Books Discussed in this Section
The IHDP working paper at http://www.ihdp.uni-bonn.de/html/publications/workingpaper/wp02m.htm
In the working paper of the IHDP (International Human Development Project) the term scale is defined as “the spatial, temporal, quantitative or analytical dimensions used by scientists to measure and study objects and processes,” and likewise I use the term levels to “refer to locations along a scale.” As this paper makes clear, there are many different types of scales and scalar relationships to choose from when doing any sort of study or analysis. Many scales are both conceptually and causally hierarchical, but conceptual hierarchy does not guarantee causal hierarchical relatedness. There are exclusive hierarchies which are not construed to be nested sets; and there are inclusive or nested hierarchies which are construed to be nested either in an inclusive or constitutive manner. Inclusive hierarchies tend to be classification systems, such as the Linnean hierarchy of taxonomic categories. Constitutive nested hierarchies are related to physical systems as well as being a characteristic of complex systems. In a physical systems, such as “cells-tissues-organs-organism,” a constructivist interpretation which is based on physical substances or “building blocks” is not appropriate, rather constitutive “units” have unitary functions as well as emergent properties. As the IHDP paper notes, :these levels are on a conceptual scale based on functional relationships rather than on a [absolute] spatial or temporal scale.”
Critical to the understanding of scales is the understanding of all the conceptual components that go into the act of identifying any given “unit” as a particular pattern—since the identification of such a “unit” is already conceptually contrived by an implicit resolution of scale –extension, duration, agency, class, and etc.—and so the recourse to the appropriate scale of inquiry is always already constrained by the conceptual derivations pre-constituting the objects of inquiry.
Consider, for example, the notion of the “domino effect,” which can be seen to be a series of causal relations that scale beyond the causal capacity of the single domino (its ability to fell one neighbor only) and its causal capacity with respect to the whole shebang. What “scales” the capacity of the unit domino at the start cannot be determined by inquiry into the makeup of the domino itself—not at the level of size, weight, shape, nor at the level of molecules and atoms. The scalar capacity of the domino only appears at the level of the arrangement of all other dominos and their pre-given relatedness. In many instances, then, the way we conceptualize units will pre-determine the way we “see” scalar relations between and among them. This is a crucial cautionary tale.
The second issue with scales is the fallacy of relation between “absolute” scales and “relative” scales. This situation results from the extrapolation of one set of scalar relations as a kind of pre-given grid along which other phenomena are viewed, without relativizing the grid to the nature of the phenomena. For example, if I use my watch as “absolute temporal grid” then it will take 6 more hours for me to fly from San Francisco to New York than from New York to San Francisco. If I use Mean Greenwich time that scale with the phenomenon of flying, then it will take me the same amount of time. If I am studying an organism’s diurnal rhythms on that flight, then it might be the case that is does take more time in one direction than the other. If I relativize time according to Einstein’s theory – I come up with another answer entirely. Spatio-temporal relations are just one example of the kind of entanglements that apply to scales and scalar relations. This too is a crucial cautionary tale.
As the authors of the IHDP working paper point out, there is a temptation to rely on the convenience of absolute scales when the goal is to design (identify) hierarchical scalar relations that require fixed grids (the absolute scales along which the units scale hierarchically). But to reiterate the problem, only with respect to space and time, illustrated in their words as follows:
… when one is dealing with processes and mechanisms, space and time become properties of those processes and mechanisms under investigation. Therefore the study of a process cannot a priori assume certain spatial and temporal scales.
Relative space is important in studies of behavioral geography that focus on individual perception of space … absolute distance rarely corresponds with the relative distance … When this happens it is not easy to “map the processes in terms of absolute scale.” Accessibility of isolation does not always depend on absolute distance, but on the cost in terms of time and resources or mileage through a transportation system. According to the IHDP working paper, the holy grail of scale analysis is the ability to generalize across levels. This would entail a comprehensive understanding of the processes and causal relationships of units that scale within each level, as well as derivative rules of order (such as in complexity theory) that allow understanding at one level to predict outcomes at other levels, through both upward and downward causation. If the rules of order that prove to generalize between levels were to become precise enough, then scalar analysis might predict unforeseen or emerging phenomena by watching the advent , development, and flux and fluctuation of scalar processes. This ability would obviously be of great benefit when having to extrapolate enormous amounts of information to study phenomenon at global or emerging levels—and trying to make current choices towards future outcomes at those levels. If so, one critical question of temporal scale then—“how do we determine what choices available to us now, in the present where these choice fields are different among different parties, move us toward or away from common future goals” —will have been substantially resolved.