8325Re: [textualcriticism] Simulation program
- Jun 10, 2014Hi Bob,I think you are right about there being broad geographic communication. However, there is something to say for the principle of least effort, too: when it comes to making a copy of a text, why go to another province for an exemplar when you can get one nearby?The simulator has geographical bias built in if the trend parameter has a positive value. One can switch the bias off entirely by setting the trend parameter to zero, in which case the data produced looks nothing like the UBS4 data for Mark. (The data set named sim-a is an example of what is produced when the trend parameter is set to zero.)Even a preference to obtain a local exemplar is not enough, however. One also needs to crank up to a high level the tendency for each region to gravitate towards its own combination of preferred readings before simulator results begin to look like the real thing. In my mind this points to a parochialism where each major production centre for copies developed its own flavour of the text.Best,Tim---In firstname.lastname@example.org, <bob@...> wrote :On 06/06/2014 10:57 PM, tjf@... [textualcriticism] wrote:
I've just finished writing a program to simulate transmission of an ancient text:
Two things worth noting:
1. To get the simulator to produce data broadly similar to actual data, it is necessary to introduce a strong local bias toward location-specific combinations of readings;
This is interesting. I've always had a little bit of uneasiness about the assumption that certain text types where tied to a geographic location. I seemed to me that the church had a pretty broad geographic communication, at least at the higher parts of the church heirarchy (e.i. the Roman church know what was happening in Corinth, there seemed to be regular exchanges between the major Christian centers of Antioch, Jerusalem, Alexandria, Rome, Corinth, etc.). It looks like your simulator supports the geographic textual isolation assumption.
- << Previous post in topic