Sometimes we seek to infer from a design what its requirements might have been, and in ID thought this question comes up. As a practitioner in the architecture of large scale computer environments (the composite set of applications, databases, and communications networks) in major enterprises, I wonder if some of the principles my profession uses in design could be useful in understanding what is going on in biology.
First a little background. What I am describing applies, in my opinion, to architecture and I would submit there is a rather considerable tension between architecture and design. But I am not going to get into that now, so let’s assume they are the same and call them “architecture”. Next, in my field you should be aware that Carnegie-Mellon University’s Software Engineering Institute (the “SEI”) has been responsible for the development of a lot of the thinking in this area. One of the SEI’s insights is that there are “Architectural Quality Attributes” (AQA’s). These are a whole set of characteristics that the global architecture of an enterprise may have.
Examples include “availability”, “customizability”, “extensability”, “understandability” and so on. A list is available here. Because the AQA’a tend to end in the suffix “ility” they are informally known as the “ilities”.
It is simply impossible for one architecture to have all the “ilities” because many conflict. For instance, if I want high “security” I am going to have to give up a good deal of “interoperability”. A large part of architecture is actually deciding what you are going to give up, which incidentally affects how the architecture can change in the future (i.e. usually it cannot “evolve” to conform to different “ilities”). This is all still fairly new, but we are now able to judge architectures in terms of the “ilities” they match and the “ilities” they do not match. A better understanding of the conflicts between certain “ilities” is gradually developing.
If we could similarly develop a taxonomy of “ilities” for biological systems we could then judge the qualities of different biological designs and understand the trade-off’s among them. They key is to have a standard taxonomy that most biologists would accept. A key issue might be that by accepting such a taxonomy biologists would be accepting that there is such a thing as “design”. Furthermore, figuring out the trade-offs between biological “ilities” would render evolution based on random chance more transparently preposterous. Lastly, this approach offers a way to infer, and perhaps, predict from an ID perspective.