A blog about problems in the field of psychology and attempts to fix them.

Tuesday, August 30, 2011

A Brief Introduction to Ecological Psychology

The deep origins of Ecological Psychology lie in the philosophies of Pragmatism, Radical Empiricism, and New Realism. But that is a much longer story...

The first key paper of the modern science is probably a paper on perceptual learning (Gibson and Gibson, 1955), in which it was proposed that perceptual learning involved better discriminating stimuli. That is, this type of learning does not involve gaining more sophisticated mental processes, but rather more sophisticated sensitivity to the details of the world. Discussion generated by this paper, and further related works, were guided by a search for the 'discriminated thing' needed to fill in the perceptual-learning theory. The most obvious candidate would be something like the stimulation created by the retinal image... but the problem with the retinal image were already well known: The retinal image is not specific to the properties of the world, and therefore it cannot provide a firm basis for accurate perception. Gibson's prior work on optic flow was working towards a solution, but something more radical was needed.

This culminated in the second key paper (Gibson, 1961), in which it was pointed out that we could study the ecology of ambient energy, i.e., the structure of the light waves, sound waves, etc. that are bouncing all around you in whatever room you are reading this in. The visual side of this work would be called Ecological Optics, and from that we get the 'ecological' in 'Ecological Psychology'. The structured energy around us is specific to the properties of the world, and so if we can discriminate this ambient energy in the right ways, we come to know the properties of the world. Or, if we want to back off just a touch: At least some of the structure in the ambient energy is specific to at least some of the aspects of the world we would like to know about, and so, in those moments when we managed to correctly discriminate the correct pattern-structures, we are accurately and unambiguously discriminating something about the objects and events around us.

A few years later this idea was put into the a book-length discussion of the evolution of perceptual systems, which include both the sensory organs and all the other parts of the organism involved in moving so as discriminate these patterns (Gibson, 1966). The evolutionary context in which human abilities are thereby put, and the developmental context that started this whole line of thinking, led to some speculation in the final chapters about what aspects of the world we would expect organisms to perceive: Evolution and development both occur in the context of behavior, and so we might expect organisms to evolve and develop so that they discriminate those aspects of the world relevant to behavior. That is, we would expect animals (including people) to be best at discriminating the world in terms of what it affords them, e.g., in terms of what future actions the current situation makes possible, e.g., in terms what outcomes can be realized by the actions the animal is capable of performing. To make this concept more concrete, Gibson invented the term 'affordance.'

Still later (Gibson, 1979), the more radical consequences of this new understanding of perception were explored at length. Gibson had been a very well reputed perceptual researcher in the '40's and '50's, before any of this ecological thinking began, but by the time of his last book there were a growing number of people interested in pursuing the more aggressive research agendas he had laid out. The most obvious research tasks were: 1) Identify higher-order invariants that specify properties of the world, 2) Show that organisms (these days, typically 'people') are sensitive to those invariants, 3) Put those together into demonstrations that behavioral success is due to proper sensitivity to invariants. Oh, and 4) don't shy away from opportunities to do this in the context of tasks not traditionally thought of as 'perceptual.'  The most important early proponents of Gibson's thinking were Michael Turvey, Bob Shaw, and Bill Mace, who played pivotal roles in founding the International Society for Ecological Psychology, the journal Ecological Psychology, and the Center for the Ecological Study of Perception and Action (CESPA) at the University of Connecticut. David Lee made the first solid demonstration of an optical invariant used to guide a critical behavior - the rate of acceleration of optic expansion (symbolized 'tau'), which can be used to regulate timing of impact with a surface. Initial studies looked at birds retracting their wings when diving into water at break-neck speeds, and modern studies span things as diverse as the control of force in punching and better understanding how people control their cars while breaking. Clair Michaels and Claudia Carello put out the first book that could be considered a textbook for the field "Direct Perception." It is now outdated not so much by advances in theory, but by lack of including the three decades worth of research success since its publication.

Today, Ecological Psychology has a small, but thriving research community using this approach to study the full range of psychological phenomenon.


Some notes (perhaps especially relevant to those interested in the Eco-psych textbook project):
  1. Many presentations of Ecological Psychology start with 'affordances', but I think that is a mistake. That organisms typically (or perhaps always) see the world in terms of opportunities for action, falls naturally out of an evolution and development focused approach to perception.
  2. There is a less extreme form of Ecological Psychology worth advocating, the one that uses the words like 'most' or 'typically' instead of 'always'. Many of the problems with Eco-psych being accepted in more mainstream contexts have to do with that pesky 'always'. At the least, get people onboard with the 'sometimes' first, and then the 'typically'. We can suggest the 'always' later... especially if your target audience is not into deep philosophy.
  3. The presentation above spent very little time (only a sentence or two) explaining alternative approaches to Eco-psych. This is similar to the format I hope the textbook can take. Lead with the ecological way of thinking, and avoid, to the extent possible, teaching students the way you do not want them to think about perception.
  4. The above presentation goes in historic order, which I'm not convinced is the best way to introduce the concepts to undergraduates. I have used the above order successfully with grad students and colleagues, but with the undergrads especially it seems something better could be done. Apparently past attempts at an Eco-psych textbook (Stoffregen tells me) were derailed before they even began by disagreements about what order things should appear in. If you're reading this, and wishing you could think of something to comment on.... What do you think is the best place to start? 


  1. Perspectives on your questions 1-4 (not in order) from a "student" recently introduced to eco-psych:

    Having had some exposure to eco-psych, I appreciated the background in the post, But in general, I don't like historical surveys as intros to new topics. Usually a general idea of the problem(s) to be addressed and the solution methodology to be presented is enough to cope with at first.

    Perhaps most important is careful definition of terms likely to be unfamiliar to students and even more careful redefinition of terms possibly familiar from other fields but defined differently. I'm hardly representative of an undergrad (sigh), but FWIW I had a very hard time with Gibson's vocabulary, as evidenced in the comments here:


    Whether to address the conventional view seems dependent on the assumed background(s) of the students. Sabrina's summary of Gibson's book was my intro to any serious consideration of visual perception, and I often found the material contrasting the eco-psych and "conventional" approaches a distraction.

    The top-level view of Gibson's model that I ended up with is a dynamic system comprising a perceiver in an environment in which the former receives information-bearing light from the latter. From the received light, the perceiver continuously extracts information about features of the environment and produces continuous action which is purposeful with respect to those features. If that description (or a modestly corrected and/or expanded version) is reasonably accurate, some of the more problematic Chapter 4 material (per Sabrina's summary, my only source) seems unnecessary. The details of the information and the processing thereof may be complex, but the system seems architecturally quite straightforward.

    Finally, I would second your position that affordances shouldn't be the kick-off topic. Even after months of intermittent discussion and reading about them, I remain confused - and suspect that would be a common reaction among even typical students. It's probably OK (necessary?) to introduce the general idea of perceived features of the environment that invite responsive action, but the technical aspects of "affordances" seem to cause problems probably better deferred as long as possible.

    All IMHO, of course.

  2. Charles, on affordances: in fairness, you came in the middle of a fairly technical argument between us and Ken. The actual basic concept is fairly straight-forward (and even more so if you start out talking about perception of action relevant properties, so I'm on board with Eric's 'no contrasting views' approach).

    I actually quite like the historical grounding, in part because the development of the theory tracks nicely as Gibson tackled specific issues and problems. He came to his conclusions for reasons, and learning about some of that helps, I think.

  3. "Whether to address the conventional view seems dependent on the assumed background(s) of the students."

    Yes! And also your assumed goals. If I were teaching a freshman great-books seminar, or something like that, where the primary goal was critical thinking carte blanche, then clearly I would need to introduce multiple views. That would be a fun class. However, if my goal is to teach students to think about perception and action like an ecological psychologist, then I want them to get a handle on the ecological framework as painlessly as possible, and then get them to think critically about specific perceptual problems.

    "a general idea of the problem(s) to be addressed and the solution methodology to be presented is enough to cope with at first."

    Agreed! How's this?
    Q: What is the problem we are trying to solve with talk of perception?
    A: How is it that organisms behave accurately with respect to the objects and events of the world?

    Q: How do we solve that problem?
    A: Perceptual systems evolve and develop to discriminate patterns of energy that can reliably guide action. (The energy patterns are a part of the ecology to which the organism is adapted.)

  4. Andrew -

    I'm not sure we disagree on affordances. I second introducing the "basic concept [which] is fairly straight-forward (and even more so if you start out talking about perception of action relevant properties". It's the details of that "fairly technical argument between us and Ken" that would concern me - eg, affordances as persistent, requiring immediate actualization, including information about structural properties. The kind of issues that seemed unresolved even after the BofC argument had endured for months.

    I was assuming that the "issues and problems" Gibson tackled while developing the theory arise because of flaws in the traditional approach, the highlighting of which would seem to clash with Eric's "no contrasting views" approach. If the background is just a framework around which to build a presentation of the "general idea of the problem(s) to be addressed", that seems reasonable.

    Eric -

    As I warned, I can nitpick any expression to death.

    In your first "A" I'd replace "accurately" with something like "effectively" or "purposefully". My problem is that talk of "accuracy" - even of behavior, especially since I'm not sure what it means to "behave accurately" - suggests faithful representation, which reminds me of the "mirror of nature". (Not that a typical undergraduate would have that reaction chain, just that the word is grating to me and seems easily avoided.)

  5. This will be a touch rambling, because I am struggling to find an ideal presentation.


    At this point I should not be surprised when people balk at the word 'accurately'... but I still think it is the correct word, i.e., the most straightforward word to describe what Gibson is up to / what the mystery is, and thus I am loath to use another.

    Any issue with 'accurately' must also be an issue with 'efficiently' or 'effectively', and 'purposefully' is a different issue entirely (c.f. Shaw's intentional dynamics).

    The question of perception is the question of how we are accurate, i.e., not wrong most of the time, surprisingly good, etc. For example, today I have locomoted several times without tripping and falling, I drove from my home to several locations without collision, I cooked dinner over a hot stove without burning myself or the food. I have grabbed, lifted, and manipulated a wide variety of objects without crushing, throwing, or destroying any of them. Sometimes I am not accurate, sometimes there are failures of perception, and I fall, or scrape the side of my car, burn my hand cooking, break things, etc. The fundamental question of perception is not a generic question about any behavior, but more specifically about accurate behavior.


    A detective named Monk walks into the room. Others have been there searching for hours. He sees that the lamp is standing straight, and finds a note tucked underneath that provides a crucial clue. "Wow," say the other detectives, "how did he see that?"

    That, is a question about perception. It is a question about seeing. The answer can be given entirely in terms of Monk being attuned to aspects of the world that others are not attuned to.

    Monk was accurate. None of the other detectives would be standing their in wonder if he was not accurate. They want to know how Monk is accurate in ways that they are not, they want to know how he perceives things that they do not -- the two 'wants' are equivalent.

  6. Uhg...

    A detective named Monk walks into the room. Others have been there searching for hours. He sees that the lamp is NOT standing straight....

  7. Eric -

    I appreciate your dilemma. My only other suggestion is the phrase often used wrt the obstacle course of life that you describe:

    successfully negotiating - or navigating in - the environment

    (I assume this general idea was behind Rorty's distinction "coping, not copying".)