I’ll Believe It When I See It
The lawyer leans over the witness stand, piercing holes of disbelief into the witness. He matter-of-factly paces towards her. She is one of the few witnesses at the crime scene. Her palms are sweaty and she’s fidgety from all the coffee from earlier. “Tell us what you saw on the night of Thursday, August 11th, 2011!” Questions flood her mind: How can I be asked to report on experiences I may or may not have had? I was present, but does being present guarantee an accurate experiential report? Maybe my opinion is tainted by others’ opinions? HELP!
It is here that she is asked to report her experiences to the jury, the judge and all those in the courtroom. It is here that an accurate report of one’s perceptual experience counts! One can imagine the tremendous weight bearing on any witness asked to testify under oath.
If questions concerning truth, perception and judgment flooded your mind after reading the above-mentioned, congratulations you are among the few who have delicately scratched the surface of an ongoing philosophical problem: the problem of perception.
There are many problems in philosophy that involve perceptual experience. The problem I’m referring to specifically involves the question of whether perceptual experience offers accurate information about the physical world.
If we’re lucky, few of us will find ourselves in a position that demands our reporting on perceptual experience. Why? Because there is an increasingly resounding sentiment in contemporary literature on perception that suggests our experiences are far from perfect.
Take for instance Daniel Simons and Christopher Chabris’s ‘dancing gorilla’ experiment*. In this experiment participants were asked to watch a video of people wearing black and white t-shirts pass a ball to one another. After the initial viewing of the video, most of the participants did not report anything strange or out of the ordinary. They were surprised to find that a person dressed in a gorilla costume was dancing during the passing of the ball. This experiment suggests that it is possible, in perceptual experience, to a give a situation your full attention, and still fail to perceive objects present in the visual field. This phenomenon is known as innatentional blindness. (Simons & Chabris, 1999)
This insight is disconcerting, to say the least. Why are we so convinced that our perceptual faculties contain lively, rich, and vivid detail about the external world? This notion stems from an old-fashioned view of perception held by Aristotle: that particles travel towards the visual system that collide with emissions from the eye. This is obviously problematic because as philosopher Howard Robinson notes, “…there is nothing to explain why such a collision of particles should constitute an experience.” (2004, Robinson) A better way to conceive of an Aristotilean account would be to draw on the idea of representation. Particles leave the object in question creating in the subject a representation. Think of a catcher’s mitt, where the mitt represents the visual system and the ball, information. The eye passively welcomes information from the external world just as a catcher’s mitt welcomes the baseball. One cannot possibly catch the ball unless it is thrown, and, in Aristotle’s case, one cannot perceive unless one obtains information from the world. It is a unilateral system.
You might be asking yourself whether the ‘dancing gorilla experiment’ threatens Aristotle-like programs. It does so in the following way: If perception plays a passive role, then how do we account for unsuccessful visual experiences? Say, for instance, I see a dog, where there ought to be a blue ball (i.e. illusion). If we subscribe to Aristotle’s view, then the blue ball is creating (in me!) a representation of a dog. See the problem? How can object x (the ball) represent itself as y (the dog) in experience? Lets take things a step further – if I perceive a dog, where there is in fact a dog, this is considered a successful perceptual experience by many a philosopher. But how accurate are our percepts? If my percept of a dog is not accurate (I perceive the dog as red vs. brown) then it would be fair to say that object x (brown dog) represents itself as y (red dog) in experience, where x and y are distinguishably different.
Some theories of successful perception assume a necessary connection between what is perceived and the thing perceived, but by what criterion do we determine the adequacy of our percepts? If what is represented, in this case a dog-like representation, only requires some correspondence between the thing perceived and the thing represented, then how accurate can our experiences really be? The connection may be a necessary one, but certainly not an accurate one. We have inevitably returned to the point from which we departed: the problem of perception.
According to the ‘dancing gorilla experiment’, even when our cognitive systems are at their best (whatever that means) we fail to perceive objects strewn in the visual field. This must be terrifying for our witness. Not only will she have to report on very specific experiences by drawing on her memory, which is already problematic (given the possibility of forgetting and distorting information), but the experiences themselves, as noted in the ‘dancing gorilla’ experiment, are prone to err. Perhaps one can claim fallibility, leave the witness stand with a sense of relief knowing full well that our perceptual systems are frail. As reassuring as this sounds, this doesn’t change the fact of the matter, and that lack of evidence, in some cases, carries with it the same force as strong evidence, and, consequently, its effects. Perhaps our witness will leave with a little more than just a sense of relief – a desire to find a way to fasten her experiences by understanding how she is situated in the world.
—————–
*Now that you’ve read the description above it may not work, but here is a great example of the dancing gorrila experiment!
21 Responses
10:45 pm
You should read Millikan’s account of Proper Function and many of the questions you are asking (i.e. how to account for unsuccessfull perceptual experience/criterion to determine the adequacy of percepts) would be less problematic, or at the very least you will have a very strong counterposition to defend against. Calling out the faillures of the perceptual system doesn’t say much against external accounts of stimuli it only shows that it isn’t perfect, just like anything else.
12:06 am
Ramon, I don’t think Polo was attempting to settle some debate between externalists and internalists though*. Rather, the sense I got was that it was simply an exploration of some quirks of perception that should force us to think a bit more deeply about our naive conceptions of the accuracy of what we see. What we experience is not what hits our retina. And as the gorilla experiment points out, it’s not just about perceptual error as the result of features of our perceptual system, but also about the way attention and other cognitive processes influence what we “see”.
*not that your recommendation to read Millikan isn’t a good one.
12:21 am
Ramon, thanks for the feedback! I’ve made corrections and clarified my position on Aristotle (which is what I think you might be referring to, right?). Anyway, I hope I’ve made it clear that while I am not providing an argument for the dismissal or affirmation of a direct correspondence theory, I am certainly applying pressure to concept of representation (broadly conceived). Thanks again!
12:28 pm
I wrote my comment in a rush, sorry about that. But the premise remains the same. I think that certain wording of mistrust towards the senses can be misleading. There is something about our ability to recognize the “flaws” of our perception that tells us that perception is trustworthy. The fact that they recur seems to be evidence that they function in a somewhat reliable manner. These “flaws” are merely technicalities of the kind of organism that we are and although awareness of such is tantamount to understanding the nature of our relationship with stimuli there is no reason to be pessimistic towards our overall ability to discern. Now, the reason I mention Millikan is because of her very eloquent account of what it is for something to function properly and what it means for something to be a normal condition for the proper functioning of that something which tells us that “flaws” such as the ones you mention (like “unsuccessful visual experiences”) are pretty much not a big deal considering that everything else in nature seems to err on the side of caution, i.e. it is better to allow for flaws and get something wrong than trying to only get something right and missing out on correct variations.
You are welcome Polo, looking forward to discussing this in person with you guys.
2:38 pm
For the most part, I don’t disagree with anything you’ve said thus far. If I were constructing an argument in favor of the classical notion of direct/indirect perception in philosophy, then drawing on Milikan would make sense. But I’m not making an argument for anything. Milikan does, in fact, maintain that we should rethink how we conceive of representation (which I think is cool), but I’m not siding with her or against her in this post. The sentiment I wanted to convey is that our perceptual mechanisms are fallible, which is something Milikan herself would agree with:
“Moreover, on every level, inner intentional representations are fallible. Nothing is direct in the sense of being epistemically “given” to mind. Interpretation of signs is always fallible, chancy, in very principle.” (Milikan, ‘Direct Perception Through Language’)
I do take issue with the overall sentiment in your last two posts, however. To say that fallibilism is a design feature and therefore entails the dismissal (or “possible” dismissal) of criterion for determining accuracy in perception is a loaded claim.
2:59 pm
On the contrary Polo, see:
“There is something about our ability to recognize the “flaws” of our perception that tells us that perception is trustworthy. The fact that they recur seems to be evidence that they function in a somewhat reliable manner. These “flaws” are merely technicalities of the kind of organism that we are and although awareness of such is tantamount to understanding the nature of our relationship with stimuli there is no reason to be pessimistic towards our overall ability to discern. ”
It is the fact that we can recognize these “flaws”, and that they recur in the same manner over and over (optical illusions) that allows us to say that there is a reliability of some kind.
3:51 pm
Ramon, I think that Polo and I could both agree with the majority of what you’ve said in the last few comments. Our perceptual systems our fallible, but because we are a certain type of biological creature, there is consistency across experiences and across individuals of that fallibility. It is reliably fallible. So if you take a look at the perceptual illusions in this post: http://cognitivephilosophy.net/consciousness/consciousness-is-weird/, yes, all humans tend to perceive the same “wrong” thing. Okay, duly noted. Or maybe I’m misunderstanding you. You’re uncomfortable with using the word ‘fallible’ or ‘flaw’ to describe our perception. You want to say that what Polo and I are referring to as perceptual error is really just a technicality about the nature of the relationship between stimuli and organisms? I’m not sure what that means. Take for example the second illusion in the link I posted, where you perceive two squares as being distinct in color, when they are in fact the same. And if you cover up the surrounding space and view just the two squares on their own, you perceive them as the same color. Are you saying this isn’t a case of perception being fallible? We can use experimentation and reasoning to discover the truth about the color of the two squares, yes, but I’m not sure how that’s relevant to the point Polo is making.
Think back to the inattention blindness discussion. What does this tell us? This tells us that certain cognitive processes (paying attention in this case) influence what we “see”. This isn’t about our perceptual system being in error or not, but rather, is something that points to the idea that representation, or experience as being constructed, is something that needs to be taken seriously. Our experience of seeing the event taking place in front of us does not include the dancing gorilla that is right in front of our visual field, and is obviously being presented to the visual system. Further, if attention is directed differently, the gorilla isseen.
You seem to have a notion of perceptual experience as follows. Object in the world -> sensory signal in the world -> transduction of sensory signal inside the brain -> perceptual experience -> cognitive process of attention -> experience of seeing. The point Polo is making in the post is that we don’t perceive the gorilla, and then subtract it out of our experience somehow…what we see simply does not include the gorilla, an extremely large object in our visual field. This leads to questions abut the nature of perception and what relationship it has with representation. Perceptual illusions are often discussed in these conversations because they are obvious examples of a lack of a direct correspondence between the object and our perception of it, but the relevance that Polo and I find in these discussions is not necessarily about fallibilism and reliabilism, but about how experience is constructed. If I perceive a gray square as white, yes, there is something about my perceptual system that causes this (actually, I think what causes this are cognitive processes far removed from the visual cortex, but nevermind), but more importantly, what is it about certain types of processes in the universe that allow consciousness and subjective experience to emerge. The difference between objects as they are and our perception of them offers clues to this constructed nature of experience.
3:29 pm
So the ability to recognize the fallibilistic character of perception entails a mechanism for tracking unsuccessful perceptual experiences? I’m afraid this begs the question: Is this mechanism fallible as well? If it is, then we fall into an infinite regress. If it isn’t then it must be perfect, which defeats the purpose of fallibilism, period.
8:26 pm
“Take for example the second illusion in the link I posted, where you perceive two squares as being distinct in color, when they are in fact the same. And if you cover up the surrounding space and view just the two squares on their own, you perceive them as the same color. Are you saying this isn’t a case of perception being fallible?”
My point is a lot simpler Greg: whenever you have the two squares on their own they look the same; whenever you put those two squares amongst others they don’t. You still had to put those two squares alone or amongst others and then you saw them this or that way. That is just the way that the squares look in those circumstances. Take the squares away and you have no squares. Put some squares and you’ll have squares that look this way or the other depending on how you arrange them. Very simplistic point if you can see.
Put a human in a hectic situation of stimuli overload and he/she will filter something out, but not just something, right? That something is quite detectable by other methods, we always miss the gorilla. Doesn’t that tell you that when your perception “fails” it always “fails” in the same way? Isn’t that uniformity of experience evidence of reliability? I would not say that my perception “fails” every time I fail to see the subatomic structure of the world. Nor would I say that I “make up” the world. If what you guys are saying is that our perception is limited and sometimes inaccurate then I think the Greeks may have beat you at it. If what you are saying is that our senses aren’t reliable then you’re also late to the party…what I am saying is that our mere survival depends on millions of years of accurate receptors accumulated in a fallible organism. Again, my point is very simple and yes, maybe Aristotelian, I just don’t think my mind creates the tables where I rest my glass of wine and I don’t think our perception is that alien to the recognition of properties in the world.
8:52 pm
On the construction part…”You seem to have a notion of perceptual experience as follows. Object in the world -> sensory signal in the world -> transduction of sensory signal inside the brain -> perceptual experience -> cognitive process of attention -> experience of seeing. ”
Yeah, I know what you mean, but I agree with you in saying that we have something to do with the way in which we perceive things, we are the kind of monkey that perceives this in that way and that in that other way, right? Our cognitive processes are engaged in the world, and I am pretty sure that they are so in a different way than those of a bee. I don’t deny that, I just don’t believe that such engagement distorts the stimuli to such an extent as to make it a complete construction. And as to the caricature of the way I see perceptual experience remember that in such simple terms I would add simultaneity to the mix and scream baam! 🙂
10:01 pm
“Put a human in a hectic situation of stimuli overload and he/she will filter something out, but not just something, right? That something is quite detectable by other methods, we always miss the gorilla (not necessarily). Doesn’t that tell you that when your perception “fails” it always “fails” in the same way? (no)”
I’m afraid I’m having trouble understanding what your claim amounts to exactly. All I’m getting is that unsuccessful perceptual experiences contain the same structure and that we can recognize illusions some of the time. What is this an argument for exactly? Philosophers like Daniel Dennett have argued against what we believe about our experiences and how they really are. We cannot just base our understanding of how we’re situated in the world by drawing on our beliefs about how the world works. We’ve operated under this flag before, and, according to Dennett, fell prey to the grand illusion, i.e. that perception involves the acquiring of rich, qualitative data that hardly fails us when forming beliefs about the world.
Again, this essay is not an argument for or against anything. The general sentiment expressed in my post is that of a sympathy towards fallibilism, and the need to investigate the extent to which our perceptual systems are fallible. My contention in this post is that a firm understanding of perception and its content has a scientific upshot, i.e. it will help us treat and understand phenomena like inattentional blindness, change blindness, unilateral neglect, akinetopsia, synesthesia and various forms of agnosia.
12:25 pm
Polo, my argument, if you can call it that, is this: one shouldn’t be so pessimistic about the fallibilistic nature of perception. From what I gather in your blog post you seem to question the very nature of witnessing to the core. I think that complete mistrust in the senses is not a great idea. Just because I missed the gorilla doesn’t mean I missed the whole thing. As you said, as soon as I was aware of the gorilla I can see it. That is my whole point. We can talk about it throroughly in person, or discuss it in a series of essays if you want…or keep informally commenting in a blogpost with blogpost-like thoughts… 🙂
1:23 pm
I agree with you Polo about understanding the limits of perception completely. I just don’t agree with the sense of mistrust expressed thus:
“The problem I’m referring to specifically involves the question of whether perceptual experience offers accurate information about the physical world?”
and
“Why are we so convinced that our perceptual faculties contain lively, rich, and vivid detail about the external world? This notion stems from an old-fashioned view of perception held by Aristotle”
Missing a dancing gorilla while asked to focus on something else (keep track of a team of 5 different people in movement, quantify those movements while a different team of different color do the same on the same area) is not a direct argument against the fact that our perceptual faculties do capture rich and vivid detail about the external world. Check: while I missed a dancing gorilla I: tracked the sudden and fast movements of 5 or 6 different people, discern those people and their movements from 5 or 6 other people also moving fast and suddenly, plus I followed carefully while counting a little ball being passed around while another ball was also being passed around by another team in the same small area… to tell you the truth it seems like my perception is not perfect but pretty amazing at paying attention, tracking, recognizing properties, dicerning detail in a very VIVID and DETAIL manner. It also seems like if hadn’t been asked to perform a very detail quantifying task such as counting the passes made by the team, chances are I would have easily seen that Gorilla, although I wouldn’t have an idea about the position of the ball…should I use your phrase and say “period”?
Greetings from central mexico guys, I can’t wait to discuss this with you and be corrected by your insight, maybe at Kinley’s? Saludos.
1:46 pm
Ramon, what I don’t quite get is that you agree with Milikan in that our perceptual capacities are fallible, yet you’re expressing dissatisfaction for a post that shares her concern. Again, I am NOT making an argument for anything in this post. Dennett’s assertions about the shortcomings of our visual system are a widely held view accepted by philosophers. As far as accuracy goes, I’m in 100% in total agreement with you in that our systems are fallible and that this is a design feature. We should definitely talk about this when you get back from Mexico! I hope you’re having a great time! Maybe we can discuss this over Huckabees, beer and some delicious food?
12:36 am
Polo, nobody is going to tell you that our perceptual capacities are perfect. Dennet’s views on the shortcomings of our visual system are very sensical too, yet, again, the part I don’t agree with is coupling these observations with a weird pessimism about the whole of our perception and then pushing it even further and making of these shortcomings into arguments for a constructivist account of perception (as Greg seems to do).
I think I’ve said this like three times already: my problem is with the whole tone in which many people seem to address the limits of our perceptual experience as: “beware all of you who trust your organism’s capacity to perceive the world, if it fails sometimes then it’s probably failing all the time and none of what you see can be trusted, the world is but an illusion of our mind’s creation (insert 50’s propaganda voice)”
Although one needs to be aware of these shortcomings one also needs to be aware of (and should be in awe, I would emotionally argue) of just how well they DO function (even if it wasn’t that often for as you very well know Millikan would say that all you need is that that they do so even once in a while to define their proper function). Successful perception doesn’t even need to happen the majority of the time, what matters is that it HAPPENS. I think that is the question of philosophical interest and one that is neglected to pursue the pessimistic path of inquiry which in my opinion seems to be something that science will gradually inform us about.
but I’m repeating myself over and over, we’ll talk about this soon and maybe we can read something together this next semester…
2:35 am
There are scientific findings that put “constraints” on classical approaches to perception, i.e. that perception involves “an internal representation of the world.” (Noë, 2010) The findings: change blindness, inattentional blindness. There are also findings that urge philosophers to develop stronger theories, i.e. unilaterlal neglect, achromatopsia, synesthesia, akinetopsia, agnosia(s) and metamers (just to name a few). Argument: If there are scientific findings that put constraints on and urge for the reconsidering of classical Aristotilean approaches to perception, then there IS definitely a problem. Focusing on pessimistic/optimistic vibes is missing the picture greatly – there IS a problem that needs to be addressed, and this truth, as the sciences have shown, is undeniable.
7:16 pm
Ramon, I feel like you’re creating this dichotomy of perceptual experience as either success or failure. If you create that sort of dichotomy, then yes, I think Polo and I would both be committed to the notion that perception is never entirely successful. But that’s an extreme version of what we’re really saying, and is misleading. There is a whole range, a grade, of the degree to which our perceptual experience corresponds to some given state of affairs in the world. Maybe we normally tend to be somewhere in the 90-99% range of accuracy, maybe in certain circumstances, less. But these are epistemological questions. What Polo and I have been discussing is more relevant within philosophy of perception and philosophy of mind. We’re not criticizing externalism about knowledge (though I would criticize it if we were having that conversation), but rather externalism about mental content (and even then, Polo didn’t really address that at all in his post, but just pointed out some factors that would lead to a conversation about the nature of our mental content, so more accurately, he’s pointing out the importance of developing accounts of representation that can handle the features of how our perceptual system does in fact function). So what we learn about the ways in which are perceptual experiences are in error (remember, all that means is ‘not 100% successful’), greatly informs issues to do with representation and mental content.
I don’t think you need to be committed to externalism about content if you are committed to externalism about knowledge, so I’m not sure what you’re views on the former are. But it does seem to me that you seem to be defending the latter, when the post was not really focused on those issues at all.
12:12 pm
“You might be asking yourself whether the ‘dancing gorilla experiment’ threatens Aristotle-like programs. It does so in the following way: If perception plays a passive role, then how do we account for unsuccessful visual experiences? Say, for instance, I see a dog, where there ought to be a blue ball (i.e. illusion). If we subscribe to Aristotle’s view, then the blue ball is creating (in me!) a representation of a dog. See the problem? How can object x (the ball) represent itself as y (the dog) in experience? Lets take things a step further – if I perceive a dog, where there is in fact a dog, this is considered a successful perceptual experience by many a philosopher. But how accurate are our percepts? If my percept of a dog is not accurate (I perceive the dog as red vs. brown) then it would be fair to say that object x (brown dog) represents itself as y (red dog) in experience, where x and y are distinguishably different.” -Polo
Greg, the dichotomy seemed to be there and it was actually what I was trying to elucidate, that it is absurd to create a simple dichotomy and even more absurd to take the pessimist’s side of it.
But what you mention about you guys talking about “mental content/representation” and me talking about epistemology does sound a bit weird. I for one thought we were talking about the same thing since the only reason I brought Millikan into the discussion is because she has this very interesting view in which she clears up the discussion on causal/informational accounts of mental content by proposing a shift from representation producers to representation consumers and then from representation consumers to normal conditions for the proper operation of such consumer’s perceptive abilities.
The paper is called Biosemantics and it’s all about representation, mental content, normal conditions for the performance of a proper function and of course language. You guys should definitely read it. It is very cool to see how she goes about making the case that representations should be of something and even when that something is to be to a certain extent constructed by the perceptual experience of the subject the subject is to that same extent environmentally/biologically/evolutionarily bound to the outside to perceive things as such. So, my point being, that I was talking about “mental content”, “representation” and “perception”.
Also, I think Polo is taking the word “pessimistic” in an unintended manner: see Pessimistic Meta-Induction argument.
2:07 pm
Pessimistic meta induction says that since previous theories haven been prove false, our current scientific theories could very well be false. My argument isnt dismissing the possibility for a variant of Arisotileanism, it is simply acknowledging real problems in science (as stated above), and recognizing that a new theory (a more robust one) is needed.
10:50 am
Well, from your piece of writing it seemed that you were making a pessimistic induction argument against perception by saying that “if percepts fail in this kind of setting then they most probably fail at other settings therefore we should actively disbelieve them.”
It’s a very basic formula that can be applied to many discussions, in this case I read it as being applied to perception and I attacked it as such. Maybe I got it wrong and your paper actually said “percepts fail in this kind of setting, always in the same way, which means they function in a reliable manner given uniformity of behavior across time, let’s keep trying to find out why.”
Isomorphism is far fetched, I would agree with that, but the alienation of our percepts from the history our evolutionary history is a remnant of that phenomenological discourse that forgets that we too are natural and not some special cognitive entity in a far away realm coping with an outside by making it up.
again, I think I’m repeating myself, so kinley’s?
6:10 am
“the question of whether perceptual experience offers accurate information about the physical world”
If they were accurate we could dismiss a lot of the methodology of science introduced precisely because they are fallible.
Put to one side the fact that we introduce instrumentation to extend the range of our perception, not only in scale (microscope, telescope) but also into ranges that we don’t have access to (infrared, radio). Even within the human scale world our perceptions are fallible with regard to precision. If they weren’t we wouldn’t need scale marks on instruments. In the everyday world it doesn’t matter. My grandmother always baked cakes by using her experience to estimate quantities. My daughter insists on using electronic scales and measures to the nearest gram.
There’s no evidence that our brains have changed significantly from the time when humans roamed about in small groups with only the crudest of tools. Our brains weren’t evolved to be precise, only to be good enough. If a predator came at you it was only necessary to recognise it and run or defend yourself. You weren’t required to identify it later in a line-up, or having speared it, to justify your self-defence to a lawyer who is suggesting the predator may have been merely asking for directions.
In the every-day world we don’t notice how flawed our perceptions are. The colour tile illusion is a case in point. As you sit in front of your computer, how often do you survey your desk and try to spot where this colour illusion might be occurring? You don’t. You don’t notice the countless times your perception gets it wrong because it doesn’t matter. The very curiosity and amusement invoked by optical illusion is due to our surprise that they are so obvious when we are made aware of them.
If you think of this in terms of information theory it goes a little like this (very crudely). Our perceptions are effectively continuous. At about 60 frames/s equivalent that’s about 5,184,000 frames a day of saccade imagery, being merged, processed, some retained and merged further, and all very crudely and variably encoded with who knows how much lossy compression. Somehow some of this imagery remains encoded in our brains and is later reconstructed as we try to imagine, in court, the events of several weeks or months ago. And we may have been playing it over in our minds countless times since then, each replay changing minor details. This is not the reliability of every day cake making. We use our brains for far more than natural selection selected.
So, while I agree with Ramon that we are adequately reliable in many respects, your post brings out a very necessary acknowledgement that we often presume that perception is as reliable as our regular experiences lead us, mistakenly, to believe, and that in special circumstances we should be allowing for that and paying attention to the fallibilities demonstrated by science.