It seems that that the stories of modularity in design and convergent designs share a great deal of similarities and provide a fertile ground for pushing the concept of modularity to its conceptual limits. Baldwin and Clark might sum up modularity by the maxim, “independent but interdependent.” In other words, just as in the story of Hora and Tempus, the parts do not depend on one another in a cumulative, linearly ordered fashion, per say, but instead, are each constructed as interchangeable and self-contained systems which then construct the larger system of the final design. Convergence, on the other hand, involves taking pre-existing systems and designs, and incorporating them into different (either new or pre-existing) systems. The modern car might be thought of as a commonplace example which contains both modularity and convergence — although Langlois is quick to point out that the original Model T was not exactly the paragon of modularity. I find the engine of a car to perfectly typify the principle of modularity, as this sub-system is by all accounts entirely essential for the functionality of the car, while the engine “module” still might be upgraded or downgraded within the designed system of the car. Even if the car in toto fails to be rigorously identified as a modular system, an engine certainly can be thought of as a modular component as its independence from and interdependence to the system define the nature of its design at large.
The car also demonstrates the principle of convergence in, say, the installation of the radio on the dashboard. Unless you drive a vehicle designed by some sort of technomasochist, your car probably still works when the radio cuts out — in other words, the radio and the car are not interdependent in the same way as the car and the engine. But the radio still seems to possess some modular qualities. For example, the car radio is specifically designed for a car, meaning that a car radio and a “non-car” radio are not quite the same thing. They may both serve the same end function of channelling signals cast over radio waves, but one is designed so as to connect with the speaker system on the interior of an automobile while the other is not (among other ways which exemplify the imbrication of the design of the car and the car radio). In this way, there seems to be a level of interdependency between the car radio and the car, albeit in one direction only — from the car radio to the car. To complicate matters further, car radios, like modules, might be upgraded or downgraded and therefore frequently subscribe to the same design specifications so as to make them interchangeable. What this also implies, however, is that the specifications of different car models themselves be designed according to even banal specifications such as depth and width so that the radio, even while non-essential to the car as a designed system, might achieve a level of pseudo-modularity.
As more and more technologies begin to be designed according to the principle of convergence, to what degree can designers or design theorists consider convergence to be different from modularity? Does modularity require interdependence from both the module and the system, or might it only require interdependence in a single direction? In other words, must a feature or function be essential to a designed system in order for it to be a module? Of course, the case of the car and car radio is a relatively simple example, however, it seems likely that this ambiguity remains relevant in more complex examples, including software design. Ultimately, I’d like to know whether the difference between modularity and convergence has to do with anything other than architectural structure (modularity) and utility (convergence). Is the difference between the two anything more than superficial?
Baldwin, C. and Clark K. (2000) Design Rules, Vol. 1: The Power of Modularity. Cambridge, MA: The MIT Press
Langlois, R. (2002) “Modularity in Technology and Organization.” Journal of Economic Behavior & Organization 49, no. 1: 19-37.