Some Devolutional Similarities to Diffusion

Abstract

Our concern in this paper is to extend some models that are descriptive of diffusion processes to the devolution of phenomena. Specifically, we search for disasopter (abandonment) groups who are the devolutional counterparts to the groups of adopters in a diffusion process and for evidence of the applicability to devolution of the wave-like model that depicts the diffusion of some phenomena. We find that extension of these diffusion concepts to the inverse of diffusion, devolution, is appropriate in the disappearance of milk cows from farms in Illinois. In the development of general theories of socioeconomic change, much attention has been given to the adoption and diffusion of innovations (Brown, 1981; Hudson, 1972). A considerable amount of time also has been spent on the devolution of phenomena, which may be viewed as the inverse of diffusion processes from both a temporal and a spatial perspective. Two examples of devolutional studies are an investigation of the demise of cotton production on the Piedmont (Prunty and Aiken, 1972) and a study of land use in Carroll County, Georgia (Hart, 1980). These latter works seldom, however, link diffusion theory per se to devolution, a notable exception being an examination of the disappearance of dairy farming in east central Illinois (Frederic, 1973). Another exception is an article which extends the logistics model of temporal diffusion processes to the disadoption and devolution of phenomena (Mattingly, 1984). As a part of what could be a first step to the development of a general model of spatial devolution and diffusion, we seek to develop two further devolutional similarities to the diffusion process. Specifically, we argue that the wave model analogy that sometimes describes the movement of innovation adoption across a surface has a spatial inverse, that is, waves of disadoption. Secondly, we attempt to identify the disadoption counterparts to the innovators, early majority, late majority, and laggards of the adoption process. In order effectively to model these presumed inverses of diffusion, our investigation will be limited to the disadoption of milk cows by farmers in Illinois. A reason for this limitation is that it has already been demonstrated that the logistics model is a good fit to the rate of disadoption of this animal in Illinois (Mattingly, 1984), and such a fit is required if one is to identify groups of disadopters who are the equivalents to the groups of innovation adopters. A further reason for this limitation is the spatial pattern associated with the disadoption of these animals in Illinois. Rather than a type that is random or more regular than random, the pattern appears to be a clustered one in which the composition of the most rapidly changing group (cluster) varies with time (Mattingly, 1984). If devolution waves exist, such a pattern appears to be a prerequisite to detecting them, and these waves would depict the opposite of a diffusion wave i.e., contracting or disappearing instead of spreading. Our attempt to extend these two aspects of the diffusion process to spatial devolution is complicated by several matters: the available data are by counties rather than by individual farmer; the exact point in time, if one exists, when disadoption may be said to have become an ongoing process is difficult to detect as there have long been some farmers who did not have milk cows, and the proportion of such farmers has differed from time to time and from one place to another even within Illinois; the individuals composing the population of farmers and the number of farmers change with time; and the U.S. Bureau of the Census data are available only on a periodic basis. These complications, however, merely make it more difficult to detect the sought for features than would otherwise be the case. The groups of disadopters In any area in which an innovation diffuses, some persons, identified in the diffusion literature as innovators or early innovators, adopt first. They are followed by a much larger group of adopters, the early majority, who, in turn, are followed by an equally large group called the late majority. Near the end of the process, only a few persons, lag* Professor Paul F. Mattingly and Assistant Professor George Aspbury, Department of Geography-Geology, Ilinois State University, Normal, Illinois 61761, USA. GEOGRAFISKA ANNALER ? 67 B (1985). 1 1 This content downloaded from 157.55.39.180 on Mon, 25 Apr 2016 07:05:30 UTC All use subject to http://about.jstor.org/terms