Ian Hacking's new book *Why is there philosophy of mathematics at all?", see here, contains many interesting ideas. One of the ideas is the dichotomy of two distinct models for the development of a science. One is the butterfy model. A biological organism follows a course of development that is predetermined to a certain extent by its genetic make-up. The example of butterfly was chosen I think because of the dramatic changes a butterfly undergoes (cocoon, etc), unlike, say, an elephant; yet even those dramatic changes are written into the script so to speak.
The other model is the Latin model of development of a natural language like Latin. Such development obviously is affected by historical factors and contingencies of human behavior.
Hacking seems to think that scholars often view mathematics as following a butterfly model, whereas he himself seems to argue in favor of the Latin model.
So to summarize, my question is: does Hacking's Latin model really apply to mathematics?