Miles of Clay: Information Management in the Ancient Near Eastern Hittite Empire

by Theo P.J. van den Hout

n the late summer of 1325, the Hittite king Suppiluliuma received a written request from the widow of the pharaoh of Egypt for a son: this son was supposed to marry her and become the next pharaoh. Suppiluliuma was suspicious and sent his chamberlain first to check out the situation in Egypt. Given the distance and the traveling conditions as they were in those days, the Hittite envoy did not return until the following spring. In front of his highest cabinet members and the Egyptian embassy that had accompanied the Hittite official, Suppiluliuma solemnly declared that he would grant the request and send one of his sons. In the biography of Suppiluliuma, his son and eventual successor would later describe the situation:

Then my father took a decision for them concerning the matter of the (requested) son and then my father asked for the treaty tablet (which described) how formerly the Stormgod had taken the people of the town of Kurustama, Hittite subjects, had brought them to Egypt and made them Egyptians, and how the Stormgod had established a binding bond between Egypt and Hatti among each other, how they were for ever friends among each other. As soon as they had read the tablet aloud, then my father spoke to them as follows: "In the old days Hattusa and Egypt were friends among each other, but now this too has happened between us, so that Hatti and Egypt for ever after will be friends to each other!"
From this and many other passages it becomes clear that Hittite kings were able to order specific tablets from their tablet rooms: in this case Suppiluliuma stresses the solemnity of the occasion by ordering and having read out loud an earlier, almost century-old agreement between the two countries. Either the original or a copy of the original must have been filed somewhere and was retrieved from the tablet rooms.

Middle column of a typical shelf list. Every paragraph marked by the horizontal lines contains another entry describing a single tablet: note that each entry starts with the numeral "1" followed by the word for "tablet."

This is quite an achievement if we think of the numbers involved: right now we have some 35,000 tablets and especially fragments of tablets. If we assume an average of 10 fragments to a tablet, we are talking about 3,500 complete tablets. However, this would only be true if all of those 35,000 pieces would ultimately all fit together as one big jig-saw puzzle which, of course, is not the case. Many fragments have been lost over the millennia and returned to dust. A lot of material, as we will see, was recycled or destroyed by the Hittites themselves. So we will certainly not be far off if we set a total of complete tablets at any time stored in the capital at roughly double that amount.

This is also an achievement if we look at the circumstances. In many cases we know how they were kept: throughout the ancient Near East we see rooms lined with shelves or with free-standing shelves. Sometimes we find walls with series of niches in them that would hold tablets. On the shelves tablets could be simply stacked or kept together in baskets. Sometimes jars would be used to file all documents pertaining to a certain period, like a fiscal year. In other cases tablets were kept in wooden chests or clay boxes. Such tablet rooms were regular rooms which must have needed light in some way. Good light is absolutely essential to read cuneiform from a tablet, for the ancient scribe as much as for the modern cuneiform scholar. The flickering light of a torch is far from ideal but having to pick each single tablet and bring it over to a window to see what it says, seems quite a hassle.

The Hittite corpus

In the Hittite chanceries scribes were continuously busy copying older compositions, often updating them. This did not happen, however, with purely administrative texts. They lost their relevance to the current administration after a certain period and were as a rule discarded, i.e., destroyed, recycled or used as fill for new buildings or walls. That is why of these genres we only have texts dating to the second half of the thirteenth century, the last period of the Hittite capital. And as a rule they exist in single copies only.

The Temple complex with, in the center, Temple 1 (red), the store rooms (yellow) and the area that may have contained workshops (green).

The texts that were copied were the ones that the Hittites themselves for different reasons wanted to hold on to, the texts they wanted to preserve for later generations. Often the older exemplars were preserved as well and stored for several centuries.

This dichotomy already suggests something of a system. But what can we say about that? We know how tablets were stored but that tells us little about the internal organization. Were records kept according to years or by subject? How did the various storage offices communicate with one another? Were their special tablet formats for different kinds of documents?

The contribution of archaeology

Theoretically, archaeology should have been able to answer at least some of these questions. We know that tablets were usually kept on shelves much like our books today. If shelves for some reason collapse, most of the books will be roughly in the same order on the floor as they were on the shelf. This is also true for tablets. When in 1906 the German Hugo Winckler started the excavations at Boghazkvy, the site of the former Hittite capital Hattusa, he found in several spots a situation which might have been ideal: working his way upwards on the west slope of the acropolis he found "ever larger pieces, the higher one came, until here and there large tablets completely preserved and often grouped together came to light."

Clearly, when Building E, as we now call it, standing on the edge of the acropolis collapsed, many tablets rolled down the slope and those that kept rolling broke into more and more pieces. Even better seems to have been the situation in the Lower City where the great Temple (Temple 1) stood. One of the assisting archaeologists later recalled "that in storage room 11 there lay . . . complete and neatly layered rows of slantways lying, completely preserved tablets."

Unfortunately, however, Winckler was interested in reading the tablets only and a wealth of information was lost. Of the approximately 10,000 tablets and fragments found during the first four campaigns conducted between 1906 and 1912, we hardly know of any findspot, let alone the order of the tablets in these "complete and neatly layered rows." This information could have told us what sort of tablets were kept in the temple and which up in the palace area on the acropolis, and it might have given us invaluable information on the system the Hittites used for putting tablets on a shelf.

click to view map Although archaeological methods had become much more sophisticated when excavations resumed in 1931, these almost ideal conditions were not encountered again. Furthermore, the dispersal of tablets over the capital has not encouraged modern scholars to credit the Hittite scribes with a talent for organization. Tablets have been found scattered all over the large area of Hattusa; we can distinguish 13 places of primary storage. Yet, if we look at the division of tablets over these places, it often is very hard to detect a sense of organization. For instance, the correspondence between the Egyptian and the Hittite court was unearthed in the Buildings A and E on the acropolis as well as in Temple 1 in the Lower City, a situation characterized by its German editor as "a bit of a mess." Such cases are very normal and some scholars have therefore denied the existence of specialized archives in the Hittite capital. But the fact that we find it difficult to detect a system in Hattusa in all probability says more about us than about the Hittites.

Moreover, there are consistencies in the alleged chaos. By way of an example we can point at the corpus of administrative or bookkeeping texts, the overwhelming majority of which stems from the acropolis with a concentration in the Buildings D and E. Similarly, most of the state treaties have been found in Temple 1, where according to the texts they were indeed deposited before the god.

Arguments for a system from the texts

The guard who brings in the defendants stands behind the gold-spear-man. When the king asks for a case (to be tried), the guard picks one out and puts it into the hand of the chief guard and tells the chief guard what the case is about; the chief guard then tells the the king.
That there must have been a system to retrieve information effectively from the tablet rooms, is an inescapable premiss of any investigation into this matter: an international power like the Hittite empire cannot have done without. As for the arguments that support this, our tablets themselves bear the marks of organization. First of all, there are the colophons. Often scribes would "sign" a tablet they had written. A typical colophon reads as follows:
Tablet no. seven of the third day; finished. "When the king celebrates the Monthly Festival." Hand of Tarhuntazidi, son of Pidda, written under the supervision of Anuwanza.
The scribe numbers the tablets in a series, indicates whether the present tablet is the last one of that series, gives the series' title and finally identifies himself and his boss. Beside these colophons, we have so-called shelf lists: inventories of tablets with all kind of information. Compare the following passage from a shelf list found on the acropolis:
Two tablets: "When the king, queen and princes give substitutes to the Sungoddess of the Earth"; finished. Its first tablet we did not find.
One tablet: text of Annana, woman from Zigazhur. "When I invoke the deity Miyatanzipa." Finished.
One one-column tablet: "When a singer brings a libation in the temple of the deity Inar, breaks thick breads and prays in Hattian." Finished.
One tablet of the Zinduhi women, how they speak in the presence of the king in the temple of the Sungod; finished.
One one-column tablet: songs of the men of Istanuwa; finished.
One tablet of a treaty: When Ispudahsu, king of Kizzuwatna and Telibinu, king of Hatti concluded a treaty. Finished.
Again, tablets are put in and identified as part of a series, titles are given and there is information on whether the entry on the shelf is complete or not. In the past these shelf lists have always been taken as reflecting the order of tablets on the shelves. But at the same time people have rightly wondered what the treaty tablet at the end is doing among the other mostly religious compositions. However, inventorying should not be confused with storing: the order in an inventory does not necessarily match the order on the place where tablets are kept.

The question posed earlier, how scribes found their way among the hundreds of tablets on a shelf, is partly answered by the labels: these are small, round tablets with the title of a composition written on it. We imagine that these labels lay on the shelves in front of the series bearing the name on the label saying, for instance,"Tablets of Mursili's manly deeds." Also, we have what we indicate with the German term Sammeltafeln, that is, collections or dossiers of shorter but separate compositions on larger tablets. All this points at activities aimed at bringing some order to the large number of tablets with which the administration dealt.

Archives or libraries?

. . . and the affair began to weigh on my conscience. . . . Then I found two old tablets.
In general the tablet collections of Hattusa have been characterized as libraries and only rarely as archives. The observation that we have so many copies or duplicates was taken as characteristic of a library rather than an archive. This was reinforced by another observation: in a state archive one would expect the originals of state treaties or royal edicts and proclamations. In fact, we hardly have such "originals," and it has even been suggested that the Hittite state archives have not yet been found. However, with the ongoing excavations now having covered most of the area of the capital, this seems hardly tenable any longer. The observation remains very important, though, and is a question we will have to come back to.

click to view map Most definitions of the terms "archive" and "library" used in ancient Near Eastern studies use contents as a guide in determining the status of tablet collections. Archives store legal, economic and administrative documents that as a rule exist in single copies only. Libraries, on the other hand, have texts in multiple copies which contain mostly "literary" material. But reducing the "archive-library" question to the distinction literary versus non-literary is shifting the problem to another equally if not more difficult issue. Modern archival science gives another, far more general definition:

An archival collection is the whole of the written documents, (. . .) officially received or produced by an administrative body or one of its officials, in so far as these documents were intended to remain in the custody of that body or of that official.
Applying this definition to the situation of the second-millennium Hittite empire, every written document that somehow belongs to the business of its administration is an archival piece. Note that nothing is said about the contents of documents. The business of a particular administrative unit will determine the contents of its archive. Since we know that in the Hittite empire religion was an integral and essential part of the administration, there is no need to be surprised at the overwhelming presence of documents of a religious character in the tablet collections or to use that observation in determining the character of the tablet collection. It is also essential to keep in mind that an archive is "an organic whole, a living organism, which grows, takes shape, and undergoes changes in accordance with fixed rules." An archive grows passively: in and of itself it stores all it produces and receives. This is in contrast to a library that actively chooses what material it wishes to collect.

Besides this we need to make a distinction between "living" and "historical" archives, to adapt a more international usage of the terms. In spite of the lack of a standard use of the word archive and the fact that languages sometimes use different terms, the basic distinction is made throughout:

"living archive" "historical archive"
English administration/record(s) archive(s)
German Kanzlei-/Registrargut Archiv
French (Italian, Dutch etc.) archive vivante/administrative archive historique

A living archive is what any administration of current affairs builds up and needs to fulfill its administrative functions. After documents have lost their immediate relevance for the present administration and have become inactive, they will be either discarded or moved elsewhere from the "living" to the "historical" archive. When that is, is each administration's or its supervising body's decision. Often a major political-administrative change will be a reason for discarding documents: a new king, for instance, would do wise to hold on to his predecessor's dossiers but might decide to use the opportunity to remove those of the latter's predecessor. Often a period of 25 years or one generation seems fairly standard. Generally, historical archives are considered a relatively modern phenomenon and almost all ancient Near Eastern tablet collections can be considered "living archives."

If then, we want to characterize most of the Hittite tablet collections as archives instead of libraries, what do we do with the multiple copies and the older tablets? The former do give a library impression and the latter should have been thrown away. Archival science deals with these problems in a very satisfactory way. As we already saw, maintaining the cultic calendar and keeping the festival scenarios up-to-date was as much government business as drafting diplomatic documents like treaties, writing instructions and letters, keeping track of incoming and outgoing goods, etc. The older manuscripts were a vital part of the archive necessary in the production of new texts that were in constant demand. In archival terms they are called "retroacta": earlier documents relating to the same matter.

Such an approach is more difficult in the case of historiography like the "Biography of Suppiluliuma" which tells the story we started out with. These works were compositions that needed no updating, yet they were a necessary source for other works like historical preambles to the Hittite state treaties. For this, archivists reckon with the existence of "archive libraries" or as they are called in German "Dienstbibliotheken." Although this combination may at first sound contradictory, it designates material that an archive needs to do its job. Every archive possesses works that are not part of its ex officio collected materials but are necessary to perform its duties. The distinction between retroacta and an archive library may at times be blurred: copies of older state treaties such as the "Treaty between Telipinu and Isputahsu of Kizzuwatna," mentioned in the shelf list above, were as much a historical source which a king might use in writing historiography as they were an essential element in drafting a new treaty.

click to view map In trying to make do with what we have, we should keep in mind a number of things. First, as was stated earlier, an archive is a living organism where documents, dossiers and parts of dossiers circulate among different offices. At the moment that the life of an administration is terminated such as at the end of the Hittite empire, we should not expect a situation where every archival item has neatly arrived back at its original or final destination. Neither is it realistic to expect, for instance, all the Egyptian correspondence to have been kept in a single place. Keeping documents together according to their contents is only one way of organizing them. Instead of interpreting the various findspots as a sign of the Hittites' inability to organize their own material, we would do better to see it as potential information on the "paper trail" of the Hittite bureaucracy.

This brings us back to the suggestion mentioned earlier, that the Hittite state archives have not yet been found. It is striking indeed that hardly any original documents have been found. But it might be that these were exactly the materials that the last king and his scribes took with them. We now know that at the end of the Hittite empire the élite abandoned the capital, taking with them their most important possessions. The originals of one's important papers proving the legitimacy of certain rights and possessions are the first things that one takes when having to move somewhere. This is probably what the Hittite king and his retinue did as well. The treaty originals deposited in the temples were in a way the warranties of their political system and especially their foreign policies, proving their relations with other parts of the empire and former allies. Moving those tablets to their new residence was proof of the intended ongoing existence of a Hittite empire.

A good illustration of the importance of such documents brings us to Troy. A king by the name of Walmu was expelled in a coup d'itat and he fled to a king in the vicinity. He was not given opportunity to gather his important papers and was unable therefore to prove that he was the rightful claimant of the throne. Fortunately, however, the Hittite king, his overlord, kept a copy of Walmu's investiture in his archives. So the Hittite king sent an envoy with the necessary documents to prove the legitimate claim of Walmu. In the letter from the Hittite king we read:

"The documents that Walmu [had ... ,] Kuwalanazidi has kept them, and Kuruntiya will now bring them to you, my son. Read them! As long as now you, my son, protect the well-being of My Majesty, I, My Majesty, will be convinced of your good will. So, my son, send Walmu to me and I will reinstall him on the throne in Wilusa."


According to the definition of archives given earlier, most Hittite genres can be classified as archival material and most of the tablet collections as archives. But not all genres of texts found in the tablet collections in Hattusa can be easily described as falling under an administrative heading. What to do with foreign myths like the Gilgamesh Epic in Hittite adaptation? Some or all of these seem to have been kept out of some kind of academic interest, either educational, aesthetic or historical, but they played no role in the administration of the Hittite state; they all represent foreign material that must have been actively collected. If we are looking for library texts, they are here. It is striking to see that the findspots of most of these are very consistent. This foreign material, of which we know the findspot, was found almost exclusively in Temple 1, with some pieces in Temples 15-16.

It would seem that the temples took an active interest in translating and adapting foreign material and kept it on their own shelves. These small collections come closest to what one could call libraries. In order to do so safely, we would of course need to know whether they were in some way kept together and/or separate from other material. But again, that information has been lost irretrievably.


In the end we can describe most of the texts that have come down to us as archival, and the rooms and buildings that housed them as archives. What emerges is the picture of a lively administration that must have been spread out over several locations in Hattusa. There were probably offices with different competencies, exchanging documents and dossiers and each performing its share in the paper trail and red tape of the empire.

Compositions of foreign origin without any clear relevance to the administration, on the other hand, seem to qualify as library material. Most of these may have been concentrated in the temples, especially Temple 1.

To refine this picture, for instance, to define the competencies of specific "offices" and to trace the route that individual documents traveled through the Hittite bureaucracy may never be possible. A lot of necessary archaeological information has been lost, disturbances dating to the days of the Hittite empire and in the centuries following its demise have done irreparable damage, and we simply cannot expect to be able to fully reconstruct the workings of a civilization that has been "dead" for so long. Still, different ways of looking at our material and new approaches in research may yet add to our knowledge of the Hittite administration.

ABOUT THE AUTHOR | Theo P.J. van den Hout

Theo van den Hout is probably best described as a philologist with strong linguistic interests. Linguistics was the initial focus of his research but it gradually shifted to more general questions of a cultural-historical nature. His dissertation (1989, published as a monograph in 1995) consisted of both a philological edition of a Late-Hittite treaty and a prosopography of the leading officials of the Hittite empire of the second half of the thirteenth century BC. That period has since then been the center of his further historical-philological research. This resulted in his second book (1998) which identified and edited a large group of oracle texts and used it a source for Hittite history. His recent philological project concerns a new edition of the extensive Royal Hittite Death Ritual.

In his linguistic research he tries to combine the Hittite data with those of the other Anatolian languages like Luwian, Lycian, Lydian and Carian. The growing importance of the not-related Hurrian has led him in recent years to get more involved with that language as well.

Together with the emeritus Harry A. Hoffner he is co-editor of the Chicago Hittite Dictionary and currently is working on the S-volume.

COPYRIGHT | A version of this article was delivered at the Humanities Open House at the University of Chicago, October 26, 2002. Copyright 2002 the University of Chicago.

(c) 2004 The University of Chicago :: Please direct questions or comments to