As a student of the MSE and a frequenter of geography courses, my understanding of scale is far more developed than the average person’s (I hope). Marceau’s (1999) article was an interesting read because it forced me to consider, in depth, the problems beyond just noting MAUP as a point of contention in your final research project. I am very curious to see what the future holds in terms of solving the MAUP—particularly the sensitivity test if we can find a way to perform it with less effort. Maybe this already exists, as it has been 15 years.
On another note, applying this reading to my own project—scale is a somewhat challenging idea to take into account when building an ontology. Marceau is very clear about the problems of the spatial aggregation of data and cross scale correlations. Scale is obviously a huge factor in farming—what one farmer produces and how they run the farm is directly dependent on the scale of the operation. I have had trouble trying to work in a varying scale for the simple notion of a farm, since I was not planning to include geometry. I have come to realize the best way to address scale in my ontology is to specify a type of farm at a specific scale and work from there (Intensive agriculture for example). In fact by trying to include multiply scales for a farm, I would be building an upper-level ontology (which is not my goal). Geospatial ontologies built a single scale, however, may be a contributing factor the MAUP because the relationships they display won’t exist on another scale, or if they do maybe they are altered? On the other hand, a good ontology should be ‘universal’ which to means it would be applicable at many scales. So is the answer many single scale ontologies or one multi-scalar one (per research topic)?