A THEORY OF NAVIGATION

This book has presented an account of Pacific colonisation based on an explicit theory of navigation. While it was clear, previously, that ethnographic voyaging was well described, its relationship to the navigation of early exploration was not understood. Vague ideas about colonisation and an inappropriate context for archaeological data were implicated in several arguable interpretations of prehistory. The first part of the book dealt mainly with issues in theory, and the second reviewed them in the detail of archaeological settlement evidence, which fitted well. So far everything was normal: argument and evidence were sitting compatibly in one another's pockets. Some conventional ideas looked a lot less comfortable, but, ultimately, there was no independent proof.

A COMPUTER SIMULATION

Further efforts to build a more tenacious case changed that. One was a computer simulation whose first purpose was to examine general propositions about colonisation and to test many detailed predictions about the order and sequence of settlement. The archaeological outlines of Pacific settlement are tolerably well-known in many island groups and, in so far as computer simulation could imitate them, it identified relevant variables and helped to explain what had happened. In so far as it produced plainly spurious patterns, it could suggest things that had not happened.

Beyond that, because simulation showed what prehistoric explorers were probably able to do, it distinguished things that they chose to do, from others that they chose not to do. Also, it provided insights into the kinds of navigational knowledge and technique that were appropriate to the colonisation of different parts of the ocean, if it were to proceed with archaeologically implied rates of speed, success and survival.