01-05-2020, 08:31 PM
A lot of us know the following graph:
![[Image: Mediterranean_shipwrecks.jpg]](https://upload.wikimedia.org/wikipedia/en/3/3e/Mediterranean_shipwrecks.jpg)
Is a graph of the shipwrecks in the Mediterraneum and is coming from the A. J. Parker's work Ancient Shipwrecks Of The Mediterranean And The Roman Provinces (British Archaeological Reports, 1992)
This work is often used to say that the roman trade was declining after the I century BC, or other similar strange thesis. In these days I have read the book Quantifying the Roman Economy: Methods And Problems (the URL will redirect you to the short review I have done here for the book), and the book review this topic.
First of all, the same Parker says that this graph is full of bias and it has to be read carefully, because its reading can easily lead to hasty conclusions (Parker himself does not suggest that there was a decline in marittime after 50 or 100). This hasty conclusion is exactly that Roman trade went in decline, or reduced, after the I century BC.
Quantifying the Roman Economy offers several ways to explain what we can see here. There are several explanation that concern the fact that effectively that graph is mainly a graph of the amphorae-shipwrecks, so the growing usage of the barrel growly falsified the result. Also, there could be an increasing size of the ships that could have explained the curve that we can see.
These are part of the first explanations that the book provide. But these are not the only, and not the most important ones. The replace of the amphorae with the barrels should be proven, and we have a similar issue for the change of the size of the ships.
More important point is that with that graph we are assuming that in every period the ships have the same probabilities to sink. Instead, this is not true at all. Ship technology progressed during roman empire. The ship pumps for extracting water is just one example (first century BC). Other points are related to the safety of the navigation, and related infrastructures, that improved during the empire (e.g. port infrastructure improved by Claudius and Trajanus, just to say some of them).
Other points, essentially we know the shipwrecks mainly according to diving abilities and turistic places, and this add other bias. Some area are overrepresented, while we have few or no data for the bigger ships, the annona ships, or other big ships, that were well able to move on deep seas on long distances (with more probability to sink in deep water).
Another point that has not been raised in the book, but I think deserves the right attention, and falls under the "probability to sink" topic, is the fact that some periods were more "turbulent" than others. For example, the great piracy period end with Pompeius in 67bc, but even after there were other issues with the second triumvirate and Sextus Pompeius' actions (I century BC).
Finally, it should has to be discussed how that graph has been draw. Apart from the date uncertainty, for each sinking the middle year between the probable extremes was considered. In order to demonstrate the further uncertainty caused by this method, one of the book author has tried to apply again the calculations and to provide a new graph, but every 50 years instead every 100 years.
The result? The peak is not any more in the first century BC, but in the first 50 years of the first century after Christ. And, instead of a constant reduction after the first century BC, an increase at the start of the third century.
I hope this pages will be useful and starting point for new insights for you
PS in the graph taken by wikipedia, 1st BCE label is duplicated, but apart this the graph seems to be the correct one.
Is a graph of the shipwrecks in the Mediterraneum and is coming from the A. J. Parker's work Ancient Shipwrecks Of The Mediterranean And The Roman Provinces (British Archaeological Reports, 1992)
This work is often used to say that the roman trade was declining after the I century BC, or other similar strange thesis. In these days I have read the book Quantifying the Roman Economy: Methods And Problems (the URL will redirect you to the short review I have done here for the book), and the book review this topic.
First of all, the same Parker says that this graph is full of bias and it has to be read carefully, because its reading can easily lead to hasty conclusions (Parker himself does not suggest that there was a decline in marittime after 50 or 100). This hasty conclusion is exactly that Roman trade went in decline, or reduced, after the I century BC.
Quantifying the Roman Economy offers several ways to explain what we can see here. There are several explanation that concern the fact that effectively that graph is mainly a graph of the amphorae-shipwrecks, so the growing usage of the barrel growly falsified the result. Also, there could be an increasing size of the ships that could have explained the curve that we can see.
These are part of the first explanations that the book provide. But these are not the only, and not the most important ones. The replace of the amphorae with the barrels should be proven, and we have a similar issue for the change of the size of the ships.
More important point is that with that graph we are assuming that in every period the ships have the same probabilities to sink. Instead, this is not true at all. Ship technology progressed during roman empire. The ship pumps for extracting water is just one example (first century BC). Other points are related to the safety of the navigation, and related infrastructures, that improved during the empire (e.g. port infrastructure improved by Claudius and Trajanus, just to say some of them).
Other points, essentially we know the shipwrecks mainly according to diving abilities and turistic places, and this add other bias. Some area are overrepresented, while we have few or no data for the bigger ships, the annona ships, or other big ships, that were well able to move on deep seas on long distances (with more probability to sink in deep water).
Another point that has not been raised in the book, but I think deserves the right attention, and falls under the "probability to sink" topic, is the fact that some periods were more "turbulent" than others. For example, the great piracy period end with Pompeius in 67bc, but even after there were other issues with the second triumvirate and Sextus Pompeius' actions (I century BC).
Finally, it should has to be discussed how that graph has been draw. Apart from the date uncertainty, for each sinking the middle year between the probable extremes was considered. In order to demonstrate the further uncertainty caused by this method, one of the book author has tried to apply again the calculations and to provide a new graph, but every 50 years instead every 100 years.
The result? The peak is not any more in the first century BC, but in the first 50 years of the first century after Christ. And, instead of a constant reduction after the first century BC, an increase at the start of the third century.
I hope this pages will be useful and starting point for new insights for you
PS in the graph taken by wikipedia, 1st BCE label is duplicated, but apart this the graph seems to be the correct one.
