Bhandarkar Oriental Research Institute

DATE OF KAUȚALYA
Author(s): D. R. BHANDARKAR
Source: Annals of the Bhandarkar Oriental Research Institute, Vol. 7, No. 1/2 (1926), pp.
65-84
Published by: Bhandarkar Oriental Research Institute
Stable URL: https://www.jstor.org/stable/44082682
Accessed: 02-04-2020 16:15 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Bhandarkar Oriental Research Institute is collaborating with JSTOR to digitize, preserve

and extend access to Annals of the Bhandarkar Oriental Research Institute

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 DATE OF KAUT AL Y A.*

BY

D. R. BHANDARKAR, ( Calcutta ).

With the advent of this century South India has become a

land of wonder to the Sanskritist. Three or four years ago the
students of Sanskrit literature were taken by surprise by the
discovery of Avantisundarï- hatha by Daņdin which seems to be
the original form of the Dasakumāra- carita ascribed to him and
which throws a flood of light on the various poets that propi-
tiated the goddess of poetry before his time. Not many years
ago a valuable addition to our list of works on Indian poetics
was made by the discovery of Bhāmaha's work on alankāra in
Trivandrum. The dramas of Bhāsa who proceeded Kālidāsa
had for a long time remained hidden from the modern world
until tbey were discovered twelve years ago at the same place,
namely» Trivandrum, by the lynx-eyed indefatigable scholar
Māhamahopādhyāya Gaņapati Sāstri. That a work dealing
with the science of polity had been composed by Kautalya had
been known for years from the references and quotations cited
from it by various authors, mediaeval and modern. These
fragments had been culled together by Aufrecht, Zachariae, and
Hillebrandt from lexicons, commentaries and other pieces of
Sanskrit literature. But nobody ever dreamt that the whole
work which was given up as lost would be recovered and made
accessible to the student of Indian history. The whole scho-
larly world was therefore seized with an extreme and agreeable
surprise when in the January number of the Indian Antiquary ,
1905, Dr. R Shamasastry not only announced the discovery of
this work at Tanjore, but actually published an English transla-
tion of some of its chapters. The whole book was afterwards
edited and translated twice by the same scholar. And as the
book is important, not only to the student of ancient polity, but
also of law and economics, it is no wonder if it has been edited
by two more scholars, by Prof. J. Jolly in the Punjab Sanskrit
Series, and by Mahãmahopâdhyãya Gaņapati Sāstri in the

* This was one of the six lectures which I delivered before the Benares
Pindu University in February last.
9

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
66 Armais of the Bhandarkar Institute .

Tnvandrum Sanskrit Series. The work is, indeed, so precious in

the whole range of Sanskrit literature that it has revolutionised
all our ideas about the culture and civilization of Ancient India.
Unfortunately, however, there is a wide divergence of opinion
in regard to its date and authorship. Some soholars are of
opinion that it is the work of that Kautalya who enabled
Chandragupta ( 320 B. 0. ) to establish the Mauryan empire.
But there are others who hold that it is a production of the
third century A. D. As the work, however, is the sheet-anchor
of the ancient history of India» it is of paramount importance to
discuss its date and authorship till a fair consensus of opinion is
reached.

In respect of style, the Arthasāstra presents a mixture of

prose and verse which is generally met with in the Dharma-
sūtras. There is, however, some difference between the two.
In the case of the latter, the kārikās are in no way intimately
connceted with the prose Sūtras which immediately precede
them. In the case of the Arthasāstra, however, there are some
instances of verses being prefaced by one or more words in
prose, either of which is insufficient by itself but which together
make the sense whole and complete. There is another difference
which is also worthy of note. At the end of the book even after
the colophon occurs a verse which tells us that the work
consists of not only the sūtra but also of the bhãsya (commentary).
I have elsewhere shown how to distinguish the sUtra from the
bhasya' The prose portion of the Dharm asütras is compos-
ed of the sūtra only, without any admixture of the bhasya.
The subject matter again of the Arthasāstra has been carefully
arranged and lucidly set forth. In fact, 4 a rare unity of plan
and structure pervades the whole work'. What is again strange
is that it commences with a carefully drawn up table of
contents, and closes with a list of devices interspersed with
many cross-references in the body of the work. The conclusion
is therefore irresistible that this is the work of one man. And
the question that we have now to decide is : who could its
author be ?

At the end of the first and the tenth chapter occur verses
which ascribe the work to one Kautalya. At the very close of
the last chapter we have a verse which tells us that the author
of the work was one who wrested the earth from the Nanda

1 Carmichaeļ Lectures, 1918, pp, 98-9.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Bate of Kautalya. 67

kings. As from the Purāņas we learn that Kau

the Nandas and anointed Chandragupta as king of their
dominions, it appears that the author of the Arthašāstra was
no other than the prime-minister of the founder of the Mauryan
dynasty who flourished in the fourth century B. C. Dr.
Shamasastry who first edited the text naturally held the opinion
that the Arthašāstra was a work of that early period- The
various political practices, social customs and religious obser-
vances detailed in this book had well nigh disappeared certainly
from the second century B. C. onwards, if not from a much
earlier time. They therefore reflected a state of society prevalent
possibly in the early Mauryan period, but certainly before the
birth of Buddha. All these practices and customs have been
beautifully culled together by that erudite scholar who has also
shown that both the Kāmasūtra of Vâtsyâyana and the Smriti
of Yajñavalkya are indebted to the Arthašāstra not only for
certain views and doctrines but also for phraseology. The late
Dr. J. F. Fleet endorsed this view unreservedly about the date
of this work in his introductory note prefixed to the translation
of the work of Dr. Shamasastry himself. " The work " says he
" accordingly claims to date from the period B. C. 321-296 ; and
its archaic style is well in agreement with the claim." "Though
the existing text " adds Dr. Fleet " is perhaps not absolutely
word for word that which was written by Kautilya, still we
have essentially a work that he did compose in the period stated
above." Among the European Sanskritists Prof. Jacobi and
Dr. F. W. Thomas are the only two other scholars who have
countenanced this view.

Soon after the Arthašāstra of Kautalya was published, it

came to be more assiduously studied in Europe than even in
India; and a view of diametrically opposite character was being
formulated, namely, that the work belonged not to the fourth
century B. C. but to the fourth century A. D. This is now the
view which is generally held by European Sanskritists, and the
late Sir Ramkrishna Bhandarkar iś perhaps the only Indian
scholar who has expressed a similar opinion. The arguments
on which this view is based were marshalled and enunciated
by Prof. Jolly2 and Prof. Winternitz3, and were briefly indicated
in a short but lucid note by Prof. Keith.4 The chief argument

2. Kautillyan Arthêastram (Punjab SK. Series, No. IV), Introduction.

3. Calcutta Review , 1924, (April) p. 1 & ff.
4. Jour . R. Ai. Soc., 1916, p. 130 & ff.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
6Ì Ànmls of the .Bhandarkar ïnsûtuïè.
of Prof. Jolly is that the Arthaáãstra bears a close alliance with
the Kāmasūtra and that if the fourth century A. D. be taken as
the possible date of the later, the former might have been com-
posed in the third century. Prof. Winternitz bases his position
on a two -fold argument. The first is that the contents of the
Arthaśastra justify the assumption that it is the work of a Pandit
and not of a statesman. It could not have therefore been com-
posed by Kautalya, as he is reported to be the prime-minister of
Candragupta and therefore to be a statesman. His second argu-
ment is something like this. The numerous discussions in
which the opinions of different teachers or of different schools are
quoted in this work as opposed to .that of Kautalya indicated
by the words: iti Kautalyah show that the Arthaśastra was a
composition not of a single author but of a school. Jaimini is
similarly mentioned in the Pūrva-Mīmāmsā sūtra, Bādarāyaņa
in the Vedānta-sūtra and Baudhäyana in the Baudhayana-
Dharma-sūtra which represent so many schools, but Patañjali,
he says, never states his opinion in the Mahãbhãshya by saying
iti Patañjalih .
So long, as these scholars confine themselves to the main
arguments of theirs, they are not open to any adverse criticism.
Of course, it is possible to hold an opinion different from theirs,
but it cannot possibly be said that there is anything objection-
able in the manner in which they maintain the views. Unfor-
tunately, however, they are not content with this and enter into
a discursive argumentation which has the air, sometimes, of
plausibility but, often, of being over-strained and far-fetched.
Thus they go evento the length of proving that there was no such
historic figure as Kautalaya at all 1 It is perfectly true that
Indian tradition as preserved in the Eurāņas and the Buddhist and
Jaina literature is unanimous in making of the wily Brāhmaņ
Cãnakya Kautalya, a king-maker, the destroyer of the Jříandas,
and the supporter of Candragupta. But, says Prof. Jolly " It
might indeed be questioned whether the prime minister of
Sandrakottos is not a figure of pure mythology, as he is not
mentioned in the Greek reports concerning Sandrakottos." 8
Prof.. Winternitz practically adduces the same arguments. The
Purāņas, says he, no doubt unanimously report that Kautalya
de_stroyed the Nandas and anointed Candragupta king. But
they never speak of his having been a teacher or an author.
Patañjali: in his Mahãbhãsya refers to the Mauryas and

5. Kaut- . Arth ( Pun. SK. Series ) Intro., pp. 33-4.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Ďate of Kautalya. 6ft

Candragupta* sabhā, but makes no mention

minister. But the Professor is not satisfied wit
and goes one step further and asserts that the very name
Kautilya raises grave doubts as to his being an author or a
minister. Kautilya means • crookedness ' ' falsehood ' ; is it
likely that Candragupta 's minister should have called
himself ' Mr. Crooked ' or ' Crookedness personified ' ?6 Surely
scepticism can not go further. It is true that Prof. Winternitz
cannot be blamed much if he takes Kautilya to mean "Mr.
Crooked'. For does not Višākhadatta in his Mudraraksasa speak
of him as Kautilyah kutila-matïh sa esa y ena. But the Professor
should have known what Samkararya has said about it in his
commentary on Kāmandaka's Nitisūra. " Visņugupta says
the commentator, " is the name given at the naming ceremony
and Cânakya and Kautilya are connected with the birth-place
and the gotra respectively. " It will thus be seen that
Kautilya was not his personal name but a name derived from
his gotra. And Mahãmahopãdhyãya Gaņapati SāstrI has
shown that not Kutila but Kutala is mentioned by Keáavar
svāmin in his Nānārthārņava-samkshepa as denoting a gotra-
rsi% that is, the sage who is the originator of a gotra . "The
word Kautilya (ikãramadhya) is a mistake" adds the Mahãma-
hopãdhyãya " handed down to us by scribes and readers. All
the ancient manuscripts of the text and the commentaries of
the Arthasāstra, available here, read throughout as Kautalya
(akaramadhya), and not as Kautilya. Kutala is the name of a
Rishi who founded the KautalaN Gotra and any descendant of
that Gotra may be called Kautalya. "7 It is thus quite clear
that the real name of the reputed author of the Arthasāstra is
Kautalya, and not Kautilya in the sense of " Mr. Crooked
and the sooner we discard the latter and adopt the former name,
the better for the promotion of truth and history.

Then, again, we are told that Patañjali speaks of the

Mauryas and even of the Candragupta- sabha , but makes no
mention of Kautalya. If the latter was really the prime-
minister of Candragupta and was chiefly instrumental in
securing for him the Nanda throne, it is inconceivable how
he is passed over although the Candragupta- sabhā is referred
to. This is what the argument comes to. But the question is
what is meant by Candragupta- sabhâ t Does it signify 4 the

6. CaU Rev., 1924 ( Apl. ). p. 18.

7. Kaut . Artha ( Tri v. Sk. SerieB ), Pt, II. p. 4.'

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
yò Ànnals of the Bhandarkar Institute .

council of Candragupta's ministers' or 'Candragupta's conclave

of literates' ? It is very difficult to say which sense is here
intended. But as Patañjali is a Grammarian, the probability
is that he is referring to the latter, that is, the assembly of
learned men. Similarly, there is no force in the argument
urged by Prof. Jolly that because Megasthenes does not
mention Kautalya at all, the latter could not have lived in his
time and could not have been a historical personage. This is
an ex silentio argument, which can never be thoroughly satis-
factory. It is true that Megasthenes does not speak of
Kautalya. But it is also true that he does not speak of the
Wandas. Are we therefore to suppose that the Nandas are a
fiction of history although they are referred to in all the early
Purāņas as having been uprooted by Kautalya for putting
Candragupta on the throne ? What then becomes of Nandrus
who is supposed to be mentioned by the Greek writers as a
contemporary of Alexander and has been identified with
Nanda even by the latest authority on ancient Indian history,
namely, the Cambridge History of India?8
The truth of the matter is that argumentum ex silentio can
never be regarded as of a perfectly reliable nature. And yet it is
this argument that has unfortunately been indulged in by Prof.
Winternitz a little too frequently. " Thus Megasthenes speaks
of mile stones on the roads, which are unknown to Kautilya.
According to Megasthenes water for irrigation is carefully dis-
tributed to private people, while Kautilya knows nothing of
such a distribution of water but mentions private water-works.
According to Megasthenes no private person was allowed to
possess elephants or horses, but they were the monopoly of the
king. Kautilya knows nothing of such a monopoly." 9 These are
some of the statements which the Professor has made to show that
Kautalya and Megasthenes are not in agreement and cannot
therefore be contemporaries. What they amount to is just this :
because Megasthenes makes mention of a thing about which
Kautalya is silent, it must necessarily follow that the thing did
not exist when the latter lived and that the two cannot therefore
be contemporaries. In the same way we may argue like this: the
Rājatarangiņī speaks of both Vākpati and Bhavabhūti as being
contemporaries and proteges of Yaáovarman, king of Kanauj.
Of these Vākpati alone refers to his patron in his Gauda-vaho

8. pp. 430 and 469.

9. Cal . Rev.y 1924 ( Apl. ), pp. 19-22.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Late of Kautalya. 71

which describes his exploits. But Bhavabhū

has composed no less than three dramas,
Yaáovarman, which is inconceivable if he is p
contrary, says that all his dramas were ac
Kãlapriya of Ujjain which is quite differe
Are we therefore justified in inferring from
part of Bhavabhūti that he did not live in
varman and that both the poets did not live
and also together ?
It has also been argued by Prof. Winternit
not only no agreement but positive discre
accounts of Megasthenes and Kautalya. The former emphati-
cally states that th9re is no slavery in India, but both the Artha-
śastra and the Dharmasūtras know different classes of slaves,
male and female. But the greatest difference between them
is found in their respective accounts of the administration.
The organisation of the financial bureaucracy, the local or
municipal administration and the military organisation as
described by Megasthenes is so essentially different from that
set forth in the Arthaáâstra that the author of the latter cannot
possibly be a contemporary of the former. Why such an
excessive and implicit credence is attached to the statements
of Megasthenes is far from olear. In the first place, it is
worthy of note that what we know of his account is not from
his original writings come down to us, but from fragments
quoted from them be other authors. And even supposing for
the moment that these fragments agree in certain substantial
items, it does not necessarily follow that they were critically
observed and correctly recorded by Megasthenes. We need
not go to the extent of accusing him of mendacity, as his own
countryman, Strabo, has done. But it cannot at all be denied
that his reports are not very often accurate and reliable. This
has been admitted frankly and unreservedly by Prof. Jolly.
And even what Prof. Winternitz has said goes to confirm it.
For Megasthenes tells us that there was no slavery in India,
and even if we set aside the Arthaśastra for the -present, this
statement is completely belied by what we find in the
Dharmašāstra as has been remarked by Prof. Winternitz himself.
Prof. Jolly, however, goes one step farther, and frankly tells us
that u the idealising tendency in Megasthenes greatly impairs
the trustworthiness of his statements, " 10 But what is most

10. Kaut» Arth -, Intro., p. 40.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
72 Annals of the Bhandarkàr Institute .

inexplicable in this connection is that while citing instances of

this idealising tendency, Prof. Jolly puts the evidence of the
Arthaéãstra against the statements of the Greek writer, though
the former in his opinion is much later in date. Two instances
ought to suffice. Thus Megasthenes says that the Indians
never took wine except at sacrifices. Prof. Jolly measures the
truth of this statement with the contents of the Surädhyaksa
Chapter of the Arthasāstra which contains recipes for the
preparation of alcoholic drinks. Again, Megasthenes says
that the houses and property of the Indians are generally left
unguarded and that theft is of very rare occurrence. This the
Professor tries to refute by drawing our attention to the chapter
on robbery in the Arthasāstra, which specifies various types of
this crime and also the punishments that are to be inflicted.
If this work is much posterior in date to Megasthenes, as Prof.
Jolly contends, where is the necessity of bringing in its
contents to disprove the statements of the Greek ambassador ?
Impartially speaking, the Hindus of the fourth century B. C.
should be given credit for the virtues of abstefniousness and
honesty for which Megasthenes has praised them, and it may
be further inferred that the Indians lost these virtues in the
third century A. D. when the Arthasāstra was put up according
to the Professor. But why call in question the veracity of the
observations of the Greek writer by setting against them the
evidence of such a late work as the Arthasāstra? But we shall
pass it over, as this presents not an antiquarian problem, but a
psychological puzzle. We cannot however sufficiently thank
Prof. Jolly for giving his frank opinion of the worth of Megas-
thenes* account. " Megasthenes " says he " was unacquainted
with the languages and literature of India and his work, though
earlier in date, is far inferior in intrinsic value to the Itinera-
ries of the Chinese Pilgrims and to the great Arabian work of
Alberuni on India. " 11 It is therefore not only idle but
uncritical to assign Kautalýa to a later date simply because
the contents of his work do not tally and are sometimes even
discrepant with those of Megasthenes' account. Again, it
deserves to be further noticed that the Arthasāstra portrays a
state of society, not so much of the Maury an as of an earlier
period. For at the very beginning of the Arthasāstra, Kautalya
tells us that his book is but a compendium of the previous

11. Ibid, ļi. 41-

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Date of Kautalya. 7 à
oris on polity. And it is only in Chapter X of the Second
A «ihikaraņa which deals with the drawing up of royal wr
at he seems to have brought its subject-matter upto date and
)ade it conform to the actual practice of the day as we lea
le concluding verse. This is exactly in keeping with
3 that some of the manners and customs depicted in th
K are of the pre-Mauryan period as has been pointed out by
... Shamasastry and as also we shall see later on.
V^e have thus disposed of most of the arguments advanced
:• i'.-->ors Jolly and Winternitz to prove that the Artha
i i nach posterior to the time of Candragupta, originato
; fh M¿iu i-yan dynasty, and that it could not have been t
>or' 1 1 , •••• prime-minister, Kautalya, supposing there wa
hist. 'C personage of that name. We shall now consider th
mai- arguments of each. We will begin with those
Pro' lš -or Winternitz who adduces two such arguments. T
first "oí hese is that the Arthasāstra is a work* not of a state
mar}, b<7 it of a Pandit, and could not therefore have be
con vf -used by the prime-minister of Candragupta. But is th
life o l' a statesman incompatible with that of a Pandit? What
ever ne case outside India, there is nothing irreconci
able be> wet7n Lhe work and function of a statesman and tho
of a i iidit, so far as this country is concerned. Here a
two instances ought to suffice. The first is that of Hemād
who 7 lo u rished during the reigns of the Yädava kin
Mahīvkvg, and Rāmacandra (A. D. 1260-1309), and w
mini; tr= to both. In almost all the copper plate inscriptio
of th< princes he is spoken of as their Šri-karaņadhipa
Chief s o? -etary who was in charge of the state archives a
issued ¿xi) borders on behalf of the sovereign. Although
thus 1 ■ r i ì. strenuous political life as minister of the Yädav
dynaf ' • a man of learning himself and wrote ma
volamii ^ works chiefly on Dharmašāstra. The seco
i s that of the two brothers Mādhava and Sâyana wh
were similarly ministers of Sańgama and Harihara I of th
Vi jay na var ćJynasty in the 14th century A. D. Althou
politics V>iid engrossed the greater portion of their time a
energy, ť- y, being votaries of learning, did not find it impo
sible to c o apöse a number of works connected with the Veda
and Indi-vi Philosophy. It will thus be seen that in India
least the r waļs no divorce between the politics of a statesm
and tlv vxíig oí a Pandit, And it is quite conceivable th

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
Î4 Annals of the Bhandarkar Institute .

the two functions were combined also in Kautalya, who though

ho was the head of the Mauryan administration, might have
very well composed a work which was somehow related to
politics though on its theoretical side.
But then Prof. Winternitz holds that the Arthasāstra evin-
ces the same predilection for endless and pedantic classification
and definition as in other scientific works composed by Pandits;
How far is this a fact ? Does Kautalya give himself up to thts-
vice to such an extent as to conflict with the mentality of a
statesman ? Personally I am of opinion that no book or. prac-
tical politics written even by a Bismarck can be entirely free
from theoretical discussions. Fortunately for us Prof. Winter-
nitz gives us four or five instances to show what he means.
Recently, however, Dr. Narendra Nath Law had occasion to
examine these, and he has done it in such fullness and with such
a lucidity that it is not necessary to take them all into conside-
ration.19 Only one of these arguments maybe here mentioned.
The sovereign, the minister, the territory with its subjects and
so forth, says Prof. Winternitz, form the seven prakrtis or
constituents of a kingdom (or state), and we have a long list
of the good qualitie * which each of them should have. This
is the first instance he has adduced to expose Kautalya's
passion for pedantry. What he means is that a statement #f
the qualities that make any constituent of the state an
ideal one has no meaning and can serve no useful
purpose to a practicial politician. But he forgets that
this list of excellences is a sine qua non in the treatment of
maņdala , whose main object is to guage the strength of a
state against the neighbouring ones, as every ruler is expected
to do. This strength can be measured only by scrutinising
how far the qualities of each constituent come upto.or fall short
of the standard suoh as that set up by Kautalya in that
specification of excellences. It is not therefore proper to assert
that this specification is but an outcome of his vice of pedantry
such as would be worthy òf a Pandit but not of & statesman .
And, in fact, any one who carefully reads the Arthasāstra
cannot fail tobe impressed by the fact that the work- combines the
minimum of speculation with the maximum of practical wisdom.
The second main argument relied upon by Prof. Winternitz
is just this : the Arthasāstra contains many discussions where

12. Cal . Rev.) 1924, ( Sept, ) p. 512 and ff. ; ( Nov. ) ]p. 228 and ff. ;
( Deo* ) p. 466 and ff.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Date of Kautalya. 75
the or i nions of the different teachers of polity are
one after another and refuted by that of Kautalya,
the wordi : iti Kautalyah. He further remarks that we
genrf .ü! find this mention of the name of a teacher in texts
ema from schools. What he probably means is that from
this of Kautalya* s name in the third person it appears that
the sul ra j of the Arthasáátra were composed, not by Kautalya
but his successors in the school. Dr. A. K. Guha had occasion
to examine this argument in connection with the Vedānta -
intra. There he has pointed out that at places where any
author his put his views against those of his predecessors if he
had expressed himself, not in the third but in the first person,
he would have laid himself open to the charge of egoism.13
The use of one's own name in the third person is* more modest
and appropriate, and is common also among writers of the
vernacular literatures in India. Nānak. Tulsldās, Kablr,
Tukârãm and others invariably refer to themselves, not in the
first but a the third person. But we need not dilate on this
matter - Prof. Winternitz himself admits that "it is, of
course, ; ¿ihle that an author may state his opinion in this way."
We : ¡ial now consider the chief argument on which Prof.
Jolly ha taken his stand. So far as the external structure is
concerncd, the Arthaśastra has a remarkably close correspon-
dence w uh the Kāmasūtra. Thus the chapters of each end
with the verses of the author, and in the case of quotations they
are invariablLindicated by a prefatory remark to that effect.
Each, agaitfB>mmences with a detailed table of contents. It
oan thus s^^ftly be doubted that both exhibit the same style
of com posi^^B and could not have been separated from each
other by an^rong interval. And if the Kāmasūtra has been
assig • che fourth century A. D., the Arthaśastra cannot be
placed ¿uure-thfon one or two centuries earlier. This line of
reasoning is practically the same as that advanced by Sir
Ramkris^a Bhandarkar, who, however, infers therefrom that
the Aril a - a is a work of the first or the second century
A. iV* I < s, indeed, the strongest argument that has been
urged against taking Arthaśastra as a production of Kautalya
and consequently of the 4th century B. 0. So far as the style
and structure of composition are concerned, there can absolutely
be no doubt tei one of these being a replica of the other. And
i
13. Jivai.Ui.attļ pp. 3-4.
14, Proc . Or . Qpngress, roona 191«», p. ¿'*

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
76 Annals of the Bhandarkar Institute .

when we bear in mind the fact that they have in common many
phrases and sentences relating to polity proper and that in one
case the Kāmasūtra clearly says that it was quoting from the
authors of polity certain passages which are found word for
word in the Arthaśastra, no question can arise as to the latter
being anterior to the former work. But, as just Rafted, the
Arthaśastra, as it is at present, cannot be much anterior to the
Kāmasūtra and cannot reasonably be taken to be earlier than
the first century A. D. This conclusion seems irresistible. At
the same time it is very difficult to avoid the conclusion that
many of the customs and practices referred to in the Arthaśastra
carry us to a period of at least the fourth century B. C., if not
earlier. Such was the practice of keeping state-owiied drinking
saloons to supply liquor to men, women, and children of all
castes, above all, of the drinking of liquor even by Brāhmaņs.
Similarly, the Arthaśastra allows the killing of a Brāhmaņa for
political offences, though through drowning, but such a thing
is unthinkable from the first century A. D. onwards. Again,
we read in this work of the exaction of religious taxes and also
robbing the temples of their money by imposing upon the
credulity and superstitions of the people. So also we read there
of the allowing of divorce between husband and wife through
enmity, and of re- marriage of women whose husbands had died
or had long been absent abroad. Similarly? the religious life
depicted in the Arthaśastra was essentially different from that
prevalent from the beginning of the Christian era onwards.
The worship of Vaišravaņa and Mahãkachchha JÉ^the practice
of Ātharvaņic witchcraft and sorcery which arv^^Jntioned in
this work seem to have entirely gone into de^^łde in this
last period.15 We are thus confronted with a piJBP- So-far as
the style and external form aTe concerned, the Arthaśastra
seems to belong to an early century of the Christian era ; but,
so far as its contents go, they reflect a phase of society which
cannot be later than the fourth century B. C.

The puzzle can be solved only on the assumption that the

work of Kautalya underwent some change of form in the early
centuries of the Christian era. And the qúestion arises: is
there any evidence to that effect ? This evidence is furnished
I think, by the concluding portion of the very first chapter of
the Arthaśastra. There we are told that the Work is composed

15. Kautilya's Arthaśastra ( trans. ) by Shftmasnatry, Preface,

DP, XVIII-XIX.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Date of Kautalya. 77
of 15 books, 150 chapters, 180 sections, and, abov
éloJcas, Books, chapters, and sections are mer
of the work and tell us nothing about its form. W
or was it* metrical in form? This question is answered
by the statement that it consisted of 6,000 slokas . This clearly
shows that the work originally was in verse. I am not unaware
that the word sloka has been taken by the translator to mean
4 thirty-two syllables' of an Anustubh. This is no doubt the
sense in which we take it in computing the copying work of a
scribe. The scribe may copy a number of even prose passages,
but we count the number of syllables they all make ; and divid-
ing it by 32 we obtain the number of the slokas which he has
copied and according to which he is to be remunerated. But
this is a most modern sense of the term, and is not warranted
by the Sanskrit literature, or, for the matter of that, any lexicons.
The natural sense of the word must therefore be taken, and we
must draw the natural inference that the original work of
Kautalya was in verse and not in prose at all. This is exaotly
in keeping with the fact that the works of the previous authors
also were in verse, so far as we know. For elsewhere I have
shown that we find verses actually quoted from Manu,
Bìhaspati, Uśanas (Bhārgava) and even Bhāradvāja in the
Sāntiparvan of the Mahābhārata.16 There is therefore nothing
unusual if the work of Kautalya also was originally in verse.
This may be proved also in two other ways. On p. 17 of the
published text there are verses, the first of which gives the
opinion of his acarya on a certain matter, and the second cites
the view of Kautalya, as is clearly indicated by the phrase
etat Kautalya-dar sanam . And the natural presumption is that
all the verses occurring in the Arthaśastra emanated from his
pen, unless the contrary is indicated. And, as a matter of fact,
there is suoh a contrary indication at least in one place. Thus
on pages 365-6 are found two stanzas with the prefatory remark:
ap=īha ślokau bhavatah. This remark unmistakably shows that
these stanzas at any rate were not of his composition. And, as
a matter of fact, we know that the first of these is met with in
the Parãéara-dharma-samhitã and the second in the Pratijña-
Yaugandharayana drama of Bhāsa.17 This trait of composition
is noticeable also in the Kāmasūtra where though there are
many verses without any such preface, there are a few which

16. Carmichael Lectures, 1918, pp. 104 and 185-91.

17. Ibid., p. 100.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
78 Annals of the Bhandarkar Institute .

are preceded with the remark bhavanti c=ãtra élokãh. This trick
of style is obviously intended to show that the verses to which
this remark is prefixed are not original, but borrowed from other
sources. Where therefore this is not noticeable, the natural
conclusion is that the verses form the author's own composition.
We can therefore safely assert that all the verses occurring in
the Arthaáâstra belong to Kautalya, unless the contrary is
indicated. This conclusion is further confirmed by the fact
that we have at least two instances of verses prefaced by some
words in the prose which each by itself are unmeaning but
which together make the sense clear and complete. This
practice of combining a verse with prose to express one idea is
often met with in dramas which are each the work of one
individual author. This also indicates that the? verses in
question pertain to Kautalya. If still further confirmation an
this point is required, it is furnished by what Daņdin says
about the Daņdanīti of Visņugupta, which was another name
of Kautalya. There Daņdin tells us that this Daņdanīti is an
abridged work of 6,000 slokas. The number of the slokas here
specified is exactly the same as that mentioned in the Arthaáã-
stra. And the word 'abridged1 ( samksipta ), which corresponds
to the term samhrtya at the very commencement of the latter
work, shows that Kautalya constructed this compendium from
previous works, consisting of 6,000 verses. In the same con-
nection. Daņdin quotes a view bf Kautalya which, it is worthy
of note, is traceable, not in a prose sūtra , but in a verse of the
Aīthasāštra18. Is it not quite evident from this that the book
of Kautalya which the author of the Dašakumāracarita had
before him was not in prose but in verse ?
No reasonable doubt is permissible that there was a time
when the Arthaéâstra of Kautalya was wholly composed of
verses. And, in fact, Daņdin is the latest Sanskrit author, who
is aware of this metrical form of the composition, It deserves
to be remembered that all the writers, prior to Daņdin, who
quote from the Arthaśastra, quote verses, and that those who
wrote subsequently cite not verses so much as sūtras from it.
This is a very important point to note. Let us now, in the
first place, pass in review the quotations made from this work
by authors who flourished before Daņdin. One work earlier
than the time of Daņdin is the Nāradasmrti , which has been

18. Daêahumaracarita ( NirņaysSgara Ed. ), p. 261 ; Arthišāstia,

p, 249.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Date of Kautalya. 79

taken to belong to the fourth or fifth centur

was the first to note that two verses almost
this work are found among the ending ve
Chap. I. of the Arthaśastra. They are as f
Dharmaś = ca vyavaharaś = ca = ca
Dharmaś~ca vyavaharaś
caritram rāja-sāsanam' I caritram rāja-sāsanam I
Vivādārthaš = catuspādah catuspād = vyavahäro = yam
paścimah pūrva-bādhakah II = uttarah pūrva-bādhkah II
Tatra satyam sthito dharmo Tatra satām sthito dharmo
vyavahāras = tu sāksisu I vyavahāras = tu sāksisu I
Caritram samgrahe pumsām Caritram pusta-karaņe
rājnām = ājnā tu śasanam II rājājnāyām tu śasanam II
Arthaśastra, BK. III. Nārada-Smrti I. 1-10-1 L
Chap. I.
To about the same period belongs Vãtsyãyana, the author
of the Bhãsya (commentary) on the Nyaya-sutra of Gautama.
While extolling the merits of his science Nyãya-sãstra, also
called Anvīksikī, he quotes one verse from the Arthaśastra.
It is as follows :
Pradīpah sarva-vidyānām Pradīpah sarva-vidyānām
= upãyah sarva-karmaņām I = upāyah sarva-karmaņām I
ãárayah sarva-dharmāņām aśrayah sarva-dharmāņām
Viddyoddeśe prakīrtitā II šašvad=Ānvīsakī matā II
Nyāya-bhāsya Arthaśastra, BK I. Chap. II.
The last line of this verse, which entirely varies and which
refers to Vidy-oddesa , the name of Chapter II, of Book I. of the
Arthaśastra, clearly shows that it is Vãtsyãyana who is quoting.
Then I have elsewhere pointed out 20 that there is one
verse occurring on p. 217 of the Arthaśastra which is found not
only in Manu (XI. 180) but also in the Baudhãyana (II. i. 35)
and the Vāšishtha (I. 22) Dharmasūtra, which pertain to the
third or second century B. C.21 The verse is as follows :
Samvatsareņa patati Samvatsareņa patati
patitena samācaran I patitena sah = acaran I
Yājanādhyāpanād = yaunat Y âjanâdhy āpanād = yaunat
taié = cânyo-pi samācaran il na tu yān-āšanād = iti II
Arthaśastra Baudh. & Vas,

15. Sacred Books of the East* Vol. XXXIII., Intro, p. XVI.

20. Carmichael Lectures, 1918, p. 103, n. 2.
21. Camb . Hist. Ind., Vol. I. p. 249

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
80 Ànnats of the Bhandarkar Institute .
Now, it seems tempting to argue that as this verse is con-
cerned more with Dharma than with Artha, Kautalya must be
the borrower, But it is worthy of note that the Vasistha-dharma-
sütra quotes this verse with the prefatory note ath=āpy = udā-
haranti. It is thus evidently a quotation. Then, again, when-
ever this Dharmasūtra cites a verse from a work connected
with Dharma- sastra, it is careful enough to specify its nameś
Thus such verses in the Vasistha are prefaced with ath=āpi
Bhāllavino Nidānegāthām = udāharanti, or Hārito =py = udā -
harati , or Manavam châtra slokami udãharanti , and so forth. It
is thus clear that the verse in question can not be a quotation from
a reputed work of Dharma-śastra. And when, on the other
hand, we bear in mind the fact that even Dharma (Law) was
originally part and parcel of the Arthaśastra as is evident from
the Book entitled Dharmastbîya comprised in it, the conclusion
seems almost inevitable that the verse is really borrowed, not
by, but from, the Arthaśastra.

Nay, there is one verse in the Arthaśastra which has been

traced in an earlier work still. There is one Chapter in the
Kautaliya which is entitled: "Consideration of Profit and
Loss in Men and Wealth." This Chapter closes with two
verses, the first of which scoffs at placing too much faith in
the stars. And it deserves to be noticed that practically the
same verse is met with in a Pāli Buddhist Jātaka. We will
place both the verses side by side for comparison :
Naksatram =atiprchchhantam Nakkhattam patimānentam
bālam =artho = tivartate I attho bālam = upaccagā I
artho hy = arthasya naksatram attho atthassa nakkhattam
kim karisyanti tārakāh 11 kim karissanti tārakā II
Arthaśastra, p. 349. Jāt. I. 258.

There can hardly be any doubt as to one of these verses

being a copy of the other. And the question arises: who bor-
rowed from whom ? If we carefully read the Jātaka, we find
that the gathā in question is by no means a necessity for the
development of the story, which is quite complete and clear
without it. On the other hand, the verse has its propriety only
in the chapter of the Arthaśastra just referred to, and explains
what immediately precedes it, that is, explains why 'faith in
the auspiciousness of lunar days and stars' is 'an obstruction
to profit'. Now, the Pāli scholars tell us that the canonical
Book of the Jātakas contains only the verses, and was composed

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Daté of Kautàlya . 81

before the time of Asoka.22 Here then we have a verse cited

from the Arthaśastra in a canonical book of the Buddhists,
which is believed to be prior to the time of Aśoka. Can any
reasonable doubt be entertained after this as to the Arthaśastra
not only being in verse originally but also being earlier than
the time of Aśoka ?

In the above attempt at collation we have not considered

the verses from the Kautalîya which we find cited in the vari-
ous Parvans of the Mahābhārata. The number of these verses
is by no means small and clearly shows that the work was
popular even in the early centuries of the Christian era. Some
of these verses have been referred to in my Carmichae! Lectures
for 1918. But the inference that is forced on us and which we
have to emphasize here is that from the time of Daņdin upwards,
whenever any quotation is made from the Arthaśastra, it is
the verses that are cited. It is curious that not a single prose
passage from it, except perhaps in one doubtful case, we find
quoted in any work prior to Dandin. The thing, however, is
entirely different, if we take the period which followed this
poet. We now find that in almost all the cases the sidras , and
not the verses, are cited from the Arthaśastra. The earliest
author after Daņdin who quotes from this work is Bhavabhūti,
who in the Fourth Act of his drama Mahāvīracarita has
laghv=api vy asariam = abhiyuktasy a krcchram bhavati , which
occurs exactly like this in the Kautalîya (VII. 5 ; p. 273 ). The
Jaina author, Somadevasūri, who flourished in the tenth cen-
tury A. D., frequently quotes the very words of Kautalya in
his work Nitivãlcyãmrta but with such cleverness that it is
difficult to discover it. Nevertheless, some of these citations
have been laid bare by Dr. Shamasastry and Mr. K. V. "Ranga-
swami Aiyangar, which clearly show that they are prose
passages, and not verses. Similarly, Medhãtithi, (9th century),
the earliest commentator on the Manu-samhita , not only
mentions Cânakya by name as the author of Nīti] but also
embodies several long pTóse passages from the latter in his
commentary. " Of other commentators drawing on the
Arthaśastra, we may mention Hemacandra's commentary on
his own Abhidhānacintāmaņi, the two printed commentaries
on the Kâmandakîya Nītisāra, Kslrasvāmin's commentary on
the Amarakoãa, Mallinātha's commentaries on the Ragbuvamśą

22. Hbya Davids, Buddhist Tndia} p. 206,

11

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
82 Annals of the Bhandarkar Institute»
and Kumārasambhava "23 and so forth. All these commen-
taries freely quote from the Kautalîya, and almost all of these
citations are from the prose portion of the work.
It is thus clear that originally the Kautalîya was wholly
in verse and continued to be so up till the time of Daņdin.
And the question now arises : when was it reduced to the
Sūtra form ? The earlist author who quotes a sütra from the
Arthaśastra is Bhavabhūti, as we have seen above. And it
seems tempting to say that this wholesale recasting of the
Kautalîya took place between the time of these two literates.
But a very short interval seems to have separated them,
Daņdin having flourished about the end of the seventh and
Bhavabhūti about the beginning of the eighth century. The
Sutra form of the work must have been in existence for a
century or two at least before it was known to Bhavabhūti.
And what appears to have been the fąct is that both the forms
of the Arthaśastra, prose and verse, must have been in existence
together for some time at least, perhaps in the different parts of
India. No other supposition can explain how, whereas the
metrical form is referred to by Daņdin, the Sūtra form was
before Bhavabhūti, when scarcely more than a quarter of a
century'-separated one from the other. It therefore seems safe to
say that the Sūtra form arose in the fifth century A. D., that is,
two centuries prior to Bhavabhūti. Certainly it must have
sprung into existence after the compilation of the Yajfiaialkya
smrti , which is generally referred to 350 A. D. If any one
carefully compares the parallel passages between this Smrti and
the Kautalîya culled together by Dr. Shamasastry and Prof.
Jolly, it gives the impression that the compiler of the Smrti is
not giving a versified translation of the prose Sūtras but rather
quoting actual verses from the Arthaśastra. This date may
seem to be in conflict with that (namely, fourth century)
assigned to the Kāmasūtra, which, as we have seen, actually
quotes from the Sūtra redaction of the Arthaśastra. But then
it is worthy of note that what has happened to the latter has
happened also to the former, namely, that originally the
Kāmasūtra also was in verse and was at a subsequent period
transformed into Sūtra. This is evident from the fact that
the very first chapter of this; work also tells us about its
close that it consisted of one thousand and a quarter verses
and that some of these are certainly the verses that are

£3« Kaut . Artha . ( Pudj, ßk. Series ) Intróf, pp. 11-2.

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
 Late of Kautalya ^
still found in the body of the book excepting those which
prefaced with bhavanti c~atra slokãhé The date, namely,
fourth century, which is attributed to this work is really
date of its subject matter and also of its original metrical fo
And the present text of the work can without any impropr
be ascribed to a period later than this century, which accord
to us is the date also of the present Arthaśastra.
But where was the necessity of giving this Sutra fo
to the original metrical composition of the Kautalîyaï W
could have been the occasion for reducing this work to a
of Sūtras ? This is the question that must arise in this conn
tion. It is very difficult to give any convincing answer to
this question. But what appears probable is that as the popu-
larity of the Arthaśastra increased, a sort of school came to be
formed and associated with Kautalya's name, to represent his
views which, just because they were instinct with practical
knowledge and political wisdom and had a unique character
of their own, were fitted to be the centre of a school of polity. It
is true that the word darśana occurs in the body of his book
in the passage iti Kautalya- dar śanam, as we have seen above.
But here the term seems to signify not 'a sohool, a system ' but
simply 'a view, an opinion*. The word is similarly found in
the Nitisāra (II. 3. 6) of Kāmandaka, but here too it bears the
same sense. But almost at the beginning (I. 1. 7) of the same
work the same word occurs, but evidently in a different sense.
Here Kāmandaka tells us that for the subject-matter of his
book he is indebted to the darśana of Kautalya, which can only
mean 'a system, a school ' It thus seems that a school asso-
ciated with the name of Kautalya came into existance, probably,
shortly before Kāmandaka wrote. All schools of Dharma-śastra
or of Philosophy had their own Sūtras which were ascribed to
their reputed founders. A necessity must therefore have been
felt of having a similar set of Sūtras for the school of Kautalya.
This in all probabltity is the reason why the original verses of
his work were mostly transformed partly into Sūtra and part-
ly into Bhãsya. Most of these must have contained exegetics,
and details which could not very well be put together as
Sūtra, and were therefore relegated to the Bhãsya. That
the work of Kautalya was gaining in popularity and held the
field for a long time has been pointed out by more than one
scholar. When this process of transformation was taking
place, it is quite possible that a few strange or late things
plight have found their way into the new book, Thus the

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms
84 Annals of the Bhandarkar Institute .

Sūtra : aurase t^ütpanne savarvãs^trtiy-amèaharãht which

really belongs to the Baudhayana Dharmasīitra ( II. 3. ii ) was
imported into the Arthaśastra (III. 7 ; p. 164 ). There were
other interpolations probably suoh as references to China,
Hārahūra, and so forth. It was at this time that the last
chapter of the Tanrtayukti seems to have been added. These
yuktis are thirty in number as in the case of the Suśruta , and
are precisely the same except one.24 They are all devices for
the exposition of a system. They are mentioned by the
Arthaśastra, but omitted from the Kāmasūtra, probably because
the first had risen to the position of a full-fledged darśana
while the latter had not. Bat these interpolations or additions
are so few and far between that the original matter of the
treatise remained practically intact, though now cast in another
form. This seems to be the only explanation that satisfactorily
unriddles the puzzle, the puzzle, namely, that the subject matter
of the Arthaśastra belongs to the Mauryan or pre-Mauryan
epoch and the style of composition to the early period of the
Christian era. The explanation is, we may repeat, that the
Kautaliya was originally metrical in form and came to be
reduced to the Sūtra in the fourth century A.D. without, however,
any violence to its internal contents.

24. My attention to it was drawn by Dr. B. M. Barna,

This content downloaded from 117.240.50.232 on Thu, 02 Apr 2020 16:15:55 UTC
All use subject to https://about.jstor.org/terms