Dignag:a. h99lb) T1Juul nul JlIm 1m Imu pill' 'trtl pll l_ Prllmli~mut
CIIJIIIJ!N;J. Tnnslated by Vasudhmrai4ita iU\d Se:ng rgyaI. &tAn Dur,
Sde dgt edition. TslttuJ m4, Vol. 0, 1..,a--8Sb. Ikproduccd in Barber,
Vol. 46.
(GautamaJ . (1912.- 19) Tht NpJ" SUmu bfGIIUtAnuI ",ith 1M BIuJuII bf
V4tsyoiJ11N111M tht Vi.rttikll sfUJdyol4lulrll. Translated by Gailginatha
Jhli. Replilll edition. S Volt. Delhi; Motilod Ban...ai..b.t.5, 1984·
Gautama. (198,) NJllyiIJiitriIJ. Reproduced in Nyiya-Tarkatinha and
T arkadnha (198s).
UMnagarbhaJ. (1987) SatytJdlNlJilvibIM,;:1I. Edited iU\d translated in Eckd
(1987).
Jliinu rimitra. (1987) ApolulprllkllTII!/1l. In jiilinlliri",itrilnibllntihillil/i.
Edited by AniU\talaI Th:lkur. Tibttan Sanskrit Wooo Series s. Parna:
Kashi PrasadJayaswal Research Institute.
KamaWila. (1968) Tht TlltllllUltlflr;r""" bfAciryll Sinumt1!itll w;th IN Com-
mnrtllry Plllfjik4 of~ri /GrmaWt"" Edited by Srimi Dvirilcidisa Sistri.
Bauddha BIar.nj I . V~a.d : &yJdlllI Bhir,lIi.
83hdingk, Ono. and Rudolph Roth. h8S1) Simsml- WirkrbNth. St. Pcttn-
burg: Ka~rlich~ Abd~mi~ der Wis..stnsch:lft~n . 6 vels. Reprinl ~d
Ddhi: Motila! Banarsidass, 1991.
Bronkhotst, Johannes. (1m) LAnlil~ tt rlali": INr un Ipiwb tit 14 pnuit
indi",1I,. BibliOlhequ~ d~ I' £col~ des HaUies eludes, Sci~nces
Rdigieuscs 10,;. Turnhom: Brepols.
BUlUnbergtt, K1aw. (1996) -On Doubting What There is NOt: The D0c-
trine of Doubt and m~ Reference ofT~rms in Indian G r:unmar, Logic
and Philosophy of Languagt." JIP l.4: }6}- ..06.
Cabcz6n, Jose! Ignacio. (I99S) · Buddhist Studies as a Disciplin~ and Ihe
Rol~ of Theory." J IARS 18: 1}1- 168.
Dreyfus. Georges. (100}) Tht SoMn4 tifT_ Hflnds Cl4pppinl: tlK EJUCIf-
tion ofII TibttJIn BwlJhist MDII!. Berkeley: Universiry or California
p<=.
Dunne. John O. (1996) M
Thoughtlcss Buddha, Passionate Buddha." jAAR
6~ 515-SS6.
BIBLIOGRAPHY 417
dcril lruu~riab.
Vol. 1: German Tnnslarion. BKGA 7. Vienna: Oncr-
~ichisch~ Akademi~ dec Wwmschaft~n.
Sille. Jonathan A_ (1.001.) ~ Pouible Indian Soun::es for the T elm Tshati 1f'UI 'j
sltyn bu as PrIlm4'!Y,"n4J1l ~ JIP 30: 111- 160.
Skinner, Quentin. (1969) "Meaning 2nd Undcnanding in me HistOry of
Ideas." Hist0'1.nt1 Thto'18: }-S}.
Sosa, Ernest. (I99S) ~Emn ce_ " In Tht Oxford CA""p,,,,io,, to Phi!4soph].
Edited by T. Honderich. Oxford: Oxford Univenity Press.
Stcinkdlner. Ernst. editor. (19913) Stwiin j" tIN BwJJhist EpistnnoUJticlfl
Trllliihon: Pro«tdinfJ Df tIN SmmJ bUtmlfh01Uli Dhtt1'11lilkirti CD"ftr-
mu: V'WI'M, lunt 11- 14. 1919. Vinlna: OstemichUche Akademie der
Wwensdu.ften.
Steinkdlncr. Ernst_(1968-69) ~ Die Entwiclclung des kpnibrvinuminam
bO Dh:arm.2l1i"i ," WZKS I t.-I} : }61- }n .
Sieinkdlner. ErnS(. (197 1) ·Wirklichkcit und Begriff bei Dharmakini."
WZKS IS: 179-1.11.
Steinkdlner. Ernst. (1974) ~On me Inletpl'Ctuion of the SlJIIbh4v.hnIl."
WZKS 18: 117-11.9.
Stc:inkdlner. Errut. (1981) ihe Spirirual Pbce of me Epistemological Tra-
dition in Buddhism." MmtD Bd1ty649: 1- 18.
Steinkdlncr. Ernst. (198.4) "SWlbhdJJtlprwtib.Nlha Again." An..lnJof4tjar 6:
4S7-47 6 .
Tille:maru. Tom J.F. (1981) "The ' Neither One nor Many' Atgumcnl for
~[inyata and its Tibcun Interpretations: Background Information and
Source: Materials." [twin tk Lmrn j. LaUWInc: University of lau-
sanne:: 10J-llti.
lil1e:mans, Tom J.F. (I98J) Ihe: ' Neimer One: nor Many' Argument for
~iinyati and irs Tibetan InterpretatioN" Omtri/"numJ fin Ti/JnA".'"
BllliJhist RtJipon .nJ Phil#uphy. Edited by E. Sreinke:llner and H.
Tauscher. WSTB II . Vic:nn:a: Arbc:ialuds Rlr Ttbe:tische: und Buddhis-
(isch~ Srudir:n: )OS-)lO.
A
.MlN.. ~ nonexmml referent Abhidhmna typOlogy. V=Sl. ~
Abh~ (Vuul».ndhu). ~ z! 81. 109087
s8nl1• •tis n}7 UIn,* 5« inll:ntion
"".t.o AMiJIM_hIMo< II~, Stt tu.biruuion
AMiJh.mwko"'~ absux cion
(V"n.~ndhv) u applyinr; 10 both luh;ect :one!
on a[>~heJuion of mliry, precl ia~. Is6nl}, H9
19 10l1S and causal charactttisria.,
on bcginningkasncss. I n! 17!=7}
on congIomer.ued partida. and COfI(leJM formation, 197--99
79 o }1 of propenies. 2i! ~ li l., 119018,
on lunna as uaruem pirical, 100-101, Hs-..9
l)OnI} and JI...bhi~ 110,
on momcnrarinuo., 80<0" .11)=11 . )06
on prim..,. cause' and supporting abstract ptcdicah~1
condition, 16S0}7 as not baring qualities, z! !!Z
Sauuintika theories in. ll1n.I2., :u not diJtina. from particulan. !L
59"'}. 79n}!, 8onJ9 !Jb IS901l, 171-7J, ~
on 5eruC pclaoption, 14ftl}. 196nl Sn .ts. abstraction
on ICnR Iphcn IMnicuJan. 79n}8 accidcnta1/cucnfial diRinctxm ,
:as JOUf'Ce for AbhkUwma 11111-91, 100, 1111-11
fYPOloIy. ,8nll Sn MJs. nccasity
1WU.t,.~ and MmA1f]"U",!", in, :Kcidcnal properties., il
"- action (nTytl)
on noo tttlitics. 40=41. IS! equated wilh wbj«five aspea or
ultimate raliry in, !l::::h awareness, 171
A~~phll.lilnhtt~hytl in U,."g 'fI'u:m. !Z=!!z lQ, 161,
(ydomitra). IZ!!!!
.'"
."
+40 FOUNDATI ONS OF D HARMAKIRTI 'S PHILOSOPHY
",
con«pnw a~ as tomning
6tyiU1U,.rWq.... Sn tramoempirical
obj<a
w poueu. ' ,I-o. }4I . }47-...s authorial intent, :i=Z
AI criterion for knowkdgc. til. 6111U/hi, S«ddimittr
Is6, )11- 11, )'6 . .. I) 61VJ6vin. Sn whole.
as critmon for thc real, .Sft64, 6tN¥ Sn ignot'ancc
-,§-n. ' I ~I . ')-&6. 1]0n04} . 6ttiNl!biIl/l,.iyfttu. Str ruk of unx·
19Jft7 ... }]S. ~ .. 1.4 companied non-aming
and Episcm-oic ldeilism. j80 6viu",1JtIJ... Sn tRlnwonhinC$l
apccted (,lIlhi_uin""!rriJl). !i.. 6l1Jilbhidr&. Strcvidcncc. rdiabiliry
~ U]--' 9 , !H, ~ 116, 1,]nl.4, of
~ ~ ~ 101, 111, l S6. 6~tII,r.ml"..phJ& Sn irutru-
"""
W .... S« exUlenl !hing
ll'
inoomplete (viJ:aJi MtNM1NItri).
Bhinvivtb, Inl 16sn}8
~WIIl",.. &>t wnmcss. of narnre-nwbh.illtf and. 161~. 170.
diffaena 17Sn.ti. !ilr 199. lQl
Mml"ti .s«
ermr participation of aggrq;ucd aloau
Biudeau, MadeI.n ...., .8n. o, 91n'., in. 103-) . 10!. 108n. <>o
.
10ln74
,,,,
&Jhk"ryi...lJln (Sintickva), i§.,
."""'"
aUlhori!)' of, I}Ml
S« .Js.definilive detamination
caulion (";MIII). 86nn. i.L 97n68
as ignoring the ultimare 10 Ic:ach S« .~ momentariness
brillgJo 110 41U orr.. S« nrkFlion
as not employing reuonint- dass-sign (j4ri). St-t univeruJ
l SO-SI cognilive im38C (~T'" ,rni/,;mlN.;
pragmalism of. }14-1 ~ ,,..tiMb.; also • appo:arancc·;
as kring the: trarucmpirical, )OS • conc:qxual image 0)
buddhahood. ~ 4S-1l as hom rnI and un real. 2Q, LI.6.
Butun~lgtl . KlaUl, 161\4, )6n47 119=11
and conceptual ilhl.lion. ~
C construed in Iemu; of I ncption,
Cabtt6n, JOK, ,)n4 94/'64. 116-+4. }40
Candralcini, rr. )61l.4i, S4n}. )16n6 correspondence: with pmkubn.
capacil)' (,JIotIt4; /Ilk,;) 100-111. u.a
of aggug:alcd partKks 10 produce ermneousl y coruuucd as oun-
a....artnall. 10M. .J96nl mental objccu, M. 141 41.
of caUJal compln:. 16)n)4, 116n}, 177n9}, l iS, ) 4. ).40,
~n l OO ,..6-47.)96. 401- 14
of iruuumallal cognition 10 make and wor in conceprua.l tOp'Iilion,
one altain a goaI. l.86-I7, 111, t!. 61l117.18)nl07. )11. HMl
"...,
!!L }I}.}SI ill irurrnm(nt oflmowledge, 1710
of iruuumtnial tOp'Iilion to
pmdua. ddinitivc .. m~nl.1 p&niculu. t!!t 116- 17.
dcielmin.:uioC'\. 194. )v. )7Ift!4 !!2: UL ) 11
INDEX .,
and pc.cePlu.aI illusion, •.1~ p,.,,~). S« pcrttption, col"lCq)-
as produccd Ihroogh panKulm.. nul
........,. u. U:2:. 1611 co ndWiion (IJipmA....... also
5t't Jw awarcnc:u: conapu; "5ummation1. )4tI.4S. )S0-4S
objcaivc image: subj«rivc imag.: conconuuncc
~nl:aria infm::nccs involving only nepli.....
reuoru for rdiulce on, 1:1 (1trWI~"",tirrh";. H0-4S. !:Q:i=:!
ccr.ai:Jb~n8 or tno:!i.ion:al, 7 - 11 , inr,,",_ involvi"s on.,. pot.ii;....
lP:- ,:}, ) 19-10 (1trWI14..,.vi"J. JSR.4S
$I)'k of rcuanill! in, 1::L )10 need for cxampla 10 (SubIWt.
commonsense: objccu. 6t-79 19=11. UYi 8
companli~ philosophical nqlli"" ("",nrrU). 11-19. )1OJ9.
roda,VOi'$. 117- 18 '48nll. ~ 1u01 04
mmpassion. S4n), 66.. lu- J6, }19 ontolopcr.l Nsis lOr, j6n67
mmpktc cr.URImmpkl.: (JllUU poiilM: (.IJ"""~), 1I-1,}, B
~-zri). S« cr.wa1 (OII\pC:.
compklc
,I. l!
as uansblion c:hoicc, lBnJ6
moapu S« Jw rndcnce-plcWalC rdation
as Cfroncow, it 61n17. !..!2. conditioning (-bh]i.uJ. S« habiiw,
L+O-4 J, ~ JOj, jll=l-4 ,~
,,,...
adudcd {1IJl",""J. Ill-H. Sec also Ilnh.kriJtl. apccttd
apec1td wic efficacy
:as mere ezduding of other (.bhiWW<irth.kriJtlJ. Sec
(.1fJtI/!dM-J, Hl-U, IU-J6 ttn~ apcctcd
~ru~IWJ,lMIM u being. l.Il1:1 ap<n<n«
narutt-moIW""" U wamlnr for ugumcnlS from , lL ~
construaing. IS9-61 ;ppcal IO. lb 11:4-2.1
as negation, & !.!i. !..!l. ~ initi31 alone as inJlTUmcnw ,
1)1-". IU46 }Ol-J. )12., Jil-IJ
as neither permanent nor ;and karma, )2.,-16
imprrmancnr. nlO-lO ROnm~ual. )OS-1
:as ob;ca of apres&ions, 9-10, of things as [hey truly lI.fC fJluhln.
118n101. !l2t!J!, 1]7=", }5)-60 ~~~h2.1},}66. )67D16
proptny or ptcdiatc u cquivaknl as .ubordinalc to InJOn. 2.S1
10. IS6-v a lcrnal obtecu. Srt ClIU1I·mmw
u qualifying univmab.. !.Ql. objccu
"7=19. I~J. ~ JJ9'-Sl. J9S Extcrnal Rcaliun (~)
role of .........c:Ul>ON in clacriplion 0(, 18--19. ft:::§2
COnsmKting. 118-19. ~ 156, Jll dM::rgenl intaprctaooru 0(, 6?-Z9
fOUN DATI ONS Of DHA1MAKIRTI'S PHILOSOPHY
Frauwallner, Erich, 1m
)4J--47. 401-14
~ or oooc:xisteoo: as irrek- Funayama, Toru, 87ns+ u6n)
YlItll to trwcwonhincss, }Io
as infinitesimal panidcs, 2!: !Q9., G
l"1NIU C indK:alor1· Strevidcna:
'"
as obj«t of ncption at f"ounh
kvd or anaIysi$, sk,. ~ ~
,6"'J1' n ndiaared j , S«plediaatc 10
be proven
66. 1'9n)8, 8SnSI, UO-lL Gancri, Jonardon. ISnl
as 5patially c:lItcndcd, 6orz9. Gauwna, !L !.i" um8. 13n u.
98=100 ~
exttcmdy temO'(c object Gillon. Brendan. ~ 1' lm7, ISu1I8.
("",~. s.. 189n67, 1, 1"""'91, 108nl00
Inrucmpirical object gill (.rrIM), .w ptupoJC
(:ZuUuic iruuumtnuJi'Y r,.nub God. ..
prbNi'!Jl'm) . .w irunumenWiry, Go rams pa bSod IUmI ten8 ge.
=N<
''''''
trilNlUUrll. .w MJbjective image
F Vllh]tUinl. S«objecrive image
f.ilic appannar;. 1S7-Si. J91 cnmmatica.l rdations.. 171,,",
.w .Js. aTOr, illusion; Ipwlow 17JA4.4, )S7n9
pcrtqXion S« .Is. pcedialc-aptalion;
hlK dnnminarion r.Jhp~, Sa.nsk.rir gnmrnatK:aitradirioo;
!!l. ) IU\J 7S. )46, 414"17 MJbject-oprcssion
.w 11M impulalion Griffiths. Paul, !..L )6~
f"onnaliun. !1. l!! 1Z G}'1Iuo. LobAng. ltl)
formal • . 5« Fonn&lilm
foundalionalism . )n6. )1)-4. )16
INDEX ..,
H HfflI.i"JM/iltJ(Arap),II1L4J.
habituation (.~) 411111, 411ft) . 40% 41-406, 4l-4J17
10 coum~nu of Raws,}6f IN,..,.".""... SN causal compkx
and inuinsk inwwncnta.liry of hierarchy ol views. St'i kvds of
pcrccpoon, 190-91, 1514""91. ...Jym
J74l14> }n nn.. 378ft!4 highest bd. of ta..l ity. &r ultimate
role in conapl formalion. IslnlS. raJif}'
,., hornologoua i....aN'" (~, ~
mk in puapnW judgment• .L j lnl!" ' 4\1014
linn. ~ IlS. ,14. jl6 Hoy. David, 7nl)
Hadcin&o lan, I nlO human aim "'M~' also
Halbbs.. Wilhelm. tIn.f) . dn.n 'h~go>I'
Hallilq. Charles, 11110.11 II focus of uealises wonhy of
happiDQI, 4h1. 1" invatigation. )61
Harvey. Van, lJ0n40 insnumc:nwiry and , ~7' 1U,
HaltOn. MuaaJc.i, lJOll. ~76. ) 11, )1', ) 14, )80
101 n.71. l,on6, 1}4n16 ~olasmmof
Hayti. Richard credibility, "S. )66n4
on ac:cidm.u.1 and (:$$Cntial ptopcr- as maw.!Cd iR$tl'Ul'lW:nw dfca,
ties, 119n67 1&1, J6~ , In, 179, )i}-8.t
on Mi.-and /fMMi .... IUm9 as ~uirinl dctenninate COntent,
on circularity ofDha.rmU.ini', 16nl, 191 )00, )0)
theory of iruuummWity. Inrus hungry gbo6t. See ,rru
on Dhumakini's prwc, j
on Dignip's UJC:$S1n(fl1 of 1
language. ~ idcntifiation. 191 , 100
on inductive UlWTIpc:aon in identiry. modc of the
DisniP" !houshl. ,,.ann 1H~"tuIh&. Stt
on inkreno: of "'pacity through swu.w,,.,ib.,,JJ,., identity.
nwMiv.t-cvidena:, 108nl00 mod< of
on inYUl'ion of puvada-pcrvadcd ignorance (Il~)
rebdon , U}l\I07 as aUK of subj«tlobject dualiry,
on llansLuion of tih.,.",;" and 1!l }IS. }4. ¥:tn..,. ¢9
JJ,."".as qlolality-pc r.:.or and .. aute ol.wreri,... W'l, 6J.
qu.ality, U .. concqxu.aliry, 61"-'7. ll2» 4
on Ifsnwrw... of equa~ with M~np. sI nn..
lW"'-wpnti.."JJ,. u ru.tunl 60-61. )71-7)
rebtion. ISlnl7 5ff.ls. internal cful:otrion
Htgd. G.W.F.. 1lt4 illusion. }6n,.1, 17'=90. Jill
hcrmc::nrutia of charity. 1L 1)9014 SN .Is. error; illUlOfY ob;«u,
bmne:nnuia of wspicion. lSO ilIUJOry ob;ecu. U 90. lJ7-SI, ln:,
hect~ut inm.nI;o: (vi~
,ft...
l!!J ) I f t " , .... 1.Ojn,l, ~ ..,8
hm.. SN evidena;
'"""
450 FOUNDATiO NS OF DHARMAK IIUI 'S PHILOSOPHY
......,.,"...,
as including oc:hcr insnumcnu of unmediated (.",.utJnu), 16 1=61,
knowledge. li b ?
instrumen talj", of. lfl,-jl4, )17, iNtn.llncnwity (pri~
)7.....,o. 4 1:a.-IS definition of. 1f)-S6
invomng only ncpl M!: cxuUuic (,.,.."'" lrimll!'1""')'
c:oncomitancJe (1m.J..IIJ<Uirtiilt). lSI, 191-97. )11 . )"snI4
HII.4S. ~ immediacy 15 3 marlt of. 111
involving only positn-c ofinfermcc, )10=14
a;mcomit;ll}(r (i~"t.tqi,,), intrinsic (~ ,,411f1l?lJlfM), 151,
3SII.4S 191=9']. j11- 16
lanJ'AlC. prtlumption. and and n(lYcity, l S4- SS, 1~
analogy as fonnf of, 14 r 4 7 of pUCC}Hion. lI7-J09
Wftd notioN; of, It=}t. 1! and pwpc»C, 1)9-)1
!W.Is. iruuumcntality; ~m....: ddlnition of, l1H
ilWrumenlJ of koowkdge; of Kripturc. 1)1- )), l)HS
nvJJw".",..riIMNiIM xemins circubrlry of. l JJ-S1
infemJ«-for-ooadf in lemu of the mediated d&a,
INDEX
'"
161-61.171-411. Jh-ls inl(flc:ztuaiiry, 1..16:. Sl
in W11U of the unmedi21ed dfc.ct. intrituic ilUtrummr;Uiry (,....ub
161-61, 168-71. ",....as '''~'''). Sit itulTUmCntalily.
at tl'UUbtion for ,'I"u~ intrinsic
12)-19 irreducibility
Sit .u. tfW(WQrUti0C$5 and COmffiOnKnK ob;ccu. 1 1-14
inwumenw object r,,..-,.; ofinfinitoimal p2rtidcs. !:i. 5b 51
snd tM rnl. 79=-14 . '4)=+4 , ~ of pan>culus. 2!
>8, of the uhimatdy real. ]!, ib !Za
shared ooOonl mncrming. 18-11,. §!, Z!! n.. !h if
lHS. ~ 1! Sit II/u reduction; miuctiY'( an.aIy.
instrwncnt ofknowkdgc (prw~. W
abo ~iNU\lmental cognition 1 fvuascna, .4Bmo. 149nl), 106n98
convenrion.al. 18)nI01. )14-18. lwara. Tabhi, lO4l!fl. !!!:2...
J~- 1 7 UII\I04, U}I\I I77. llJ-16 . llln u j
ddinition of. 1S4--,:6
and insmlnwn w dfc.ct J
(p,.."up*,n.J..). ~1 JartlC$, William. u.I
bngu2lgt. ptcsumption :and jlri (cbss-si&n). &runivttW
anaIos7 as. 14HZ Jayatillch. J.N .• 111'16
U!nOSI p~t f.ad:or in Jh.a, Ganganatha. l,nu
produaion of aaivily. lOnl S. jijUL Sit <!din:: 10 know
1l..4.161. 169, JIIs }IWIIA.Su~
Q motivator of :IClion (pnNuU). JliInalrim.im. 1J:llI.llI
16t-66, 168n1. 19i-}09. }ion16 j "", s.. '""'""'" obi«<
rdation to action. 49:10. 110-11 joog.mcnl of IiU1KnCIS (dApt1fIJilN~
:and ptUpOSC'. f H9 PNI1Iitj~), 1I'e16. uk!
KriptUft: as. llbS. )61-'7) judicious penon (p~1I#11It;
sharaI norions concnning, !1:::l1o prJt,4",nwhlri,,)
i!! !J:j, 110 at Kling for the sake of goW, 1751.
ul!im2le. )'4- 18 )'.
u unmcdiated with rcgud 10 u aaing on pankulan. )90. 41)
activiry. 111 at acting through doubc. JiSll4
as wh21 make. one obuin a pi Q acti ng through inwurncmal
-'
Kampab. )z.6n6
Kapstc-in. Mallhew. 4OfISl. 41n·s;
' 74l'I,
on rdauYe dadng of South Asian
pbiJo.ophen. Inl
1}Of11). 1Son41 Irri]i. 5« acrioa
Urili I)'Ilcm. 16=10, 161. 170-71 ~!,il.t1J4. 5« mommw1nc:A1
u~ 5« inurumem Kumirila
bnn., m as Dhannaklni'. intmocutor, J1i
5« .1r4lwmic world on dcmenl$ of:an in~. )Sf\-iS
.brmu,,!u motionl. J01 on aampic:s, JOn)9. }In}9•
unu" lu objea]. 5« patinH )1.n-i1
Iwmic 'II'Orid (J.M). 186061. )1,--)0 on ilUtrumcmal knowledge ;II
Unr. 5« Igem dUtina from action. ~
IUryorhmI. SNdiKl-cvidma: on insrrumenta oflc.nowkdr;r, ~
K..uura. SMryil. s'n7 . l09ni7. on ;'\I,.;.nne ;n.. n..,n",n... liry.
u6n99. l?9n8l1. 101-1. 171n19 )nm). }800l6
IGty.iyalU,9}n61 on pc:rttpllon, l.o4lI1S
KdIner. Birgjl. 86n5). 1.46"4, on predia.l~ncc ..dation.
316mb 18nn . !!z }On}.
krv,JJ"HJi'" 5« inFerence. on purpose and inmwncnllll
involving only pomive kncwIoI&<,o!
colKOmit<tnO:
~"Y'ftirdin.. S« inl'CKI>CC.
involving only Iq2live
. ...
un IoCIIX urpla. ,).IIU. ',Il', .
N non~ion (."lIp.1.6.hi),
Nigifjuna, 1m, +4n61., H 91n60, llO-1.I, JsSnu, )88
Nagalomi, MuatOlhi novelty of insuuma'lfal cognition,
OD .rrhUriJtl. 8+n.49, l S7nSf, 1H-SJ. 198-m
1,!r60, 171-n. 17S, 171. ",.,... Sr.. ratOning
OD circularity in Dharnukirti', Nyiya, !L 17n)1
theory of instrumentality, _w..u. (Vi..,."""
on analogy, 14~ n1
....,
l}}fUS, l}7nl.l
on imptrmanen« and change, on desire to know and doubl •
liin.., 1.4fl1S
~idmrJL S« tdfks.mea on r;oaIs. ~
n,u ullll marion. Set on gnmlTW' and ontology, }8nS1
1IIIIM.i"",nuiIMMhtt on indisperuibility of c:ocrea
nanlll: (,'Urri: ~,.. IfMWIItt), knowJcd&c for libcntion, 11n18
See ,.."u...... ,..
"'"~
on ~-predtcar'" marion,
nondisposilional proptny. 179-80 18n)s,19n jl, )On)9
Sr.. "Js" XCidentallcssmliai on importance of knowkdgt for
distinction; nca:ssity spirirual libcntion. 11m8
~rtffI. ,efi"",nf (.IJv..). on in ~ u grnund..:l i n
IJ7=lfI pctapuon. j1ll41
INDEX
."
on the object ofbnguagc, 101n74. ofknowkdge (prllftl4twMwu),
lJ7nlH J.Jf=IS. ~I . 14S
on marion between action and Sn Ill. ctedibili£),
inslfumem, 49n74. 49nn- ontological miuction. Sn n:auaion
on ~ perception, 1Jnu, llIUl, orden of contq)lJ, ,~U)'
1.41114, LpllS, losn79 ordinary persons (p,rthtt:j.tw.
on Iharat philosophia.l principia:. Iln.c,brli,,)
:u nl 7 bdidi of as ob;«t of rd'uu.OOn,
on the whole (IIIIIIJ"'"" ), J9nS4. 14- 17. fu. ~
I01n72.. IIOn90 as capable of derermining the ~
lrine of Iwm:i. 1}C)=\1
o ill ca~ble of determining tbe
""-
object (hrm.1I). SnPUlml
cognitions of as having quotidian
instrumentali£)" )14-1.8
obj«tive ima&e ~r.. .Ju ill iIw::apabk of direa.ly pera:iving
".~.~IJ),."ikllJ&. Sn
ddinil~ dnerminalion. 1S.5U'*'-
quent ~ua1 judgmenl
wbjea of I proposition
propaty-IllllbJM_ Sec IllllMiIlll, IS
proptfty, sIWMi...
Plopat)' WI puvadc:s
(lIJIIp"ulb."",,). .sn ~ing
,,.l1li",,"
Struuuumenl of propc:rty
knowitdgt. IS motivalor of action propertY 10 be prtJ\'al
,n""';" &r activation; activity; (t.#; 'VmI4). S« prtdiate 10
pracria.l Klion; refem'ltial "'P~
runaion prop06ition 10 be- p~n r,rlUijiU;
'~NL 5«puIpOlC ,./qA}. $no tha.is
predicale (tlh.nrt4). 5« propaty ,.rrhtttJiI-$no ordinary persons
predialt-aprtuioO (JlN,,,,",,,,~
r/M,.",.,vid/II/¥J,). 117n1l9.
l13nn9. Is6nl). 171n44. lO)nSl'"
...
psychologism , 14n1S, 17n)1, lli !L.
.
quaiiry-pouasor. 17=11, n nso aha ~reduaiYc rcuoningl
Stt ..huubjec1 of a proposition of diKributed (fuilia. til 6t:79,
quali[f/qualificd tdalion. !.!J., ~
I1SnIl9. HB-Sl.)H and kvds of analysu. ~ a.
..,....
Quine, W.V.• 2uml7 70=7 6
~, ~ .p=-;).1t 6z.-6).
R
Ram-I'ruad. OWuavanhi. ll9mo oflhc real. l!t 2io.1!. i8=-99.
Ra.tnaltini, l}uull
R..ml,.,w (N1prjuna), W') Sa ttl. irreducibiliry: ~
real (S/l.I) anaI)"iI; reduaion
as dwxtcriud by Idie dfx:xy, ItfdcntW fi..maion r,"'''!"i), ih
4sn6.t, ~ 195, I11n1l8
as COII t (tll oiindubitabk Stt (1M semantic convention
knowing. JH6, t!.o Lt-lz, rdino:mem. }10. )11-)0
91-1I} rdkx.ive awarmcss (1lIIW'!'wtLut..)
as inapret.5ibk. 12. i!.::iZ as arising as the: subj«tiYc asp«!
as momentary. i!:::!Z of awum.el$. )9lnl
as simpk, }Z- i S, 11 79-81. 98-11 ) insuummr.ality of as sheer ausal
s«..Is. particular, IWO rcalilia: dJic:ic:ncy, 160. 176, 178
rnliml.. )61l4i. S!. S9nIJ. 'ZL n"...dUl li.,. of, +07"IS
u6-1z. !l!, }lJ role in UlXnaining
lUIOnins (1IJIIJIa.' ,un) in.n rumemaliry of pc.ception,
as Buddhul practice. 1SI. )18-1.9 191, 195, )78",+ 3iS-36
corxfilionals in, 1 ~15 sellks.snas as, 401n1 ) , .. 06nlS.
and ICYds of analysis. 61 ,.oSnl5
pwpoM' as a ~ ooor in, fl u uilimale p,,,1N1fU, ) 11
SN "'- inkn:ftCC1 redlK'ti"" an&iy- reI.lion., t!=1,.!1
w: 5I}'k o f rnsoning rdiPOW subscratum, 10nl4
..60 FOUNOATIONS OF OHARMAKUtTrs PHILOSOPHY
><mO<' ""'"
the c:onuadictoty of the prcdicalC'
(¥J'I'lNIhilll!"'m4!'A}IMJ.r). S« to be: provC'n
instrumental dfcct. mediated ~ Srrlimibrity
ranocc object f/'i',."q.; ,,;,~, ";"Ml"", Sncopl'C'SC'nOr;
I snp., 91n6o, 1)Onll. ~ 1D11 w..1tbip,..".".. S« supporting
Srr Ifiso lnnscmpirial objca condition
rtpeatability (IIIIIItItJtIl. Srr Ulti.S«apacity
diuribulion IUM~. Srr ratric:rion. in
fUlriction ("9wIJU; abo - invuiabk causal polC'nliW
ruk") S1Icya Mchos klan. 9n17. ~ 76-78
in caUAI potentials (/Mti"iJtt-). Mm4"rLr~!,& Srr univaul
16oIU·9. 161-6~. !M! 19S. )44
inlcQubje«iVC'. j16n7
of pervasion !'dation. :zBn}6. !2z
149='" ZO'I J.06n,8. 2.09=10
of the rea/lo a spuiotCtnponi
--
U"""'''' S« imputation
,,.,
Oharmaklni'.IJIIlrin! citation of, infinitesimal pa.nic:Ia. l!!
IlOnl07
Dhannkinr. aniNde loward in :If rooted in odwion, )+4. )H~
&:bate. Innl4 for Alb;ca and predicate. 117n1l9.
eight demen~ of Dharmakini'. l1.8n1l9. 17lIl44
dixwt.ion of, 40-41 ACNe cilium. ~
instNtDenuliry of, lL 1)nlO, 11-4. Sit .w. cognitive image
lll- n . l)HSt )61-7) ICnK faculry (i~. abo -1CftJC
:If not of hl.UlWl origin ocpn', 1'=14. 12.4. 116-17, 1.61,
(""~. 1J) US , 169-7'0. }I)-Is
thredOkt analysis of. 14~ sensible dement (l)Ioru..... abo
5« .w. credibility; Kriptunl infer- -JeNe sphere', ~ ZO=71, 79nJl.
~ 10l nn. I09 n87. J97""
tetf (brrfIut) JCmOf)' oontaCt (s.1r1rj~ sp.rt.),
:If agenlofiruuumental knowing !J., 1!! 169",,0, )44. )6)
in Brahmankal thouJht. H! Shah. Nagin. }<tSnlf
beliefin as diminated in 1I;nliflll, Sidmu, Mart. 116n99. 1}1Jl116.
""""
l..,nll, ls oru s
,,.
.II distributed entiry, +41161, 61n17. Silk. Jonathan, 1)6n19
similariry (~
as object oIbdicfin ~i. :If basis for error. ,6n7. 11-8g. !:i!
iL~ )71- 11 of images to their obj«a. &.cls.
Iin9c
:If a eum«. )4). mini}
... uninr.,...,.hk. j6) .;mplici.,. ( .....6<1; ...... - ;
tdf-dinging (h1tvIlSttthc). ~ dwcd notions cona:ming. JZ=-J9 .
J71-1l
sdJIesmeu (.I/IlhruI1I; ""i~)
:u antidote to Alffaing. 6onlS.
"
Buddhist Yin.,: on, iCt"i'. 1!:
79nJ!. &0
191n1)l. )6In1O, }69n1l. }71 Sit.u. Nngulariry
all a dist:ribuled entiry, ~1 linsuJariry ('''u.' ,fun",)
AI inHnunc:no:ai ob;ca, ",. li"!,,ill;'; "k"1KC ....;thoul.
4C)lnl}, 4O'Snl4, 408m1 m-60
,.61 FOUNDAT IONS OF DHARMAKI II.TI' S PH I LOSOPHY
_.,
on elcmCfltl of an jnfcrcnc:c. lS"'4S
on grammatical instrumentality.
of inference. 1-+8010 , 1-+8011
on swJHMlIft., 1St=14, Is6nl),
,6,")0. 16~ ' 70n'tl-. l86n61,
,...,.
on Knse organs. Ip'U, 1)0l.), 191-96, 101-1. 1090101, Ulnll l
on 1'IIIIJM4U11pnriiMlfIiht" lS I nl7
Sou. Ernest. 18}1lsl Slhiramati, !1
IOtcriolozy, ..,.n61, i!:!.1t:::H! ,rbiit.tL S«spalial ClllcrWon
60-61, 191. ~-8. )17 uyle of R'IJOnins. }, 1: u. !.L tl. H
sfN'rlA. S«acnsory contxt !ubhagupta. "nl)
ipUia.I mension (~). 12.. ~ IUbJ«l-a:prcsuon
s6-1l, 6l. 6crz<J. ~J, 98-111, (tihtI""il/tJdiJJbttII). 117f11l9 ,
.!1.L. ,.o,.. ,.osIU,. ,18nIl9. 10)094, )S1019
S« ~ dUtribution; whole lubjective image (,zrihtdliHrII; also
,pctdl of 1iI credible pcrIOn ~wbjective aspca of awarencss~).
(4pllJlld.). &racdibi.lity; 171, 176n9), J91, 407
KriPlure aubjcalobjcct duality. S«duality
'pirilual adept ~1f). 1.4016, 1§.. subject of a pr0p05ilion (J),.,mri1.;
!!i. l86n61. 116-17. 19)01)0, ·idh],Jh.f1IIilt; , . . ). 1901 ).
)01-6, )18n8 lkll, ~ 16cJ-n, 199, lOOn&,.
$pirinaal excrciJcs, !lQ. '40 '"'='
Ipiriulil frttdom (~), 11. lsn18. aubjcct-predic::ue reblion. !£.o !..1§..
4 . ~ ~ ll9nlO,
. )14 , )2.4 171"'4-4, ).46-S1. 3S7f17
IPUnow pcroeptlOn sutJ.;tanI;C ('/'IIIIJII). It-1+, J!. l2!
~,.~bhtls.). }6n.d. tzus •. J.4!>-SO. 1Q-S9. }6}. 196-99
m su£rcring (~. ..6-..7. ""
&rIlWcrror, illusion 60-61. 110, 19801)8. 199, )19
tquan: bnckcu, 11-1, supcrimpolilioo (SlUlWI'OpII). Sn
S.etnkdlnC"f, Ems. imput3tion
on the argumcnl for lupcncnsibk objcn (116,",*)
momentariness from exUtena', as amcnabk 10 perception of
97 n68 pti... z..tn~6
on IInIMhi]tl. 178-79 as the domain of Kri~. 1.4). }67
INDEX ,6,
as ntH anvnablr: to inr~ Dignlga'J rhmry . ........9
}11\41. ~ 366n.4- J68 Str ,Js, tvidtnu-predicttC
as DOl: charaa~rWng infinitcsimal mation: ncaviry
panidcs. l09nl] nwld!ll!lll- Str panicular
one who JCCI (.tiMriJiUMr/i,.). 11JIInh4,.M""'IUI. Stt inr~rmtt.
)OS. )66n ll ror oncselr
SIr "'- rnnxmpirical object IWIU,!,uwWlUI. Stt ~
IUpponing condition
(.."d4n"6lJ9tI). '<4n6). 16,-61. 1UIIt4 prhul!'Ji'm. Str
11'='1. Il4n~" lOImoo inlCrumC"nwiry, inrriruic:
mpponi"& enm.pIc ryllogUm. !l
(~"fi) "/."H4I1I11). lQ., WSn,s I)'It~matKiry. klL ~ a z!.
-"",* ''''
as C'.OfI.Kructcd. US-:t6. u6-p.
136-117. )1). ))9-1:1 T
as nature-IUIIM.f,., ISJ- n . 1sS-7J. tMUlmJtL Stt IUllMiI14PrlltilMtUiJM.
19)=1.01 idmtiry·modt of
and nccasiry. Illdl uJ."..,tti. Sec IlWhbllWprlltihMht&.
as propc:ny' IUIIMrI,., '1i=9'j. production. mode: of
11HZ. '70-7), 181.-91. 19t-101 rAnlWbiMV. ("tMr·"",M.lJ).
two IICIUQ' of, ISH', 164nl4. 111- 11
170-7). 1,)-U1 Tllm...sII,!,phiI (Stntanlqira).
univaWs :and. ~ ,nliS. l09 nI7. l11nIlO. 1)lnu6,
Sec abo IUIIMrllltlprlltiluMh. }¢n6. mn l)
IUIIbhl,,,,MrJ.. StrdiflUmcc in TlIItMU1!'rlliMpdj iU
nature (KamalaIala), I lllllIo. 1)lnn6,
IUIIMl~oa' (1UIIbh4~). }16 n.I) , ,Hon6
'1n~, I 11:1l, f79 , 1O)-11 tdK;: c:lJic:::lq/funaion. Scc
Stt 11M ..-viden«: IIrtJ,.~
IUIIMJl1IIp,.tiIMtuih&, idmtiry. trW. Stt purpose
modC'of (C'mporal alC'lUton. ih ~ il!
IUII~ Sec IUIIhbl~ntt ski, 6l.1l2. '4JliS), 111). 379
wdlMwpr.tilMMh. r n.anual Str.JJ. distribution:
rel.arion' ~=-
gcntnl ddinition.. r;&;1 theory of unconsciow: aror. ~
i .... ntiry.......... (wt.".".) o(. )n . ),'n,' }
' SM), lO t-l l Str ills. aror
and inUtMic: insuumentaliry or theW (priltijU: ~ abo
infCfcntt, 1,6. 197-98 "propos.irion to be provtn "), !L
maning of IJlllMolu. in. IU-SS. ~.H! 106-7
19)- 101 thrWold (VideOtt (",,j~nJitit-J.
production· mode (~) or. Str (Vidmtt, thfC'C'foid
')~-SJ , I] M~ Tibnon W "UlK"llIIQn, b l! 1l!
U response to problems in 69-79. Ibn<4J, )OOnl<4). JI,
.6. fOUNDATIONS Of DHAII.MAKIlHl·S PH ILOSOP HY
as
conditions. 16Snp
rqJramQ.ri~ or Abhidlwnu
typoIogy, ~
-=
"J#Irim.. SNcxmcomitanct,