Monday, November 06, 2006

The Rock Pile Site Prediction Experiment

I claim that rock pile sites are most likely either on a hill shoulder looking out over water or at the headwaters of a brook - where a spring or break-out zone produces water coming out the side of a hill. To prove this point, in the past, I proposed the following experiment:
1. Take a fragment of topo map, of someplace remote and not yet explored. [This is how I came to choose Leominster State Forest.]
2. Circle a number of likely places on the topo map.
3. Go to those places and locate rock piles.

It turns out this is a flawed experiment. First, it turned out to be impossible to get to most of the locations I had circled - because of mountain laurel. Second it was hubris to think I could guess the conditions on the ground or the type of rock pile sites to be found at this "remote" location. Also, you need some rule about how big to make the circles and about how thoroughly to cover the available map area with circles.

So what was the result? When I first tried it, I was only able to get to two of the circles - with no rock piles in either. Each time I tried to get to one of the other circles, I had to give up but ended up finding rock piles in places I should have circled.

Knowing this, how can we improve the steps of the experiment? Well, be more systematic about circling all the likely places and pay more attention to accessibility. (Ironically, last weekend I did get to a third circle - the one second from the top right, and I did find rock piles there.) I tried to follow the modified guidelines a few weeks after the first experiment - I made a largish circle around the area near Redemption Rock (just south of Leominster State Forest), and ended up finding a number of rock pile clusters. In the end, I do not think the experiment is too successful. It is hard to be sure it evaluates site prediction when it could just be that there are lots of sites and you stumble into them regardless of what prediction/search strategy you are using. However the experiment does have one clear beneficit - it gets you looking at the topo map and gets you out exploring new territory. It does result in locating new sites.


Anonymous said...

I don't see any problem with your experiment. You're not going to find stonework at every site you circle, but I believe you are on the right track by suggesting that cairns are going to be found at a brook source (spring), etc. Haven't you proved this anyway?

pwax said...

Yes, but this experiment is one of the basic research tools that we have - optimally it helps confirm hypotheses, minimally it stimulates site discovery.

Tim MacSweeney said...

Some of the most interesting things I've come across are often associated with water features that aren't on the map. I've followed little tiny streams that I could put on the map by following the contour lines - sometimes more or less because the maps aren't always accurate. And I've often found zigzag rows that border the riparian zone of these little brooks that protect them from burning, perhaps prevent flooding. I've come to realise some streams around me have been diverted. And where the stones are missing there often is a nearby modern construction.

Tim MacSweeney said...

A clue in a name?
I stumbled upon this one:
"Naromiyocknowhusunkatankshunk Brook
Sherman is the location of the Naromiyocknowhusunkatankshunk Brook (29 letters), on the north end of town near the New Milford border. The Indian name means "water from the hills," (the English wording is 17 letters long). The Naromi Land Trust in Sherman derived its name from the brook.[1]
For some time the brook was officially known as "Morrissey" Brook (9 letters), but an official name change was put into Public Act 01-194, "An Act Concerning Certain Real Property Transactions," which was approved July 11, 2001.[2] The 29-letter name was noted in an 1882 book, History of the Towns of New Milford and Bridgewater, 1703-1882, by Samuel Orcutt.[3]"

pwax said...

I would love to learn how to say that but I am still working on "Chaubunagungamaug"