Posted on December 23, 2015 @ 04:59:00 AM by Paul Meagher
In T.T Forman's book, Urban Ecology: Science of Cities (2014), he argues for a multiscale approach to understanding objects or patterns:
This hierarchy of scales highlights another key dimension in understanding urban ecology. Typically the object or pattern of interest at a particular scale is strongly affected by characteristics at three scales. First, characteristics at the next broader or higher scale encompasses and tend to control the object. Second, characteristics at the next finer or lower scale help control and explain the internal mechanisms or functions of the object. Third, other objects at the same scale interact competitively or collaboratively with the object considered. These three hierarchical interactions effectively mold the form of an object and determine how it functions. (p. 12)
In Urban Ecology, for example, Forman posits these 9 different scales:
- Urban Region
- Metro Area
- Major Land Use Type (e.g., residential, commercial, etc...)
- Micro-site (e.g., wall, roof, basement, and so forth)
So if you are trying to understand some pattern at the city level then you should also think about the factors that influence and determine that pattern at the Metro scale and at the Major Land Use Type scale in order to better understand the drivers of that pattern and how it might evolve over time.
One aspect of this technique that should be noted is that what constitutes a scale or level of analysis is not always clear and the art of using this technique is in large measure defining the macro and micro
scales that are appropriate to understanding the object or pattern you are interested in. In the simplest cases you might be able to invoke some domain of physics or chemistry to define the micro scale of analysis. In the case of understanding how mosses grow (a recent interest of mine) the micro scale might be the soil profile under the moss and the macro scale might be the topography of the landscape that a particular type of moss is situated in. What you are trying to understand may suggest what micro and macro scales might be appropriate. The benefit you might derive from multiscale thinking may reside in defining what micro and macro scales might be appropriate to use.
Another aspect of this technique worth noting has to do with the reason for engaging in this type of analysis. My own feeling is that there is no direct benefit in adopting this approach unless you consider intellectual understanding a benefit. Otherwise, adopting this approach is not for the direct purpose of solving a problem or designing something. Prior to solving a problem or designing something, however, we need to understand the patterns we are observing and it is in this prior stage that a multiscale approach to understanding might be useful. Without a good understanding of the domain our problem solving or designs are likely to be less than optimal.
In the case of investing in a business, for example, a multiscale approach might be useful in understanding the viability of the business better. The business operates in a particular niche and we might evaluate the business with respect to competition and collaboration with other companies servicing a similar niche (level n). To follow this multicale approach we would also have to bring in the microscope and analyze the business at finer scale, and a macroscope to analyze the business at a broader scale. Once we have an understanding of the business at these three scales we might be in a better position to decide if we should invest in the business. So the investment heuristic here is only to invest in a business after you have performed an analysis at three scales.
I'm not sure at this point how the micro and macro scales should be defined for businesses in general. It would be nice to have set of scales like Forman does for urban ecology to utilize for understanding business patterns for the purpose of investing. I'm still trying to master multiscale thinking and would not want to try to define such a scale this early in the game. Studying some micro and macroeconomics texts would probably prove useful in such a quest. Perhaps in a future post I'll revisit this topic with some suggestions for scales that might include time and motion studies all the way up to global economic forces that can be used to better understand business patterns.
Fields such as multiscale modelling might provide some useful formal techniques.
Allan Newell in Unified Theories of Cognition (1991) was very interested in understanding cognition at multiple scales and has one of the best extended discussions of multiscale modelling. Below is Newell's time scale of human action. Newell's book didn't address the social scale which is why it is not very details at this scale.