Web Magazine for Information Professionals

Ancient World, Digital World: Excavation at Halif

Paul Jacobs on how field and research strategies were impacted significantly by the use of digital technology in the 1999 field season at Tell Halif, Israel (the Lahav Research Project).

The 1999 excavation at Tell Halif in southern Israel by the Lahav Research Project was the culmination of a long development in our approach to primary research in the ancient world under the influence of digital technologies. Like other excavation projects in the region, the Lahav Research Project, too, had turned already in 1977 to record-keeping and maintenance of databases by means of "portable" computers. And while the introduction of the computer to our excavation and recording procedures was an important step, it was not until the 1999 season that we found it necessary to change our field procedures significantly in order to make use of the broader capabilities of digital recording. That is, our experience with digital recording in the field leads us to believe that the equipment itself has altered significantly the way material is (ought to be) recovered, recorded, and reported.

Chart of Halif strata

Both that conclusion about the changes digital equipment brings to the discipline and the gradual movement by the Lahav Research Project to full use of digital equipment in doing archaeology are based on factors that characterize the field of study in general. The first of these is that excavators at ancient near eastern tell sites suffer from a wealth of data; for example, even the moderate-sized Tell Halif consists of some seventeen major occupation strata, one built atop another to a depth of more than six meters. Since all the depositions that make up the mound of Tell Halif resulted from human activity, a seven to thirteen acre site (size variations in different periods) will yield thousands of artifacts in a single six-week field-season. Even extracting from consideration the most common artifact at a tell site, the potsherd, which in our system is treated separately from other artifacts, nonetheless, the wealth and diversity of artifacts may be overwhelming to a recording and reporting system. Indeed, some older excavation reports do little more than indicate the number of artifacts of a given type with no further analyses presented or attempted. In the past, archaeologists–even the most responsible ones–have found it imposible to do more than to publish representative examples from the large numbers actually recovered.

A second factor–clearly related to the first–is that final reports in this branch of archaeology are often delayed by decades. Indeed, it is not uncommon for twenty years or more to elapse between final excavation season and final publication. Often there are good reasons for the delays; time is needed to process materials in laboratory settings, and many project directors earn their wages not from archaeology directly but from teaching and administrative positions in universities. Yet, the delay in the availability of excavation data means that other scholars cannot make use of information critical to their work.

A third factor in our exploratory use (for archaeologists) of digital technology is that costs associated with standard means of publication prohibit even the best-intentioned archaeologist from displaying data in but the smallest or representative way. And the selections of materials to exhibit in monograph or journal will likely be those that are the "most interesting" or "rare," skewing unintentionally the depiction of the data towards those interesting or rare and away from the common. One of the unfortunate results of this limitation of print media is that the main body of collections remains essentially unavailable to scholars and to alternate interpretations.

Recognizing these factors the Lahav Research Project, first under the direction of Dr. Joe D. Seger, engaged in a type of experimental "pre-publication" of raw data via a website titled DigMaster. In the wake of the 1992 and 1993 seasons, in which some 545 artifacts of just one type (ceramic figurine) were recovered in Field IV, DigMaster was Tell Halif Field IV Aerial View created to allow the early display in graphic format of all of the data, not the figurines alone, from these two seasons. As a result some 1895 color photographs, 331 ink drawings, 85 QuickTime VR movies, 68 plans and soil profiles (sections), as well as summary reports for each of the fifteen 4 m. X 4 m. excavation units were placed on-line as early as 1995. While this effort does not represent "final publication" of the field data, it does make available all of our data in an early "pre-publication" stage. The DigMaster project successfully demonstrated that early distribution of data is possible and that it could be done in such a way that the resulting "database" was not skewed by selective publication schemes; we also concluded that early distribution or "pre-publication" of "raw" data is an obigation to the field of research. Nonetheless, we also came to see that the DigMaster "pre-publication" we employed was dictated by the fact that we, too, had not been prepared to manage successfully such a large number of artifacts; only "after the fact" did it occur to us that we had an obligation to make all of the data available and to do so early in the processes of analysis. That recognition led to the experimental season of 1999.

PART I. A Digital Season.

In the 1999 season, then, we determined to test the limits of the digital equipment and our newly devised recording system in a modest-sized field season, a "digital dig." Based on past experience and the numbers of finds during the 1992 and 1993 seasons, as well as on needs for additional stratigraphic and architectural information, we selected a few small segments to excavate in Fields I and IV. In doing so we planned to establish digital stations at several places on the tell in order to learn more about various strata. I had estimated that we would (in a shortened, four-week season) recover perhaps as many as 125 examples of the ceramic figurines in Field IV and as many as 100 other artifacts, such as flint blades, from Field I. We intended to have all artifacts processed (photographed, drawn, and described), all architectural drawings and profiles/sections inked and scanned, and a final field summary prepared by the end of the fifth week. In addition, our plans called for us to make regular postings of data, photos, and drawings on our website, so that experts not in the field could scrutinize and evaluate work as it appeared on daily postings.

The benefits of this digital digging proved highly beneficial, recommending the procedures to other excavation projects, though not without significant alterations to planning and approach to a field season. Of course, while much of what we planned was successful, in reality the experiment was somewhat adversely affected by the failure of communications via modem from the field in Israel to the server located in the Cobb Institute of Archaeology in Mississippi; we were unable to make daily postings as we had planned, because older equipment at Kibbutz Lahav broke transmission after 7 minutes on-line. Furthermore, contrary to expectations, we actually discovered more than three times the predicted number of artifacts; the numbers of artifacts created a backlog in processing that was not completed until our return to the United States.

The first benefit to be realized is that of a "paperless excavation"; team supervisors recorded all data on laptops which were, in turn, hard-wired to a laptop-server that assured shared information at all the various stations and assigned registration numbers as required. A typical entry by a team supervisor includes description of the activity undertaken, three-dimensional location of the activity (or artifact/object, feature, soil layer), standardized technical descriptions, and a list of relationships between the soil layer, feature or architectural element and others (below, above, abuts, cuts, etc.) These–and others were available to the supervisor on drop-down windows in the Microsoft Access application written by Christpher Holland. Record-keeping became simpler, in part because the on-screen report form provided all of the "blanks" that needed to be completed, as well as options to fill the blank or box, but also because within a few days most of the students and volunteer excavators had been trained on the entry of data. (An unanticipated benefit of the digital recording on laptops was that the newly trained students soon learned to look for relevant data as they excavated.) More importantly, the freshly entered data was immediately available to all wired locations in our excavation system: field, lab, photo studio.

Because the manual entry of data–even using a laptop computer–involves inadvertent recording errors, we also needed some check on the procedures that would reveal problems while the excavation teams were still in the field, Screen of DigDug indeed while excavation of a feature, architectural element, soil layer, or object was still in process. The creation of a three-dimensional grid by John Vander Zwaag served this purpose exceptionally well. His program (called by him "DigDug") automatically retrieved data as it was entered and represented them as points (for objects recorded in situ) within three-dimensional field and area grids, or as three-dimensional cubes, lenses or spherical or globular shapes, depending on whether the excavated locus (three-dimensional entity) was a layer of soil, a pit, a wall, a floor, etc. From moment to moment the on-screen representation of all of the excavation procedures would appear as most recently recorded or changed.

By looking at the available screens at any laptop in the system, one would know activity occurring at any other place in the field, just as one could know the disposition of an artifact, for example, since team supervisors, lab personnel, photographers, and artists entered data into the common database controlled and distributed by the server to all other stations simultaneously. Similarly team supervisors could determine immediately if incorrect measurements were entered, since "DigDug" would represent the mis-measurement as actually keyed; a soil layer that appeared on-screen a meter above the actual surface of the tell informs the supervisor that an error in addition or in entry had occurred. With zoom and rotating capabilities built in, "DigDug" became a vital on-line aid to accuracy in recording, but also to the process of excavation. An added, unanticipated benefit of the three dimensional plotting of the locations of artifacts (visible on-screen by types, represented by color assignments) was the fact that it became possible to "predict" the discovery of artifacts of the same type based on the "clustering" of artifacts that appeared on-screen. For example, as a result of observing the clustering of a large number of ceramic figurines along the western limit of Area J7 (a 4 m.X 4 m. unit of excavation) we predicted that a still larger number would be found in the same soil deposit immediately to the west; expansion of excavation into that soil layer verified the prediction. (A screen of DigDug is displayed at the left of this page, above.)

Furthermore, because some of the laboratory activities were moved to the tell itself, a minimum amount of data entry (e.g., artifact type, field identification, sizes, etc.) was completed even before the objects were sorted for assignment to the conservationist, photographer, artist, etc. The photographer, or other staff member, could then see the number and types of material that would soon appear at her station for treatment, just as supervisors could determine that photography of an artifact had been completed, because photographs were posted in the same database and were seen at all other stations, including those in the field. The immediate availability of all data was an important gain for the field procedures of the season.

Part II. The Site

Tell Halif, whose ancient name is unknown, is a medium-sized, 7.5 acre site in southern Israel, with a history of occupation which began in the Regional map Chalcolithic era (ca. 3500 BCE) down to the modern settlement of Kibbutz Lahav (founded in 1963 CE). This record of occupation is marked, significantly, by long periods of abandonment, which correspond to the re-settlement of nearby Tell Beit Mirsim (8 km to the north.) Human exploitation of the resources of the territory surrounding Halif and Beit Mirsim, particularly throughout the Bronze ages, attests that the region was environmentally incapable of supporting more than one major settlement at a time, at least until external national government support (specifically the kingdom of Judah) maintained both Halif and Beit Mirsim as fortified border towns; during the Bronze Ages, human settlement alternated between the two sites.

Because the town was located at the juncture of the Judaean hills, the Shephelah and the Negev, people at Halif (and at Beit Mirsim) faced annual rainfall of 300 mm., a fact which necessitated dry-farming techniques and, when crops failed, temporary abandonment of the environmentally marginal region. At other times the destruction of Halif or Beit Mirsim resulted in abandonment and precipitated movement to the sister site across the valley.

Early and Late Bronze Halif: Excavations in Field I

The importance of Halif resides, then, in large part in the levels and means of adaptation to environmental and political conditions. Occupation of the mound site began with the construction on virgin soil of a heavily fortified town of the Early Bronze III (Stratum XV), though an earlier village settlement (Stratum XVI) of some size, ca. 15-20 acres, had been located at the foot of the hill on which Halif was later founded; this non-fortified Chalcolithic-Early Bronze I village had been abandoned already for some three centuries before new arrivals to Halif began to clear soil to bedrock in order firmly to foot the heavy stone and brick fortification walls which they Early Bronze III house built to protect the newly established town. The curtain wall, still standing 4 m. high when exposed in Field I, varied in thickness from 3.5 m. to 7 m. and was incorporated into a protective system of outer walls and a sloped ramp of crushed limestone (to the full 4 m. height of the fortification’s stone base.) At places an additional 1.2 m. of brick superstructure survived atop the stone base. This heavily fortified town, founded probably to engage in the production and trade of lithic tools, nonetheless suffered destruction by fire less than a century later. The Stratum XV fortified town was succeeded by three non-fortified villages (Strata XIV-XII) on the tell, eventually to be abandoned until Early Bronze III chert core the beginning of the Late Bronze (Canaanite) era, ca. 1550 BCE.

In the 1999 season a house of the final Early Bronze III town (Stratum XII) was excavated as part of the experiment in digital recording and reporting. The house sat on one the highest of a series of three terraces that had been built to take advantage of the eastern slope of the tell. The terrace-wall was also the retaining wall for a narrow street that led up the slope between houses. Houses stood well above the street and were constructed of fieldstone walls topped with a brick superstructure. The compacted earth floors were covered with pottery and other household objects smashed in place in the destruction of the house. A deep stone mortar A11105 set into earth floor A11104 attests to domestic, food-production activity.

In order to establish patterns of activities in these 24th century BCE houses, artifacts were plotted on all floors. In situ objects were photographed as discovered, often drawn in place by an architect. Each artifact was then assigned a unique, identifying number and was sent the the Object Registrar to begin the process of description, photography and drawing, all on-site. The same treatment was afforded architectural features–photography, drawing, measurement and description of physical characteristics.

Following a destruction by fire of the final Early Bronze settlement (Stratum XII) the tell appears to have suffered abandonment, remaining essentially uninhabited for the next eight centuries, while neighboring Tell Beit Mirsim reflects an active settlement throughout the Middle Bronze era. Finally, Tell Halif was once again the center of occupation following the burning of Beit Mirsim, probably at the hands of avenging Egyptians. This Late Bronze (Canaanite) village, non-fortified throughout its long history (Strata XI-VIII), soon reflected the dominance of Egypt, which controlled all of southern Syria-Palestine as part of its northern empire.

The final Late Bronze occupation of Halif marked in itself a major change in the nature of the architecture and of the function of the site. Strata XI-IX contained houses typical of the era, whereas the architecture of Stratum VIII consisted primarily of stone-lined silos sunk into a specially-prepared, deep fill of wet clay, loess, and general debris. In Late Bronze IIB (str. VIII) Tell Halif functioned as a depot for the collection and redistribution of grain, probably as taxation to be used to support the royal city of Gaza, some 35 km. to the west.

Iron II and Persian Halif: Excavations in Field IV

Iron Age pillared house

In the 1999 season excavation was also conducted on the western edge of the tell in Field IV, where in earlier seasons substantial remains of the 8th century BCE Judahite town were uncovered beneath soil layers rich with hundreds of ceramic figurines which derived from a 5th century Persian age cultic site once located there.

In late Iron II (8th century BCE) Tell Halif, probably the Biblical site "Rimmon", served the kingdom of Judah as a border-fortress. The heavy fortifications (including casemate wall and stone-covered glacis) enclosed approximately 7 acres, a limitation which forced houses to be built against each other or to share walls. In Field IV a row of three pillared houses were first uncovered in the 1992 and 1993 seasons, and more fully exposed during the 1999 season. The houses were destoyed probably in 701 BCE by a raiding party from the main army led by the Assyrian king Sennacherib in siege against nearby ancient Lachish.

That destruction sealed in these three houses, beneath fallen bricks, roofing material, and support beams, nearly two hundred whole and restorable ceramic vessels, household items such as loom weights and stone tools, metal weapons, two iron plows, balance scale weights used in local commerce, beads and other jewelry, as well as thousands of samples of animal and fish bones (plus a few fish scales), charred grain and grape pips, legumes and other cereals as evidene of the diet. Finally, the "long room" of the northernmost pillared house enclosed a ‘house shrine,’ which featured the molded head of a ‘pillar figurine’ (the goddess ‘Asherah), two carved limestone blocks (possibly ‘masseboth’), and an incense stand.

Horse and rider figurine

In the debris layers above the Iron II houses of Stratum VI, just as encountered in earlier seasons, in 1999 we recovered another large corpus of ceramic figurines that belong typologically to the Persian era of the 5th c. BCE. Though they were found in soil that had been long ago disturbed by people seeking buried building stones, they nonetheless attest to an active religious ritual at Halif. Similar to rituals practiced elsewhere in the eastern Mediterranean and Egypt, that represented by the figurines at Halif probably found the ceramics useful as votary objects to stand in the presence of the deity or deities. The figurine finds from the 1999 season range from horse and rider to fragments of feet and pedestals. The entire corpus of figurines may beviewed on the DigMaster site.

Part III: Conclusions

The value of the 1999 experimental field season, however, derives less from the physical discoveries, and more from the insights this experience with digital equipment in the field bring to the processes of recovery of past cultures. Though the 1999 season was both experimental and limited to four weeks, several observations and conclusions may be drawn from it. We believe that these should be generalized for other excavation projects.

For the first time in the history of modern archaeology it is now possible to manage and to report successfully relatively large numbers of artifacts and data. The "wealth" of artifacts the excavator encounters may now–with modern digital technology–be recorded in increasingly sophisticated ways and displayed as graphic illustrations in databases readily available to numerous people simultaneously. We conclude that, because it is possible at a relatively inexpensive fare, it is also obligatory that the excavator report data-sets fully. But, in fact, the potential of digital recording/reporting also changes one’s perspectives on the entire excavation-collection process. Because so much more is possible, so much more is necessary; how and what one collects is genuinely affected, as it should be. Even how one sees the soil layers and the collectionof data is altered by the capacities to record and report so much more. Technology, in this instance, necessarily structures field and research strategies.

The objection that bodies of material are sometimes too numerous to record and report fully should, we believe, bring pause to the excavator to reconsider priorities. If we take as axiom that it is the responsibility of the excavator to report fully, the converse is that this obligates the excavator to plan in order to be able to record and to report fully. The gains in digital management, recording, and reporting, while simplifying some aspects, actually lenthens the recording aspects. Digital photography, for example, shortens delivery time of the image and eliminates costs of film; it also makes possible/necessary that much more photography be made part of the plans of an excavation season. The LRP photographic record, for example, increased by thousands in the 1999 season over those of pre-digital seasons. And, while digital photography permits considerable latitude in documentation, it is also true that most often each digital image requires manipulation in a computer program (such as Adobe Photoshop), sizing for easy transmission or for appropriate appearance on monitors, and archiving. In addition, our VR movies consist of a minimum of 36 still photographs, all of which receive batch and individual preparation prior to processing to a VR movie.

The second conclusion from the attempts of LRP in managing its excavation data is that the projects also are obligated to ‘pre-publish’ data. The era of holding information for decades must end; the responsibility to reveal data before final interpretation has arrived with decisive force in the application of digital technologies to archaeology. No longer may important collections of material disappear into archives scarcely ever to be seen again or to be available to alternate interpretations. Early dissemination (‘pre-publication’) of excavation data and graphic illustrations of artifacts will assure the usefulness of the excavation and will add dramatically to archaeological databases; such early and full publication will also address a significant need in near eastern archaeology, the virtual inaccessibility of comparative material because they are hidden behind necessarily limited traditional publications or may be found only with extreme effort in the archives of countries of origin. Digital databases, though not as ‘good’ as the objects in hand, offer a workable solution to the problem.

Finally, the capabilities and the demands that digital recording/reporting bring to the excavation project require a staff that are knowledgable about various aspects of digital technology. For example, in the 1999 season, two staff members were full-time computer facilitators. However, the project needed another full-time computer person (even though the team numbered only 37) and twice the six laptop computers for daily work. Staff makeup must change to include technical people to maintain digital equipment, and start-up costs must dedicate sufficient funds for basic digital tools.

Abel, M. 1938. Geographie de la Palestine. II. Paris.
Aharoni, Y. 1967. The Land of the Bible: A Historical Geography. London: Burns and Oates.
Kloner, A. 1980. Hurvat Rimmon, 1979, Israel Exploration Journal 30: 226-228.
Seger, J.D. 1984. The Location of Biblical Ziklag: An Identity Crisis, Biblical Archaeologist 47: 47-53.

Author Details

Dr. Paul F. Jacobs
Cobb Institute of Archaeology
Mississippi State University
Email: pfj1@ra.msstate.edu
Website: www.cobb.msstate.edu