I've been sitting on this for 15 years.
The project got its start in 1972, when I first arrived in Nova Scotia to do research on a region (the Annapolis Valley) and a process (agricultural transformation, from export to local supply), and it began in the most mundane way imaginable. I opened the telephone book and spatial patterns tumbled out. The dominance of D'Entremonts in the Pubnico exchange, and Hiltzes and Keddys in New Ross, the concentration of Ellses around Canard and Saulniers near Saulnierville --these observations meant something, and were parts of larger patterns, but what? And how to get at the patterns and what they could say about constitution and process in regional systems?
I played with these data off and on for a number of years, as I gradually came to understand more about how Nova Scotia worked, but it wasn't until I bought a microcomputer in 1984 that I had the wherewithaltic study of surname distributions. A blitz in the summer of 1985 produced the maps collected here, but various stones in my passway kept me from "finishing" the project. Now I see that some of the impediments were technological: the equipment and software available to me in the mid 1980s allowed me to produce precise analog maps, which one could compare and admire, but I couldn't see how to do the quantitative analysis of spatial distributions that would have allowed me to go beyond anecdotal and impressionistic generalizations about the aggregation of surname territories into empirical regions. And my own interests continues to morph and ewander, drawing my attention away from the interesting but obscure questions of Nova Scotia's subregions and internal migration history. By 1990 I had left Nova Scotia and soon after I embarked on a new career as a librarian.
Recent acquisition of GIS software reawakened the human geographer, and now I have the tools that would have allowed me to make the quantitative analyses that eluded me. Perhaps unfortunately, the digital data which pushed the plotter pens in 1985 are in a format which is probably unrecoverable, or at least practically unrecoverable. The maps were made with an early version of AutoCAD, saved on disks formatted for the obsolete TI-Pro, and the original intent was to produce analog maps. In theory I could re-enter all the data from the original coding sheets, but the end result wouldn't justify the effort. After all, who really cares about Nova Scotia's constituent regions, or internal migration, or the distribution of surnames?
So this corpus of maps is a curiosity. Keddys, Handspikers, Bishops and MacDonalds are predictably fascinated by the pairs of maps that bear their names, but aren't all that interested in distributions of Hiltzes, MacLeods or LeBlancs. Students of human geography may find the method quaint or impossibly compromised (what about the women?), and Maritime historians would find much to criticize in the original data. But there are patterns, and processes can be inferred by comparisons of pairs of maps, or sets of maps with similar spatial concentrations. In the more innocent and perhaps less rigor-obsessed atmosphere of the 70s and 80s, when I used these maps and other surname data in teaching, the provocative question seemed a sufficent goal, and these maps certainly pose lots of questions, especially if one is talking to classrooms of Nova Scotians. But they don't speak to the problems of the late 1990s, or not with the same study-your-surroundings voice.