03-13-2008, 10:20 AM
Of course it is difficult to quantify such a question - far more difficult for the ancient world than the medieval world for the very reasons the survival rate is set lower by the good doctor Kwakkel - but that doesn't mean it isn't allowed to try to make an educated guess. Of course, many historians wouldn't want to, as they are well aware of the problems that can arise if their educated guesses are taken as "fact" by laymen, but I for one welcome it. While I much doubt there can be any statistical uncertainty percentages attache to such figures, I am pretty certain they would be bigger than the percentages given themselves. 
There are a lot of problems involved in such estimates. The issue of libraries being lost due to invasion, as Timotheus points out, cannot really be taken into account - as far as I know, the fate of the collections of Seville and Toledo after the arab conquest of the visigothic kingdoms are uncertain even today - they (or rather their later copies or translations) might have ultimately ended up in the collections of al-Hakam II or they might have just gone to rot - we cannot really say, as far as I know - al-Hakam's efforts of collection seems to have been directed into the larger arab-muslim world, but that doesn't exclude survival of visigothic collections from his own lands. Another major problem is that the numbers given for library collections come from so many different sources - from the 12th century or so it (to me) seems to be more common with precice lists of exactly what codici are in a collection, and what texts they contain, whereas many earlier estimates are taken from estimates made by chroniclers or ethnographers describing a collection, typically as part of a broader description of a city, ruler or region. We know, for example, that descriptions of "tens of thousands of books" in the university library of Paris around 1200 are inaccurate, because the inventory list survives (as I recall it) - something we lack from a great deal of other sources. To use al-Hakam II's libraries as an example, it seems kind of suspicious that the 400,000 volumes reportedly in it around 950AD required a reported staff (of every level from caretakers to copiers to curators and researchers) that was three times smaller than the collections of Paris around 1200, which had at least 40 times fewer books...
One can just go on creating problems like this for an "estimate of survival" or any estimate of volume numbers or works for that matter - such an estimate can never be "certain" ( nothing in history can be entirely certain, after all) but it shouldn't prevent informed academics from making such estimates if it is an useful tool for research.
There are a lot of problems involved in such estimates. The issue of libraries being lost due to invasion, as Timotheus points out, cannot really be taken into account - as far as I know, the fate of the collections of Seville and Toledo after the arab conquest of the visigothic kingdoms are uncertain even today - they (or rather their later copies or translations) might have ultimately ended up in the collections of al-Hakam II or they might have just gone to rot - we cannot really say, as far as I know - al-Hakam's efforts of collection seems to have been directed into the larger arab-muslim world, but that doesn't exclude survival of visigothic collections from his own lands. Another major problem is that the numbers given for library collections come from so many different sources - from the 12th century or so it (to me) seems to be more common with precice lists of exactly what codici are in a collection, and what texts they contain, whereas many earlier estimates are taken from estimates made by chroniclers or ethnographers describing a collection, typically as part of a broader description of a city, ruler or region. We know, for example, that descriptions of "tens of thousands of books" in the university library of Paris around 1200 are inaccurate, because the inventory list survives (as I recall it) - something we lack from a great deal of other sources. To use al-Hakam II's libraries as an example, it seems kind of suspicious that the 400,000 volumes reportedly in it around 950AD required a reported staff (of every level from caretakers to copiers to curators and researchers) that was three times smaller than the collections of Paris around 1200, which had at least 40 times fewer books...
One can just go on creating problems like this for an "estimate of survival" or any estimate of volume numbers or works for that matter - such an estimate can never be "certain" ( nothing in history can be entirely certain, after all) but it shouldn't prevent informed academics from making such estimates if it is an useful tool for research.
