An HCI colleague today asked me to define the study of Complex Systems – especially as it relates to the the study of HCI (human-computer interaction). Now the definition is still pretty hotly contested by mathematicians, cognitive scientists, etc. But this conversation occurred between two HCI folks. Here was my answer, as concise as i could give it (not because my colleague is simple, but rather because we didn’t have a lot of time). I think it bears writing down here for other folks who might be wondering about this. The reasoning works in four basic steps:
- A system, as defined by some notable systems scientists, beginning with Marchal  is simply S=E,R. In other words, a system (S) is a set of elements (E) and a set of relations (R) between those elements. So for example, as far as an HCI researcher would care, the MySpace “system” might consist of the users, the administrators and the software and the relations between them all.
- HCI has primarily concerned itself deterministically with the elements of a system – in this case the users or the administrators or the software. This is to say, that, at its core, HCI has assumed that, if it could know the elements of the system well enough, the outcome of that system could hypothetically be known.
- Complex Systems has primarily concerned itself probabilistically with the relations of a system – the relations between the users and the administrators and the software in the MySpace example. So at its core, Complex Systems, because of the complex, unpredictable nature of those zillions of relations, has pretty much thrown out the hope of exactly knowing the outcome of the system, but instead employs its own methods to figure out the probabilities of different outcomes.
- Complex HCI (term coined, i think, by Kevin Makice) will try to simultaneously understand both the elements and the relations in systems such as MySpace, etc. – and the ways that one affects the other.
To be complete and just to both disciplines, this post would need to be a great deal longer, since, of course HCI is not completely deterministic, and there are, um, complexities to the Complex Systems side that this simple 4-point post does not address. But the general gist is here, i think – at least as it currently appears to me.
..And from this, it is easy to see the potential synergies between HCI and Complex Systems, for the design of just about any software, web platform, mobile device now must consider not only each user or device (element), but the huge numbers of connections (relations) that will occur between that user or device and the millions of other potential users and devices with whom that person or device will connect.
 Marchal, J. H. (1975). On the Concept of a System. Philosophy of Science, 42(4), 448-468.