JONATHAN BUTNER

The notions of complexity theory and dynamical systems are the bedrock of science.  In the sixties psychology embraced the idea that the whole can be more than the sum of its parts.  This was a straight steal from emergence and self-organization wherein behaviors are a function of many lower order parts interacting to make higher order behavior.  Simultaneously, the higher order behavior constrains the activities of these lower order parts. This stands directly contrary to reductionist notions that we can experimentally manipulate parts of a system and learn something that has large scale impact in human behavior, for example.  My work is based on the notion that if emergence/self-organization is true, then we need to carefully consider how we can conduct research methodologically and statistically in a way that moves science forward.

Dynamical systems can be thought of as a subset of complexity theory in that it is the parts of phenomena we can depict using equations, usually from calculus.  I like to use the deceptively simple definition of dynamical systems, 'given where we are, were do we expect to go next?'  The basic tenant for systems theory is that the qualitative descriptors we use in the social sciences are actually depictions of temporal dances variables enact through time as a byproduct of emergence.  For example, depression my be something we can depict akin to homeostasis wherein environmental events and contextual cues knock a person around over time, but the person keeps returning to a set value, resilient to those events.  Dances can be relatively simple and quite complicated while always taking into account the environment as opposed to isolating or controlling for everything.  

Much of my work then is innovating and applying advanced statistical and methodological techniques assuming we can write equations to depict changes through time, even though those equations are unknown (and even the underlying principles that would drive the equations would also be unknown; we cannot replicate physics at this stage where we can define forces, laws, etc.  We can only approximate them).  Psychology is messy and thus there are many things an engineer can do that we cannot.  At the same time, theory only goes so far. And thus I am just as active working in teams applying these ideas to a variety of phenomena.

If I were to stipulate what aspects of systems theory I am most focused, it would be in the nonlinear and multivariate nature of systems. If we have many qualitative descriptions for phenomena, these would translate to different dances in time.  More than one dance always requires nonlinearity in our equation forms to capture them.  Further, I am not one to belive that our tenants of measurement from psychometrics automatically carry over in systems. Thus, we must think about how to deal with many variables at once.  This is sometimes depicted as coordination, synchrony, data reduction, and high dimensional systems (though I usually work with low dimensional systems to illustrate the concepts).