Tosanaji Keris
by  A.G.Maisey


There is a widespread belief amongst students of the keris that the keris is a representation of a
serpent or a naga. It is not difficult to see a similarity between the waved form of the keris and a
serpentine shape, and the belief is that this waved form represents the active naga, whilst the straight
form of the keris represents the naga at rest.

The word “naga” is in fact Sanscrit for “serpent”, however, when we speak of a naga in the context of
the ancient belief systems which surround the keris, we are not speaking simply of a serpent. We are
referring to a class of semi-divine beings, half human and half serpent who were relegated to an
underground kingdom by Brahma, when they became too populous on earth. There are a number of
individually named nagas, who carry individual responsibilities and are capable of supernatural acts.

The keris in its present form has a history which reaches back to probably the 14th Century, and in its
initial form to at least the 9th Century. If it is indeed a representation of a naga, was it always believed
to be this, or has the belief developed, along with the development of the keris itself?


Perspective and Period

When we consider the keris it is necessary to consider it from a chosen perspective.

If we attempt to consider all aspects of the keris at one time, it is very difficult to arrive at a clear
determination in respect of any single aspect. Similarly, because of the seemingly changing nature of
the keris throughout its history, it is necessary to adopt a time frame into which to place our

With my “Origin” paper, I presented two separate arguments: The first proposed a line of descent for
the modern keris which commenced in pre-history and completed at the point where the weapon that
we choose to call the keris, was being used as a thrusting weapon.
The second argument was clearly identified as hypothetical and in this I adopted what I think of as the
pragmatic line of development, which is one of several possible lines, for presentation of a theory to
account for development of the form of the keris. The time frame in which this was set was the period
prior to the mid 14th Century.

In the present commentary, I will attempt to look at the keris as an object encountered for the first
time, that is, without preconceptions, and try to form an opinion in respect of the place it occupied in
society during its early history, and in contrast, the place it occupied in a similar society 500 years
later. The context for this commentary is early Java, with an emphasis on the Kingdom of Mojopahit in
the 14th Century, and 19th Century Bali. This is to be an attempt to compare the probable nature of the
keris in 14th Century Java with what it appears to have become in 19th Century Bali.


When we construct a hypothesis we take what appears to be fact, along with half understood ideas, half
truths, lies, misunderstandings, we colour all of this with our own 20th Century, Western European,
Judeo-Christian interpretations, and we construct an argument, which, if we are clever, will convince
people that we are presenting a truth.

However, the "truth” that we have presented is a "truth" which is in danger of being generated by the
culture and society from which we come.

Further, this cultural perception can be modified by our professional background. A problem may be
addressed by people from varying disciplines, the answers to the problem may vary according to the
discipline applied. However, if it were possible to test those answers against what was regarded as
truth in the time and place that generated the question, all the answers could well be regarded as
incorrect, by a person from that earlier time and place.

If this is so, then it is probably not possible for any modern man - and in this I include post-colonial
Indonesians - to understand the keris.

Not what the keris is now, but what the keris used to be, perhaps the best we can do is to try to come
close to an understanding, in our terms, of what the keris might have been for people with a different
frame of reference.


When we consider the keris:naga relationship, we are forced to draw upon sources which, for the most
part, do not go back to the origin of the keris.

We draw upon the verbal advice of people whom we believe to be reliable informants; we draw upon
old Javanese and Balinese manuscripts. We draw upon the work of earlier researchers. For one reason
or another, all of these sources are deficient.

The reliable informants present a cultural perspective which is relevant only to the time at which it is
presented. This perspective cannot be the same as the cultural perspective from an earlier period, and
as we move further from that earlier period, the possibility increases of a cumulative distortion of
cultural perspective, relative to point of origin.

Even where we are dealing with a conservative society that values its traditions and which practices a
way of life that appears not to have changed for generations, we cannot necessarily accept that no
change has occurred. The people who make up a society, and who live within that society’s culture are
not frozen in time. The combined experience of all the preceding generations within a society are what
moulds the culture of that society, as we observe it at any point in time, small changes in the way
things are done, or thought about, can occur, and within two or three generations nobody knows that it
was ever any different.

The problem with old Javanese and Balinese manuscripts is well known: a very short life requiring
repeated transcription, which can result in substitution of modern words for older words, modern
interpretations for original text, and additions by the transcriber to the text. The theme of a work may
remain the same, but detail and presentation can vary. This being so, there is the possibility with old
manuscripts that views and opinions current at the time of transcription are reflected in the new

The researchers of previous times used the same sources that we must use today, but
with the advantage of having informants who were closer to their cultural roots. However, even though
they were closer to their roots, the way in which they viewed their world cannot have been exactly as it
was several hundred years earlier. Time and the human experience alter the elements which fit
together to constitute a society and its culture.



So we are left with a seemingly insoluble problem. However, perhaps it is possible to come to an
understanding of the problem, even if we cannot provide a definite solution.

To assist in coming to this understanding, I would like to commence with what are perhaps the most
reliable sources of information on old Java. These sources were not subject to the factors which have
tended to corrupt and distort what we find in early Javanese and Balinese writings. Although these
sources probably contained error, it is unlikely that they contained any intentional fiction and their
purpose was straightforward reporting. The sources to which I refer are the old Chinese reports.
Groenveldt brought these together, and I submit hereunder an extract from the Ying-Yai Sheng-Lan

QUOTE: The king goes bareheaded or wears a cap with golden leaves and flowers:
he wears no garment on the upper part of his body, but around the lower part he has one to two
flowered cloths (sarong), and he uses a piece of flowered silk-gauze or linen to fasten these around his
loins, for which reason the latter is called loin-wrapper (slendang). He carries one or two short daggers
called pu-lak (see footnote) and always goes barefooted. He rides on an elephant or sits in a cart drawn
by oxen.
The men in this country have their long hair hanging down and the women wear it in a knot; they use a
kind of coat, and a wrapper round the lower part of the body. The men have a pu-lak stuck in their
girdle, everybody carrying such a weapon, from the child of three years up to the oldest man; these
daggers have very thin stripes and whitish flowers and are made of the very best steel, the handle is of
gold, rhinoceros horn or ivory, cut into the shape of human or devils faces and finished very carefully.
The men and women of this country take great care of their heads; if another touches it, or if they get
into a quarrel in trading, or if they are drunk and insult each other, they draw their dagger and begin
stabbing, thus deciding the question by violence. If one is killed, the other runs away and conceals
himself for three days, after which time, he has no more to account for his opponent’s life. When on
the contrary a murderer is caught on the spot, he is also stabbed to death immediately.
They do not know the punishment of flogging with whip or bamboo; for great and small offences the
hands of the culprit are bound on his back with a thin rattan and, being led away a few paces, he is
stabbed with the dagger in his side or between the ribs once or twice, until he dies; not one day passes
without a man being killed, which is very frightful.
End Quote.

in Groenveldt’s work, his footnote(3) presents an argument that the word ‘pu-lak' is a
reference to the keris.

The report from which this extract was taken refers to a visit to North Java and mentions Tuban,
Gresik (Ts’e-Ts’un), Surabaya and Mojopahit.
This is the most relevant quote to the present discussion, however, examination of Groeneveldt’s work
will provide additional information on Javanese weaponry, and importantly the nature of the people
themselves. The Chinese report native Javanese as being very black, very quick tempered, prone to
violence, lacking in personal hygiene, inclined to eat (to the Chinese) rather unusual food, such as
worms and beetles, and to sleep with their dogs.

To move from the Chinese sources to 14th. Century Javanese sources:


Canto 54, Stanza 2, Verse 4:
‘Exterminated were the animals, thrusted, lanced, cut, crissed, dying without a gasp”
Javanese: "--- Tinumbak, Inirás, kinris, pjah tanpagáp---"
This is the only reference to keris in the Nagarakertagama. (The word "kinris" = kris
with indicative infix “in”.)


•        “The criss, a token of manfulness, has its place at the front”
This is relevant to the progress of a king.
The Javanese word used for keris is curiga.

•        "---a gilded steel criss”
Javanese: “twék melela, hinémasan”.
The word used for keris is twék.
This as the reward to a man for bravery in combat.

•        "---a criss as he is pleased to use---"
This relevant to the dress of a king.
The Javanese word used for keris is curiga.

Pigeaud (Java In The 14th Century) gives two further probable references for keris in the
Rajapatigundala. and Sarwadharma 5 Verso 3.
The Javanese word used in both works is “pakris”.

In the above 14th Century writings we find three words which have been translated as keris (criss):
• Curiga: associated with a king.
• Twék: As an honour given to a common man for bravery in combat.
• Kris:         1). associated with use in a hunt
2). associated with a common man.

This use of different words all translated as keris (criss) raises several questions:
• Is the variation due to linguistic convention?
• Do the three different words refer to the same physical object?
• Do the three different words refer to the same physical object which has three different natures,
depending on its use or association?
• Are the words used original to the manuscripts?
At the present time I have no answers to these questions.


There are repeated instances of the use of a keris as a weapon in this work. There is one passage that
describes an exhibition of keris play staged as an entertainment. Only when Ken Anggrok needed a
special keris to kill Tunggul Ametung do we find mention of the keris associated with supernatural
powers. Tunggul Ametung was “Sakti”. Mpu Gandring made keris which were especially “Sakti”. So
Ken Anggrok was advised by Bango Samparan to order a keris from Mpu Gandring so that he could be
sure of killing Tunggul Ametung. (Sakti: having magic or divine power.)
(I do not have access at the present time to a copy of the Pararatan in Old Javanese, so am unable to
give the word(s) used to refer to keris in the Pararaton.)



From the sources, both Chinese and Javanese, I do not believe it is possible to draw a conclusion,
based upon the available evidence, that in 14th Century Java, the keris was looked upon as anything
other than a weapon and a male symbol.

Further, if we look at the way a keris is shown in the Panataran reliefs (Circa 1347) what we see is a
keris that looks quite similar to what we now know as a Bugis keris, both in blade form, and style of
dress. The keris shown in the famous Sukuh stele, dating from the mid 14th Century, is also a short,
straight blade, and shown together with other everyday products of a smith.

We do not find anything in these graphic representations of the keris which could suggest anything
other than simply a weapon. We also do not find representations of waved keris, but according to
entrenched Javanese belief, these did already exist in the mid 14th Century. Since the artists who
executed these carvings tended to idealise the subject matter, is it not just a little strange that we find
no waved blades? Is it possible that the waved blade did not appear until sometime after the mid 14th
Century? However, what we do find from examination of old Javanese works is mention of a rural
territorial organisation known as “dapur”, which is also the name applied to the shape and features of a
keris. Logical argument could perhaps extend this relationship to provide a link to the various keris
styles and their supposed efficacy in certain applications.


We move now to a more recent source, Wiener’s 1995 publication “Visible and Invisible Realms”. In
addition to Balinese manuscripts, and general reference material, Wiener has drawn heavily upon
information obtained from personal informants in the preparation of her text. This work gives the
appearance of having been meticulously researched, and it is an essential reference for anyone wishing
to gain an understanding of the keris.

However, the understanding which will be gained will not be the understanding of 14th Century Java.
It will be the understanding of present day Balinese, reinforced by information taken from Balinese
manuscripts, most of which were probably last transcribed in the 18th or 19th Centuries, but which
originated much earlier.

Analysis of the data on keris presented by Wiener creates an entirely different picture of the keris to
that which is to be gained from 14th Century sources. The conclusion that I draw from this analysis is
that the heirloom keris of any particular descent group was, in 19th Century Bali, to the Balinese, a
materialisation of the Cosmic Naga Basuki, which within that descent group tied the present members
of the group to their antecedents and assisted in the maintenance of an established hierarchical order.
It had the binding power of the naga, so it bound each custodian to its previous custodians. However,
this effect of binding to previous custodians was only relevant within the descent group where the
keris was an ancestral heirloom, or where the keris was given to someone outside the descent group.

Where it was given, the binding effect of the naga bound giver to receiver, and vice versa. Such a gift
was evidence of the love of the giver for the receiver, and could be a grant of power to the receiver.

The link which the keris established between ancestral custodians and present custodian provided
protection against enemies, both seen and unseen, and this protection could take effect at a distance.
Thus the keris acted as a unifying force, bringing together past and present in one dimension, and
dispersion over distance in another dimension.

But here we are speaking only of the keris in general. When we consider the reason for being of a royal
heirloom keris, I am forced to the opinion that such a kens was in fact superior to the ruler himself.
The ruler owed his origin to the royal heirloom keris, and with his acceptance of kingship came the
responsibility to maintain order within his realm.

In effect the royal heirloom keris was a manifestation of the power that embraced all dimensions of
time and space, and placed that power in the hand of the king. This manifestation of power was not a
mere symbol or representation of the Cosmic Naga Basuki, it was the Cosmic Naga Basuki, in the sense
that as Basuki flowed through the cosmos, Basuki also flowed through the keris.
Basuki existed in the cosmos as a spiritual entity, and as this entity, was responsible for the unity and
maintenance of order of the cosmos. The royal heirloom keris was a materialisation of Basuki, and in
this form, the king could approach him. Basuki maintained stability in the cosmos, and the king’s
antecedents were a part of the cosmos, thus the royal heirloom keris provided a conduit to these
antecedents, and from the antecedents, to the king. The king’s hierarchical superiors (Kings of
Mojopahit), and his hierarchical inferiors (the subjects of his realm) were also a part of the cosmos,
and the royal heirloom keris, manifestation of the Naga Basuki, unified the king and this hierarchy.

Within this belief system there are three nagas: Anantaboga, Basuki, Taksaka. If these three nagas
merge to become one naga, then that naga is also Basuki. Anantaboga is Brahma, Basuki is Wisnu, and
Taksaka is Siwa. One of Wisnu’s incarnations is Kresna. From the perspective of this belief system, the
keris becomes something quite different from that which it appeared to be in 14th Century Java. If all
the possible relationships of Basuki within the system are charted, the ramifications appear to be
extensive. It is probable that when a late 19th Century Balinese King held the royal heirloom keris in
his hand, he had full faith that he was in direct contract with the forces of the universe.

So, quite a disparity between what the keris appears to have been in 14th Century Java, and what it
appears to have become in 19th Century Bali. Bali is widely regarded as the inheritor of Mojopahit, and
pre-colonial Bali was probably as close to 14th Century Mojopahit as we are able to imagine. However,
as Pigeaud warns, “Bali should  not be regarded as a replica of Mojopahit.”

The Chinese sources identify cultural differences between early Bali and early Java, and there is the
additional factor that the view from 14th Century Bali was upwards to the King of Mojopahit, a diety,
and the only representative of divine order on earth. The view of the King of Mojopahit, and his court,
was downwards to an agent ruler. The court was staffed with politicians. Is it possible that the 14th
Century Balinese view , which provided the roots for the 19th Century Balinese view, was the same as
the 14th Century Mojopahit view? I rather think not.
Thus it is not reasonable to project the 19th Century Balinese view of the keris backwards in time and
assume that it is a view which was held, in all particulars, at any previous time or in any different
place. It is a view which is relevant only to the time and place where it is found. However, the roots of
this view probably already existed in Mojopahit, or perhaps even a little earlier.



Mojopahit was an inland, agrarian kingdom. The power of the king depended not on rights of property
ownership, but on acceptance by his subjects of his divine right to rule. His power was over people not
over property, however, it appears that this power was limited and that he had difficulty in getting his
subjects to always follow his wishes.

The wealth of the Court of the Kingdom of Mojopahit rested firmly on agriculture. Agriculture requires
a civil and natural environment which is in a state of order. War, civil disturbance, seasons which do
not fall when they should, insufficient rain, plagues of insects, all these things, and more, can disrupt
an agricultural society. It seems probable that the function of the King of Mojopahit was to provide
protection against these disruptions.

For an agent ruler to be able to maintain order within his realm, it was necessary for him to be able to
demonstrate a link to the King of Mojopahit. This link would bring the subjects of the agent ruler
within the sphere of influence of the King of Mojopahit, and once within this sphere of influence, the
agent ruler was able to ensure a state of order for his subjects, as agent of the King of Mojopahit.

Often, when a man was sent to act as agent ruler in a particular area, the King would give him a keris
to help him establish his authority to rule. This keris was then the link between the King of Mojopahit
and the agent ruler and his subjects. Since the success or otherwise of agriculture undertaken by the
agent ruler’s subjects depended on the agent ruler’s ability to maintain order within the civil and
natural environment, and since the agent ruler’s power to do this was linked to the power of the King
in Mojopahit through the physical agent of the gift kens, it is not difficult to understand the
importance of such a kens to the agent ruler’s realm.

Whether the keris:naga association was initiated by Mojopahit politicians, whether it was initiated by
an insecure agent ruler, whether it grew as a folk belief within the society itself, or whether it
originated from another source, we do not, and probably cannot ever know. What we do know is that
14th and early 15th Century sources do not indicate a keris:naga relationship in Java, whilst later
Balinese sources provide plentiful evidence of such a relationship. It is probable that the Balinese
sources do not reflect accurately the beliefs current in the 14th Century in either Java or Bali, in
respect of the nature of the keris.
We also know that agricultural societies throughout history, and worldwide, have been very
superstitious and given to ritual. The agricultural societies of Java and Bali were, and are, no different.


Perhaps the two dominant characteristics of primeval Javanese religion are the necessity for ancestral
worship and the concept of duality. Ancestor worship was considered essential to the maintenance of
cosmic order. It was the link which bound together the various parts of humanity, the living and the
dead, the visible and the invisible worlds. The concept of duality was evident in all aspects of the

The Naga Basuki was essentially a binding agent, capable of creating and maintaining unity, by flowing
through the cosmos and binding together all things in it.

It is possible that in the kens these elements were synthesized into one entity. The keris became the
place at which those ancestors who had previously had custody of the keris came together with the
Naga, linking past and present and in the present unifying the custodian with the cosmos. But the keris
was primarily a weapon, and as a weapon, it had the nature of a destroyer. So, the nature of the keris
was to unify and to destroy. A nature which is not in conflict with the Javanese concept of duality.
Destruction need not necessarily be a negative phenomenon, if by the act of destruction, the harmony
of the cosmos can be maintained, or restored.
If this is in fact what the keris became, then it follows that it was not an inanimate object, but a being
in its own right. The mantras which were used to bring the keris into existence, the ritual which
surrounded its care and custody, its protective and consultative essence all tend to support this.


I am of the opinion that the concept of keris as Cosmic Naga was a concept which had achieved full
development in Bali prior to the end of the 19th Century, but that elements of the concept already
existed in an earlier time, and were possibly used as a political tool by Javanese courts from the time of
Mojopahit. I can find no evidence to support the view that at the time the modern keris began to
appear in Java, that it appeared as a conscious expression of the Cosmic Naga.
I am of the further opinion that in the period between the founding of Mojopahit and the conclusion of
the 19th Century, political, social and cultural elements all played a part in the formation of the nature
of the keris as it was in 19th Century Bali.


This commentary is not intended to be an exhaustive examination of all factors relevant to the keris:
naga relationship. My research has been minimal, and what is presented is merely a reflection of my
present limited understanding of the question.
I would welcome comment on what I have set forth above. I do not pretend to be an expert in Javanese
or Balinese literature, history or anthropology. I have merely taken the work of other people and
subjected this work to a process of logical analysis. If the conclusions which I have drawn can be
shown to be invalid, it is because my use of logic is flawed. If my conclusions can be considered as
valid, it is because of what I have drawn from the work of others.

This article first published in "Arms Cavalcade" Vol. 1,No 3-December 2000,the official
journal of the Antique Arms Collectors Society of Australia.

This article or any part thereof may not be copied or reproduced in any form without
the express written consent of the author.