And just to finish off our preliminary and notional analysis of the causes of the Mycenaean collapse. In a previous post I presented what I supposed to be the notional probabilites of several different causes resulting in the Mycenaean urban destruction. That chart looked like this:
|
Invasion by Dorian Tribes
|
0.1
|
|
Systems Collapse
|
0
|
|
Drought
|
0.025
|
|
Disease
|
0.025
|
|
Inter-polity warfare (Civil War)
|
0.53
|
|
Over-exploitation
|
0.05
|
|
Sea People invasion
|
0.05
|
|
Earthquakes
|
0.23
|
 |
| Throne room hearth. Nestor’s Palace at Ano Englianos, Messenia, Greece. 1200 BC. Courtesy of SquinchPix.com |
In a second post I presented the notional probabilities of the several different causes resulting in a mass migration and/or depopulation. That chart looked like this:
|
Invasion by Dorian Tribes
|
0.164
|
|
Systems Collapse
|
0
|
|
Drought
|
0.076
|
|
Disease
|
0.013
|
|
Inter-polity warfare (Civil War)
|
0.669
|
|
Over-exploitation
|
0.019
|
|
Sea People Invasion
|
0.003
|
|
Earthquakes
|
0.057
|
We notice from these tables that things which have a high explanatory power for a result such as urban destruction (Earthquakes with P of 0.23 or almost a quarter) have a different and lower probability of causing a resulting depopulation (Earthquakes with a P of 0.057 or about 6%). In this post I wrap up the process by computing the probability of both outcomes resulting from the same cause.
The method is simply to multiply P(Cause | UrbDest) * P(Cause | Migration) for each cause because the probability of one specific thing causing both events is just the product of the two probabilities for that one specific cause. That result is presented in the following table:
|
Invasion by Dorian Tribes
|
0.042333
|
|
Systems Collapse
|
0
|
|
Drought
|
0.004904
|
|
Disease
|
0.000839
|
|
Inter-polity warfare (Civil War)
|
0.915244
|
|
Over-exploitation
|
0.002452
|
|
Sea People Invasion
|
0.000387
|
|
Earthquakes
|
0.033841
|
When we combine the probabilities in this way we see that some causes have either become vanishingly small (Sea People, P ~ 0.3%) or quite insignificant (Earthquakes: 3%) while other things have become almost certain (Inter-polity warfare, P ~ 92%). I present the results in this pie chart:
 |
| P(Cause | UrbDest) * P(Cause | Migrat) or Probability of a specific cause resulting in both outcomes. |
Here we have a pictorial representation of the probability of urban destruction and depopulation/migration given the several causes. Clearly, under the assumptions made so far, the leading candidate for causing both these outcomes is Inter-polity warfare among the several Mycenaean states – civil war.
I take this opportunity to warn the reader that all the numbers used in this multi-post exercise were notional only. They were inevitably biased by my own thoughts about what causes were likely to have what outcomes. The numbers were not the result of counting or measuring anything.
The value of this exercise lies in decomposing the ‘Mycenaean Collapse’ into two component parts and attempting to evaluate each part separately. I suggest that when we do this we have a better result than when, as many authors do, we attempt to lump everything together.
squinches 