The dynamical stance laid out by Chemero in the previous chapter has a potential flaw (besides being a bit weak-ass) - it's not clear how it can serve as a guide to discovery. How do you do productive science taking this approach? Chemero is going to make two suggestions, only one of which I think works: first, he's going to suggest dynamical models such as the Haken-Kelso-Bunz (HKB) model can serve to stimulate empirical work even when they are entirely phenomenological. This approach is, I think, entirely incorrect, and this chapter is full of serious problems (only some of which are unique to Chemero). Second, he's going to suggest that Gibsonian ecological psychology can actually solve the problem much more robustly anyway, by serving as an underlying theory of behaviour. This will work better, and I would advocate Bingham's model of coordination as an exemplar of this, more promising route.
But first, the HKB model as guide to discovery (this chapter is largely the material from Chemero, 2000; I intend to turn this post into a paper to rebut that paper and point to the Bingham model as an alternative, so comments are especially welcome on this one). Time to get a little critical, I'm afraid.
But first, the HKB model as guide to discovery (this chapter is largely the material from Chemero, 2000; I intend to turn this post into a paper to rebut that paper and point to the Bingham model as an alternative, so comments are especially welcome on this one). Time to get a little critical, I'm afraid.
I've described the HKB model in detail here; it's Kelso's potential function model which describes the basic phenomena of bimanual rhythmic movement coordination. The key feature of the approach is to treat limbs engaged in coordinated actions as nonlinearly coupled oscillators, where the coupling requires energy to maintain and so would dissipate without maintenance (Kugler, Kelso & Turvey, 1980). Chemero's suggestion is that if this model can lead to useful experiments and be adapted to cope with new findings without losing this essential character, then this would demonstrate the HKB model and approach can serve as a guide to discovery for a science, specifically RECS.
Problem 1
One initial problem with this analysis crops up while Chemero is describing the HKB model. He goes over the basic features (c.f. Kelso, 1995) and then says
Importantly, the HKB model makes a series of specific predictions. First, it predicts that as rates increase, experimental subjects will be unable to maintain out-of-phase performance. Second, even at slow rates, only relative phase of 0 [0°] and .5 [180°] will be stable. Third, the behavior should exhibit critical fluctuations: as the rate approaches the critical value, attempts to maintain out-of-phase performance will result in erratic fluctuations of relative phase. Fourth, the behavior should exhibit critical slowing down: at rates near the critical value, disruptions from out-of-phase performance should take longer to correct than at slower rates.
Chemero, 2009, p. 88
The first two are not predictions of the HKB model: they are experimental phenomena which the HKB model was built to describe. This matters, because Chemero wants the HKB to serve as a guide to discovery and counts these as evidence that it can do this. In fact, they simply reflect the phenomenological nature of the model and the modelling approach.
The next section is a series of eight cases, seven about the HKB model, which supposedly show how it can act as a guide to discovery:
Case 1: Modifying the HKB model with a noise term to fit the data
Schöner, Haken & Kelso (1986) added a noise term to the HKB data. It's not at all clear how this is supposed to count in favour of the HKB as a guide to discovery, given that noise terms are standard in models. All that happens is the model fits the data better; this is still, therefore, a development driven by the data, and not the reverse.
Case 2: Reconceptualising learning as a phase transition
Chemero then (incorrectly, as far as I can tell from reading the paper) describes the first extension of the HKB to include learning (Schöner & Kelso, 1988) as allowing the B/A ratio to vary over trials; this control parameter dictates the shape of the attractor layout, and learning then becomes a change in this layout - a phase transition. The article (and all the later empirical work with Pier Zanone) actually models learning as the imposition of a required relative phase, via a third term in the model (p. 86 of Schöner & Kelso, 1988). Learning is indeed a phase transition for the HKB approach, but one brought on by competition between the existing and target landscapes. Note that this led to various predictions which fell through when tested, and although incorrect predictions would presumably be a cause for concern in any guide to discovery, Chemero does not discuss any of this learning work.
Chemero then makes another error, and cites a paper (Amazeen, Sternad & Turvey, 1996) which he claims demonstrates a phase transition due to learning. He describes the paper as training participants to wiggle their fingers in a difficult 5:4 poly-rhythm, which then led to a reorganisation of the attractor layout such that 180° became more stable than 0°. This paper has nothing to do with learning, however. There was no training involved, and the reversal in the stability layout was actually the result of detuning (if the two oscillators have different preferred frequencies, then (roughly) the HKB layout remains the same shape but is moved along the relative phase axis to a degree proportional to the frequency difference, hence the empirical result). I have no idea what happened here, but this case is a good lesson in the importance of reading the primary literature.
Case 3: Modelling perception-action couplings
Chemero is clearly familiar with detuning, because this case is about how extending the model to include a detuning term (Kelso, DelColle & Schöner, 1990) bolsters his case for the model as guide; it can now cope with new phenomena (coordination between a person and a metronome beating with a variable frequency). Incorporating detuning was a perfectly interesting addition, but describing this as modelling 'perception-action couplings' is entirely incorrect (and note the modification is data driven again). First, even 1:1, within person coordination entails perception, so detuning cannot be the move that incorporates perception-action couplings. Second, detuning doesn't actually impact the coupling at all, except indirectly by altering the motion to be perceived; detuning does not change the basic HKB shape, which emerges from the coupling requirement. Of course, the HKB model can't tell you about this; only a fully perception-action model can. This problem doesn't stem from Chemero; it's typical in the literature to not notice that within-person coupling entails perception. In essence, this is still model fitting, not guided discovery.
The next section is a series of eight cases, seven about the HKB model, which supposedly show how it can act as a guide to discovery:
Case 1: Modifying the HKB model with a noise term to fit the data
Schöner, Haken & Kelso (1986) added a noise term to the HKB data. It's not at all clear how this is supposed to count in favour of the HKB as a guide to discovery, given that noise terms are standard in models. All that happens is the model fits the data better; this is still, therefore, a development driven by the data, and not the reverse.
Case 2: Reconceptualising learning as a phase transition
Chemero then (incorrectly, as far as I can tell from reading the paper) describes the first extension of the HKB to include learning (Schöner & Kelso, 1988) as allowing the B/A ratio to vary over trials; this control parameter dictates the shape of the attractor layout, and learning then becomes a change in this layout - a phase transition. The article (and all the later empirical work with Pier Zanone) actually models learning as the imposition of a required relative phase, via a third term in the model (p. 86 of Schöner & Kelso, 1988). Learning is indeed a phase transition for the HKB approach, but one brought on by competition between the existing and target landscapes. Note that this led to various predictions which fell through when tested, and although incorrect predictions would presumably be a cause for concern in any guide to discovery, Chemero does not discuss any of this learning work.
Chemero then makes another error, and cites a paper (Amazeen, Sternad & Turvey, 1996) which he claims demonstrates a phase transition due to learning. He describes the paper as training participants to wiggle their fingers in a difficult 5:4 poly-rhythm, which then led to a reorganisation of the attractor layout such that 180° became more stable than 0°. This paper has nothing to do with learning, however. There was no training involved, and the reversal in the stability layout was actually the result of detuning (if the two oscillators have different preferred frequencies, then (roughly) the HKB layout remains the same shape but is moved along the relative phase axis to a degree proportional to the frequency difference, hence the empirical result). I have no idea what happened here, but this case is a good lesson in the importance of reading the primary literature.
Case 3: Modelling perception-action couplings
Chemero is clearly familiar with detuning, because this case is about how extending the model to include a detuning term (Kelso, DelColle & Schöner, 1990) bolsters his case for the model as guide; it can now cope with new phenomena (coordination between a person and a metronome beating with a variable frequency). Incorporating detuning was a perfectly interesting addition, but describing this as modelling 'perception-action couplings' is entirely incorrect (and note the modification is data driven again). First, even 1:1, within person coordination entails perception, so detuning cannot be the move that incorporates perception-action couplings. Second, detuning doesn't actually impact the coupling at all, except indirectly by altering the motion to be perceived; detuning does not change the basic HKB shape, which emerges from the coupling requirement. Of course, the HKB model can't tell you about this; only a fully perception-action model can. This problem doesn't stem from Chemero; it's typical in the literature to not notice that within-person coupling entails perception. In essence, this is still model fitting, not guided discovery.
Case 4: Social coupling
Schmidt, Carello & Turvey (1990) showed that the basic HKB phenomena persist when the coupling is between people; Chemero describes this as social coupling and is excited by the fact that the HKB model could be extended to not just perceptual (Case 3) but social phenomena. This is entirely the incorrect way to analyse these results; the Schmidt et al paper is, in fact, perceptual coupling. Just because there are two people involved does not change this fact. In fact, the kinematics from this experiment were used by Bingham to drive the coordination displays in his first visual perception studies, effectively using them to make a point-light display. (There is an extensive research literature on biological motion perception which uses these types of displays to explore the motion based information underpinning our knowledge of people's gender, age, psychological state, etc; Nikolaus Troje's BioMotion lab has numerous excellent references and displays). Again, only a fully perception-action model, with specific hypotheses about the information involved, could enable you to understand this fact; the HKB approach simply doesn't provide the tools.
Case 5: Asymmetries other than detuning
Treffner & Turvey (1995) expanded the model with some additional terms to account for symmetry-breaking, i.e. asymmetries between the oscillators caused by things other than detuning. They coped with handedness and attention by adding two additional sine terms to the model, with parameters which could be set to weight their contribution to fit data from handedness and attention experiments. Again, a perfectly sensible exercise in model fitting, but there is still no guide to discovery: the asymmetries were phenomena the model could not cope with, not phenomena revealed by the model, and the model fitting exercise was not constrained in the way I've suggested is critical (the added terms were simply additional sine components with their own parameters, one of which was just labelled 'attention' or 'handedness').
Case 6: Speech production
Port (2003), a linguist at IU, has suggested you could extend the HKB model to serve as a general model of meter in speech production. Chemero suggests that this work shows how the HKB model can be extended beyond it's original scope, a key feature of any guide to discovery. It's not clear how Port's work helps, however: his data from a simple speech task showed four attractors (at in-phase, anti-phase, and two other locations). You can indeed model this by adding an extra Fourier term to the basic HKB model, but again this is mere data fitting, and the HKB model has not served as guide to discovery other than loosely inspiring the experiments. Essentially, Port's model isn't a modified HKB, it's just an HKB-style model of a different dynamical regime.
Case 7: Cortical coordination dynamics
Kelso got quite interested in later years in describing patterns of cortical activity in terms of the HKB model. Coordination of activity between different regions seems to vary with task, most interestingly in that changing task seems to be associated with in-phase activity. Chemero is correct in noting that this work is quite useful to counter the claim that RE cognitive science ignores the brain. However, nothing in his review of this work suggests this work is anything more than an interesting way to describe brain activity. In addition, the work which has revealed the task dynamic from the HKB pattern emerges in the first place entail a specific mechanism which does not obviously generalise to cortical activity. The general principle, of modelling complex phenomena as non-linearly coupled oscillators, seems to be of preliminary use (the HKB model, for all it's limitations, has led to a lot of activity) but there's still no mechanism to constrain future work.
Case 8: Solving (representational hungry) gear problems
Chemero describes some work (e.g. Stephen et al, 2009) in which a classic problem solving task has been studying in terms of dynamics. Specifically, this work goes looking for various dynamical signatures (critical slowing down, fluctuations) in measures of problem solving behaviour. This seems to be quite interesting work, but to say it's inspired by the HKB model is misleading. While the HKB model is the most common example of this type of dynamical systems psychology, it is not the source of predictions about slowing down or fluctuations - dynamical systems theory is, and Stephan et al draw all their predictions directly from there, only citing Kelso (1995) once! Even if you allow for a more charitable reading, of research using an 'HKB-like' model, it's still not the model that's done any guiding to discovery; it's the principles of dynamical systems theory.
Some thoughts
I was really disappointed with this chapter, and as I said I'm inclined to turn these concerns into a paper. The section on the HKB model is, as far as I'm concerned, an unnecessary tactical error. The HKB just doesn't do what Chemero needs it to (serve as guide to discovery) and this is made all the clearer by the presence of an alternative (a theory, ecological psychology). The history of physics he refers to in the previous chapter (Mach's phenomenalism vs. Boltzmann's atom theory) was decided convincingly in favour of the theory; RECS has a theory, and so it just doesn't need to try and lever anything out of the (phenomenological) HKB approach. This is especially relevant given a) the errors in Chemero's treatment of the model, which are not unique to him, and that b) it took a theory based approach to actually start to address the mechanisms underlying coordination.
To his credit, Chemero notes at the end that this type of approach won't satisfy many people, and that the better solution is, indeed, the theory based approach to science which he develops in the next two chapters. But he still saw fit to include it in the book, and so I'm assuming he still thinks this is a sensible strength of the approach he is developing. It isn't, and I think it represents a flavour of science that is causing nothing but trouble for psychology. A lack of a guiding theory in psychology has left the field chasing phenomena and unable to begin developing much in the way of a coherent account of human behaviour. Social psychologists barely speak to cognitive psychologists who know nothing about perception or motor control, and we rattle along in our separate little domains. Toy models designed to merely describe a limited set of phenomena, as the HKB is, cannot serve as guides to discovery.
To his credit, Chemero notes at the end that this type of approach won't satisfy many people, and that the better solution is, indeed, the theory based approach to science which he develops in the next two chapters. But he still saw fit to include it in the book, and so I'm assuming he still thinks this is a sensible strength of the approach he is developing. It isn't, and I think it represents a flavour of science that is causing nothing but trouble for psychology. A lack of a guiding theory in psychology has left the field chasing phenomena and unable to begin developing much in the way of a coherent account of human behaviour. Social psychologists barely speak to cognitive psychologists who know nothing about perception or motor control, and we rattle along in our separate little domains. Toy models designed to merely describe a limited set of phenomena, as the HKB is, cannot serve as guides to discovery.
References
Amazeen, E. L., D. Sternad, and M. T. Turvey (1996). Predicting the nonlinear shift of stable equilibria in interlimb rhythmic coordination. Human Movement Science, 15, 521 542. DOI
Chemero, A. (2000). Anti-Representationalism and the Dynamical Stance Philosophy of Science, 67 (4) DOI: 10.1086/392858 Download
Kelso, J. A. S. (1995). Dynamic Patterns. Cambridge, Mass.: MIT Press.
Kelso, J. A. S., DelColle, J., & Schoner, G (1990). Action perception as a pattern formation process. In Attention and Performance XIII, ed. M. Jeannerod, 139-169. Hillsdale, N.J.: Erlbaum
Kugler, P. N., Kelso, J. A. S., & Turvey, M. T. (1980). Coordinative structures as dissipative structures I. Theoretical lines of convergence. In Tutorials in Motor Behavior, ed. G. E. Stelmach and J. Requin. Amsterdam: North Holland. Download
Port, R. (2003). Meter and speech. Journal of Phonetics, 31, 599-611. DOI
Schöner, G., & Kelso, J. A. S. (1988). A synergetic theory of environmentally specified and learned patterns of movement coordination. II. Component oscillator dynamics. Biological Cybernetics, 58, 81 89. DOI
Schoner, G., H. Haken, H, & Kelso, J. A. S. (1986). A stochastic theory of phase transitions in human hand movement. Biological Cybernetics, 53, 247 257. DOI Download
Schmidt, R. C., Carello, C., & Turvey, M. T. (1990). Phase transitions and critical fluctuations in the visual coordination of rhythmic movements between people. Journal of Experimental Psychology: Human Perception and Performance, 16(2), 227-247. Download
Stephen, D.G., Dixon, J. A., & Isenhower, R. W. (2009) Dynamics of representational change: entropy, action, and cognition. Journal of Experimental Psychology: Human Perception and Performance, 36(6), 1811-1832. DOI
Treffner, P., & Turvey, M. (1995). Symmetry, broken symmetry, and handedness in bimanual coordination dynamics. Experimental Brain Research, 107, 163 178. Download
Chemero, A. (2000). Anti-Representationalism and the Dynamical Stance Philosophy of Science, 67 (4) DOI: 10.1086/392858 Download
Kelso, J. A. S. (1995). Dynamic Patterns. Cambridge, Mass.: MIT Press.
Kelso, J. A. S., DelColle, J., & Schoner, G (1990). Action perception as a pattern formation process. In Attention and Performance XIII, ed. M. Jeannerod, 139-169. Hillsdale, N.J.: Erlbaum
Kugler, P. N., Kelso, J. A. S., & Turvey, M. T. (1980). Coordinative structures as dissipative structures I. Theoretical lines of convergence. In Tutorials in Motor Behavior, ed. G. E. Stelmach and J. Requin. Amsterdam: North Holland. Download
Port, R. (2003). Meter and speech. Journal of Phonetics, 31, 599-611. DOI
Schöner, G., & Kelso, J. A. S. (1988). A synergetic theory of environmentally specified and learned patterns of movement coordination. II. Component oscillator dynamics. Biological Cybernetics, 58, 81 89. DOI
Schoner, G., H. Haken, H, & Kelso, J. A. S. (1986). A stochastic theory of phase transitions in human hand movement. Biological Cybernetics, 53, 247 257. DOI Download
Schmidt, R. C., Carello, C., & Turvey, M. T. (1990). Phase transitions and critical fluctuations in the visual coordination of rhythmic movements between people. Journal of Experimental Psychology: Human Perception and Performance, 16(2), 227-247. Download
Stephen, D.G., Dixon, J. A., & Isenhower, R. W. (2009) Dynamics of representational change: entropy, action, and cognition. Journal of Experimental Psychology: Human Perception and Performance, 36(6), 1811-1832. DOI
Treffner, P., & Turvey, M. (1995). Symmetry, broken symmetry, and handedness in bimanual coordination dynamics. Experimental Brain Research, 107, 163 178. Download
Posts
Copying this comment from Tony to here so my reply makes sense :)
ReplyDeleteTony said:
Ouch!
Let me begin by copping to some of the mistakes you point to. First, yes, Amazeen, Sternad and Turvey was the wrong paper to cite. Nia Amazeen has already (politely) yelled at me about this. The backstory is that I wrote this chapter while sitting in on Turvey's seminar in 2003, and many of the papers discussed in the chapters are from Turvey's reading list. Somewhere in the 5 years between writing the chapter and the page proofs, the wrong citation from that list got cemented into my text. I plead guilty; I plead guilty also to not having gone back to check which article I read while writing this section. Second, I'm not sure what I'm supposed to have missed about Schoner and Kelso, but I don't dispute your superior knowledge of this literature. Third, my claims about Stephen, Dixon and Isenhower were based on discussions with Damian Stephen, rather than the published paper. In fact, I wrote the section of the book before having seen the paper. It is better to view the work in the 2009 paper, and the work that has followed, as you suggest: as stemming from the same sources as HKB model (i.e., synergetics and the dynamics of phase transitions) than as stemming from the HKB itself.
OK, I suck. I wish I could say there weren’t other mistakes in the book, but there are.
(I also would have written some of this differently had I read your blog posts on coordination dynamics beforehand.)
There are some things I do stand by, though. For one, Schmidt et al has been understood by most everyone as genuinely social coupling. There is a cottage industry on social coordination dynamics that has grown out of this work, as you know. I think they get it right: interpersonal perceptual coupling just is a variety of social coupling. For another, while I agree that a theory is a better way to have a guide to discovery than a set of flexible models, I think it is a mistake to say, as you seem to be saying, that good science can’t be done without explicit commitment a theory. Look, for example, at the Stephen et al work. As you correctly point out, this work is inspired by the dynamical models of phase transitions. That’s not a psychological theory. The experiments are inspired by the models, as attempts to see if there are more phenomena that they apply to. So, as a matter of empirical fact, lots of successful science is inspired by models, without commitment to any particular theory.
That said, I do agree that theory is a better guide to discovery than models. It is more flexible, and not only because theories are generally compatible with a wide array of models. This is why the next few chapters are in the book.
Thanks, again, for spending so much time on the book.
I admit this chapter got me fairly grumpy. I wrote most of it a little while back, while I was working through some of the coordination stuff, and my critique of this chapter is grounded in my frustration with our inability to break through the dynamic pattern stuff, no matter how often we break it. Your chapter happens to have a lot of this kind of thinking in one place, so it gave me plenty of focus :)
ReplyDeleteSecond, I'm not sure what I'm supposed to have missed about Schoner and Kelso, but I don't dispute your superior knowledge of this literature.
It's a dense paper, and I stared at it for quite a while. As far as I can tell, the B/A ratio isn't what they tweak to model learning; they simply impose an external dynamic to add a third attractor at, say, 90°. This would be consistent with the Kelso & Zanone literature. This approach is a problem because it's predictions don't pan out (learning is easier near 0°, not 180°), but worse, Zanone now just claims it predicts what happens and never refers to that early stuff. Drives me mad.
(I also would have written some of this differently had I read your blog posts on coordination dynamics beforehand.)
Actually, I've been thinking about this. As I say I'd quite like to turn this into a paper; your paper and this chapter put a lot of things I'd like to address in one convenient place and so it's a useful starting point. But the paper needs to be more than just me being mean about some errors! The plan is to use it as an excuse to lay out a few problems with the HKB approach but then direct attention to Geoff's model and the work that backs it up, to try and take the point but move it forward. It'd be nice to get a dialogue going on this stuff.
Part II:
ReplyDeleteThere are some things I do stand by, though. For one, Schmidt et al has been understood by most everyone as genuinely social coupling. There is a cottage industry on social coordination dynamics that has grown out of this work, as you know.
Well, the original paper talks about social coupling but it's not really. That said, the later stuff Richard and Mike did was genuinely about social phenomena; that would be the literature I'd go to to make this point, not the original paper. But that's just a preference.
think they get it right: interpersonal perceptual coupling just is a variety of social coupling.
Really? I'd be inclined to say it's the other way round, given the biological motion perception literature. I'm surprised you go this way, actually.
For another, while I agree that a theory is a better way to have a guide to discovery than a set of flexible models, I think it is a mistake to say, as you seem to be saying, that good science can’t be done without explicit commitment a theory. Look, for example, at the Stephen et al work. As you correctly point out, this work is inspired by the dynamical models of phase transitions. That’s not a psychological theory.
Well ok, to a point. I think they took a problem and used tools from the dynamics bag to tackle it, and that did indeed take measures you wouldn't normally think to take, etc. That's fair; and to be honest, that's the kind of work the HKB model did early on too. In my more charitable moments I do remember that the HKB did actually kick start a whole field. My problem is still that it was really only a decent first swing at the problem, but it quickly became the entire way in which you talk about coordination. Geoff's approach is much more grounded in the actual mechanisms, and is I think a much more fruitful path. Even for Stephan et al, their work will only really pay off if they keep moving. If they spend all their time tweaking their model, they'll never get anywhere.
That said, I do agree that theory is a better guide to discovery than models. It is more flexible, and not only because theories are generally compatible with a wide array of models. This is why the next few chapters are in the book.
I'm on your side with this bit of the enterprise, although as a heads up, I'm not convinced by the specific 'shoring up' you advocate.
I was also a bit frustrated with this chapter, and not fully convinced by the discussion of the HKB model. However, I do not think the situation is a bad as Andrew describes. The two inclinations which are put at odds are this: 1) We want to favoring approaches that lead to new discovery, by which we usually mean that the approaches make clear predictions, which are then confirmed. (In my favored technical language, "the intended implications of the model are verified.") But... 2) many things lead to new discoveries, often quite exciting ones, which are not really predicted in the classic sense of the term, and we don't want these discoveries to count as much.
ReplyDeleteSo, the HKB model was there, and using it, in a Radical Embodied Cognitive Psychology kind of way, new stuff was done. Much of it is somewhat mundane, but still definitely counts as new discoveries (well, we know it works with index fingers, what about middle fingers? what about ring fingers? etc.). Other studies were sort of inspired by the model, e.g. the hypothesis that the model could fit the data, but not with any prediction intrinsic to the model. Often, this required adding new complications to the model, which is awkward, because more parameters fit data better (duh!), and because at some point it is just a different model.
All this is a very messy story. I don't think we get to count a more complicated model as a "discovery", so the question is whether attempts to apply the model led to things that do count as "discoveries." Certainly the discovery of similar phase shifting in interpersonal situations count, for example. Thus, I don't think it is terrible that this example was used: it is an example of a type of discovery-inducing activity using the approach championed in the book. However, ultimately, such discoveries due to experiment-inspiration will always be less satisfying than good, old-fashioned prediction from first principles.
I'm already rambling too much, time to stop.
I am an undrergraduate and as I work through this book, most of what I read has sort of been over my head and I was really scared when the equations came up in this chapeter. As we go on though, I'm beginning to understand much more than I initially did. Especially helpful in this chapter was definately the cases and examples in them. For example, in the 8th case the gear problem help me better understand what was trying to be said.
ReplyDeleteEric, I think my main problem is while that the HKB was indeed a short term shot in the arm for the field, that's all it was. It did get people asking questions they hadn't done previously, and measuring things they hadn't thought to look for previously; but it really quickly stopped being able to account for the data it was producing.
ReplyDeleteAt that point, the sensible thing to do would have been to jettison the HKB model itself but keep the basic idea (modelling perception action as a non linear dynamical system). This was the thing it had actually contributed: a foot in the door to dynamics. But instead, everyone deified the model itself, which has led to the field chasing it's tail for a long time. Only Bingham has learned the lesson and moved on, and it's been an uphill battle.
So right; my concern with using models like this as guides to discovery is to give too much weight to the model.
To Eric's students - if you're a little weirded out by the maths in the HKB model, you may want to play with this Excel file I made to plot the model out. If you change the A and B parameters, you change the frequency of the movement being modelled.
ReplyDeleteI should expand this to include the learning prediction sometime.
One thing to keep in mind here is that the guide to discovery problem would seem to indicate that it is impossible to have a progressive scientific research program based on modeling phenomena. The decade+ run of HKB, and the fact that, as Andrew put it, got “a foot in the door for dynamics” shows that a family of mathematical models can be a guide to discovery. That, really, is the whole point of the chapter.
ReplyDeleteAndrew is surely right that HKB has run its course. Even Kelso admits that. At the (profound) risk of downplaying the personal hardships that he endured, we might compare HKB to Jackie Robinson. Jackie Robinson was a great baseball player, but he wasn’t the best even during his own era. He wasn’t perfect, for example he was a prominent supporter of Richard Nixon. But he is really important and rightfully deified because he “got a foot in the door” for African-American baseball players, something which played an important role in the Civil Rights movement in the US.
Don’t get me wrong: HKB is no Jackie Robinson, not even close. But the strained comparison might make sense of the way many people (including me) hold HKB in such high regard. HKB was the first of its kind, and it showed you could do scientific psychology by applying nonlinear dynamical models to human action. Lots of good stuff wouldn’t have happened without it, including what Andrew is calling the Bingham model, which of course, Andrew had no small hand in.
OK, I look forward to further bashing in Chapter 6!
Fair comment. I'm inclined to think the model waaaay overstayed it's welcome, and it really got into people's heads as somehow something real (see: all the attractor dynamic stuff on learning that ended up making incorrect predictions). Watching people try to contort the data into the HKB approach has been a worry.
ReplyDeleteBut I must remember not to down play the fact that the foot in the door was real, and counts as real progress.
Andrew,
ReplyDeleteI don't know the dynamic systems world as well as the eco-psych world, but it seems to me that everything in the eco-psych world is continued past the point of usefulness.
For example, the stuff on detecting the length of wielded rods was great, but it is time to put it in a text book, leave it as foundation, and get on with life. If I see one more talk showing that you can detect the length of a rod by wielding it in a slightly different way, I am likely to scream (probably I'll have the good sense to walk outside first). You can detect it free wielding the rod, if the rod can only rotate in a plane, if the rod is attached to a glove, taped to your back, if it is under water, if it is in oil, if you are holding it generally upright or generally horizontal, etc., etc.
Don't get me wrong, the initial work on this was foundational, and great. It's just that its long past the time (in my opinion) that anything new is learned by continuing with it.
Eric
P.S. Contorting data to fit a model is a different sin.
Eric, I agree with you to a point. I've blocked pointless and badly run affordance studies at review only to see them show up unchanged in the Ecological Psychology journal, and I find that a worry.
ReplyDeleteThe dynamic touch literature is another example of getting stuck chasing their tail. I'm still waiting for someone to take the next step, namely try to identify the (kinematic) information specifying the (dynamic) inertia tensor. People clearly perceive it - but how?
I do agree with Tony's take on the dynamic touch literature, that it's an important part of the story because it's rigourous and not about vision. Those are no small things.
> Case 1: Modifying the HKB model with a noise term to fit the data
ReplyDeleteSchöner, Haken & Kelso (1986) added a noise term to the HKB data. It's not at all clear how this is supposed to count in favour of the HKB as a guide to discovery, given that noise terms are standard in models. All that happens is the model fits the data better;
Objection, m'lud. The noise does a hell of a lot more than improve data fitting. It is absolutely necessary to ensure symmetry breaking, to ensure that the system never gets stuck in an unstable equilibrium, and its effect is predicted (predicted, not observed) to get larger as an attractor becomes shallower (critical fluctuations). Noise within coordination dynamics is terrifically important.
This comment has been removed by the author.
ReplyDeleteFred,
ReplyDeleteI'm not as familiar with these models as I should be. From your comment I gather that the "noise" term in the HKB model is not equivalent to the "error" term slapped on the end of most any statistical model? The error term in the statistical models simply states "and sometimes our model is wrong", that is, it does nothing but help the modeler save face in light of an incomplete or incorrect model. In contrast, you seem to suggest that the "noise" variable in the HKB model indicates adaptive variation that aids the organism, with the model merely reflecting that.
Am I getting this right? If so, could you elaborate a bit?
The noise term in the HKB is simply parameterised white noise; perfectly sensible, but nothing specific to coordination. Yes, it was predicted to increase prior to a phase transition, but that's a standard feature of this type of dynamical system.
ReplyDeleteContrast this to the noise in Bingham's model; proportional to relative speed to create the differences at 0 and 180 (recently backed up by data). Very particular to coordination, and grounded explicitly in perceivable information.