Donald Hoffman, The Interface Theory of Perception

After reading The Interface Theory of Perception, I was wondering if you can clarify what the author is trying to point out, "Natural Selection optimizes fitness, not veridicality" (5). Also, would you pinpoint what the author is trying tosuggest through various researches done on other species that not necessarily H. Sapien? Would any further researches based on this idea be developed to have less complexity and be able to generalize directly to human perceptions?
This paper discussed many relevant researches in other species as well as H. Sapien, but I was not quite sure if I understand how the point that vision can get the evolution wrong by conflating fitness and accuracy. I learned a lot from this research, however, the conclusion drawn by Donald Hoffman was somewhat confusing because from my understanding, the perception of an organism cannot be directly compared with another species/organism. My argument was also mentioned in the paper to show that the examples are self-refuting in author's argument and discussion. –David Chang

I feel that the Interface Theory somewhat challenges everything that I have so far learned in psychology.
From my knowledge, I thought that Hoffman's mention of the conventional evolutionary argument (for the most part) held value.
I have learned that our sensorimotor abilities and brain help us to reconstruct the outside environment - ultimately creating an inner perspective/replication of the world.
However, if Hoffman's proposal is truth, does this mean that our perceived reality is completely different from objective reality?
How do we then know if our perception is even close to reality?
His argument does seem somewhat convincing, but I don't know if I can agree that "simple drives truth to extinction" and that perception solely serves as a guide (pg.22).
I think that if anything, factuality plays a factor in our survival and perception (contrasting from his main claim - pg.23)
Ultimately, I feel that Hoffman's theory is too extreme and that perhaps a fusion/merging of both theories would be better.
Even though parts of the interface theory seem promising and makes sense in the direction that natural selection rewards fitness, it is hard to believe that it completely dismisses accuracy...how else would we know that what we know is what we know?
Although perception may not paint a wholly veridical picture of the world, I would like to think that truthful representation plays at least a connecting role.-Rebecca Chung

What makes species-specific guided perception more feasible than truth guided perception; it seems evolution would equally drive both. Donald Hoffman ignores, perhaps, the possibility that perception seeking the truth is what guides a species in its niche, since, it would seem, that veridical perception would be most advantageous, most fit (Pg. 1).ERNESTUS DELA PENA

Great article, although not concise— it is interesting and even funny; as the author argues his point, dissolving the conventional beliefs about perception, he makes use of a witty double entendre (page 6). One point, on page 10, about super normal stimuli in the case of the jewel beetles and the 'stubbies' got me thinking about humans and their attraction to parallel super normal stimuli, the sort of which plastic surgeons create. It was interesting to note that, in my conversations with men on the aforementioned topic, they state that although they recognize the stimulus/i as super normal, they cannot, so they say, help but feel an uncanny attraction toward the said stimuli. Perhaps the beetles and the gulls can be cognizant of their error, yet, propelled by something (but what?), be forced to copulate with rubbish and peck at painted sticks. Nicole de Faymoreau

In p.10 of the "interface" reading, Hoffman discusses the idea of "supernormal stimulus" that refutes the reconstruction thesis because of these stimuli's tendency to deceive an organism's perceptual categories. Hoffman then gives the example of the jewel beetles falling for the stubbies because the stubbie serves as a supernormal stimulus. What are some examples of supernormal stimuli in humans? Neil Ponce

pg. 13 of text: "When a file icon is dragged to the trash and disappears from the screen, is the file itself destroyed, or is it still intact and just inaccessible to the user interface?"This quote begged me to think about things that our interface has yet to encounter. What of things that we have no conscious knowledge of yet? How does the interface come to deal with these new things in the first place? Do they exist if they don't yet exist to our interface, or no longer exist to it?~Amanda C.

Can you please go over the Principle of Faithful Depiction? (Section 1.2, page 3-4) I don't really understand it. Also, the reading gives a lot of support for this principle, but none against it. Are there any arguments against it? If so, what are they? Jewel Adams

A thought: our perceptions are within our observable window. Is the human observable window larger than those of other animals? As humans, we create, manipulate and understand the world in a way unmatched by any other animal. We understand there are light waves and sound waves beyond our reach but we know they exist. I assume other animals do not wonder on the things they cannot perceive. Animals perceive fruit, mates, predators, prey, shelter, things essential to their survival. But humans perceive things essential to their survival and much much more. Can this then be an argument for the human ability to perceive the world more accurately? We study illusions and errors in our perception but can our perception still be that far off from the actual physical world? Monette Leyva

1. I have trouble buying into the idea that our visual perception is merely a representation or, as Hoffman puts it in his analogy, an icon of the real object. On page 12, he argues that he "doesnt take [his] tiger icon literally but [he does] take it seriously" lest there be consequences. I don't understand how we don't take it "literally". In our every day life, what is the difference between taking an object literally, and seriously. 2. I have difficulty understanding why an interface is necessary for all objects. Again in his analogy, he says that many people would prefer not to deal with the complicated computer, and rather, simply just have an icon represent what an object is or can do. My question is, does this imply that every single object in our life serves a large purpose, and is so complicated that we must have an icon represent it? Syliva Wong

1. There seems to be a great deal of ambiguity withing Hoffman's writing due to the unnecessary jargon. It is perhaps unwise to examine an undefined system (mind) with a defined system (software).Reasons for that are that it is not obvious to what extent the two can be compared and surely manes of Alan Turing and Kurt Godel and are bound to take front/center stage in such debates. Since my knowledge of the The Church-Turing Thesis and the both Godel's Theories of Incompleteness is quite elementary, I can only speculate.2. I also cannot bring myself to refer to this Theory of User Interface as a scientific theory, since it seen to be much more of an analogy. Now analogies can be great if proposed say by Douglas Hofstadter (e.g. the mind as an infinite loop of feedback mechanism).3. Personally I much rather hear of the faithful depiction in terms of the Bayesian Ideal Observer Theory. It seems to be a perfectly sound idea that the mind itself works as the ideal observer doing it's best to optimize a performance in a task given the physical properties of the task and some biological constraints.

Some crucial questions that arose from this reading:

Can our mind be quantified? If not all then to what degree?Is it perhaps a quantum phenomena? and if so then at what point does the transition from classical occur? What falls under conscious and unconscious? and if mind is quantum, then what interpertation of it is to be used? Can those aspects that are conscious be tied in to the conscious effect on the wavefunction collapse? Do we follow formal logic all the time of just most of the time? And why unlike a computation machine we are so easily unfazed by paradoxes and contradictions?

And one that is most interesting to me is the origin of mathematical logic in relationship to the perception and other processes of the mind. Sometimes one explains the other and other times they are complementary. Alissa Egorova

Because humans can more or less navigate everything in our world, perhaps our evolutionary solution is the reconstruct the world as it is. Because of society and medicine, the idea of fitness is somewhat unimportant, so maybe we don't require a particular interface that creates restrictions on our vision. And because of this, perhaps this is why humans have advanced so much further than any other animal in history. Chris Silver

Don’s questions relating to Hoffman and first class discussion (these could turn into exam questions):

  1. Is it valid to consider perception as the construction of an internal model of the physical world as it relates to the observer? Is Hoffman justified in rejecting the idea of a straightforward correspondence or ‘isomorphism’ between the model and the modeled physical reality?
  2. Show how the phenomena of phantom limbs and/or phosphenes can support the explanation of perception as a causal chain.

Topic for Tuesday: Vision and Optics. On-Line material:

  1. Brian Wandell on Retina and Brain:
  1. Austin Roorda, Optics of the Eye:

Austin Roorda, Optics_of_the_Eye.ppt

3.My lecture slides:

Why isn't vision perfect?htm

Why isn't vision perfect?Ppt

Contributed questions:

If there are nine known visual field maps, is there a central location where all the information is summed, forming the image we "see"? –Amanda C.

I didn't quite understand the slides from the Wandell lecture, mainly because I felt like I didn't have much previous knowledge of the topic before. Can you elaborate more on the concepts of color aberration and the CIE-XYZ functions, and also explain how to read the graphs presented with these concepts? –Neil Ponce

I love the graphs and the images that give us amicroscopic view of the pigments and neurons.
On Slide 21 and 22, there are an abundant numbers of red cones.Does that mean that normal human eyes have a higher ratio of the red cones? -David Chen-Tao Chang

Is disparity of different cone distributions between individuals the cause of vision related qualia (ie. the redness of red)? =ERNESTUS DELA PENA

- I find it odd and interesting how the pupil still manages to do it's job, changing the depth of focus of the eye, despite having many outside factors affecting it's size.

- How do we cope with so much chromatic aberration? Is it due to parts of the eye working harder? Or to do with us not actually being needing or being able to cope with that much information. i.e. even if there was no chromatic aberration, our sight would not be perfect anyway. –Sophie

Foundations of Color Vision (Wandall):
What exactly is a CIE XYZ function and what is its significance in regards to color matching (29-31)?
Also, would you explain the graph concerning pattern color sensitivity (36) and what exactly Y, I, Q stand for in the following slides (37-38)?
Optics and Optical Quality of the Human Eye (Roorda):
What is the purpose behind the Shack-Hartmann Wavefront Sensor (99-109)? Is it a measuring instrument for wavefront aberrations?
I was quite overwhelmed by the amount of information covered in this slide set; are we responsible for all the material?
Comment:
I thought it would be relevant and especially interesting to mention that the application of established knowledge of the human eye has recently been taken to a new level. Bionic Vision Australia (BVA) has publicized that they have created a prototype of a bionic eye. It will be exciting to wait and see if testing and further research will reap an actual bionic eye that may completely reshape the lives of those suffering from visual loss. ( Chung

- What is it exactly that causes optical abberations in the eye, and in particular what is the difference in the cause of chromatic and achromatic abberations? Can you have one without the other?

- Similarly what causes a wave abberation? Does it occur in everyone to different extents or is it an abnormality within the eye? Is it corrected by laser-eye surgery, and therefore the reason for long/short-sightedness?

- In the retinal sampling image (slide 132 of 161 in "Optics of the Eye") how can the seemingly random variations in gradient between the cells be accounted for? Is it truly random? -Izzie Sharman

Don’s questions relating to Vision and Optics (again, these could turn into exam questions):

  1. How closely does human visual resolution approach the physical limits set by the nature of the light distribution on the retina?
  2. How closely does the quality of the retinal image approach the limits set by the nature of light?
  3. What are the factors that make the retinal image imperfect?

Tuesday April 13: Visual Sensitivity/How photoreceptors respond to light

How does light suppress the circulating current in the eye and what does that have to do with phototransduction, synaptic transmission, and neural coding? I'm not sure how in depth we need to understand but i would appreciate it if you could go over some of the important graphs and charts about reproductability and rod sensitivity. ~Ashley Colantuono

Why can you see a star better when you're not looking directly at it? -Kelly Svoboda

In order to perceive light we need at around 9 photons, however I've seen this number range from 5 to 10 photons depending on the source of material.

Does this number change due to varying ratio of photons actually reaching the retina for different tested subjects? Presumably this isn't such an important question. But it did stir a question of the efficiency of the eye Since the retinal blood vessels are located in front of the photoreceptor layer, they must account for some energy loss. How much do the vessels supplying the retina play a role in normal perceived vision?And why can't we see them? Alissa Egorova

Comment:
I find it extremely interesting that although the actual human eye cannot consciously see just a single photon, the rods and sensors in the retina can respond to a single photon. However, if we all could actually see single photons what exactly would happen? What would we experience? Would there be an overwhelming amount of visual noise, especially in low-lit settings?
Question:
Would you be able to elaborate on the graphs on Amplification By The Phototransduction Cascade (22) and Candidate Mechanisms For Reproducibility (37)?-Rebecca Chung

How exactly do convergence and sparse signaling work? Does each indepdendent visual image from the small fraction of rods come together to form an image which ultimately forms our visual perception? –Niel Ponce

Don’s questions relating to Visual Sensitivity:

  1. How closely does human visual sensitivity approach the physical limits set by the nature of the light distribution on the retina?
  2. Why can’t we detect and individually perceive every photon of light that is absorbed in the rod photoreceptors?

Thursday April 15: A newly discovered visual system

In the 'Irradiance detection' section, it states that "detailed studies in rd/rd cl mice have shown that although there is diminished sensitivity, a full PLR can still occur in mice lacking rods and cones". I was just wondering if there were studies that were unable to show such results and/or if a full PLR occurred majority of the time? Also, has there been further study concerning the PLR being a two-stage process?Even though "the results from rd/rd cl mice demonstrate that rods and cones are not required for photoentrainment", it mentions that they still play a role. What kind of role do they have in photoentrainment (besides the findings of partial compensation for the loss of pRGC photosensitivity)?. The article states that "the specific roles played by rods, cones and pRGCs in irradiance detection remain unresolved"; is there any new research concerning these roles/interactions? I thought it was interesting when the article noted that "melanopsins may subserve several different roles". But, following it stated that certain "results suggest that the apparent diversity has been considerably overestimated." Does this mean that it is no longer a possibility? If melanopsins can still potentially serve more than one role, are there any speculations as to what they may be? -Rebecca Chung

Specific reasons to have rod and cone input to the melanopsin ganglion cells not yet clear.

Pupil (‘Steady state’), discomfort (photophobia), diurnal rhythm pacing, mood/season?

How is it that this new retinal pathway is only being discovered now? could this have new implications on how we regard light in relation to sleep? –Amanda