SCIENCE

“[O]ne of the most interesting things about archaeological material, is you get the same object, you get three archaeologists, you come up with three different interpretations based on that” – Vance Watrous

This essay is a critical evaluation of archaeology as a science.  In what sense is archaeology a science? How does it compare with the other families of sciences?  Is it an empirical science or does it belong to the hermeneutic disciplines?  I look at archaeology as it is practised, with special attention to methodology, interpretation of the archaeological record and the role of historical texts.  Also discussed is its relationship with other hermeneutic disciplines (especially textual studies) and how this impacts on the archaeological endeavour.  The eventual aim is to establish the limits of the archaeological reach.

Archaeology was born during the modern era – just like all the other well-known disciplines.  During the modern epoch, all “disciplines” strove towards respectability – to be recognized as disciplined sciences. The ideal was to be objective, to reach final results, to establish truths. There was an optimism that humans will eventually be able to obtain certain and final knowledge. This perspective has, however, changed. During the last half-century philosophers of science have levelled severe criticism against the possibility of obtaining final results and in our present post-modern era, it is recognized that these goals are for the most part unreachable. Although all sciences will always be “disciplined”, using a strict methodology, with various techniques and methods, their ability to obtain final results are limited.  This is also true of archaeology.

Although I am not an archaeologist myself, I have a live interest in this field of study (I come from a philosophy of science background). In my reading, I find that many researchers still operate in the modernist frame of mind – trying to establish certain “truths”.  It seems that a debris of positivism is still silently present in their midst. This brings the question to mind: What type of science is archaeology? and what are the limits of its reach?  Although there are many good articles by philosophers of archaeology that focus on these matters, these philosophers can obviously not remove themselves from the scientific “paradigm” of which they are part (Kuhn 1962). This could easily result in a tempered criticism, where the (unconscious) objective is to defend the basic tenability of the status quo. I recognize that my own critique is just another perspective, but I believe that it could be of value as an external critique of archaeology as a science.

In this short essay, I focus on those aspects of archaeology that I see as critical to the evaluation of archaeology as a science.  These include methodology (with special reference to inductive and deductive methods), interpretation of the archaeological record as well as the role of historical texts.  I elucidate my arguments with examples from archaeological experience over the last few decades.  These include material related to various excavations and finds, mostly from the ancient Middle East (since this is where my own interest lies).  Throughout the discussion, I compare archaeology with the various families of sciences, namely the natural and social sciences, as well as hermeneutics.

Methodology

In the effort to establish archaeology as a scientific discipline, archaeologists use the tools that are traditionally associated with both the empiricist and the rationalist approaches to science, namely inductive and deductive methods. Inductive methods start from the empirical data and try to establish generalized laws. Deductive methods start from theoretical hypotheses and test this against the data. Although researchers try to use one or the other method, in practice it is very difficult to keep these methods apart. The one feeds into the other and the other way round.  Data is used to formulate hypothetical models which are tested and then (using the information) revised to establish better models.

The only science that uses purely deductive methods is physics.  In physics, it does happen that the mathematical equations that are deductively derived, produce new insights that do not originate with any data. A good example is the Dirac equation which predicted the existence of antimatter.  Mathematically constructed deductive theories are powerful in the sense that they give precise predictions and allow for the possibility of falsification (which go much further than mere “verification”).  The science philosopher Karl Popper recommended this “falsifiability” in physics (Popper 1957).  But most other philosophers of science recognize that this could not be established as a general rule to define a science.

During the modern epoch, and even today, many researchers view deductive and inductive methods as tools that would eventually establish what reality is like. Many researchers held to the correspondence theory of truth and believed that with better and better models, the essence of reality will eventually be laid bare. This optimism was punctured when physicists came up with two different theories describing the same reality, namely Einstein’s general gravitational theory and quantum physics.  Although both theories could in a remarkable way predict the correct outcome, they describe reality very differently (Pine 2006). One could view these as complementary theories modelling the same reality. But this implies that, although reality exists, we do not know – and would probably never know – what it truly is like. There is absolutely no way to establish whether any model describes reality as it really is. A similar situation exists in the social sciences, where the two complementary theories of behaviourism and psychoanalysis are used to describe the complex reality of human behaviour.

Although archaeology uses these same methods, there is obviously no corresponding existent (historical) “reality” against which models could be tested. The process of time has destroyed the historical reality that once existed.  Even in cases where a city was suddenly destroyed and the artefacts are well preserved (as in the case of Pompey), the archaeological record is incomplete. All archaeological records are incomplete. What are the implications of this?  It implies that archaeologists are stuck with an incomplete “reality” – the historical situation or “reality” does not exist anymore. There is an unbridgeable distance between the historical reality and the archaeological reality which is present in the archaeological record. From this, we can deduce that no historical “reality” exists (comparable to the existent but unknowable reality underlying nature) against which theoretical models could be tested.

Various models could be produced, all of them corresponding to the same archaeological data, but with nobody knowing which model (if any) corresponds to the historical situation. The fact that a model is confirmed by the data says nothing; it only confirms that an archaeological correlate for the model has been found. It does not in any way establish that the model corresponds to the real historical situation on the ground. In this case, the models are not complementary, they are competitive. And it is impossible to decide which one (if any) is correct (We find the same problem in other related disciplines. In textual studies, for example, the real textual histories of many ancient texts are forever lost and no amount of hypotheses and testing could ever reproduce it – see below).

To illustrate this problem I can mention the difficulty of distinguishing between trade, emulation, and migration in the archaeological record. Even with lots of data, it often happens that archaeologists cannot decide among themselves which model (or combinations thereof) is applicable. Archaeologists of different backgrounds adhere to different models to explain the same data. A few examples will suffice. Some believe that there is abundant evidence that people from Mesopotamia migrated to Egypt during the period of the Uruk expansion (Redford 1992:24); others believe the data only shows emulation by a local elite (Wilkinson 2002:241, 245). Some use a model of migration to explain the cultural change at the beginning of the Bronze Age in Cyprus (Frankel 2000), others apply a model of emulation by an emerging elite to the same data (Knapp 1993a).  Some believe that the cultural change in the coastal areas of southern Palestine at the beginning of the Iron Age prove the arrival of the Philistines (Dothan 1995); others believe the change was due to trade (Vanschoonwinkel 1999). The list is endless, so to speak. Bernard Knapp said with reference to the Minoan “colonies”: “[T]here is no way to settle it one way or the other” (Knapp 1992).

Some archaeologists are quite straightforward in their assessment of this problem. David Anthony, for example, writes the following about the tendency to prefer models of trade or emulation over migration: “Migration has been demonized and has mystified Western archaeologists since the rise of ‘New Archaeology’ in the late 1960’s… Several writers have noted that the rise and fall of the popularity of migration and diffusion in western archaeology seems closely linked to the prevailing milieu in politics, national interests and intellectual trends” (Anthony 1997:21).  It seems that we can conclude that it is not only “scientific study” that decides which model is accepted; things like political correctness, national interests and what is trending, play an important role. One can hope that the incorporation of other disciplines like genetic studies will eventually lead to more substantial results, but the applicability thereof is often restricted by the lack of relevant data.

One can compare the archaeological study of historical data with the study of the origins of the universe. The universe still exists and scientists study light that originated during the early stages after it came into existence. Although it is impossible to establish exactly how the universe originated, scientists were able to show that the astronomical data corresponds well with the predictions of the Big Bang model – allowing them to eliminate the competing model of a static universe. Although the biological sciences to some extent share the problem of incomplete historical data, it can be assumed that the laws of nature that gave birth to the species are still valid. From this one can propose that eventually, the biological sciences would also be able to “predict” (as we find in physics) and gain a better understanding of the processes involved. At this stage, there are (at least) two competing models of evolution, namely the synthetic model and the punctuated equilibrium model.

Interpreting the archaeological record

As in the case of the empirical sciences, data collection forms a central part of the archaeological endeavour. But does this imply that archaeology is an empirical science? Although archaeologists carefully prepare their digs, using all sorts of techniques to establish a good definition of the excavation, the dice is always loaded against them. Although data constitutes “evidence”, one cannot compare it with any experiment in the natural sciences or controlled study in the social sciences.

There are two reasons for this. In the first place, the dig is a non-repeatable experience. Although it could be compared with similar excavations elsewhere, it could never be the same. The fact that it is a one-off experience implies that it is after all also a personal (individually and socially) experience for those involved – it is not necessarily so that other archaeologists would have arrived at the same conclusions. One can mention many examples of different interpretations of the same archaeological excavations and data. Vance Watrous’s observation says it all: “[O]ne of the most interesting things about archaeological material, is you get the same object, you get three archaeologists, you come up with three different interpretations based on that” (Watrous 1998).

The other aspect in which archaeological excavation differs from any controlled experiment is that one can never know if the accessible data is a representative sample. In most cases, it obviously is not. Any conclusions drawn from the data are therefore open to criticism. In the days of positivism, it was often assumed that no “evidence” is proof that such evidence does not exist. But this is just plain wrong. Since it can never be shown that archaeological data is representative of any particular historical situation, how can any scientific conclusions be drawn from it?  It is not only possible but inevitable, that evidence has disappeared with time or that it has been severely damaged or reduced.

Many examples could be given to illustrate the non-representative nature of archaeological data.  Giorgio Buchner found one small precision balance in the refuse of a metal workshop, together with a miscast bronze fibula and a few related items, as the only possible evidence for Strabo’s statement that goldware production was one of the main activities of the early Iron Age Euboean settlement off the Italian coast (Buchner 1979). A while ago it was reported that the organic layer (probably due to a tsunami) covering archaeological remains from 6000 years ago in the Burren More in North Clare in Ireland, dissipated when exposed to air. The same is true for papyrus and other materials. This again shows what I have demonstrated above (although this time from the inductive side), namely that archaeologists are stuck with an incomplete record of historical “reality”, which severely restricts their ability to establish which model corresponds to the historical situation.

Although archaeology involves “experience”, archaeologists do not conduct “experiments”. Excavations are neither repeatable nor representative. Archaeology is therefore not an empirical science. Even if we accept that excavations are empirical in nature, this is not enough to establish archaeology as an empirical science. This was recognized decades ago by philosophers of science. Even when archaeological excavations are conducted in a carefully controlled manner, the archaeological record is always open to different interpretations. Although interpretation cannot be excluded from any data (even in physics there exist various interpretations of quantum physics), in archaeology interpretation is not grounded in empirical experimentation.

What was proposed, is that the archaeological record should be viewed as text (Schiffer 1985) or at least as having symbolic meaning (Guarinello 2005). As with textual studies, archaeology is essentially an interpretative (hermeneutic) discipline. Various interpretations of a “text” are always possible and no single interpretation could be established as the right one. Bernard Knapp wrote in this regard: “Archaeologists of different backgrounds, however, read such ‘facts’ in different ways and assigned alternative meanings to the same material. In other words, the archaeological record, as ‘text’, has its own autonomy and is open to multiple interpretations on the part of its readers” (Knapp 1993b). Seen from this angle, it makes sense that theoretical models, which are in fact nothing but sophisticated perspectives, will always be competitive even if it incorporates some complementary aspects.

But is the analogy of the archaeological record with texts really a good one? Although it captures the fact that the interpretation of archaeological data is unequivocally perspectivist, it does place that record on the same level as texts. Is that really the case? What distinguishes texts is the information contained therein – information about historical events and situations. This is exactly what non-textual archaeological data lacks. Although such data could give some basic information, for example of catastrophic events (both natural and human-produced), this information is severely restricted because its relation to known history is not explicitly given. For the most part, non-textual data merely gives partial access to background information, for example about material culture.

The enormous difference between texts and non-textual data in its ability to provide information could be easily illustrated. The presence of foreign communities, for example, is very difficult to establish solely from the archaeological data. The Old Assyrian colonies in Anatolia would have been “invisible”, had it not been for the discovered tablets. Malcolm Weiner said in this regard: “As Machfeld Mellink, James Mellaart and others have observed, had the tablets not survived little else would suggest the existence of an Assyrian colony since the colonists adopted local architecture and pottery. The situation with regard to Karum Kanesh is not unique. Tablets tell us of many Assyrian trading colonies in Anatolia. At other major excavated sites as Bögazkoy and Alishar, again only the tablets give any clear indication of the presence of an Assyrian trading colony… It is clear that there are differing views among specialists as to whether an Assyrian custom can be detected at Kültepe… I think it would be very difficult to identify something which is specifically Assyrian without the tablets” (Weiner 1984:17, 26). The same can be said about the Ugarit tablets which changed the whole academic perspective on Canaanite culture.

There is, however, a certain type of information that only archaeological data, and more specifically organic material, could provide. This is information about the age of the data. For many archaeologists, this unique ability of archaeological data to accurately date structures and catastrophic events provide the scientific respectability that the data could otherwise lack. For this, they are dependent on other disciplines, for example, dendrochronology. But how scientific are the statistical techniques used for this dating? Contra the general opinion among archaeologists, this is not a rigid statistical science.

In his book A Slice through Time the dendrochronologist M.G.L. Baillie acknowledges that the master chronologies “are not 100% matches” and that the application of the technique is based on subjective judgement: “The practised dendrochronologist is looking for matches that he/she is willing to accept, based on experience, as correct matches between long ring patterns”. In his review of this book, Ron Tappy wrote: “This subjective intuitive aspect of dendrochronology might easily fail to satisfy the tolerances and significance levels expected by statisticians… Recognition of this subjective human element and the inconclusiveness of many of the case studies introduced in the course of the book dampen somewhat one’s appreciation for the purportedly absolute precision of the science. Various factors, such as the loss of the outermost layers of unconsolidated sapwood from a collective sample, seem to compromise the accuracy of the overall method” (Tappy 2001:215). Clearly, dendrochronology is not as “scientific” as is often assumed.

Dating archaeological layers have other problems as well. Sometimes the dendrochronologically derived dates for samples from the same archaeological layer differ substantially. So, for example, the grain and charcoal samples taken under well-controlled circumstances from the destruction signifying the end of layer 6 at Tell Brak (this is the period just before the Naram-Sin palace) gave dates of 2023 BC and 2662 BC respectively (Oates 1985:144)  Archaeologists normally assign the reign of the Akkadian ruler Naram-Sin somewhere in the middle between these dates.

Another problem is that the presence of organic material does not necessarily determine the age of the structures in which it is found. The structures could have been in use for long periods before the people used those products. Even when organic materials form part of the structures, it could have been incorporated during repairs or rebuilding. Pottery does provide a basis for relative dating, but in many cases, such finds are insufficient to clearly establish the period to which the associated structures date.

The role of historical texts

Archaeology shows the most promise when artefacts and texts are found together. When texts are inscribed on structures, as is often found in Egypt, the possibility of obtaining substantial information about historical events is the best. This is probably one of the most important reasons why Egyptian archaeology is generally viewed as secure (if only for the main dynastic periods). When the texts are found in well-preserved layers, the prospects are also good. When the provenance of the texts is, however, unknown or when other historical texts are used in the interpretation of the archaeological data, the situation is much more complicated. But even in the best of cases, both the archaeological artefacts and the texts (or the combination thereof) are open to various interpretations.

Furthermore, the overall interpretation of the situation involves other archaeological data and historical texts. Although archaeology has considerably enlarged our knowledge of the past through the continuous production of new artefacts, the practice of archaeology always includes historical texts, except for the early periods. This implies that archaeology is an empiric-hermeneutic discipline which is (as far as texts are involved) very much dependent on other hermeneutic disciplines.

The historical texts used in archaeology are also studied in disciplines like philology, history, literature, religious and textual studies etc. These fields have their own particular ways in which they deal with texts. Of special importance is the reliability of the historical information contained in these texts. This depends on many factors, one of which is the textual history of the texts. All historical texts have a certain textual history, how it came into being. Texts like the Bible are typically compiled from other older texts and traditions and then edited and copied in the process of being handed down.

So how is this textual history determined? Theoretical models are constructed, using methods typical of these fields of study, like source criticism. These disciplines, however, face the same barrier found in archaeological excavation – the real textual history is lost. Given the impossibility of substantially (empirically) verifying many of the assumptions used and the fact that new information could always upset the carefully constructed models, there is no hope of ever arriving at the real textual history. The data used to test such hypotheses is always provisional; their verification does not establish facts. Even if one allows for the possibility that the real textual history could be reconstructed (which I do not), it is still impossible to know if it has been achieved. This implies that many possible interpretations present themselves, resulting in the formulation of various perspectives. The hermeneutic disciplines can therefore also do no better than developing (ever) competing perspectives.

In their effort (especially during the modern epoch) to get these disciplines recognized as sciences, scholars have historically adopted a “critical” approach towards the information in historical texts. This included the use of the rationality criterion, assuming that we can rationally determine the likelihood that historical events have taken place. But this assumes that human behaviour is always rational, which it is not. Some confirmed historical events are very unlikely to have happened – but it did. There is absolutely no way to rationally determine the historicity of events.

Furthermore, this approach assumes that without supporting evidence, the information could not be trusted.  In practice, this often leads to a positivist rejection of textual information that has not been verified. But on what basis is this verification done? On the basis that archaeology provides sufficient evidence to decide what should be accepted as reliable information. But archaeology is not an empirical science, dendrochronology is not a rigid statistical science and all archaeological data is provisional in the sense that there is always the possibility that new data will be found that could (even dramatically) alter the picture.

There are many examples where textual information that was at some stage distrusted, was later confirmed by new evidence. A few will do. There was a time when the historicity of the “Great Rebellion” against Naram-Sin (and even his foreign conquests) was doubted (Cooper 1993). The discovery of a school text from the period of Naram-Sin himself, however, confirmed that this event did in fact happen. Joan Westenholz commented: “This fragment of a student’s poor exercise is the only extant proof that literary works were composed on the theme of contemporary historical events. The triumph of Naram-Sin over the rebellious city states was probably celebrated in pomp and circumstance… In the city of Esnunna… a teacher made this subject the topic of an assignment of a written composition for a student” (Westenholz 1997:223).

Another example concerns the Biblical figure of David, whose historicity was widely dismissed in scholarly circles before the discovery of the ninth century Tell Dan Stella on which his name appears. In this case, some scholars were so convinced of their case, that they initially rejected the discovery as fraudulent. It is interesting to note that both the Akkadian Empire (which was remembered as the greatest empire ever to have ruled over Mesopotamia) and the Davidic kingdom share the same lack of physical evidence from the original period. Mario Liverani writes the following about the Akkadian Empire: “If we didn’t know from the texts that the Akkad empire really existed, we would not be able to postulate it from the changes in settlement pattern, nor from the evolution of material culture” (Liverani 1993). The opposite is also true – in traditional circles (which are largely excluded from the “scientific” endeavour) the information in the texts is often accepted too uncritically.

The fact that disciplines like textual studies are grounded in the modernist-positivist methodologies of the past, which had and still have a great impact on the development of thereof, seriously undermines its credibility. The use of such approaches in the practice of those disciplines (which in my experience is very much alive, for example, in textual studies) is a matter of serious concern. This is all the more problematic since archaeology has a relationship of reciprocal dependence on those disciplines. These disciplines are dependent on archaeology to interpret the archaeological data that is used in the interpretation of texts; archaeology is dependent on their analysis of historical texts to interpret the data.

Clearly, both the archaeological data as well as the texts are open to widely different and conflicting interpretations. No wonder that the formulation of coherent narratives using the interpretations developed in the various disciplines has in the past lead to heated debates. In this regard, the different views on the destruction of Ebla (during the Akkadian period), on the Troyan War, on various Biblical traditions and on many other issues can be called to mind. Scholars who give more credence to the archaeological data, typically follow a “minimalist” approach, whereas scholars who give more credence to the texts, typically follow a maximalist approach. Between these, many approaches are possible. The simple fact is, however, that none of these will ever be able to arrive at any final conclusions. In most cases, one cannot even accept that the narratives proposed by the academic community are a “true” approximation to the historical events.

Conclusion

In this essay, I posited a critique of archaeology as a science. Archaeology is not an empirical science similar to the natural and social sciences. Although there are some researchers – especially those who were educated during the final days of the modernist epoch – who are deeply influenced by positivist ideas and who sometimes give the impression that they regard archaeology in those terms, this is obviously not the case. Although archaeology mine and analyze data under controlled circumstances, this could not be compared with repeatable experiments in the natural sciences or controlled studies in the social sciences.  Archaeology is much more restricted in its reach.

One can never go beyond a perspectivist view on any historical situation.  Archaeology could be described as an empiric-hermeneutic discipline. A distinction should, however, be made between hermeneutics as the interpretation of texts in combination with other data, and hermeneutics as the interpretation of non-textual artefacts. The archaeology of the pre-textual age is not “hermeneutic” in the traditional sense. I have shown that non-textual data should not be given the same status as texts.

Although archaeology is primarily concerned with physical excavation, it involves many other disciplines. Many of those disciplines are concerned with texts and are also open to widely different interpretations. The process of determining the trustworthiness of information in texts is fraught with problems. When one comes to the eventual formulation of coherent narratives that incorporate both archaeological data and texts, one could never go beyond the practical limits of incomplete and under-represented data and information.

The result is a range of ever competing narratives, which can never arrive at the “truth”. So, of what use is archaeology then? Although archaeology cannot provide final answers, it does increase our knowledge. Its ability to produce new data allows us to construct more sophisticated narratives than ever before – but nothing beyond narratives. In the final instance, archaeology contributes to our enjoyment of life in the same way that literature, art and religion does. It enables us to live the “good life”.

References

Anthony, David W. 1997. Prehistoric Migration as Social Process. In Chapman, J. and Hamerow, H. (eds.), Migrations and Invasions in Archaeological Explanation, BAR International Series 664.  Oxford: Archaeopress, Oxford.

Buchner, G. 1979. Early Orientalizing: Aspects of the Euboean Connection. In Ridgway, D and Ridgeway, F.R. (eds.), Italy before the Romans, The Iron Age, Orientalizing and Etruscan periods.  London: Academic Press.

Cooper, Jerrold S. 1993. Paradigm and propaganda. The dynasty of Akkade in the 21st Century. In Mario Livernai (ed.). Akkad. The First World Empire: structure, ideology and traditions. Padova: Sargon srl.

Dothan, Trude. 1995. Tel Miqne-Ekron: The Aegean Affinities of the Sea Peoples’ (Philistines’) Settlement in Canaan in Iron Age I. In Gitin, S. (ed.), Recent Excavations in Israel.  Archaeological Institute of America.

Frankel, David. 2000. Migration and ethnicity in prehistoric Cyprus: Technological Proof of Migration. European Journal of Archaeology 3:167-187.

Guarinello, N. L. 2005. Archaeology and the Meanings of Material Culture. In Pedro Paulo Funari, Andres Zarankin and Emily Stovel (eds.). Global Archaeological Theory. Contextual Voices and Contemporary Thoughts. New York: Kluwer Academic/Plenum Publishers.

Knapp, A. Bernard. 1992. Bronze Age Mediterranean Island Cultures and the Ancient Near East. The Biblical Archaeologist 55(3):112-128.

Knapp, A. Bernard. 1993a. Social complexity: incipience, emergence and development on pre-historic Cyprus. Bulletin of the American Schools of Oriental Research 292:85-106.

Knapp, A. Bernard. 1993b. Thalassocracies in Bronze Age Eastern Mediterranean Trade: Making and Breaking a Myth. World Archaeology 24(3):332-347.

Kuhn, T. 1962. The structure of scientific revolutions. Chicago: University of Chicago.

Liverani, M. 1993. Akkad: An Introduction, in Mario Livernai (ed.). Akkad. The First World Empire: structure, ideology and traditions. Padova: Sargon srl.

Oates, Joan. 1985.  Tell Brak and Chronology: The Third Millennium. Mari Annales de Recherches Interdisciplinaires 4:137-144.

Patrik, Linda E. 1985. Is There an Archaeological Record?  In Schiffer, M. B. (ed.). Advances in Archaeological Method and Theory.  New York: Academic, New York.

Pine, R.C. 2006.  Science and the human prospect.  Internet: http//home. Honolulu.Hawaii.edu/pine/book1-2.htm/

Popper, K. 1957. Philosophy of Science: A Personal Report, in C. A. Mace (ed.). British in Mid-Century. London: George Allen and Unwin.

Redford, Donald B.  1992.  Egypt, Canaan, and Israel in Ancient Times.  Princeton: Princeton University.

Tappy, Ron E.  2001.  Book Reviews: A Slice through Time: Dendrochronology and Precision Dating by M. G. L. Baillie.  Journal of Near Eastern Studies 60(3):215-218.

Wilkinson, Toby A. H.  2002.  Uruk into Egypt: imports and imitations.  In J. N. Posgate (ed.). Artefacts of Complexity: Tracking the Uruk in the Near East.  Wiltshire: British School of Archaeology in Iraq.

Vanschoonwinkel, J. 1999. Between the Aegean and the Levant: the Philistines. In Tsetskhladze, G. R. (ed.), Ancient Greeks West and East.  Leiden: Brill, Leiden.

Watrous, L. Vance. 1998. Egypt and Crete in the Early Middle Bronze Age: A Case of Trade and Cultural Diffusion. In Cline, E. H. and Harris-Cline, D. (eds.), The Aegean and the Orient in the Second Millennium, Liège: Université de Liège.

Westenholz, Joan Goodnick. 1997.  Legends of the Kings of Akkade.  Winona Lake: Eisenbrauns.

Wiener, Malcolm H. 1984. Crete and the Cyclades in LM I: The tale of the conical cups. In Hägg, R. and Marinatos, N. (eds.), The Minoan Thalassocracy, Myth and Reality.  Stockholm.