Wednesday, April 2, 2014

Quantified spatial analysis of tectonic boundaries vs. ancient civilizations


https://rowman.com/ISBN/9781498514279/Impact-of-Tectonic-Activity-on-Ancient-Civilizations-Recurrent-Shakeups-Tenacity-Resilience-and-Change

This posting has been revised and incorporated in a published book as of August 2015 (link above)


Quantified spatial analysis of tectonic boundaries vs. ancient complex cultures

Eric R. Force (eforce@email.arizona.edu)

 I have previously described the tectonic environments of the “great” ancient civilizations in the old world (1).  It is impressive how closely associated most of them are with tectonically active plate boundaries, despite the seismic and volcanic dangers involved.  The association is most obvious in the spatial relation at several scales, between the originating sites of these civilizations and the tectonic boundaries (Force, 2008, fig. 1; Force and McFadgen, 2010, figs. 1, 2), though its significance is supported in other ways. In othert posts I analyze various aspects of the dynamics of the relation (2). 
First, however, I need to quantify the strength of the apparent spatial relation, in order to address the possibility that it is some sort of coincidence. Then it may be possible to select subsets, both geologic and cultural, and compare their behaviors in order to provide clues to the meaning of the relation and link these to archaeological-historical observations. 

Quantification of the basic spatial relation.--Since the originating sites of the ancient civilizations are map-points whose distance from the nearest tectonic-boundary line can be approximated, we have the basis for finding an approximate probability that this distance is random for the available space where such sites are possible.  In my paper of 2008, I performed this calculation excluding two ancient civilizations, China and Egypt, as exceptions because I did not see any particular connection to tectonic boundaries.  Subsequently (3), I placed these civilizations in tectonic environments that permit inclusion, albeit at greater distance, and  omitted Assyria as a separate civilization, bringing the number of treated ancient civilizations to twelve.  In this post I redo the calculation to incorporate these changes in variables. 
Note 4 lists the distances used to calculate average distances of ancient civilizations from active tectonic boundaries.  The average distance is about 100 km.  In other words the civilizations lie in a belt averaging only 200 km wide, with the southern boundary of the Eurasian plate and related plate-tectonic boundaries defining the center of this belt, which contains an area of 4.4 million square kilometers. In contrast the theoretically available land area for origination of these civilizations in Eurasia and Africa was 85 million square kilometers, suggesting that the ancient civilizations crowded into the 5.2 percent of available land that is adjacent to tectonic plate boundaries.  This might seem an unrealistic percentage given that northern Siberia and other extremely unsuitable areas were included as available areas, so to exclude these I have used a figure of 77 million square kilometers in the calculation as being available (5). 
It may also be misleading to measure to the nearest fault that represents a plate boundary where the boundary is partitioned among two or more main faults, as in the Zagros Mountains and Baluchistan.  To accommodate this ambiguity, I have used a double-width of 290 km, which along most of the boundary is a generous overestimate.  This combination gives a percentage of 8.3, or about 1 in 12 (6) as the fraction of available land where ancient civilizations originated adjacent to active continental-scale tectonic boundaries. 
Next is the calculation of the probability that this distribution is random (7), given that the area of the belt is about 1/12th of the available land areas.  Because there are now no exceptions, this is an easy calculation.  The chance of random distribution given the observed distribution is one in twelve to the twelfth power, using the binomial probability distribution formula (given in Force 2008).   Technical points of the calculation need not be repeated here, but readers who do some gambling may enjoy an analogy that makes it easier to visualize.  Computing this probability is the same as computing the probability in roulette of a ball landing on the same number twelve times in a hypothetical roulette wheel that has twelve numbers.   These readers will not be surprised to learn that this calculated probability is one in 8.92  trillion.  Thus the probability that the civilization distribution is random relative to tectonic boundaries is so small that the spatial relation is like those we accept instinctively.  It is far beyond any need for statistics.
In my 2008 paper I also varied another factor in order to see whether the probability of random distribution could be increased, in a sort of sensitivity analysis. Four civilizations IN the belt were omitted on the grounds that some authorities would lump them with other civilizations or not recognize them as “great.” In the equivalent calculation of this “worst-case scenario” using the above figures, the probability of random distribution was still one in 282 billion.  The conclusion that these ancient civilizations are spatially related in some way to tectonic boundaries is robust.  In other posts I present probability calculations that confirm various different aspects of the relation (8).


Looking at assumptions, constraints, and definitions.—Perhaps the outcome of quantification was pre-determined by the design of the experiment, which involves a certain time period and a certain half of the world.   I think it’s pretty obvious that “constraining” to the period of antiquity from the beginning of the Bronze Age to about A.D. 500 (about 4000 years) and “constraining” the area considered to the whole eastern hemisphere gives ample scope to variation!  Times and areas outside those boxes are considered in other posts.  They present interesting stories of their own, consistent with the conclusions drawn thus far, suggesting in fact that they have broader application. 
Perhaps most critical is the term civilization, in the sense of “great ancient civilization.”  This is a term not much in favor among anthropologists of our day, and for some good reasons.  Cultures can be complex in different ways, and the complexity of “GAC”s as conventionally defined requires recognition of characteristics that can be preserved in archaeological and historical records, with particular emphasis on public architecture and preserved writing (10).  Postings listed above have recorded some of the accomplishments of GACs that have convinced previous authorities (and me) of their greatness, in the sense that their complexity has many dimensions and far exceeds that of the neighboring cultures of their times.  But it would be desirable to get beyond both the term “great civilization” to give cultural complexity a more rigorous basis.
Some historians and anthropologists think that attempts to quantify cultural comparisons are to be avoided (but see note 11). For my purposes, a list of the characteristics of GACs by Childe (12) that is still highly regarded (and which has been discussed and slightly modified for 60 years) seems the best basis.  The assemblage of ten characteristics listed by Childe is that which more or less defines GACs as they are conventionally accepted and which seems to correspond to the spatial relationship I’ve documented. Thus a rough “quantification” of my use of GAC corresponds to Childe-scores of eight or more. 
The correspondence invites, indeed demands, exploration independently of reactions to “great civilizations”. That is, the correspondence of Childe-scores of cultural complexity with tectonic boundaries is telling us something about the emergence of this assemblage of cultural traits in a tectonic context.   In Mar. 4, 2014 I aggregate Childe’s criteria to construct cultural complexity “transects” at continental scale; the resulting distributions justify confidence that the “GAC” designation is meaningful and that it corresponds spatially to tectonic boundaries. 
            It seems safe to conclude that ancient cultures sufficiently complex to have been called “great ancient civilizations” show a remarkable spatial relationship to active tectonic boundaries, especially the southern boundary of the Eurasian plate and closely associated structures.  Though apparent in map view, the relationship is resoundingly supported by quantitative probability analysis.  The next and more interesting steps are to probe the meaning of the relationship. 




Notes

1.   June 27, 2014, June 16, 2014, May 18, 2014, and May 8, 2014)

2. Mar. 4, 2014, Feb. 1, 2014, Jan. 7, 2014, Dec. 11, 2013, Nov. 26, 2013, Oct. 16, 2013, Aug. 29, 2013, July 17, 2013, June 3, 2013, and May 10, 2013

3. May 8, 2014

4. Distances in km used to calculate average—Rome 130, Etruria 150, classical Greece 10, Mycenaeans 25, Minoans, 60, SW Asia 25, Mesopotamia 150, Indus 90, Persian 30, Aryan India 100, China 300, India 130.  These values are discussed in posts describing individual civilizations, and in some cases are themselves averages for civilizations of composite origination. 

5. Subtracting areas north of 60 degrees and a few extremely mountainous areas.

6. cf. Force 2008, rows 3 and 4 of table 1)

7. To calculate probabilites, I used the binomial distribution formula, which requires that the outcome of each “experiment” be independent of previous ones. The people founding ancient civilizations of course knew nothing about plate tectonics, but they were cognizant of earthquakes and volcanoes.  If we’re trying to increase the probability of random distribution, the question is whether such phenomena in previous civilizations influenced them in founding new sites of civilizations in similarly active locations.  It seems unlikely that they would choose locations on this basis, which to them seemed to hold only disadvantages.  A related consideration is whether the location of the previous civilization has preempted a significant share of the length of the plate boundary.  On this subject remember that the occurrence in question involves only the founding sites for each civilization, i.e. only a point on a line.  But if such preemption should occur, the probability of random occurrence would be decreased, not increased.  Thus if exceptions to independence occur, they would seem to strengthen the spatial relation. 

8.  Jan. 7, 2014, Dec. 11, 2013, and Nov. 26, 2013

9. Jan. 7, 2014, and Oct. 16, 2013).

10. Questions remain where the evidence has been erased.  For example, only scraps of either the civic architecture or the language and literature of the Etruscans have been preserved.  Would we regard Etruscan civilization as “great” had it not been for their funerary arts and architecture, and the evidence from other cultures (and the plaques of Pyrgi) that they did have a literary corpus?

11. Ian Morris (2010) on the other hand, has tried to use quantification methods that avoid cultural baggage, with an index that includes energy capture, urbanism, and ability to wage war.

12. Childe (1950).  These are cities, occupational specialties, concentrated food surpluses, monumental public architecture, capabilities for planning and organization, record-keeping and writing, practical science, arts, trade, and some sort of ideological solidarity. 




References

            Childe, V. G., 1950, The urban revolution: The town planning review, v. 21, p. 3-17

Force, E. R., 2008, Tectonic environments of ancient civilizations in the eastern hemisphere:  Geoarchaeology v. 23 p. 644-653.

Force, E. R., and McFadgen, B. G., 2010, Tectonic environments of ancient civilizations: opportunities for archaeoseismological and anthropological studies, in Ancient Earthquakes, M. Sintubin, I. S. Stewart, T. M. Niemi, and E. Altunel, eds.:  Geological Society of America Speical Paper 471, p. 21-28.

Morris, Ian, 2010, Why the West rules—for now: Ferrar, New York