Archaeology in Three Dimensions - Computer-Based Methods in Archaeological Research

Journal of EastErn MEditErranEan archaEology and hEritagE studiEs

t h E p E n n s y l v a n i a s t a t E u n i v E r s i t y p r E s s v o l . 2 n o . 1 2 0 1 4

JEMahs

Journal of EastErn MEditErranEan archaEology and hEritagE studiEs

V o l . 2 n o . 1 2 0 1 4

iv From the Editors

f E at u r E a r t i c l E s 1 3d Archaeology: New Perspectives and Challenges—The Example

of ÇatalhöyükMaurizio Forte

30 Virtual Heritage: Researching and Visualizing the Past in 3dDonald H. Sanders

48 Archaeology in Three Dimensions: Computer-Based Methods in Archaeological ResearchLeore Grosman, Avshalom Karasik, Ortal Harush, and Uzy Smilanksy

B o o k r E V i E w s 65 Roman Palmyra: Identity, Community, and State Formation, by

Andrew M. Smith IIReviewed by Bettina Fischer-Genz

67 Textile Production and Consumption in the Ancient Near East: Archaeology, Epigraphy, Iconography, edited by Marie-Louise Nosch, Henriette Koefoed, and Eva Andersson StrandReviewed by Deborah R. Cassuto

the j o u r n a l of e a s t e r n m e di t e r r a n e a n a r c h a e ol o g y a n d h e r i ta g e s t u di e s (j e m a h s) is a peer-reviewed journal published by The Pennsylvania State University Press. JEMAHS is devoted to traditional, anthropological, social, and applied archaeologies of the eastern Mediterranean, encompassing both prehistoric and historic periods. The journal’s geographic range spans three continents and brings together, as no academic periodical has done before, the archaeologies of Greece and the Aegean, Anatolia, the Levant, Cyprus, Egypt, and North Africa.

As the journal will not be identified with any particular archaeological discipline, the editors invite articles from all varieties of professionals who work on the past cultures of the modern countries bordering the eastern Mediterranean Sea. Similarly, a broad range of topics will be covered including, but by no means limited to:

Excavation and survey field results;Landscape archaeology and GIS;Underwater archaeology;Archaeological sciences and archaeometry;Material culture studies;Ethnoarchaeology;Social archaeology;Conservation and heritage studies;Cultural heritage management;Sustainable tourism development; andNew technologies/virtual reality.

Appearing four times a year in February, May, August, and November, the journal will engage professionals and scholars of archaeology and heritage studies as well as non-practitioners and students, both graduate and undergraduate.

In addition to combining traditional and theoretical archaeological data and interpretation, the journal’s articles may range from early prehistory to recent historical time periods. It also aims to publish accessible, jargon-free, readable, color-illustrated articles that will be informative for professional and non-professional readers. The journal does not publish unprovenanced artifacts purchased on the antiquities market or objects from private collections.

submission informationDigital submissions should be sent to: www.editorialmanager.com/JEMAHS. All correspondence should be sent to: Dr. Ann E. Killebrew ([email protected]). By submitting their work to JEMAHS, authors agree to editorial modifications of their manuscripts that are designed to help JEMAHS fulfill its mission.

Articles should be submitted as a MS Word file together with all illustrations (1200 dpi for black and white; 600 dpi for grayscale; and at least 300 dpi for color) referenced in the manuscript. Permissions to use photographs and copyrights for all illustrations are the responsibility of the authors and need to be included when the manuscript is submitted. (For more information regarding copyright issues for authors, go to: http://psupress.org/author/author_copyright.html). Papers should be limited to not more than 20–25 manuscript pages or ca. 6,000–7,000 words. Shorter papers are welcome, but authors wishing to submit a paper longer than 25 manuscript pages (including endnotes, references, and appendices) should consult with the editors in advance.

For complete author submission guidelines, please visit: http://www.psupress.org/journals/jnls_JEMAHS.html

subscrip tion informationThe Journal of Eastern Mediterranean Archaeology and Heritage Studies is published quarterly by The Pennsylvania State University Press, 820 N. University Dr., USB 1, Suite C, University Park, PA 16802. Subscriptions, claims, and changes of address should be directed to our subscription agent, The Johns Hopkins University Press, P.O. Box 19966, Baltimore, MD 21211, phone 1-800-548-1784 (outside USA and Canada: 410-516-6987), [email protected]. Subscribers are requested to notify The Johns Hopkins University Press and their local postmaster immediately of change of address. All correspondence of a business nature, including permissions and advertising, should be addressed to The Pennsylvania State University Press, www.psupress.org.

The Pennsylvania State University Press is a member of the Association of American University Presses.

rights and permissionJEMAHS is registered under its ISSN (2166-3548 [E-ISSN 2166-3556]) with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 (www.copyright.com). For information about reprints or multiple copying for classroom use, contact the CCC’s Academic Permissions Service, or write to The Pennsylvania State University Press, 820 N. University Dr., USB 1, Suite C, University Park, PA 16802.

Copyright © 2014 by The Pennsylvania State University. All rights reserved. No copies may be made without the written permission of the publisher.

journal of eastern mediterranean archaeology and heritage studies, vol. 2, no. 1, 2014Copyright © 2014 The Pennsylvania State University, University Park, PA

Archaeology in the twenty-first century is in the midst of a revolution. Geoinformatics and geomatics, which includes GIS, geospatial technologies, surveying, remote sensing, photogrammetry, and virtual reality reconstruc-tions, are at the forefront in transforming how archaeol-ogy is practiced. These new technologies are leading to the increased engagement of excavators and making use of existing excavated data by analyzing it to the fullest extent possible with today’s technology. This not only helps to preserve sites and artifacts for future genera-tions—it increases on-site learning exponentially. In rec-ognition of the rapid change in archaeological technology and methodology, the first two issues of Volume 2 are devoted to the use of new technologies in archaeological documentation and research.

The three articles in Issue 1 were first presented at the 2013 Society for American Archaeology Annual Meeting in Honolulu, Hawaii, in a session co-chaired by Ann E. Killebrew and Brandon R. Olson. They represent some of the most innovative approaches to twenty-first-century archaeology today. “3d Archaeology: New Perspectives and Challenges—The Example of Çatalhöyük” by Maurizio Forte describes a ground- breaking project that aims to record every phase of an archaeological excavation in 3d while going completely paperless. This is achieved by utilizing new technolo-gies such as laser scanning, computer vision, and photo-grammetry to recreate a virtual excavation process that is reversible in a simulated computer environment and

can be used in real-time, day-to-day analyses. Donald Sanders, well known for his digital reconstructions of archaeological sites, presents several of these projects worldwide in “Virtual Heritage: Research-ing and Visualizing the Past in 3d” and provides new insights resulting from the work on these interactive 3d reconstructions. His most recent efforts focus on the 3d integration of various aspects of archaeological documentation and reconstruction in REVEAL, a new software system that has recently been developed.

“Archaeology in Three Dimensions: Computer-Based Methods in Archaeological Research” represents the work of Leore Grosman, Avshalom Karasik, Ortal Harush, and Uzy Smilanksy of the Computerized Archaeology Laboratory at The Hebrew University of Jerusalem. This truly remarkable laboratory is utilizing optical scanners to create highly accurate, rapid, and economically viable 3d digital models of artifacts, whose properties are then analyzed by computer programs developed in-house. Two book reviews, one by Bettina Fischer-Genz of Roman Palmyra: Identity, Community, and State Formation by Andrew M. Smith II, and the other by Deborah R. Cassuto of Textile Production and Consumption in the Ancient Near East: Archaeology, Epigraphy, Iconography complete our first issue of 2014.

We hope this issue and the next will encourage discus-sion and development of new technologies in archaeo-logical research and cultural heritage management.

f r o M t h E E d i to r s

ann E. killebrew and sandra a. scham

journal of eastern mediterranean archaeology and heritage studies, vol. 2, no. 1, 2014Copyright © 2014 The Pennsylvania State University, University Park, PA

into the study of past civilizations. This process continues today, and a prominent example is the adoption of computer-based methods and technologies in contem-porary archaeological research. Besides the standard applications (printing, data handling, communications, etc.), novel digital technologies, based on innovations in computer technology, have paved the way to previ-ously untapped approaches and perspectives (e.g., Bruno et al. 2010). We will briefly review the development of advanced optical equipment and specially constructed algorithms, which have been successfully integrated into archaeological research, serving a large community of expeditions and archaeologists.

The Computerized Archaeology Laboratory (CAL) was established in 2010 at the Institute of Archaeology at The Hebrew University of Jerusalem. Here, modern methods, techniques, and ideas from computer science, in particular from computer graphics, data processing, robotics, and statistics, are integrated into archaeologi-cal research. The laboratory operates three optical scan-ners, providing 3d digital models that can be enhanced using additional methods developed in our laboratory. We would like to point out that the methods and applica-tions described in this article have been used successfully in the past three years on more than 30,000 pottery frag-ments and 3,000 stone tools (to mention only the most abundant artifacts). This level of integration is proof of the practicality of a 3d approach and its advantages, as

ArchAeology in Three Dimensionscomputer-Based methods in Archaeological research

leore grosman

Avshalom Karasik

ortal harush

Uzy smilanksy

This article reviews the activities at the Computerized Archaeology Laboratory at The Hebrew University of Jerusalem, where techniques and ideas from computer science (e.g., computer graphics, machine learning, etc.) are integrated into archaeological research method-ologies. The laboratory operates optical scanners which provide 3d digital models that are then analyzed by com-puter programs developed in-house. These programs address research issues and needs which could not be tackled without the availability of digital 3d models, thus broadening the horizons of archaeological research. To date, these methods and applications have been applied successfully to more than 30,000 pottery fragments, 3,000 stone tools, and many other archaeological finds submitted by more than 100 expeditions. This level of integration provides daily verification of the 3d approach and its intellectual and economical advantages.

abstract

Since its early days, archaeological research has taken advantage of methods and techniques from various disci-plines of the natural sciences, and incorporated them

JoUrnAl of eAsTern meDiTerrAneAn ArchAeology AnD heriTAge sTUDies | 49

tested daily on a large set of archaeological data. Below, we review our interdisciplinary activities, hoping to share our experience and enthusiasm with the archaeological community at large.

Our discussion will start with a brief overview of how optical scanners work and the resulting digitized 3d models. The applications of this 3d technology in archae-ology are further elaborated at three levels of process-ing: (a) digital conservation and display of archaeological finds, which requires a 3d model as produced by the scan-ner software, with hardly any further processing, and is appropriate for display and presentation purposes; (b) computerized processing of 3d models for print-ready archaeological drawings in standard format and the extraction of simple measures; and (c) development of new computer algorithms in response to research ques-tions, which cannot be addressed without the availability of digital 3d models such as the computation of the cen-ters of gravity and moments of inertia in the production of prehistoric stone tools.

3d Scanning of Archaeological Artifacts

Scanning and recording a 3d image of an object can be thought of as transforming an archaeological arti-fact from the physical, tangible world to the digital, virtual domain. The digital model provides a precise image of the surface of the scanned object. It is acces-sible to all imaging and display programs, allowing for the archaeological information stored in an item’s shape to be studied and extracted by simply using a computer without the object being present. The hard-ware and software that constitute the scanning system were purchased from Polymetric Technology, located in Darmstadt, Germany. The scanning technology is based on a simple optical principle: stereoscopic information can be derived when the object is viewed simultane-ously from two directions. This is manifested in the optical scanner by two cameras that are attached to a rigid pole on either side of a projector illuminating the object (Fig. 1A). To increase measurement precision, the projector illuminates the item, which is on a turnta-ble, using light bands and the reflected light is recorded by the cameras (Fig. 1B). The resulting information

provides a precise measurement of the coordinates of a dense set of points on the item’s surface (Fig. 1C). The scanning protocol is automatically controlled by a dedicated computer. Once sufficient information is accumulated, the software merges all the views into a 3d digital model of the object (Fig. 1D). The precision of the digital model depends on the size of the object and the amount of effort and time invested in its scanning. Accuracy on the order of 20–50 μm can be achieved, which is required for the study of small objects (e.g., coins or scarabs). Texture and color information can also be obtained. Thus, the 3d model presents a complete, reliable image of the original item.

Applications in Archaeology

Different applications require different levels of sophis-tication in processing the data provided by the scanner. Below, we will divide the discussion accordingly, by reviewing separately the three levels which were outlined in the introductory section.

Digital Conservation and Display

Archaeological finds often undergo continuous dete-rioration due to changes in their storage environment. In some cases, the finds are at risk of damage due to natural causes or vandalism. Digital documentation of these items in 3d guarantees that the precise shapes and textures of the items can be recovered independent of the actual state of the object. A typical example of a digital conservation project of this type was the scanning of approximately 600 Lower Palaeolithic handaxes from the Upper Jordan Valley site of Gesher Benot Ya‘akov, whose exposure to the air caused them to slowly crumble (Saragusti and Goren-Inbar 2001). Archaeological relics of the utmost importance to research (such as cunei-form tablets that were destroyed in the museums of Iraq) would not have been lost forever had they been docu-mented digitally in 3d.

The dissemination of information within the pro-fessional community and to the general public is one of the most important tasks for an archaeologist.

50 | A r c h A e o l o g y i n T h r e e D i m e n s i o n s

Many excavations undergo digital restoration using 3d imaging for the lithic assemblage (Malinsky-Buller, Ekshtain and Hovers forthcoming; Ben-Shlomo and Van Beek forthcoming). 3d models of small finds may be inserted into virtual site models to assist in the archaeo-logical reconstruction of the site, as well as to intensify the visual experience.

A museum visitor is often barred from approaching precious exhibits (e.g., coins or jewelry) and therefore

cannot see important details. The projection of computer models on screens and using standard accessories (e.g., a joystick or mouse) for controlling the images may dra-matically improve the accessibility of a precious minia-ture object to the general public.

Finally, 3d digital models offer new venues in the teaching of archaeology, particularly for lessons in which the original objects cannot be shown in the classroom.

f i g . 1The four stages of 3D scanning: (A) the scanner consisting of two cameras and a projector mounted on a tripod; (B) an object positioned on the turntable and illuminated by light strips; (c) the skeleton of triangulated vertices representing the surface of the object; and (D) the full 3D model. (courtesy of the computerized Archaeology laboratory.)


Print-Quality Drawings from 3d Models

Archaeological publications (excavation reports, articles, and books) rely on time-honored traditions and conven-tions that are practiced by the archaeological community world-wide. They offer a standard way by which 3d objects are projected onto 2d pages with maximum detail and fidelity. Keeping this in mind, we have developed several algorithms, which are now available as user-friendly applications and produce print-ready drawings from 3d models, in accordance with the traditional standards. However, these standards were accepted long before computers were popular and ubiquitous. We therefore offer optional improvements to standard displays, which are easy to implement and versatile, providing additional relevant visual and metric information. Below, we will illustrate our methods for three types of archaeological findings: potsherds, stone tools, and items decorated with reliefs (coins, stamps, cuneiform tablets, ostraca, etc.).

Documentation of Potsherds

The common documentation of potsherd fragments assumes that the original vessel possesses cylindrical

symmetry, which certainly applies to vessels formed on potters’ wheels. The shape of such vessels is completely rendered by a planar cross-section—the profile—defined as the intersection of the potsherd with a plane that passes through its axis of symmetry. The axis of symmetry coin-cides with the axis of rotation of the wheel on which the vessel was produced. Most of the finds in archaeological excavations consist of pottery fragments, and the deter-mination of the axis of symmetry is not an easy task. The problem is compounded when the fragment forms only a small part of the complete vessel, or when there are accessories to the vessel such as handles or spouts. Until the introduction of computer technology, potsherds were drawn by hand, based on rather primitive measure-ments, and resulting in inaccurate and biased results. In the Computerized Archaeology Laboratory we have devel-oped a program called Pottery3-d that finds the axis of rotation of scanned potsherds with extreme precision and produces a drawing of the pot which adheres to the tradi-tional conventions used in archaeological literature (Fig. 2; Karasik and Smilansky 2008). Figure 3 demonstrates the steps that are automatically taken by the computer from the 3d model to the final print quality drawing.

f i g . 2A screen capture of a “hippo” jar in the Pottery3-D program. (courtesy of the computerized Archaeology laboratory.)


The conventional display of pottery profiles incorpo-rates several visual aids, including marking the symme-try axis or indicating wheel lines. Handles or spouts are also documented in a standard way. These features are included in the drawing produced from the 3d model. The software, which is user friendly, enables quality control of the drawing, adding or removing details according to the requirements of the archaeologist, and printing of potsherd plates to scale. An optional improvement is the addition of two snapshots of the object from directions which can be chosen by the archaeologist. This option goes beyond the traditional practice. However, it is wel-come by many who use 3d technology for potsherd docu-mentation (Fig. 4).

The statement that wheel-thrown pottery has cylin-drical symmetry is only approximate. As a matter of fact,

due to various reasons, ideal symmetry is only rarely achieved. This is most apparent in vessels which are pro-duced on a slow wheel, or are built from several sections that do not fit well together. Uneven heating in the kiln and pressures due to congested loading of the kiln or to connecting handles to the unfired vessel might also introduce deformities. 3d information allows for quanti-tative characterization of the deformation, which attests directly to the technological parameters characterizing the ancient workshop.

Precise measurement of the entire profile of a vessel (i.e., the profile describing the intact vessel) allows for an exact calculation of the vessel’s capacity (without having to resort to the antiquated method of filling the vessel with rice) and the total volume that the vessel occupies in space.

f i g . 3The different stages of digital documentation for a potsherd from the 3D model (top left), finding the optimal axis of rotation (top right), testing the quality of the overlap between the various sections (bottom left), to the final illustration compatible with standard archaeological practices (bottom right). (courtesy of the computerized Archaeology laboratory.)


Based on the tens of thousands of sherds modeled, we feel confident in recommending this method not only because of its accuracy and the research opportunities it provides, but also because of its low cost relative to the conventional hand-drawn practice.

Documentation of Stone Tools

Most stone artifacts, particularly ancient ones, are not simple in shape and the surface features (scars, ridges, tubercles, etc.) provide important and exclusive evidence of the possible manner of use, its production stages, and the train of thought that its design involved. The hand-drawing of a stone tool is complicated and lengthy, and the finished product is often biased by personal interpre-tation (Fig. 5).

The drawing of a stone tool depends on the manner in which the tool is positioned and the result may vary sub-stantially between different draftsmen. Moreover, dif-ferences in the positioning of the tool by the researcher lead to differences in the extraction of even the simplest metric information. 3d scanning and the resulting digital

model are a good basis for the development of computer software called Artifact3-d that positions the tool,

f i g . 4An example of an illustrated plate along with two 3D views of each item. (courtesy of the computerized Archaeology laboratory.)

f i g . 5A hand drawing of a bi-face stone tool discovered in Abu ghosh. (courtesy of the israel Antiquities Authority.)


considers its geometric properties, and generates views, dimensions, and sections that have been selected algo-rithmically without concomitant interpretation (Fig. 6; Grosman, Smikt and Smilansky 2008). The software that has been developed is user friendly and after providing the “objective” positioning, it allows the archaeologist to examine alternative ones. The code allows for a large rep-ertoire of measurements, production of sections, or the addition of visual aids (Fig. 7). It must be noted that the software calculates quantities that can be measured only from 3d information, such as the location of the center of gravity and the inertia moments.

In the last three years, approximately 3,000 stone tools have been scanned and documented. Some have been published in professional literature using the com-puterized method described above (e.g., Goren-Inbar 2011; Malinsky-Buller, Grosman and Marden 2011).

A detailed documentation of the scars and ridges on the surface of a stone artifact is essential for enabling its

complete documentation. Recently, we initiated the first step towards this goal. An algorithm was developed for an automatic, objective, and precise segmentation of the tool surface to scars and ridges (Figs. 8–9). It is interest-ing to note that this segmentation allows a quantitative analysis of the scars and ridges networks including the number of scars, their surface area, depth, and mean cur-vature (Richardson et al. forthcoming). Currently, we are in the process of integrating the algorithm to Artifact3-d.

Documentation of Items with Reliefs

The surfaces of many objects are decorated with recessed or projecting reliefs, which provide highly important archaeological information. In many cases, the reliefs lose their original acuity over time, impeding reconstruction and comprehension of their content. We use software that was developed by our colleagues at the University of Haifa and the Technion in Haifa, Israel, for processing

f i g . 6A screen capture of a lithic tool in the Artifact3-D program. (courtesy of the computerized Archaeology laboratory.)


f i g . 7Automatic drawings of lithic borers that were discovered at nahal ‘ein gev ii. They were scanned in 3D and processed using Artifact3-D. each artifact consists of five views and a section through its wide part. (courtesy of l. grosman.)

f i g . 8An example of a lithic tool (a levallois point) from Abu-Zif marked by scars and ridges. (courtesy of the institute of Archaeology, The hebrew University of Jerusalem.)

f i g . 9A 3D model of the levallois point from Abu-Zif seen in fig. 8. (courtesy of the institute of Archaeology, The hebrew Universityof Jerusalem.)


and improving the image based on 3d scanning (Fig. 10; Gilboa et al. 2013).

Innovative Archaeological Research

In this article, we present studies for which novel meth-ods that answer archaeological questions using 3d data have been developed. They make use of the unique advan-tages offered by the combination of precise information provided by the optical scanning and the development of dedicated algorithms to address the hitherto inaccessible research directions. The selected studies presented here will discuss the detailed analyses conducted on Lower Palaeolithic stone handaxes and the typology of potsherds.

“The Rolling Stones”: Quantitative Study of Post-Depositional Processes on Prehistoric Stone Tools

Sometimes, stone tools undergo substantial wear and tear due to climatic or geological process. This is true in particular for tools that rolled around in water—either in river beds or sea shores—and due to the harsh environ-ment, their original shape underwent significant changes. This effect can be observed even in areas that are now deserts but where paleolakes and paleorivers have been identified (Ginat, Zilberman and Saragusti 2003). Indeed, a significant number of stone tools that are discovered

these days bear damage due to such processes. Therefore, any archaeological research based on the morphology of stone tools must establish if the tools are from an in situ context or the extent to which they have been damaged. It is essential to take into account the effect of wear upon the morphological parameters used for further study (e.g., length, width, symmetry, etc.).

In “The Rolling Stones” experiment (a name borrowed from the famous rock band, whose members have also sustained a certain amount of wear in the last 50 years), we measured the wear process in the following manner: a group of eight flint handaxes were knapped by a contem-porary knapper and were subject to a controlled series of processes simulating natural wear by rolling in a turbu-lent environment (Fig. 11). The damage to the handaxes was inflicted by rolling them in a barrel that contained a few basalt river-bed pebbles (Fig. 12). The eight handaxes were optically scanned and weighed before the start of the experiment, and after 5, 10, 20, 40, 60, 100, and 200 rolls in the barrel. Thus, at the end of the experiment we could follow the degradation of each of the handaxes by studying their 3d models. Figure 13 reveals the findings: the profiles (i.e., projections on the plane of symmetry) of each of the handaxes are shown, clearly indicating that material loss and shape changes are a function of the number of rolls. The salient aspect is that fracturing occurs mainly at the active ends of the tool and they affect the profile in a rather disorderly fashion. We used two comprehensive measures

f i g . 10An example of surface relief on two decorated candle fragments from gamla in a 3D model. (courtesy of D. Adan-Bayewitz and s. Terem.)


to study quantitatively the profile damage: roughness and asymmetry parameters, which were previously intro-duced in prehistoric literature (Saragusti et al. 2005). These parameters were derived from the 3d models and have allowed us to track the wear history as recorded in the shape parameters of each handaxe (Grosman et al. 2011).

The first use of these results was in estimating the age of a handaxe assemblage that was found in Wadi Zihor in the southern Negev Desert, where geologists suggest that Pleistocene fluvio-lacustrine deposits indi-cate the existence of a paleolake named Lake Zihor, which existed during the early Pleistocene (Grosman, Goldsmith and Smilansky 2011). Two assemblages were discovered in two distinct find-spots in proximity to

the paleolake: one at the edge of the lake and the other from the terraces surrounding it (Ginat et al. 2003). A careful examination of the groups clearly showed that they had undergone completely different wear processes. Analyzing the mean asymmetry of the two assemblages and taking into account the effect of wear in one of the groups, which showed definite signs of shape deterio-ration due to rolling damage, allowed for a consistent picture to be drawn. Our main conclusion is that the date of the handaxe assemblage from the ancient lake shore was potentially contemporaneous with the Lower Palaeolithic site of ‘Ubediya, one of the oldest known sites (ca. 1.4 million years ago) inhabited by the first hominids migrating from Africa (Bar-Yosef and Goren-Inbar 1993).

f i g . 11The eight handaxes knapped and rolled in a barrel during “The rolling stones” experiment. The Artifact3-D program automatically positioned the handaxes, rendered the views, and produced the plate. (courtesy of the computerized Archaeology laboratory.)


f i g . 12implements used during “The rolling stones” experiment, including (A) the barrel and (B) the pebbles inserted into the barrel to simulate the high-energy rolling process. (courtesy of the computerized Archaeology laboratory after grosman et al. 2011.)


f i g . 13The results from “The rolling stones” experiment. (A) one handaxe successively scanned after 0, 5, 10, 20, 40, 60, 100, and 200 rolls. The arrows indicate the place in which the stone was damaged at each stage. (B) eight contour lines for each handaxe representing the different rolling stages and tracking the wear processes. (courtesy of the computerized Archaeology laboratory after grosman et al. 2011.)


Computerized Typology of Potsherds

Potsherds are the most common finds in excavations in our region, and for many periods they hold important, unique information on the chronological, spatial, and cultural contexts of a site. This information is derived by sorting sherds and comparing them with pottery types known from different sites and periods. The process consumes considerable time and human resources, and it significantly delays the publication of excavation reports and the conclusion of archaeological research. In order to overcome this difficulty, a computer program that sorts potsherds by shape based on parameters measured from the profiles of the fragments has been developed. It is worth noting that these characteristics can be obtained not only from drawings that have been produced from 3d scans as described above, but also from hand-made draw-ings, which have been scanned and digitized.

The idea underlying the computerized sorting algo-rithm is that the potsherd shape is determined by various details which appear on different scales of size—from the maximum vessel diameter or maximum height, through the cants and angles of its body or neck, to the fine curv-ing of the rim. All are important for sorting the fragments

by vessel type (jug, cooking pot, bowl, etc.) and size, but especially the rims when attempting to differentiate subtypes.

Given an assemblage of potsherds (i.e., numeric data describing the profiles), the software generates a clus-tering tree (Fig. 14). The tree in this figure stands at its top, and each “leaf” at its bottom corresponds to a single potsherd in the assemblage. The leaves connect to form branches, which also connect to higher branches and so on. Each branch indicates a clustering of similar pot-sherds. The height of the split between the branches indi-cates the significance of the sorting into sub-branches. The tree structure allows for a consistent hierarchical numbering of the branches. Thus, Branch 321 belongs to Branch 300 at the highest (most crude) level, to its sec-ond sub-branch, 320, which is itself sorted into two sub-branches of which 321 is the first. The automatic sorting of vessels into types (branches) as produced by the soft-ware is certainly compatible with the traditional archaeo-logical judgment. Had we concentrated on the grossest sorting level, namely Branches 100, 200, and 300, we would find open vessels in Branch 100, semi-closed ones in Branch 200, and closed vessels in Branch 300.The greater the resolution in the tree branches, the finer the

f i g . 14A cluster tree for 360 potsherds from Tel Dor. The branches of the tree are marked at three levels of division into groups. (courtesy of the computerized Archaeology laboratory.)


distinctions that are found, to the point that nuances focus on a vessel’s rim, such as between Branches 121 and 122 (Fig. 15). Indeed, an individual comparison of comput-erized sorting and manual sorting of the same complex, as done by experienced ceramicist A. Gilboa, showed an impressive correlation between the two methods (Karasik and Smilansky 2011).

Another example of the success of computerized typology is provided by the ability to differentiate two vessel assemblages from two different workshops in

Kfar Hanania (Adan-Bayewitz et al. 2009 and references therein). The software could distinguish between bowls that originated from two different production sites in the village, but were all assigned by the site’s archae-ologist to the same type, namely 1E. The computerized analysis indicated that minute, but significant, morpho-logical details unique to each manufacturing site escaped the attention of traditional sorting, but once pointed out were admitted to be present. In both groups the vessels had a thickened rim, but the thickness differed,

f i g . 15hand drawings of vessels belonging to the branches sorted in the cluster tree. (courtesy of the Tel Dor excavation Project.)

f i g . 16examples of Type 1e bowls from two production points in Kfar hanania, each row matching other sources. (Adan-Bayewitz et al. 2009.)


and proved difficult to discern without the help of the computerized method (Fig. 16).

A recent test case of typological clustering of vessel types was conducted on “Hippo” jars, named after their large swollen bodies (Fig. 17). This is one of the ceram-ics vessels that typify the Iron Age during the late tenth through ninth centuries BCE in northern Israel. The geo-graphical distribution of “Hippo” jars is along the Beth Shean Valley, Jezreel Valley, and in the Lower Galilee (Alexandre 1995: 81). In the current study, an assemblage of 158 “Hippo” jars from several sites (Tel Beth Shean, Tel Rehov, Tel Amal, Tell el-Hammah, Tel Megiddo, and Horbat Rosh Zayit) was measured in 3d. An automatic typological classification was done based on mathemati-cal representations of accurate cross-section profiles of the 3d models. The results of this grouping were com-pared to other independent features of the “Hippo” jars: petrography, potters’ marks, and sites. The goal of the study was to examine aspects such as centralization and

standardization of production in order to draw regional, historical, and socio-economic conclusions.

Preliminary results show that there were at least four morphological subtypes of “Hippo” jars and that there is a strong correlation between these groups, the petrographic sources of production, the different types of potters’ marks, and some of the sites where “Hippo” jars were found. Based on this analysis, it is possible to reconstruct important con-clusions about the nature of this unique assemblage and the process of production and distribution of the jars.

In the Future

There is no doubt that 3d technology has proven itself as a research tool of the utmost importance in coping with current and future challenges of artifactual archaeology. User interfaces developed at CAL, which are available to the archaeology community at no cost, enable easy and

f i g . 17A 3D model of a “hippo” jar from Tel rehov. (courtesy of A. mazar.)


convenient access to computer files for precise measure-ments or the preparation of materials for printing. The use of digital information to address more complex issues requires continued computer advances and CAL is working hard on such developments and introducing them to the archaeological community as a user friendly and convenient software tool.

Additional studies requiring specialized research and development are now being conducted, and, for the sake of completeness, we will mention only those which are now in their final stage. A project of great importance is the formation of a database that will fulfill a number of basic requirements. A database should enable convenient access to finds for the excavator and other archaeologists who need data for further research. It will allow the addi-tion of data both by including more items in the database and by adding information to each item once it becomes available (e.g., petrographic, chemical data, or verbal descriptions). This will be followed by the integration of 3d data into a national archaeological database that will include geographic data, area mapping, and plans of the various sites. Finally, one of our important goals is to include advanced data querying tools for the special needs of the archaeological community.

Another project is to add all the information gathered in past excavations that was documented by hand draw-ings into the digital database. This integration of old and new data will allow comparative studies between sites and periods, enabling studies that were almost impossible

before because of the prior need to rely on printed reports that appeared in various books or periodicals. And who knows, maybe we will see in the future the computerized version of the late Ruth Amiran’s book Ancient Pottery of the Holy Land: From Its Beginnings in the Neolithic Period to the End of the Iron Age, which would allow for ceramic parallels to be found using computer technology?

noteThe Computerized Archaeology Laboratory would not have been formed and prospered had it not been for the support of institu-tions and persons who contributed by directing resources, allocat-ing personnel and equipment, thought, encouragement, trust, and willingness to take risks and pave the way through the terra incog-nita of the computerized approach to archaeology. We thank The Hebrew University of Jerusalem and the Institute of Archaeology for adopting the laboratory and for productive cooperation; the Weizmann Institute of Science and the University of Haifa for pro-viding equipment and hosting the laboratory in its initial stages; the Israel Antiquities Authority, which has supported the develop-ment of technology and participated actively in examining it in various archaeological contexts; and the Yad Hanadiv Foundation for a significant contribution without which the effort would have been prematurely curtailed. There is not enough space to list the names of all the persons to whom we are grateful. We mention the students, who are running the laboratory on a daily basis while learning to adopt the computer as an organic tool in their archaeo-logical research; and the directors and staff of the Institute of Archaeology, past and present, who were willing to integrate the laboratory as an organic section of the institute. Special thanks go to our colleagues Ilan Sharon, Ayelet Gilboa, Gonen Sharon, Anna Belfer-Cohen, and Naama Goren-Inbar for the accompaniment, trust, and joint work since the beginning.

leore grosman is a senior lecturer at the Institute of Archaeology, The Hebrew University of Jerusalem, Israel, and the head of the Computerized Archaeology Laboratory. As a prehistoric archaeologist, she is engaged in research exploring the history of humanity, particularly the transition to an agricultural mode of subsistence. She is the director of the Terminal Palaeolithic excavations at Hilazon Tachtit Cave and Nahal Ein-Gev II.

avshalom karasik is the head of the Computerized Archaeology Laboratory at the Israel Antiquities Authority. His PhD dissertation explored mathematical methods and computer applications for the analysis of archaeological artifacts, with a focus on morphological classification and typology.

ortal harush is a graduate student at the Institute of Archaeology, The Hebrew University of Jerusalem, Israel. She is a staff member in the Computerized Archaeology Laboratory, a member of the Tel Rehov excavation publication staff, and an area supervisor at the Tel Abel Beth-Maacah excavations.

uzy smilansky is professor emeritus in the Department of Physics of Complex Systems at the Weizmann Institute of Science, Israel, and a visiting professor at the Institute of Archaeology, The Hebrew University of Jerusalem. He is the founder of the Computerized Archaeology Laboratory.


referencesAdan-Bayewitz, D., A. Karasik, U. Smilansky, F. Asaro, R. Giauque,

and R. Lavidor. 2009. Differentiation of Ceramic Chemical Element Composition and Vessel Morphology at a Pottery Production Center in Roman Galilee. Journal of Archaeological Science 36:2517–30.

Alexandre, Y. 1995. The Hippo Jar and Other Storage Jars at Hurvat Rosh Zayit. Tel Aviv 22:77–88.

Bar-Yosef, O., and N. Goren-Inbar. 1993. The Lithic Assemblages of the Site of ‘Ubeidiya: A Lower Palaeolithic Site in the Jordan Valley. Qedem 34; Publications of the Institute of Archaeology, The Hebrew University of Jerusalem. Jerusalem: Institute of Archaeology, The Hebrew University of Jerusalem.

Ben-Shlomo, D., and G. Van Beek. Forthcoming. The Smithsonian Institution Excavation at Tell Jemmeh, Israel, 1970–1990. Smithsonian Contributions to Anthropology 50. Washington, DC: Smithsonian Institution Scholarly Press.

Bruno, F., S. Bruno, G. De Sensi, M.-L. Luchi, S. Mancuso, and M. Muzzupappa. 2010. From 3d Reconstruction to Virtual Reality: A Complete Methodology for Digital Archaeological Exhibition. Journal of Cultural Heritage 11:42–49.

Gilboa, A., A. Tal, I. Shimshoni, and M. Kolomenkin. 2013. Computer-Based, Automatic Recording and Illustration of Complex Archaeological Artifacts. Journal of Archaeological Science 40:1329–39.

Ginat, H., E. Zilberman, and I. Saragusti. 2003. Early Pleistocene Lake Deposits and Lower Paleolithic Finds in Nahal (Wadi) Zihor, Southern Negev Desert, Israel. Quaternary Research 59:445–58.

Goren-Inbar, N. 2011. Culture and Cognition in the Acheulian Industry: A Case Study from Gesher Benot Ya’aqov. Philosophical Transactions of the Royal Society B: Biological Sciences 366:1038–49.

Grosman, L., Y. Goldsmith, and U. Smilansky. 2011. Morphological Analysis of Nahal Zihor Handaxes: A Chronological Perspective. PaleoAnthropology 2011:203–15.

Grosman, L., G. Sharon, T. Goldman-Neuman, O. Smikt, and U. Smilansky. 2011. Studying Post Depositional Damage on Acheulian Bifaces Using 3-d Scanning. Journal of Human Evolution 60:398–406.

Grosman, L., O. Smikt, and U. Smilansky. 2008. On the Application of 3-D Scanning Technology for the Documentation and Typology of Lithic Artifacts. Journal of Archaeological Science 35:3101–10.

Karasik, A., and U. Smilansky. 2008. 3d Scanning Technology as a Standard Archaeological Tool for Pottery Analysis: Practice and Theory. Journal of Archaeological Science 35:1148–68.

Karasik, A., and U. Smilansky. 2011. Computerized Morphological Classification of Ceramics. Journal of Archaeological Science 38:2644–57.

Malinsky-Buller, A., L. Grosman, and O. Marder. 2011. A Case of Techno-Typological Lithic Variability & Continuity in the Late Lower Palaeolithic. Before Farming 2011:1–32.

Malinsky-Buller, A., R. Ekshtain, and E. Hovers. Forthcoming. Organization of Lithic Technology at ‘Ein Qashish, a Late Middle Paleolithic Open-Air Site in Israel. Quaternary International.

Richardson, E., L. Grosman, U. Smilansky, and M. Werman. Forthcoming. Extracting Scar and Ridge Features from 3d-Scanned Lithic Artifacts. Proceedings of Computer Applications and Quantitative Methods in Archaeology.

Saragusti, I., and N. Goren-Inbar. 2001. The Biface Assemblage from Gesher Benot Ya’aqov, Israel: Illuminating Patterns in “Out of Africa” Dispersal. Quaternary International 75:85–89.

Saragusti, I., A. Karasik, I. Sharon, and U. Smilansky. 2005. Quantitative Analysis of Shape Attributes Based on Contours and Section Profiles in Artifact Analysis. Journal of Archaeological Science 32:841–53.

Archaeology in Three Dimensions - Computer-Based Methods in Archaeological Research

Documents

Transcript of Archaeology in Three Dimensions - Computer-Based Methods in Archaeological Research