Isoglosses do not accurately reflect linguistic usage in a region, because the isolated conservative forms they are based on do not represent actual variants extant in the population. The isograph enables researchers to find more representative dialect trends. Canadian and American data from the Dialect Topography database are submitted to isographic analysis of linguistic boundaries at the provincial, national, and cross-border levels. Topography provides a multi-dimensional picture of how variants are used. Variants occur in different proportions, so analysis is quantitative and implies the abandonment of the isogloss because the discrete datum-points of dialect geography are unavailable. The isograph compares adjacent regions, and plots potential channels for language spread. Percentage differences between neighbouring regions are calculated, and a line is drawn between neighbours with the least difference. When all lines of minimum distance have been drawn, the result is a constellation of the most similar regions.