registered authors login here For more on PMWiki, see pmwiki.org |
This reading is in the coursepack. If you like it, see also "Mind Time : The Temporal Factor in Consciousness" by Benjamin Libet The synopses of the human brain have often been compared to binary language. For example, the human brain perceives a loud sound as compared to a quiter sound, not because it sends a stronger electrical signal but simply sends more "bits" to let the brain know that the sound is a loud one. The interesting part of the reading was mapping bandwidth requirements to various brain functions. Defining conciousness in neat little placeholders seems to miss the point, though. For me, the practical nature of human interaction and the limitations of our perception are easily assessed with common sense. gjh235 . i mostly agree. though seems like the hoffman reading, is saying not all limitations can be accounted for with common sense (if there were, we'd have no bush supporters in the US) people really do eat sand in the desert believing the mirage. but thease are extreme cases, and in daily life, you're right, a few missing bits is easily overcome. but the reason i had brought up glen gould before is that actually musicians hear with a different part of the brain. in the same way many native asian speakers don't hear "l"'s the same way europeans do. it's not so much mispronunciation, but that they never developed the same wiring. just like we can't perceive the intonation differences in chinese, or miss the nuances of vowels in french. it's nothing to do with our ears, it's that our brain has to be tuned for what to look for. hence, getting back to glen gould, his sensation hearing a loud note is both quantitatively and qualitativly coming from a different place than if george bush heard the same thing. (a loud bang would be "assembled" using a different part of his brain) it's a little bit of a "tree falls in the forrest' argument, but the sound really doesn't exist (except maybe as chaotic, pulses of air waves), it is formed by whoever comprehends it. and different listeners, build it with different methods. the input isn't a fixed external thing, as norretranders describes. we can't really even make much of an educated guess what the input is like in raw form (or if it even exists). but this is a long winded way to say, i think you're right in the end, that ultimately successful interaction is more important. and our limitations seem to work out fine in daily life. we really only have our common sense to work from, but that seems to get us fed most of the time. norretranders opens up a can of worms with this bit/perception idea, when really the end point is just that we aren't conscious of most of what we think we are. hoffman goes on to show how, in the process of comprehending, in our heads we fill in the missing info. we mostly see what wee expect to see. mostly make sense of what's new, by projecting and seeing an image assembled from experience. our senses supply a very small portion of the picture, but by updating random elements fast enough, we end up seeing most of it after a few seconds. for example, it's common to see/hear the chaotic sensory info that accompanies a migrane headache and believing that it's angels visiting with a message. to someone like Joan of Arc, angels with messages are well within her repetoire of comprehensability but migranes aren't. though, your point sort of trumps it all, since most of us aren't going to get messages from angels in our daily routine. (wow. i could max out this server just with raw text about this book. it's one of my favorites. and not that i whole heartedly agree with every point, or even most, but that it talks about many of the aspects i find most fascinating in this sort of neuro-tech hybrid world.) the first thing, that stands out like a sore thumb seems to be that perception (and the portion that ends up in consciousness) is given a bit value. not at all that it can't possibly be a numerical value, but that it is measured as described, which no human could (probably ever) do. even how a computer image can be a bit map or vector art, with very different file sizes seems to make this obvious. one can be conscious of a red room (vector-like, a solid color) or visualize a scene of that room with the window open in the late afternoon (bit map-like, gradations and areas of hues). we can generally remember what a fire engine looks like. but there are also artists like breuggel(sp?), matisse, albers, etc. who studied light, who could only be motivated by our ability to switch between mental image rendering schemes. no fixed value being possible (why one point on a graph and not a multidimensional shape (simultaneous but discreet points)?), it seems to set us for pages afterwards on a wild goose chase. most of us are comfortable with a 7 digit number, but in new york it went to 10 a few years ago. a person still may have a dozen numbers memorized. many of the area codes will be the same, so perhaps it is really only adding one more bit. but are there now significantly fewer numbers memorized in new york? probably not. part of what alters the bit value of perception, that makes it dynamic, impossible to call static, is that most things set off associations and have resonant emotional values that change. truly arbitrary visual stimuli is incredibly rare. thus no stimuli can possibly have a fixed bit value. a number 295-3547 may be random to only a few, but for a person who onc e had a crush on someone who lived at 29 e 3rd street, or another who's mother was born on march 5th, 1947, everyone has a different bit value. different things are certainly noticed, but even subconscious things register differently. ("why does 3547 seem so familiar?") or imagine a neuro disorder where frequencies are accurately reported to a person's brain, but a malfunction makes them functionally tone deaf, in that tones never make it past short term memory. they would throw off the measuring of pitch sequence recall completely. as would someone like Glenn Gould who could surely recall probably over 100 tones. not simply that they only use different neumonic methods in the consciousness, but actually different tones and particularly relationships between tones exist for very different parts of their brains. if you watch the neural paths triggered, they are completely different. musicians listen with a different part (simplified to a phrenoloy-like model of distinct modules, but a close enough to illustrate) of the brain than most. sound is no longer purely sensory/pleasure/informative information. similarly, you may hear 7 notes of an Eastern scale (24 tones compared to our 12, approximately but not consistently divided in halfs), be able to distinguish each consecutive pair as higher or lower in pitch, but have no idea if the 2nd was higher than the 6th. not that 7ish isn't an entirely peculiar number in the way described, that microstates are often (but hardly always) neccessary to get beyond powers of 7ish-ness. but that perception could have any one fixed value. and then particularly astounding that anyone could consider that value independant of a perciever? (even the criticism by the bell labs engineer either misses this or the author didn't really understand his point (possible, but i doubt). the way babies learn to see/use their eyes/configure the neural paths they will use later during those first 4 weeks is far different than what info that person sees (even pre-consciously) at 30. consciousness is entirely relative but still also has a certain influence on what information the senses take in. facial expressions are prioritized over dull grey shapes. those priorities are hardly fixed or universal. they too need never be registgered consciously but organize sensory input. there is no practical way to speculate about sensory information prior to it's becoming sensory information to someone. an a priori world may or may not exist, but we can't possibly measure it. and certainly not without projecting somebody's subjective consciousness onto it. granted norretranders eventually aludes to a paper that says "the tragedy of kupfmuller's publication consists in ...the fact that cognitive variables are quantifiable ... " but of course this cuts out most of the actual quote which appears about 17 pages into the argument. too little, too late. still a wild goose chase. though the models of communication that follow, and what we do beyond out limited consciousness within our minds, can be just as insightful. His description of the processes of gaining from each other's behavorial/cognitive "trees" seems no less valid. non-conscious communication can still be extremely viable, without assuming any precisely measurable limitations to some potentially perceivable environment. even at the most extreme, we could just as well learn from each others' entirely imaginary experiences. (and there's no way to proove we do or don't) is language that isn't expressable by non-verbal means of any actual "tree"-like use to us? why couldn't non-verbal but cultural significant cues be the cultural info shared that makes "trees" communicative/sought after? why can't that bit value be continually subject to a subject's cultural influences, rather than visa versa? If you liked this chapter, definitely check out Benjamin Libet's "Mind Time". Also, on the subject of 7-ishness, check out Lakoff & Nunez' "Where Mathematics Comes From" which explains how they see mathematical concepts as arising from embodied minds. Definitely a good read. What appeals to me about the material in the Norretranders book is that he points out how little we do with our conscious mind. In fact, we do many of our everyday activities without being conscious of them. This is relevant when you consider any high-performance instrument or tool: you don't want to have to concentrate on using the tool, you want to concentrate on what you're doing with it. Glenn Gould had to focus on the music, not on the piano. A well-designed tool allows for this. See Paul Dourish's book, "Where the Action is" for more on this. For example, think the saws in the shop. The chop saw, the radial arm saw both have well-placed cutoff switches, because they allow you to focus all of your attention on the blade and the wood. The scroll saw does the same, to a certain extent, because the switch is very close to your field of vision when you're cutting. The bandsaw, on the other hand, has a switch well away from the area you're supposed to concentrate on, so when you want to stop, you have to shift your attention momentarily. For me, that's not ideal. that's exactly it. sort of like the phenomenon of people getting their best ideas in the shower, has to do with activating that auto-pilot. maybe it is like a catalyst? maybe like neurons getting excited because of neighbors becoming active (it is electricity)? but odd that he makes a huge effort to quantify sensory info. isn't at all necessary to say "more/less", as much as consciousness either lets in or monitors a "different" set than other parts of the brain. Could there be info that gets to the conscious but not subC? Probably not, but there's no way we could find out. Great points. I agree that the first thing that stands out for me is the idea of quantifying perception in terms of the number of bits processed.
I found the graph of Helmuth Frank's observations interesting. It shows this "bandwidth of consciousness" peaks in late adolescence and slowly declines. As we age, we lose out ability do distinguish subtleties (16 bits/sec). I wonder if there are ways of seeing or hearing that can be learned or practiced that can off set this decline. And once this decline happens, can we reverse it (maybe by practicing some for of meditation or heightened awareness exercises) Here's my notes on the piece: The bandwidth of conciousness is far lower than the bandwidth of our sensory perceptors." - So how do we widen this effectively? Is it wise (or even laudable) to do so? More happens in your head than you know, unless you stop and think about it." - Isn't it better that way? The narrow focus of conciousness allows us to navigate and act, yes? All these exercises require no more ofyou than moving your attention around your body or your surrounds or the inside of your skull" - This suggests that training would enable more, no? Shouldn't we be studying *that*, if increased efficacy is our goal? It is important to have entropy in teh macrostates: They must permit many different microstates." - is this to allow mnemonics? There are many skills that we best exercise when we do not think about what we are doing" !! The concept of outside of consciousness... *that's* a new way of thinking. And yet if we look at the study, it indicates that this is most efficacious! In regards to excellent authors: "They are masters at stagng plots that usevery small amounts of information to make the entire register of previously produced exformation grow forth in people's heads." - We need to use this in designing interfaces? Isn't is extremely culturally and contextually bound? THIS seems like smashing nuts with sledgehammers, to me... Live concerts are not about music but about what music does to people" - This was powerfully enlightening for me; I hate going to live music performances for precisely this reason - music makes me feel intimate, open, and vulnerable, and for that reason I love to listen to it, either with my wife or with headphones. Rocking out with other people, though? No thanks! I don't *know* those other people, or event he audience; being able to listen to it is a private experience. Concious language manages very little of what goes on in a social situation. Far too much information is discarded before we get to the information." - the best way to lie is to believe the lie... ;) But for this, isn't telepresence (or communication mediums like IM or IRC or even Email) valuable because it is a simulacrum of real comminication, except stemming purely from conciousness? I.e., are we seeking it (and using it) because it gives us an illusion of control, when really it's preventing us from communicating more effectively (i.e., honestly)? Wow, this stuff blows me away, the study of perception is incredibly fascinating. I really enjoyed Nørretranders discussion of "chunking", the ability of people to take in combinations and patterns of images to identify and make sense of the world is something machines haven't quite made yet. As a graphic designer by training I have spent many years working with type, kerning pairs, noticing shapes and ligatures. It may be my imagination but when I am driving and looking for an address I always seem to be able to pick out a street name even before I can focus on the individual letters by the shape of its silhouette. It is an interesting mental game to percieve the small gap between that type of recognition and when your eyes actually focus. The chart for "Information Flow in Sensory Systems and Conscious Perception" was interesting in the ratios of total bandwidth to conscious bandwidth. The largest ratio of total to conscious bandwidth at 100,000 to 1 is smell, which explains why we can all put up with the subway system (to a point), the converse is hearing at 100,000 to 30, the lowest ratio which explains why we can forgive bad visual quality more easily in multimedia than bad audio. Nørretrander finally uses the Three media example which made me think of my experience with each medium and which I prefer. In an interview situation I would much rather listen to it on a radio, I am not as distracted by the visuals on TV and can just concentrate on what I am hearing, nuance and intonation bring more meaning to me than just the written word in print. I am really interested in this whole field. I will pick up this book if the rest is as good as this chapter, any other related readings I would love to find. david, tom has some recommendations above, also see stuff by A.R. Luria, "visual intelligence" by donald hoffman, "image & logic" by peter galison, "exploring consciousness" by rita carter. good stuff in oliver sacks books, too (a lot of his best stuff is in his footnotes?). I'm not sure how to react to Nørretranders' use of Miller's information bit analysis because it doesn't seem coherent to me. Consciousness is considerably more fluid and complex than what Miller's description of short-term memory span would suggest. It is affected by not only attentional configuration but also by factors as diverse as endocrine states, blood flow and sleep history. All of us have had the experience of both reduced awareness when driving a familiar route, or heightened awareness following a stressful event. Nørretranders is certainly correct that sensation does not equal conscious awareness. Even beyond the limits of attention, much of the information that arrives at the brain is totally inaccessible to awareness. We process much more than we are able to report verbally. In fact for the most part, verbal reports are a poor measure of perception. A different way to quantify perception is through measuring action. As interaction designers, we need to be aware that we will primarily influence human action. It's less important that chop saw users be able to verbally report the location of the cutoff switch. What counts is that they accurately activate it during use. I usually get skeptical reading scientific analyses of fundamentally immeasurable human phenomena, but I really appreciated Nørretranders' approach to laying out how our brains process information. This being my first reading dealing with information / perception theory, I was astounded to learn about the division of our perceptual environments into numerical bits of data. In all of the examples of other researchers, it seemed that the size or organization of the bits that a scientist constructed depended more on the theory they were trying to prove than the actual ability to divide say, a letter, into a certain number of bits. So, I was very pleased when Norretranders admits that the exact amount of data we perceive "really does not matter. What matters is that we admit far more bits to our heads than we ever become conscious of." This generalization is much more exciting to me than trying to pinpoint exactly how information our brains are perceiving. Its funny because I've been thinking about this for a while now but didn't have the terms to describe it. The fact that we construct ourselves by what we ignore just as much if not more than what we pay attention to. Exformation. I've been thinking of it more in terms of an evolutionary defense mechanism (survival depends upon focusing on pertinent information), but what interests me here is how the author hints at its effect on the subconscious. Particularly, I am intrigued by the discussion of how lying is made possible by the limits of conscious perception. I'm not sure if I'm fully understanding this section, but it makes me wonder how technology could be funnelling or filtering exformational dynamics to bring about more lies. For instance, take an air force pilot looking at running dots on a video screen from thousands of feet in the air and deciding to shower them with missiles. In this setup, the technology does not communicate whether the dot is civilian or enemy combatant, or even arguably, whether it is human at all. Thus, the humanity of the individual of the ground is exformized (is that a feasible verb?) so that the pilot doesn't have to question his orders and essentially lie to himself about what action he is taking against another human being. I also feel that this reading puts a few new tricks up my artistic sleeve. The experiment of closing your eyes and opening them just for an instant really allows you to take in a scene as a whole rather than its individual components. His discussion of archetypes in successful literature was also extremely useful to me. I've always felt like I tend to barrage the viewer / participant with too much extraneous information in a lot of my work. It's so much more effective to rely on what the person already knows and is familiar with, to get them quickly engaged and emote him/her from there. For a while at the beginning of the chapter I was fairly skeptical about the notion of representing the information we take in as bits. It seems as though many of our senses are more analog in nature, and as such the nature of the data we take in is richer than a small handful of bits. For instance, how many bits are in the color red? And we measure many physical sensations such as pain by the number of nerves sending pain signals to the brain (pin prick vs. touching a hot stove). It seems to me that the measurement of pain is a very fine gradient that indicates just where the pain is and how much it hurts; and that this information cannot be represented in 30 bits. I can get behind that this massive information processing cabability happens subconsciously, but there is still part of me that feels as though you are still aware of it. Whether or not certain experiences "count" as conscious (vs. sub-conscious) is an open question for me. The part about the importance of exformation cleared up some of my concerns, though. I'd been thinking about how our experiences allow us to encapsulate the information that we see and thereby process more of it (the analogy in my head was actually a zip file, believe it or not). With the concept of exformation, I think one can imagine that we learn how to subconsciously summarize what we are perceiving and then deliver a simple notion to our conscious mind. So perhaps, the routing of information for things that are not being learned in the moment is external environment -> sub-conscious -> conscious. So perhaps the concious mind is primarily concerned with learning and absorbing new information, while the sub-conscious processes things that have already been routinized and hands them off in turn to the conscious as appropriate. The reading reminded me of an interesting example of my own subconscious processing that I became aware of earlier this year. I'm an avid basketball fan, and when I would have friends over to watch basketball games on TV, I found myself both unable to hear them or to speak to them during a particularly engrossing segment of action. The interesting thing is, even though I would have 0 recollection of actually hearing them speak (or seeing them speak, etc.), as soon as the play was over and I relaxed my attention, my subconscious would remind me that they had said something (and I could ask them to say whatever it was again). I developed my own theory of information buffering to explain it. The reading made me think of this (though it isn’t that connected): when I was little we sometimes had a substitute teacher who spent the majority of classtime showing off his amazing ability to remember everyone’s name and date of birth, after having seen the info only once on the roster (he was later fired cause he never did anything else). His technique for remembering so much involved taking in elements that we might not normally notice from different sensory inputs and making connections between them. As an obvious example, a kid in the class had the last name of Phillips, and the substitute told us the kid’s nose (which was long and bony) had reminded him of a Phillips screwdriver. And, more interestingly, a girl’s high-pitched “here” after he called her name had reminded him of some animal which had a name similar to the first syllable in her last name. While we would normally try to remember someone’s name based on their face, he was remembering it based on a feature of their face, or their voice, so that he was deciding what sensory input he took in and retained based on if he could use it to make a connection in his mind. This cross-sensory memory technique was fascinating to me because the memory techniques that the other teachers taught us were all based on making connections in our mind between words or concepts we’d already learned, rather than what we were currently taking in. Rob brings up a very good point above: "As interaction designers, we need to be aware that we will primarily influence human action. It's less important that chop saw users be able to verbally report the location of the cutoff switch. What counts is that they accurately activate it during use." Unlike Libet or the other neuroscientists mentioned in this reading, we don't get to devise experiments that can somehow quantify mental phenomena, we have to design for the givens. We can, however, take away the general principles that come from their work, as Fiona's teacher example shows. We can design interfaces that allow people to memorize frequent actions, so that they can be done without conscious thought, or we can design controls that only afford certain actions, for the same reason. On the other hand, sometimes we want the physical equivalent of the "ARE YOU SURE?" dialog box. In those cases, we can design controls that require concentration in order to use them. To do this, though, we have to think through the activities we're designing for, even act them out as we plan, so that we can experience for ourselves which steps seem to happen unconsciously, and which steps seem to interrupt the flow of the action. It doesn't matter if a control's position makes sense logically if it doesn't make sense physically. The band saw controls in the shop are one of my favorite examples of a control that makes perfect logical sense, yet always leads to fumbling in actual use. Because you can't rely on seeing the stop control before you hit it, and because it's right next to the go control, which has the same physical behavior, it's easy to hit the go instead of the stop. You can't use that button set effectively without both seeing and thinking, or without a lot of practice. How might it be re-designed? A single toggle switch, maybe, or placement right above the work surface, as is done on the scroll saw. The QWERTY keyboard is an interesting example in light of this reading. It was originally designed as a way of slowing typists down, so that they didn't jam the keys of typewriters. Typing teachers developed methods to train typists to speed up their use of the QWERTY keyboard, and typewriter mechanics improved to handle increased typing speed, yet QWERTY remained. Even better keyboard layouts like Dvorak couldn't unseat it. Now we have things like the nubs on the F and J keys on the home row to help us improve typing speed. All of this as a way to get around the effects of QWERTY, which were intentional to begin with. This is such an interesting topic. I'm feeling a bit bummed because you can only do so much with a Wiki... We should definitely pick this topic up in class if we get the time. I think it is interesting to think about all of the mental processes that happen in our minds throughout the day. Above all, anything to do with the functionality of the brain is mind-boggling. I imagine that the next IMPORTANT 'technological leap' will have something to do with the human brain, thinking that we all feel that there is so much more going on in there. I want to add to what Fiona wrote about memorization techniques. I think that memorization also has a lot to do with emotions / feelings. Not sure if I can explain this completely. For example, when memorizing a script, one familiarizes himself so much with the text and the character that the real self is almost forgotten when the character is played. Personally, I tend to lose myself into that world, if math, physics or music. I don't know much about the topic of Emotional Intelligence, but it would be great if someone could write about it, or recommend a good reading. Somehow I find it hard to accept that we have this "bit per second" rates that are related to our bodies. I think that the binary way is obviously not the way our brain works, and I guess we have yet to figure that one out. But one thing is certain, that we do have selective sensing. What truly interests me is how the "sensing levels" are different between people. Just like the way I perceive the color blue might not be the way anyone else in the world perceives it. For example, when meeting someone who is mentally-disturbed, the way their sensors work and the world they are in intersects but doesn't align with my/our world and sensational boundaries. What intrigues me is how we can both stand in the same room but still be in completely different worlds. Not sure if I'm clear. |