NavList:

I've have had both "The Raft Book" and "We, The Navigators" on my bookshelf for many years and I just reviewed them. In our terms, Polynesian navigation constisted of dead reckoning and latitude sailing, all done without instruments. The navigators learned what heading to steer from exploration and experience and used stars and constant wind directions and constant wave patterns to extablish the headings. Since voyages tended to be done seasonally, these heading references tended to be reliable. Since the heading to be sailed includes a correction for current, the stars don't necessarily show the actual azimuth to the destination but the heading to be sailed that includes the cross current correction angle that a modern navigator would work out with a plot or, a pilot with the trusty E-6B. Then, by using zenith stars, they could establish a latitude and follow it to the destination. 

So, how could such a system provide for accurate and safe landfalls? I have pointed out many times that you always know where you are but the germane question is "do you know where you are to the precision needed at the moment?"  Modern CN is an "enroute" navigation system and it only has to be accurate enough to get you to your destination within the range of shore based navaids so that you can identify that shore based navaid and then switch to using that shore based navigation for the final approach to the destination. In the middle of the ocean you don't need to know your position to an accuracy greater than "I'm in the middle of the ocean." But, as you approach land then you have to get out the sextant and get a more precise location so that you can divert your course to come within range of the lighthouse light and then use the bearings on that light to then shape you course so that you get close enough to spot the sea buoy and then follow the other buoys up the channel. Each time you shift to the more precise system as you get closer to the harbor. 

The Polynesians were doing the same thing but without sextants, chronometers or lilghthouses. How could they accomplish this? Simply by using their knolwedge of the natural world to expand the radius of the landfall navigation system so that even with a instument-less dead reckoning that they still came close enough to the destination to pick up the information that allowed then to correct their heading and find the land. They were very knowledgable about the birds that that would see out on the ocean and know which ones would fly towards land in the evening and then follow them. According to Gatty, some land birds can be found 75 miles at sea during the day and then return to the land in the evening. So a Polynesian navigator didn't have to hit a one mile in diameter island dead on but had a circle of 150 miles to hit, hit it anywhere and you can find your island. They also recognized wave patterns that were distorted by nearby land and could detrermine the direction to that land.

The same goes for zenith stars and latitude sailing. Though they couldn't determine the exact zenith as we can with our sextants they could get it within a couple of degrees, latitude uncertainty of 120 NM, so precise enough to intercept the 150 NM circle of land birds. 

Lewis has a chart in his book showing the islands with 30 NM circles around them (conservative detection range for polynesian navigators) and they make chains that are hard to miss in their entirety. It was llike early pilots flying to Hawaii. The islands are close enough together that the chain presents a very big target more than ten degrees wide and hard to miss. 

gl