Hi Sean, Thanks for taking the time to comment on my post. I am sympathetic to a lot of what you’re saying. I don’t think this will alleviate much of your concerns, but upon retrospect, I think I should have titled my post: How to not make illusions a big deal for direct perception. At the core of a lot of what I said was an attempt to motivate the claim that illusions need not be a big deal for theories of direct perception. Moreover, those disjunctivists of the direct perception persuasion are going about it wrong. With that said, let me respond to some of your other points. I mostly agree with your point that a well-developed theory of visual perception can give a physical account of illusions, but will not do much to account for the phenomenal experience of said illusion. I think we both agree that the physical account makes illusions not a “big deal” in the sense that we know why they are experienced as they are (e.g., light refraction at the air-water boundary in the case of the stick in water), and knowing why makes them less mysterious. What I want to get away from is appearance/reality distinctions in matters of perception. I think that’s where a lot of our disagreement lies. “Sure, we have a physical account of how the stick looks bent in water, which could explain why the sense data seems the way it does. But it’s still an illusion—i.e., a big deal—because the appearances (“seeming”) are not revealing of the reality.” I guess I want to skip around the point and respond in a deflationary kind of way: “Well, if perception is direct, then the appearance is the reality. As a source of visual perception, it doesn’t just seem to be bent, but the stick really is bent in water.” Moreover, if the stick were to serve as a source of other kinds of perception, then those need to be involved as well, which would change the nature of the perception. For example, if the stick will be engaged with visually and haptically, then grabbing the stick, manipulating it, and placing it in and out of water will reveal that the stick—as a source of combined visual and haptic perception—is straight. This is a very systems-based approach to perception, and is inspired by the ecological psychology approach to perception, which says that all perception requires taking into account that the organism-environment is a single system, where the features of the environment and capacities of the organism constrain what is perceived, and where perception and action are inextricably connected. In this way, perhaps we can save “seemings” if we allow that perceptual appearances are the reality—that is, until more perceptual information is available; and at that time the appearances will change, as will the reality.

Please forgive the reposting of my previous reply. The hyperlinks did not show up, so I reposted with the full URLs listed at the bottom. Rookie blogger mistake, I know. Dear MrSkimPole: Thanks for the comments. I am very sympathetic to what you’re saying, namely, that “complexity” refers to phenomena that we as observers have lost track of the cause-effect chain. I think this is probably true of many cases and is consistent with what many refer to as a form of “weak emergence” (1). Termite nest building is likely a case of complexity qua weak emergence in that as observers we do not have a clear perception of the causal chain—which is something like tracking the pheromone patterns of the termite “doo-doo” (technical term for feces) that eventually lead to mounds, which eventually leads to more complex tunnels, etc. In that case, I think more suitable terms that ‘self-organization’ is indeed recursion or feedback. Nonetheless, I think there are some cases of self-organization that are more ontological and are congruent with ideas like “strong emergence” (2), whereby the emerging properties are novel in relation to the properties of the constituents in isolation or as linearly related (e.g., additive causal powers). I think these kinds of “strongly” emergent properties that arise via self-organization can be epistemically novel (i.e., from the point of view of observers like us) and/or ontologically novel (i.e., not merely the sum of the parts). A classic example of such self-organization is the Rayleigh-Bénard convection (3). A way I’ve started to think about emergence in relation to complex systems is via the concepts “component dominance” and “interaction dominance.” In short, systems are component dominant when the properties of that system reduce to the properties of the parts and the linear and additive interactions among the parts. Systems are interaction dominant when properties resulting from the dynamics of the whole system override the properties of the parts. Some excellent papers that discuss component and interaction dominance include: Ihlen & Vereijken 2010 (4), Van Orden et al. 2003 (5), Van Orden et al. 2010 (6), and Wagenmakers et al. 2005 (7). In the end, MrSkimPole, I think we are mostly in agreement. Were I a gambler, I’d bet that much of our disagreements are merely verbal disputes. Thanks again for the comments. (1) http://www.iep.utm.edu/emergenc/#SSSH2ai2 (2) http://www.iep.utm.edu/emergenc/#SSSH2ai1 (3) http://www.scholarpedia.org/article/Rayleigh-B%C3%A9nard_convection (4) http://www.ncbi.nlm.nih.gov/pubmed/20677894 (5) http://web.haskins.yale.edu/Reprints/HL1319.pdf (6) http://www.tandfonline.com/doi/pdf/10.1080/10407410903493145 (7) http://www.ejwagenmakers.com/2005/pileofsand.pdf