Possible earthquake history of Burnt House at tell Arpachiyeh, Iraq; or did Agatha crack a prehistoric case?

Possible earthquake history of Burnt House at tell Arpachiyeh, Iraq; or did Agatha crack a prehistoric case?

Eric R. Force

The excavation and description by Mallowan and Rose (1935) of the little mound of Arpachiyeh (near Mosul and Ninevah) first established the sequence of prehistoric cultural artifacts of Halaf and Ubaid assemblages.  Particularly rich were  Halaf remains of the so-called Burnt House at level six from the top (TT6), where diverse remains were preserved under a fallen burnt roof.  This horizon is thought (Campbell 2000, p. 1) to be about 5800 B.C.

The destruction and preservation of Burnt House is ascribed by Mallowan and Rose (1935) to invasion, the invaders’ scattering of artifacts, and fire set by them.  This genetic model follows description of only three paragraphs within a single page (p. 17), which with further comments on p. 106 remains all we know of the destruction itself.  I suggest that Mallowan’s genetic model (delivered in first-person singular) should not be taken for granted (as suggested also by other authors below), and propose here that earthquake damage is more probable.

Indeed the published journal of a photographer/conservator of the expedition (Christie 1977, p. 449) noted that preservation of the assemblage was due not only to fallen roof but also to fallen walls.  We now know the latter is an indicator of earthquake damage, though it can have other origins.  On one hand, this was not a critical point to Christie and we perhaps should not give it too much weight.  On the other hand, none doubt that she was a keen observer, and since she mentions fallen roof and fallen walls separately, she knew the difference.

Other authors such as Campbell (2000; see also Campbell and Fletcher 2000) also look askance at the invasion model.  Campbell conducted an extensive inventory of artifacts from the Burnt House (TT6) wherever accessible. Campbell’s body of information includes many intriguing relationships otherwise unclear.  His frustration with Mallowan and Rose shows as: “Little is known of the exact distribution of the objects in the Burnt House. Insufficient information is available on the nature of the burning” (Campbell’s p. 4) and “Unfortunately it is now impossible to reconstruct the original location of more than a few of the objects.” (his p. 7).   Campbell suggested a ritual aspect to the destruction of Burnt House.  I find no additional information in Hjjara et al. 1980 that add pertinent data on this subject to those of Campbell. 

I briefly suggested (Force 2017) that damage at Burnt House could plausibly have resulted from earthquake followed by fire, and in the same paper proposed criteria for earthquake damage as opposed to other causes in mud-brick structures with flammable roofs.  There is no doubt that fire can follow earthquakes where flammable roofs collapse into rooms where fires were lit. 

Four of my earthquake criteria (in table 2 of Force 2017) are, first, that unburnt debris should underlie burnt debris, some of the former resting directly on the floor; second, that only the upper margins of the fallen debris should be burnt; thirdly, that fallen debris (and mortality if any) should extend well beyond evidence of fire; and last, that fallen walls suggest seismic activity.  Admittedly in this case the evidence is tenuous for this last criterion, though I consider it suggestive, and it inspired this re-examination of the overall logic.

Let us look more closely at the other lines of evidence at Burnt House:

1. Mallowan and Rose (1935, plate XXI b with caption) present a photo of complete though probably fractured bases of pots “found on the floor of the potter’s shop in the burnt house TT6”.  Campbell (2000 pp. 7, 10) notes that this unique photograph, apparently from near the center of the room, is puzzling but suggests in situ damage, in this case directly on the floor.  That is, damage preceded fire.

2. Campbell (2000) describes artifacts burnt only on one side, suggesting to him their projection from unburnt debris at an angle (p. 12 and fig. 8, #2 and #3).  He wonders “Was the Burnt House in some disarray before the fire took place?”  He notes also many restored artifacts “where conjoining sherds have been burnt to radically different temperature and in different atmosphere,” (as noted by Mallowan and Rose 1935, p. 106 and plate XIX) suggesting to me different positions in the debris as well as the scattering proposed by Mallowan and Rose (1935, p.  17, 106).  That is, the ceramics were fragmented and incorporated in debris before the room was burnt.

3. Both Mallowan and Rose (1935, p. 106) and Campbell (2000, pp. 8, 10) note that debris at the north end of the main room contains unburnt and little-burnt artifacts.  The fire must have been more local than debris fall. 

Thus three of my criteria for earthquake damage are handsomely met in the Burnt House horizon.  Earthquakes there of course are consonant with the region’s historic seismic activity and current seismic risk.  In contrast to these criteria consistent with earthquake damage, I consider the evidence for invasion per se as very weak (also based on criteria of  Force 2007, table 2).  No evidence of weaponry or related mortality is presented by Mallowan and Rose (1935). 

Musings about blind spots of Mallowan  

Mallowan’s many writings suggest an intent to be thorough and an open-minded attitude, so his leap to an invasion theory for damage at Burnt House is puzzling to me.  Of course, genetic models based on incomplete descriptions are a common failing of early and less-exacting work in many disciplines, and would be rejected today.  Mallowan was tutored by early archaeologists who had no formal training in their field, and who in addition had preconceived ideas of what aspects of their excavations were of importance—Biblical in the case of Woolley and epigraphic in the case of Campbell-Thompson. 

Mallowan (1977) in the course of his career attributed at least six destructive horizons with or without fire to invaders (p. 113, 138, 153, 157, 160, 253) without mention of possible earthquakes, even though the last-listed was first described as earthquake damage by others of Mallowan’s staff at Nimrud (Oates and Reid, 1956), acknowledged by him in 1966 but not mentioned in 1977.  Several of the six sites (e.g. Brak) are in locations where damaging earthquakes are to be expected in sequences representing more than 500 years, as is Arpachiyeh itself.  Indeed, for Mallowan’s sites in the upper Balikh valley the expected seismic intervals are considerably shorter than I thought in Force (2017; see note below and my posting of 11/13/18).

It’s simplest I think to assume that Mallowan was uninterested in earthquakes, even though he knew something of earthquake damage (Mallowan 1977, p. 108).  Other authors such as Kenyon and Schaeffer were suggesting earthquake damage at sites to the east, but not in northern Mesopotamia proper, a situation that has continued to the present day. 

Note: For Force (2017),  I was blissfully unaware that the seismic history recorded by Ambraseys (2009) for Edessa should be added to that for Urfa/Sanliurfa, as these are alternate names for the same place under different regimes.  In my fig . 3 of that paper, therefore, the seismic recurrence interval shown at latitude 37-37.5 degrees in the Balikh valley should be about 250 years rather than 500 years.

REFERENCES

Ambraseys, N.,  2009, Earthquakes in the Mediterranean and Middle East: a multidisciplinary study of seismicity up to 1900:  Cambridge

Campbell, S., 2000, The Burnt House at Arpachiyeh: a reexamination:  Bulletin of the American Schools of Oriental Research v. 318, p. 1-40

Campbell, S. and Fletcher, A., 2000, Questioning the Halaf-Ubaid transition, p. 69-85 in  Beyond the Ubaid, R. A. Carter and G. Phillip, eds: Studies in Ancient Oriental Civilizations #63, University of Chicago.

Christie, A., 1977, An Autobiography:  Dodd, Mead/Collins (New York/London)

Force, E. R., 2017, Seismic environments of prehistoric settlements in northern Mesopotamia: a review of current knowledge:  Bulletin of the American Schools of Oriental Research v, 378, p. 55-69.

Hijara, I., , Watson, J. P. M., and Hubbard, R. N., L. B., 1980, Arpachiyeh 1976: Iraq v. 42, p. 131-154.

Mallowan, Max, 1977, Mallowan’s Memoirs: Dodd and Mead, New York

Mallowan, M. E. L., 1966, Nimrud and its Remains:  Dodd and Mead/Collins, London

Mallowan, M. E. L., and Rose, J. C., 1935, Excavations at Tell Arpachiyeh 1933:  Iraq v. 2, p. 1-178.

            Oates, D., and Reid, J. H., 1956, The Burnt Palace and the Nabu Temple, Nimrud Excavations 1955,: Iraq 18, p. 22-39.