It sounds like you have a strong grasp on the ubiquity (I love that word :) of nonlinear dynamics. I've wondered about fashion trends myself, but don't have a clear model in mind for how they evolve. I bet someone has done it. Anyway, thanks again for responding Bob, and please do stay in touch. I'd love to hear about any ideas that emerge about these topics or others as the blog continues here. -Dave

Hope its okay to call you "Gomez," you can call me Pincus (friends do)... Anyway - very glad you are intrigued. We'll try to keep it that way. I agree, the momentum thing is amazing, and almost always left out of research. One interesting point of fact, in side by side tests of nonlinear and linear models, nonlinear models on average account for twice the variance - this means the models fit the data about twice as well. Steve Guastello did this comparative analysis in his last book - "Managing Emergent Phenomena." My read on this has always been that the main reason for the better fit is that nonlinear models include momentum. They also include curves, and so on, which helps too. Your question about timing is very interesting. My best answer right now would be "yes" it is hard to capture timing effects in research, even nonlinear research. What we tend to do in general is try to sample as much as a system's behavior as possible, and then we can see sensitive states or regions in which a system is relatively stable versus unstable, or succeptible to small influences. Speaking of Guastello - he has done a lot of research using "Catastrophe theory" which is a good general example of this. In a recent study he published with one of his former students, Kelly Smirz (in the journal Nonlinear Dynamics Psychology and Life Sciences), they found that they could predict patterns of bing-drinking in college students using the interaction of attitudes toward drinking and peer pressure. In terms of timing and sensitivity, what they found was that under high social pressure, you get 2 stable regions of binge drinking - basically partiers and abstainers. The students who do the occasional "beer bong" (moderate users) dissapear. So you are looking across drinkers to figure this out, but what this suggests is that a drinker who is moderate in attitudes toward drinking will have a strong timing effect if they walk into a situation with heavy social influence, even if that influence is against drinking. This person may look back at college as the start of his or her alcohol problem, even though paradoxically he or she entered an anti-drinking college situation with a moderate attitude toward drinking. This person just needs to stumble into the "Animal House" during orientation week and momentum will take it from there! I can't think of a research situation that captures that exact moment for a single individual. We still build our research understanding out of many observations in nonlinear research. We just allow the outcome observations to be curvy, branchy, or to have different levels of chaos, instead of fitting onto strait lines. I have considered doing a study of relationship process where I could give a person feedback through bug in the ear (i.e., these people don't like what you are saying) and then seeing what happens to the group dynamics - the fractal (branching structure) would most likely lose lots of branches and get quite rigid (and uncomfortable). This would be a way to experimentally look at a single moment's influence. But it is fun to speculate about the butterfly effect (i.e., sensitive dependence) in our day to day lives - if I had gotten better grades I wouldn't have gone to this college, and I wouldn't have met Josh, and I wouldn't have tried Sushi, and I wouldn't have taken Kristen out for sushi, and she wouldn't have fallen for me, and we wouldn't have gotten married, and my kids wouldn't be here...So my kids should be grateful that I slacked in high-school. This would be hard to model, but fun to try... Okay Gomez - I've gone on too long here. Hope I didn't scare you off. -Dave

Hi will try to hit on what the traditional models can do and how the nonlinear ones improve on those. This is a key consideration. -Dave

Cool Taufik. I'll try to cover the crossover stuff as we go - areas where traditional psyc theory has been advanced by adding NDS theory. This fall we'll be having a book come out (just finished edits today) called: Chaos and Complexity in Psychology: Theory of Nonlinear Dynamical Systems, by Cambridge U press. (Shameless promotion I know - but 100%of proceeds are going to SCTPLS, so...). This book is going to be the best source for this - with each chapter author taking existing theory in each area of psyc (from neurons to organizations) and presenting the move from traditional research to the latest in NDS. Glad to have you here too. -Dave

Hey Aaron, Great to have a therapist on board too. Answer - yes, work has been done here. But there is lots more to do. I'll try to weave this issue into today's post - which will start out being about the Joker, from Batman, so it will be a fun weave to try to make. Off the top of my head, the best research in therapy outcomes has been done by Wolfgang Tschacher, a psychiatrist from University of Bern, Adele Hayes from U of Delaware, Tony Tang at Northwestern University. I'd include myself, but I've only done more basic research this far, and each of these folks has actually looked at nonlinear patterns in actual therapy cases over time. Nutshell descriptions of their primary contributions thus far - Tschacher has found that greater coherence (less entropy) in self-report measures of therapy process between client and therapist predicts better outcomes. Essentially, he has come up with a novel way of studying the self-organization of shared mindset - the therapeutic alliance - that develops between client and therapist, which is so important for positive outcomes. Hayes has found evidence that clients will often 'fall apart' or disintegrate, with symptoms getting worse, more eratic shifts in mindset and mood, less coherence in behavior, etc. just before they make sudden gains in treatment. This suggests a sort of - route through chaos and on to a higher level of self-organiztion as people make adaptations within therapy. And finally, Tang has looked a lot more at 'Sudden Gains' in treatment. Which appear rather common, both in CBT and in psychodynamic approaches. Linear research, based on dose-response curves at best, and simple pre-post averages in treatment responses, tend to miss sudden gains and the like, which hides potentially interesting nonlinear mechanisms that underlie positive responses to treatment, regardless of the particular treatment you're using. There are many more, like Terry Marks-Tarlow, for example, who has a new book out from Routledge called: Psyche's Veil... This is more germain to contemporary analysis, not tx outcomes really. But she is a great writer and extremely knowledgable about analytic therapy and NDS. So I'd recommend her book to any therapist. Thanks again, -Dave Thanks again, -Dave

Your welcome Facc. I first got into this 'stuff' in late highschool and college too. Good time. Open mind and all. -Dave