COMPUTERIZATION OF THE THAMMASAT VERSION OF THE KOTMAI TRA SAM DUANG
Yoneo ISHIII

The Kotmai Tra Sam Duang or the Laws of the Three Seals
1.
The Kotmai Tra Sam Duang (hereafter Abbreviated as KTSD) is a popular appellation given to a corpus of traditional laws of Thailand compiled in 1805 by the order of King Rama I of the reigning Chakri Dynasty. The importance of the KTSD lies in the fact that it is virtually the single source for those who wish to study the pre-nineteenth century legal texts of Thailand. It is generally accepted that the Ayutthayan royal archives were almost completely destroyed when the old capital was sacked by the invading Burmese army in 1767. In fact, it is stated in a royal decree issued in 1794 that only a ninth or tenth of the Ayutthayan code of laws survived the catastrophe brought by the Burmese invasion. 'Phraratchakamnot Mai 28' King Rama I who ascended the throne in Bangkok in 1782., should have felt necessary to review the existing legal texts which obviously were not only far from complete but also should have understandably been corrupted in the war-stricken confusion.
The preamble of the KTSD refers to a divorce case which eventually came to reveal the inappropriate state the existing legal texts. This led the king to order a prompt and total "cleansing" or kaan chamra or the corpus of laws. The task was commissioned to a group of specialists who started the operation on January 31, 1805 and ended, it is believed, either on September 3rd December 16th of the same year. It look, therefore, nine or eleven months depending on which is to be taken as the of its completion. the results of this enterprise can be seen today in the from of Chabap Luang or the Royal Copy of the KTSD. This Royal Copy is said to have been made in three sets, each consisting of forty-one volumes of samut khoi. They were kept in Ho Khruang. Sanluang samrap Lukkhun, and Ho Luang, respectively. Besides, another copy known today as Chabap Rongsong or a secondary copy was produced.
Due to unknown reasons, however, all of these copies seem to have scattered and lost by 1980, when a systematic attempt was made to reclaim the missing copies. As a result, 80 volumes of Chabap Luang and 17 volumes of Chabap Rongsong were recovered. Thus we are able today to have some idea as to how the whole the whole corpus originally looked like.

2.
As is well known, the name KTSD or Three Seals Law is derived from the fact that on each volume of the corpus are stamped all three official seals the Ministers of the North, the South and the Central, who were respectively in charge of the three provincial divisions constituting the traditional kingdom of Thailand. The impression of the three seals may be taken as indicative of its universal application within the realm.
Furthermore the exclusive nature of the corpus is indicated in such a statement found in the Preamble that from thenceforth any text without the Three Seals should never be used in the administration of justice. It is not unlikely that all of pre-KTSD legal text should have been left abandoned at the time of or after the completion of the whole corpus, if not intentionally been destroyed. So far no legal text predating the KTSD is reported to be in existence.

3.
It is regretful that little is known today about the actual process of the 'cleansing" in 1805 A.D.. But considering the relative shortness of time spent to finish the whole operation.. it seems unlikely thet the whole texts under examination be completely rewritten. More likely is the case in which the texts should have undergone partial revision by way of interpolation, omission, addition, renovation, etc... This being the case,the finished work should most probably retain not a small of the original features. Without pre-KTSD urtext in hand, however, it is not easy a task to delineate the modified portions from the original ones. It may be hoped that the internal text critique methodology, if properly applied, could reveal the complex structure of the KTSD texts. And for this, a comprehensive index of total glossary, if any, must be extremely useful. It is with this convicion in mind that a computreization project of the whole text was planned and launched in 1978 at the National Museum of Ethnology where I then held a position of visiting professor ship.

4.
Before going into the detail of project, let me describe the structure of the KTSD.
The original corpus which is said to have 1,603 matras, consists of the following titles. For convenience sake they are grouped into the following four groups:

"Group 1" contains (1) Preamble, (2)Phrathammasat, and (3) Inthaphat of which Phrathammasat provides with an inventory to the subsequent "laksana" or titles. A legend of Manosan found in Phrathammasat is to give authority to whole texts.
"Groug 2" which lacks the Pall introduction as fonud in the texts of "Group 3" seems to require separate treatment.
"Group 3" are a body of laws, known as "laksana". Each "laksana" is usually prefaced by a pail sentence, the corresponding Pail of which is to be found in Phrathammasat.
"Group 4" are royal decrees promulgated by kings of late Ayutthyan and Thonburi-Bangkok periods covering the period the between 1527 through 1805A.D.

5.
First, the problem of eras used in different texts of the corpus. Many have been said about sakarat or era used in KTSD. In addition to the most commonly used Chulasakarat, are found Mahasakarat as well as Phutthasakrat. Beside, there is another era for which interpretations have been proposed. It is called Sakrat Kotmai by Prince Damrong, while Phiphat called it Chulamanisakarat. However, both are, rightly or wrongly, refuted by Chit Phumisak who believed that it be no other than Mahasakarat. Secondly, it is to be noted that there is no guarantee that any sakarat shown in a given text should correspond to the date when the text in question was actually made. This is easily seen from an example taken from one matra of Phrathammanun which refers to Krung Kao. The said matra is in the section of the laksana dated 1743 A.D. A.D., if the era being Chulamanisarat, or 1633 A.D., if it be Mahasakarat. In neither case, reference to Krung Kao which must be the appellation of the old capital of Ayutthaya is unnatural and must be an interpolation in a later age.
Not a few examples of other interpolations, charges, additions etc. can easily be pointed out.Thirdly stylistic differences in the time of composition to the given texts or in provenance of the given texts.
These problems shall be more appropriately approached if a relevant tool which should enable us to make comprehensive cross reference of the whole texts.

Computerization of the KTSD: an attempt to a KWIC Index

The computerization project of the KTSD was planned and started for the first time in 1978 at the National Museum of Ethnology, Suita, Osaka.
A study group was organized at the laboratory of Dr,Shigeharu Sugita, Associate Professor of Computer Science with a wide participation of computer scientists and Thai language specialists.

A problem of segmentation

Before being input to a computer, a text must be properly segmented. At first sight, the segmentation of the Thai text seems to have no problem.
But as the project went on, it was found that the operation was not as easy as it was thought at the beginning. In fact, even after nearly ten years of trail and error, we are still unable to claim that all problems have been solved.
Compared with such language as Sanskrit in which the law of snadhi often disguises the appearance of juxtaposed vowels, segmentation of Thai where each word retains its distinct from in a sentence, hardly causes problems.
But it soon came to be noticed that with mechanical segmentation alone linguistically satisfactory results were difficult to be reached.At the same time, how a given word be segmentation often depends upon the need of a user of the index which is one of the most important aims of the computerization of the KTSD. While a lexicographer may be concerned about frequency of use of such simple word as "ko", "dai", "laew", "pai", "ma",etc., a historian shall be more interested in such a combined expression as not easily be bridged by a single principle for segmentation of words.

A KWIC index and its problem

KWIC is the abbreviation of a method of index production. In this system, a chose word always appears at the center of CRT, being sandwiched by preceding and following chains of words. The length of one line is determined by the size of a given CRT and nothing to do with the meaning of the sentence or sentences of which the indicated word constitutes a part. For the users of the index, this is not always the best way of actually appears is not necessarily an appropriate one.
Oftentimes unnecessary content is given, while proper content is outside the frame of the CRT.
In the revised version which is almost completed by now, the KWIC has been replaced by another method which is widely adopted by conventional maker of a concordance, for example, Bible concordance, Sahkespere concordance, etc.,: namely, a word is accompanied by a complete sentence in which it appears.
In order to reduce the size of the index, a device is worked out. For the words whose frequency is extremely high, only its location is indicated without giving its context at all. In this, we have followed a well designed computer index the Greek new testament published in 1983 in Germany.
It is hoped that our new KTSD Index may be reduced in its size from the original 70 volumes to 3 volumes of manageable thickness.