John - thanks for your thoughts, and for the offer of textile material, I will contact you offline about that ... and re fuzzy sets, it seems like an interesting idea, you may have to teach me!
Pamela - The Batak question is an interesting one. As you say, the reason that it is not included in the analysis is that I was not able to find enough characteristics to make the results statistically valid, mainly because Batak weaving uses relatively few motifs. This is a pity. Looking at supplementary weft motifs might be more fruitful, but these do not necessarily come from the same source as ikat motifs as you rightly say. Looking at loom design, names of loom parts and technique might also yield insights.
Putting statistics aside, some time spent with Batak cloths (and with Sandra Niessen's monumental study on the topic) reveals some differences (to me at least) in Batak warp ikat weaving versus warp ikat from other regions of Southeast Asia. Most other warp ikat in SEA is woven in narrow strips with the decoration organized in hierarchical fashion: there are principal bands of ikat, subsidiary bands and finally minor bands with ikat dashes, at least on the most elaborate cloths. This organization is found on Hainan ikat, Mindanao ikat, Indonesian ikat … in other words it is a near-universal and probably ancestral feature. Even the most highly evolved variants such as Iban and Toraja ceremonial cloths retain vestiges of this organization in the form of smaller ikat motifs in side-bands. Batak cloths are different in conception: the ones with warp ikat decoration tend to use the same motif distributed in bands of equal width across a relatively wide cloth. They are more similar in this respect to weft ikat cloths from neighboring regions, many of which also feature relatively simple motifs distributed evenly across the cloth, without the hierarchical bands. Which begs the question of whether the Batak tradition should be more properly grouped with those weft ikat traditions rather than with other warp ikat traditions in SEA. I have posted an example of a Batak cloth as an illustration... compare the layout of this cloth with the Hainan tubeskirt in the first posting in this thread.
As I mentioned in the paper, you can pretty much draw a line from north to south on the map, dividing the warp ikat traditions of the eastern regions of SEA (backstrap looms with continuous circular warp, no reed) from the weft-oriented traditions (reed looms with divided warp) to the west, with some areas of overlap in the middle, and some outliers (eg Lampung ikat weaving). This is normally attributed to "Indian influence" from the west, which presumably replaced older traditions. If this is correct, then Batak weaving might be the result of adopting the styles of this tradition (wide cloths with simple, evenly distributed ikat motifs and supplementary weft embellishment) but without making the switch to weft-orientation, perhaps due to the relative remoteness of Batak weavers?
All this is very far from "proven", and parts of it are not well-defined (what does "Indian influence" mean, exactly?), but it does make an interesting discussion.