To download a copy of this paper, click here: mitrovic.doc

gmslobodan mitrovic Dept. of Anthropology, CUNY Graduate Center

THIS IS A DRAFT VERSION

Over the past decade, Geographical Information Systems (GIS) greatly influenced the way archaeologists manage and manipulate spatial information. Landscape archaeology has been transformed through the dramatic growth and widespread introduction of GIS data applications, as these began to reach maturity. The role of GIS as a flexible mechanism for the articulation, exploration and analysis of landscape data became fully realized.

In my paper I would like to look at possibilities for analyzing physical changes that occur on archaeological sites through time. In addressing this issue, my goal would be to look at ways of finding answers on these questions (drawing from Langran 1992):

•	Where and when did change occur?
•	What types of change occurred?
•	What is the rate of change?
•	What is the periodicity of change?
•	What can we get from Virtual Reality Modelling Language (VRML)?
•	Human-landscape interaction

The data sets that I would preferably like to rely on are the ones treating Neolithic sites. However, even though GIS offers an exciting and powerful research tool to archaeologists, there are not many definitive dates in archaeological data. Overall paucity of GIS-archaeological analyses concerning change in time and space (here time should be read as on equal footing with space) is somewhat depressing, but to use another cliché: it looks like positive changes are afoot.

In the following text I also look at what archaeology could gain by gazing at time sequences, as well as at changes from the base state. This, in my opinion, is particularly important when analyzing tell-sites (mounds made by successive building and destructing of dwelling structures one on top of another), and even more important for tell-sites situated near rivers and prone to river erosion. Archaeology as a discipline is often borrowing models from other fields, and here potential for implementing temporal GIS (TGIS) will be presented.

Time and Temporal GIS

Such is the nature of temporal GIS analysis that most of the articles from the nineties, as well as from this decade (and overall in anthropological science these only count in low tens) begin with stating that: the amount of temporal (historical) and especially spatio-temporal data has dramatically increased. Archaeologists proudly agree that integrating time into GIS is one issue where they can make a significant contribution (Wheatley & Gillings 2002:242-3). Indeed, as Wheatley and Gillings state (2002:242), “a discipline whose subject matter is thoroughly temporal, it is surprising that full and explicit discussions of time within archaeology and anthropology are a relatively recent phenomenon”. But what comes forward from various studies – GIS as well as others – is recognition of complexity of social understanding of temporality and the ways in which time and space are deeply entangled within one another.

In broad terms temporal GIS is a discipline avenue which attempts to store and analyze spatial objects and attribute changes through time. A TGIS specialist has to decide, among other things, what size of time slice she will attempt to use in a respective model. One particular point of exploration here is the worldly nature of data: how does a spatial feature act at certain times and how does it develop over time, due to inevitable temporal uncertainty, we may not know exactly when an object had a particular form or was located in a certain place. Furthermore, differential rate of change emerges as a result of the intersection of temporalities, ranging from the slowest processes to micro-level phenomena (McGlade 1999).

Presentation of temporal data

The concept of time-space is rather common, as many Indo-European languages show in expressions to articulate space relations temporal terms are used and vice-versa: the principle is embedded. We use it constantly in our daily lives and the passage of time is normally understood via perception of changes of objects over time and their movements in relation to one another. These cause-and-effect relationships are crucial to our understanding of nature and structure of various elements in the world around us.

In representations time can usually be seen as linear (Newtonian time line, with time stretching forward into the future and backwards into the past), cyclic (appropriate for the seasons of the year and class schedules), branched (hypothetical paths which time might follow in model of future or past), discrete, or Einsteinian dimension which interacts with space (Hazelton 1991, Langran 1992, Peuquet 2001, Freelan 2003, also Gosden 1994, Shanks & Tilley 1987). Hahn also informs us of another conception of time: kairological (ancient Greece), which was experiential time, accorded to the duration of human-centred activities (McGlade 1999: 144). Probably the two most frequently used metaphors in Western cultures are linear and cyclical.

The central issue is: how can time be plausibly represented, in terms of creating a database and visualization? It is accepted that databases for many GISs have been designed to represent static situations, not change (Peuquet and Qian 1997) as presenting time is problematic for several reasons. Firstly, commercial GIS packages do not account for time and process, focusing on space and its attributes. Next, spatially defined date have three components: time, location and attribute - and most mapped data fix time (Langran 1992: 11-12): "To create a temporal GIS is to dislodge the time component from its fixed position and thus free this log-jam in geographic enquiry." Langran (1992: 28-9) also addressed the issue of designing a database that can easily store and retrieve temporal data as another big challenge (Green 2002). Given that archaeologist is interested (volens nolens) in interpretation of reality, rather than reality itself, linear and cyclical time are the conceptions within TGIS that I would like to focus on.

In many applications GIS operators ask what the consequences of various choices would be. Similarly, when archaeologists engage with sites to reconstruct pieces of the past into a form that carries significance in the present, they have to make choices “of what elicits attention, of what to record, of what to conserve” (Schnapp 2004: 11). Modelling effects of a variety of strategies and comparing their outcomes would be the role of GIS, as GIS would not only be used to describe and understand processes but also to possibly direct by making decisions. Langran (1992) has foreseen that representing spatio-temporal data in GIS would be a problem of implementation (and after this seminal work, studies only elaborated on the topic, without original contribution). Overcoming the obstacle of representing time process in archaeology will never offer definite spatio-temporal solutions, rather more or less well-grounded narratives, as archaeology on the whole, is a narrative discipline (and in that sense my view is largely influenced by post-processualist thought: Hodder 2001, Shanks & Tilley 1987…). Finally, possible approach in regard to this story-telling quality of archaeology would be a tree model of time. Every decision making point is branching point, and each choice is a branch. Thus a point on the tree represents a time and a sequence of decisions taken up to that point (much like DNA studies when tracking genes down the timeline). This approach incorporates both linear and cyclical concept and would be useful on a large scale syntheses of possible ramifications of narratives (if one ever gets there). One part of the problem of mapping time is that it is only possible to progress forward in time with no turning back. Other issues concern topology of the time-axis, dating of time, vagueness of time, and complexity of changes themselves. Referring to this complexity, in general terms and with appropriate data-sets, a TGIS should be able to answer queries such as the following (listed in a rough order of increasing complexity, from Freelan 2003:49):

•	When did feature X exist or cease to exist?
•	What existed at location A at time t?
•	What happened to a given feature or location between time t1 and t2?
•	Did event A exist before or after condition X (or event B)?
•	What patterns exist between events A-B-C and features X-Y-Z?
•	Given data for Feature Y at time t1 & t3, what was the likely state of this feature at time t2?
•	What would be the state of feature X at time t?
•	What is the predicted outcome following event A after time t?

Continuous and discrete

As Lock and Harris put it (Lock & Harris 1999:10): “landscape is continually evolving three dimensional space where one of dimensions is time. The representation and compression of temporal axis upon essentially two-dimensional landscape surface creates a palimpsest which is the data for a landscape archaeologist”. That is why any study attempting at grasping changes needs a temporally-explicit methodology for recording information about changing features and changes in the site. Furthermore, it needs means of interpolating intermediate conditions between known points in time and means of recording uncertainty and diffuseness in the spatial and temporal extent of features (Halls et al. 2000).

My attempt to summarize a model for possible future work is based on enterprise of other projects. Namely, these are Stephen Freelan’s Fairhaven map archives case study (Freelan 2003), ongoing TimeMap project - Mapping the fourth dimension (Johnson 1997-2002), and Halls et al. (2000). For the two case studies of well-known archaeological sites (which will be described in more detail further below) I suggest the combination of linear, non-linear (McGlade 1995) and cyclical notions of time, as well as combining raster and vector GIS in terms of conceptualization of space and time. Fundamentally, the view would be that of series of snapshots at known points in time, and a series of transitions between these snapshots. In a way archaeological snapshots are very much what we have on the tell-sites. Laid out like this, this model is not a topological model – it does not record spatial or temporal relationships between features, only their position in space and time.

The ever-changing environment means change involving daily human actions to interactions between atmospheric, terrestrial, fluvial and geological processes. It is becoming increasingly important to understand the processes behind these changes in order to better manage effects on the environment. Central to this is the storage, access and visualization of all environmental data, to facilitate improved understanding of the data and better predictions through modelling (Peuquet and Qian 1997). In between our snapshots of knowledge, in what we are terming a transition period, in most of the times we have no specific knowledge about the feature. However, we often have some idea on how one snapshot turns into another, because we know something about the processes acting on the object, or because we can reasonably assume uniform rates of change – and that is how we account for interpretation, too.

But, conceivably, if we are to delve into changes from the base state (and in the grand design of interpretation, any state can become a base one), the snapshot model would have to develop into space-time composite model (Langran 1992:86). Every individual feature has to have a record of beginning and ending dates as attributes. That is how each portion of the continuum can break from the whole and become a discrete object with its own distinct history (Langran 1992:41-42). The biggest issues here would, again, be storage of data and error proneness due to many interconnected levels of information. This topological model could easily be the ultimate archaeologists’ tool to grasp spatio-temporal relations, for excavation methodology favours discrete work on features. Excavators can never phase together features ideally to create continuous picture.

As the described approach would be combination of two (plus non-linear) notions of time, it would also have to be a combination of two spatial representations. As presented above if we imagine at each point on a line a 3-D snapshot of an area being studied, snapshots may change continuously or suddenly as the time coordinates change. For example, geological structures change continuously during sedimentation processes but earthquakes cause change that is effective immediately (on time that passes at different rates, ref). Thus both discrete and continuous concepts of space and time are equally important: certain information is perceived as spatial feature while other as continuous fields. Time can be seen as a continuum or not, but in many applications we can only study series of instantaneous events, assuming that ideally no change occurs between them. Although raster systems as relative conceptualization do not easily give such information as perimeter, shape and contiguity, they are data intensive as they have to record data at every cell location at continuous surfaces. So while raster systems are analysis oriented, vector systems tend to be database management oriented. They are more efficient in storage of map data because they only store boundaries of features and their geometries, and querying of databases is much quicker. .

Presenting two potential case studies

As this paper was originally envisaged to be the synopsis for work in the near future, the two sites from the title of this section are Vinca in Serbia and Catalhoyuk in Turkey (Vasic 1932-6, Srejovic et al. 1984, Mellaart 1967, Hodder 1998). Vinca and Catalhoyuk are tell-sites, made by successive building and destructing of dwelling structures one on top of another forming a hill like structure. People have been invoking Catalhoyuk and Vinca in the same breath for a long time (well, at least archaeologists from the Balkans), and in some ways it is an apt comparison. Long lasting occupation was favoured by the positions of these sites and significant events and changes in material are mirrored in individual levels of their vertical cultural layers.

The first archaeological excavation at Vinca was undertaken by Miloje Vasic in the nineteen-twenties (Vasic 1932-6), and at Catalhoyuk by Mellaart in the nineteen-sixties (Mellaart 1967). The works are still going on both as newest excavation commenced in the nineties when concepts of citation, trace and repetition are singled out as elements of “material memory” on our sites (Boric 2002, Hodder 2001). To go back to concepts of time, linear time is an ordinary concept of time-line with the past and future separated by present. Secondly, myriad of processes are cyclic, or nearly so. In historical archaeology these are daily or yearly cycles, whereas in prehistoric these can easily be as imprecise as decades or hundreds of years even. It is the archaeological time resolution that is ever poorer as we go down the time-line to deeper prehistory. For the deep past we have to rely on known referential phenomena in order to fill in the gaps. In order to avoid numerous mistakes, which are inevitably along the path one way or another, spatio-temporal databases have to be constantly maintained and updated to reduce risk of mistake. Now, at this point notions of uncertainty and diffuseness need to be introduced (as it is inevitable in archaeology). We may not know exactly where an object was located in the past. Spatial uncertainty may arise from inadequate data or imprecision in the records (such as generalized statements of location in historic records without specific geographic references). Moreover, position of an object can be loosely defined for spatial diffuseness reflects natural and cultural phenomena, such as cultural influences – and these have everything but sharp boundaries.

For the reason of “poor resolution”, linear time model would not prove very useful for many archaeological and geo-morphological processes. Also, data collected at the same point of different rounds of a cycle do not provide further and possible crucial information about fine detailing of an erosion or decay processes, for instance. So naturally we come to the conclusion that time model which would be worth pursuing in anthropological disciplines is open ended rather than a finite circle. Long term trends and inaccuracies need to be taken care of with an utmost proficiency, especially with a cyclical phenomenon - eliminating mistakes that would otherwise show on other cycles. But for the time being, I will focus on linear and cyclical models as these seem especially attractive for applying on the two tell-sites.

The two tell-sites

The archaeological record is fragmented as a result of changes in the patterns of human activity over time, and the destructive activities of populations developing previously-used sites. Material traces of social practices in the past are patchy at best and archaeologists must piece together the history of a site from fragmentary evidence. In that sense temporal geographic information systems are very similar to archaeology as there are no satisfactory ways of assessing change and discontinuous tradition (McGlade 1995: 114). In addition, McGlade says that “it is not change per se that is important, rather as archaeologists we must shift our focus to questions which deal with (i) the rate of change and, perhaps more important, (ii) the change in the rate of change. It is these attributes which above all define system dynamics (1999: 156).”

Changes in site disposition result from large- and small-scale disturbances, such as fires and storms, as well as human activities such as building, destruction and re-building. Life cycles of occupation and abandonment seen through repetitive practices of construction, rebuilding and destruction at the same place characterize these tell-sites (Chapman 2000, Matthews 2004). At Catalhoyuk, similar ground-plans, internal division of space and seasonal consecutive layers of plastering strongly emphasize reiteration through a thousand of years. The site also served as burial ground for Hellenistic population. Vinca, too has been occupied and re-occupied continuously for several thousand years and served as a burial ground for medieval peoples (Srejovic et al. 1984). From this perspective, tells are always incomplete and in a perpetual state of becoming - as floors, walls, burials or pottery are being displaced and refigured over the long term (Tringham 2001, Boric 2002).

“It is a paradox that a tell cannot by definition begin life as a tell – its earliest incarnation is as a flat site, like other flat sites in its vicinity. Why did a community stay on the same site at all, and why was a particular place selected for this reoccupation? … The physical transformation of a tell-to-be into a tell depends upon two long-term physical concentrations – people and house daub “(Chapman 2000: 207-8). To evaluate changes in settlement patterns due to natural or human activities would be the goal of the proposed study. I would like to argue that creating a spatio-temporal database and temporal GIS is a great possibility as Vinca and Catalhoyuk have facilities to allow spatial queries. The ability to add to the database is a great benefit here, but also as access to present-time, historical sources is increasing, particularly by way of archive digitization projects. The two sites are special in one more aspect – they have been interpreted differently (and drastically different) in the past than in the present studies and that adds to their peculiarity.

Digitizing of Vinca archives from the 1920s campaigns is done and presently the 1970s excavations’ archives are being digitized (Tasic 2003). The wealth of ritual artifacts at Vinca is one of the site’s best known characteristics, but the size of Vinca tell is unclear due to modern settlement. Change of thickness of cultural deposits is apparent, but unknown throughout the whole settlement. It is ironically so well-known, but so little known. Functional differentiation of space is hazy, the sequence of construction is not easy to establish, and the stratigraphy has always been somewhat dim through the decades of investigation. Additionally, the same stratigraphy is radically different with the introduction of metallurgy (Copper Age) in the overall periodization (Stevanovic & Jovanovic 1996:203-4) and with changes in the architecture. Stevanovic described how successful building and destruction of houses marked the Neolithic period in the Balkans (1997). She suggested that house construction and destruction were complementary architectural practices, and argued for “the possibility that a deliberate social strategy existed during the Neolithic according to which, houses were built with the intent to be destroyed at a certain point in their use-life, moreover by an act of conflagration” (Stevanovic 1997: 335-6). The original excavator of Vinca-Belo Brdo, Miloje Vasic celebrated the site and in the 1920s and 1930s extensively excavated. However, in the years after the Second World War he suggested that the tell-settlement had the character of Black Sea Coast Ionic (Greek) colonies the Danube was connecting it with. Via Danube and Black Sea Vinca was communicating with Ionia and other sites in the east basin of Mediterranean - all the way to Egypt. Pieces of cynabarit ore and kilns for smelting were found in the deeper levels of Vinca settlement and interpreted the site as a colony of skillful miners and metallurgs. Although he initially published the finds as Neolithic (Vasic 1932), in the article from 1948 Vasic stated that Vinca must not be dated before the beginning of Iron Age. A deep Iron Age ditch that was dug through upper layers of the tell exposed material from the Metal Ages, as well as so called bucchero pottery, finely polished Neolithic black ware that he believed to represent Greek imports (Chapman 2000).

On the other hand, walls decorated with paintings and skulls stimulated Mellaart to make a distinction between shrines and houses (1967), for central idea seems to have been the binary distinction between sacred and profane space (Eliade 1959). As it appears however, all houses excavated so far had some evidence of art or ritual, even the least-elaborate houses contain some paintings (Farid 2004, Hodder 2004). Micro-morphological analyses of floor deposits show that buildings Mellaart called shrines also included mundane domestic activities and were all lived in (Matthews 2004). Profane, domestic, dirty areas were spatially separated from ‘ritual’, sacred, clean areas, but only by a series of ridges, steps and platforms within a house. Yet it is Mellaart’s shrine that is exhibited in Ankara Archaeological Museum, whereas at the actual site of Catalhoyuk a house (albeit with special role in mortuary practice) is conserved and opened for permanent exhibition.

Continuity and change

Extensive excavations from the initial campaigns destroyed the sites during the archaeological process, just as modern work continues to destroy, albeit on little less scale. Temporal trends are observed from data obtained, so for instance, it would be possible to give an absolute representation of scrapings performed each year, but the successional outcome is uncertain at this time. Less than spatial uncertainty, temporal uncertainty is something of a given in archaeology, but can also stem from inadequate data (like chronology estimated from relative dates), or inadequate archive reporting (generalized statements about when something occurred or first-hand accounts with rough dates). These occurrences are common in archaeological practice and most definitely every archaeological project has to deal with uncertainty and fuzziness of data. Solutions need to be sought and errors suppressed but often corrections can only be implemented on new data or data that we had direct access in the past to. For temporal fuzziness we could record limiting time values and a descriptive transition function (linear, early, central, late etc.). For spatial fuzziness best estimate of the boundary and as many descriptions towards the interior and exterior as possible would be essential for changes from one condition to another often occur gradually rather than abruptly with no identifiable transition point.

"{At the Neolithic Catalhoyuk]} there is a continual tension at the site between continuity and change. On the one hand, the increased materiality of life leads to a greater potential for constructing continuities, especially based in the house as it is reused and rebuilt over centuries. There is repeated replastering, there is burial beneath floors, there are installations of animal heads and sculpture on walls that last through the occupation of houses and may be taken for reuse in later buildings. On the other hand, the increased materiality of life allows that life to be endlessly re-crafted by shifting an oven here, by changing foodways, pot and obsidian types there, and by beginning to accumulate beyond the domestic sphere (Hodder 2004 in press)."

At Catalhoyuk and Vinca excavations have been taking place during summer every year, much like on any other long-term project. If we consider sites’ continuity and diversity in such a way tha identity of tells is not fixed, because the past and its mythologies themselves are unfixed (Massey 1994: 168, Chapman 2000), always open to interpretations: we have an opportunity to investigate the role that certain events play in the bigger historical picture of the two sites. Since the data exist from different decades, we may have an opportunity to correlate changes on many levels. Excavating every day starting at seven o’clock in the morning every summer we commit ourselves to long-term association with the site. It is possible to envisage that peeling and scraping off layers of white plaster off of a platform at Neolithic Catalhoyuk house is a cyclical process - or better - reversed cyclical process. If changes are detected, studies could begin to investigate the role of these specific events or cycles. Also in the light of different perceptions in the present – the experimental house model at Catalhoyuk exhibited within the complex (Stevanovic 2003? Ref.) has been serving different purposes over the years. The Catalhouyk compound areas may not currently represent a habitat type, and the reader should acknowledge that the comparisons made in this document reflect a snapshot in time. James Mellaart put Catalhoyuk back on the maps, but the new project under Ian Hodder attracts buses full of tourists on a daily basis and continues to build up the set of buildings. So for modern day excavators involved in interpretation of the past, as well as for the Neolithic dwellers the nucleation of people in households living close to one another is a part of the process of increased sedentism, recognized in the archaeological record by the construction of larger, more durable houses and the occupation of larger settlements (Tringham & Krstic 1990, Chapman 2000). Additional analyses incorporating sophisticated statistical time series could be employed here too, but these are beyond the scope of the current demonstration.

Using this new data source, archaeologists can consider old and new snapshots of area or object to understand broad structural changes that may be occurring. The archaeological process relates to the notion that “nothing in the past was essential, fixed, systematized – everything was about transformation and change… Interpretation is always doubly embedded in the changing contexts of the present and of the material past (Hodder 1999: 179-180).” Again, it is of great importance that these data accurately reflect the site in geospatial terms, as any modelling effort is only as strong as its data constituents. These data should be reviewed closely, verified for accuracy, and rigorously analyzed. From these analyses, decisions about continuity can be reviewed as often archaeologists talk about almost abrupt changes in an environment. However, the temporal trends, spatial patterns, and numeric statistics would provide a more complete package upon which to evaluate the scene. Fully understanding the patterns that are occurring on our sites might lead to question the effects of various trends and take action if required (conservation, preservation). We can also consider the implications of activities on the landscape as a whole – both in the past and in the present.

Preservation

More recent excavation projects on our two sites have been aware of erosion and off-season processes to take great care in protecting structures. Project leaders state that one of the central aims of current projects is protection and conservation of site’s architecture should be properly conserved and protected (Hodder 2004 in press, Tasic 2004 in press). Both Catalhoyuk and Vinca have their on-site museums to present archaeological structures. “At Catalhoyuk the whole part of the site has been covered so that the ancient houses are protected and that visitors can gain some sense of walking around a Neolithic village” (Hodder 2004 in press). However in Vasic’s and Mellaart’s time cultural horizons were not being taken care of in such a fashion.

In terms of monitoring the abrasion of sites, the well developed database is definitely the key prerequisite of any study that aims at explicating such phenomena. In some cases, a whole context can be dramatically modified in a matter of hours, as a result of an extreme storm event. For instance, the Danube took away a significant part of the Vinca tell revealing 50m wide section. Understanding soil erosion is therefore essential for explaining geomorphology of such areas. The use of soil erosion models has become increasingly prevalent. Realizing detachment of sediment particles and realizing the transporting medium - in this case water - are on the time-scale as important phenomena as dating the first occupation horizon. Functionally interrelated database with a common interface, able to produce complex queries dealing with temporal and spatial data would be a huge modelling step towards understanding processes.

Still the most important question is, of course, how can we develop space-time query interface that can satisfactorily handle inexactness (Langran 1992, referenca)? As much as the question of inexactness could be posed to archaeology in general, if temporal GIS is to be used for predictive modelling or any other study of change in time and space, development of a means to deal with missing data is fundamental. There is a twofold use of referencing and placing of data on a timescale: the study of an entity’s history to recognize temporal trends or previous states and to store the results of predictive modelling for analysis and display. Before that time and accommodation of incomplete temporal information we have to use other kinds of interpolation procedures that would be appropriate for various situations to fill in missing data. Multiple times and alternative histories need to be considered, as well, that is why in this and many other contexts dealing with space-time information, I see simulation as an important analytical tool. The results of multiple simulations, as alternate histories or future-histories, need to be recorded. Simulations in turn could provide means of displaying information which have a temporal component, in addition to the normal three spatial components, on a two-dimensional screen.

Individual features are archaeologically discrete within a space (Neolithic house) and are discretely excavated. We can only roughly phase them together by assigning numbers to phases marked by an individual event (such as destruction of an oven, Farid 1999). Spatio-temporal database would be an attempt to smooth the progress of looking through different perceptions of (feature’s, building’s, site’s…) given past, and facilitate phasing processes and contexts together. That is where power of GIS lies - in the ability to link objects on a map to a database, combining the graphical display of data with database query and analysis functionality. Ultimate goal is to allow spatial and temporal queries, which in turn would allow manipulation with spatial information. Introduction of the temporal component would definitely work towards improvement of those breaks between snapshots, as well as, overall understanding of the site dynamics. Implementing the model laid out above on these two sites will not be an easy task, but data is almost endless and the project members are willing to make an effort. At Catalhouyk and Vinca we already have ever growing geospatial information, databases already serve as per excellence sources of information, merging together information from various specialists, digital archives, videos, and enormous amount of photo documentation. All this makes these two sites very practical for insertion into a TGIS and incorporating animation and Virtual Reality modelling to generate metrics for use by archaeologists.

Bibliography:

Apan, A.A. 1996. "Tropical Landscape Characterization and Analysis for Forest Rehabilitation Planning Using Satellite Data and GIS." Landscape and Urban Planning 34, pp. 45-54

Barrera R., A. Frank & K. Al-Taha 1990. Temporal relations in geographical information systems: A workshop at the University of Maine, Orono, 12 - 13 Oct., 1990. Orono

Berggren, A. & Hodder, I. 2003 Social practice, method, and some problems of field archaeology, American Antiquity 68.

Boric D. 2002. The Lepenski Vir conundrum: reinterpretation of the Mesolithic and Neolithic sequences in the Danube Gorges. Antiquity 76, pp. 1026–39.

Burrough P.A., R. van Rijn & M. Rikken 1996. Spatial data quality and error analysis issues: GIS functions and environmental modeling. In Goodchild M.F. et al (eds) GIS and Environmental Modeling Progress and Research Issues. Fort Collins, pp. 29–34

Chapman J.C. 2000. Fragmentation in archaeology. People, places and broken objects in the prehistory of south-eastern Europe. London and New York

Dykes J.A., K.E. Moorerb & D. Fairbairn 1999. From Chernoff to lmhof and Beyond: VRML and Cartography. In VRML 99. Paderborn

Eliade, M. 1959. The Sacred and the Profane. New York: Harcourt Brace.

Farid S. in press Volume 4 in the sequence of monographs on Catalhouyk Research Project

Fernie K. & J. D. Richards 2002. Creating and Using Virtual Reality. http://vads.ahds.ac.uk/guides/vr_guide/sect21.html

Fischer M., H. J. Scholten & D. Unwin, (eds.) 1996. Spatial Analytical Perspectives on GIS. London

Freelan S. 2003. Quasi-Temporal GIS & Archival Maps, http://gis.esri.com/library/userconf/proc03/p0987.pdf

Gell A. 1992. Art and anthropology. London

Gillings M. & G. Goodrick 1996. Sensuous and reflexive GIS: exploring visualisation and VRML. Internet Archaeology, 1: http: /www.intarch.ac.uk/journal/issue1/

Gosden, C. 1994. Social Being and Time. Oxford.

Green C. 2002. An experiment in the creation of a temporal GIS for archaeology. Unpublished MsC dissertation, Leicester University. http://www.zen26819.zen.co.uk

Halls P., Polack F. & O'Keefe S. 2000. A new approach to the spatial analysis of temporal change using todes and neural nets. Cybergeo 139

Hazelton N.W.J. 1991. Integrating Time, Dynamic Modelling and Geographical Information Systems: Development of Four-Dimensional GIS. University of Melbourne

He, H. S., D. J. Mladenoff, V. C. Radeloff & T. R. Crow. 1998. "Integration of GIS Data and Classified Satellite Imagery for Regional Forest Assessment." Ecological Applications 8, pp.1072-1083.

Hodder I. 1993. Narrative and Rhetoric of Material Culture Sequences, World Archaeology 25, pp. 268-282

Hodder I. 2001 (ed.) Archaeological Theory Today. Cambridge, UK

Hodder, I. & Cessford, C. 2004 Daily practice and social memory at Catalhoyuk. American Antiquity 69

Hodder I. in press, Volume 4 in the sequence of monographs on Catalhouyk Research Project

Johnson I. 1997. Mapping the fourth dimension: the TimeMap project. Sydney

Jorge L. A. B. & G. J. Garcia 1997. "A Study of Habitat Fragmentation in Southeastern Brazil Using Remote Sensing and Geographic Information Systems (GIS)." Forest Ecology and Management 98, pp. 35-47

Langran G. 1992. Time in Geographic Information Systems. London

Lock G., & T. Harris, 1999. Analysing change through time within a cultural landscape: conceptual and functional limitations of a GIS approach

Massey D. 1994. Space, Place and Gender. Cambridge

Matthews 2004. in press, Volume 6 in the sequence of monographs on Catalhouyk Research Project

McGlade J. 1995. Archaeology and the ecodynamics of human-modified landscapes, Antiquity 69, pp. 113-132

McGlade J. 1999 The times of history: archaeology, narrative and non-linear causality, in T. Murray (ed.) Time and Archaeology. London and New York.

Peuquet D. 2001. Making Space for Time: Issues in Space-Time Data Representation. GeoInformatica 5 (1), pp. 11-32

Shanks M. & J. Tilley 1992. Reconstructing archaeology. London

Shanks, M. & C. Tilley 1987. Social Theory and Archaeology. Cambridge.

Shepherd I. 1996. Putting time on the map: dynamic displays in data visualisation and GIS. In Fisher et al. (eds.)

Srejovic D., Stojanovic D., Tasic N., Krstic D., (eds) 1984. Vinca. Beograd.

Stevanovic 2003 The experimental house at Catalhoyuk?

Tasic N. 2003

Tasic 2004

Thorn C. E. (ed.) 1982 Space and Time in Geomorphology. London

Tringham R. 2001 “Household Archaeology” in Smelser N.J. & P.B. Baltes (eds.) International Encyclopedia of the Social and Behavioral Sciences. Oxford, pp.6925-6928

Vasic M. 1932-6 Preistorijska Vinca I-IV. Beograd

Wainwright J. 1994 Assessing the impact of erosion on semi-arid archaeological sites, Past and present soil erosion. In Archaeological and geological perspectives. New York

Watson A. 2001 Composing Avebury. World Archaeology Vol. 33(2), pp. 96–314

Wheatley D. & M. Gillings 2002 Spatial Technology and Archaeology. London and New York Worboys 1995

Yeh, T-S, & B. de Cambray, B. 1995 Modeling Highly Variable Spatio-Temporal Data. Proceedings of the Sith Australasian Database Conference (ADC’95), Adelaide.

return home