I’m doing some intellectual scratching about re: the nature of pictures and pictoriality. I think one of the best philosophy books on the subject is Flint Schier, <em>Deeper into Pictures</em>
If you can recognize a dog, when you see one, and a picture of a dog, when you see one, then (ceteris paribus) you can recognize a picture of a cat – or of a candle, a house, a tree – provided you can recognize those things themselves. Schier calls this ‘natural generativity’. Your ability to recognize dogs and cats ‘naturally generates’ an ability to recognize pictures of dogs and cats. Pictures are hereby contrasted with natural language: being able to recognize dogs, and knowing they are called ‘dogs’, in English, does not provide any leg up regarding what the English word is for other things you can recognize. Making the semantic connection between ‘dog’ and dogs is not ‘naturally generative’ of any connection between ‘cat’ and cats, so forth. Each entry in the dictionary needs a separate convention. Not so with pictoriality, which, roughly, require only one convention: if a picture, P, admits of a naturally generated interpretation, I, then P means I. Pictures are supposed to represent what they look like they are of. That’s it (except for some twiddling).
This will seem grossly to underestimate the role of culture and convention in picturing, but I think it does not. The thing you have to get is that something can be completely, utterly tip-to-toe conventional – which all pictures of course are – without it being the case that picturing ‘works by convention’, as opposed to ‘naturally’. Take a hammer. Hammers are completely conventional, in that there are different ways you can design every last little bit of the thing. For any hammer you show me, I can show you a different hammer that would do the job, at least tolerably. But it doesn’t follow that hammering works ‘by convention’, i.e. not naturally. An Austinian, ‘performative’ account – How To Do Things With Hammers – would exhibit notable incompletenesses. Pictures are more like hammers than you might think. The key is to stop thinking of ‘natural’ and ‘conventional’ as having an innate tendency to exclude each other. Above all, do not note that a thing is ‘completely conventional’ and infer its workings are, therefore, ‘by convention’. (I didn’t learn this lesson from Schier, in particular, but he does a good job of not committing this common fallacy.)
A lot of you will be sure this is quite naive, so you have your say in comments. Meanwhile I, who know better than to think this is quite naive – yet will indulge your commentary to the contrary with avuncular good grace! – am pondering puzzles further down the line.
Schier considers cases that pain me for their borderline inconceivability. “Suppose I walk up to what I see as a blank canvas and find myself (to my surprise) inclined to say ‘This canvas represents Marilyn Monroe'” (203). And suppose others have the same response to the canvas. Specifically, those who can recognize her in life (and film and photos) are inclined to say the blank canvas is ‘of Marilyn’. Those who can’t, aren’t. So the conditions of the picturing relation are met, without us being inclined to say this is a case of picturing. But how does it work, this stupid canvas? Schier does not really say. But let’s have a go. If I can have a Jennifer Aniston neuron, I can presumably have a Marilyn Monroe neuron, tip-most taper of some vast, neuronal candelabrum of natural capacity to recognize Marilyn Monroe. (Not that anyone ‘naturally’ recognizes Marilyn Monroe without having ever seeing her. I’m talking about the natural capacity to recognize her after having seen her before.) So fine. We hypothesize that there is some subtle feature of this canvas, some through-the-eye effect it has that we aren’t aware of, that can cause that neuron to fire. Of course, the eye bone’s connected to the brain bone, so the effect has to ramify through that vasty, submerged iceberg of neuronal Marilyn recognition machinery. Which is conceptual processing, not visual processing, per se. The neuron fires when the eye sees either a picture of Marilyn or the words ‘Marilyn Monroe’ (the Jennifer Aniston neuron fires in response to ‘Jennifer Aniston’, so why shouldn’t the Marilyn neuron be able to read, too, in effect?) But the words ‘Marilyn Monroe’ are not a picture of Marilyn. So how can we say whether the canvas, in having the effect it does, is more like a word we aren’t aware we are reading – in which case it’s not a picture? – or is more like a picture we aren’t aware we are seeing? – in which case it’s not like a word? A Schierian account would seem to lean towards regarding it as a picture, since the effect is ‘naturally generated’ rather than being the result of some conventional training in associating this canvas, arbitrarily, with Marilyn – as when we teach children that dogs are called ‘dog’. But, then again, it still doesn’t seem much like a picture of Marilyn. So maybe this just goes to show that natural generativity is insufficient for pictoriality. The whole thing seems a bit like synaesthesia, of some weird sort. So then maybe we should say that picturing itself is a kind of synaesthesia that all humans share. As Nelson Goodman always points out, pictures don’t much resemble the things they are of – they are flatter, often much smaller, etc. But (here is the seed of rehabbed ‘resemblance’) that doesn’t prove they don’t ‘look like’ the things they are pictoriallty of, merely that the looks-like relation can’t be cashed out in terms of resemblance, if that is understood as a quantity of shared visual properties. The looks-like relation is a function of recognition, and how that is effected is an open question. A synaesthete may be very strongly and systematically inclined to say that the number three is blue, hence the numeral three looks blue? Is that right? Synaesthetes think the numeral gets its color from the thing it is of. (You can even have a community, hence a degree of public meaning. Nabokov and his mother were both synaesthetes, and they supposedly agreed about which numbers are which colors.) So a series of numerals (or letters) may function as a ‘picture’ of a color collage. It’s no good arguing against this effect on Goodmanian grounds: the colors and the numbers have completely different properties, hence cannot be said to resemble, simpliciter. The synaesthete may quite agree (or does this contradict a synaesthetic sense that three really is blue even though this aspect of mathematical reality is never tested in math class?) Nevertheless, the strong sense of a visually-affordable internal relation is there. You can’t argue people out of the reality of this ‘looks like’ phenomenon, by demonstrating the absence of resemblance, by Goodmanian standards. The one thing ‘looks like’ the other, reliably. But in a picturing sort of way? Do synaesthetes think (feel?) a series of numerals is both a word-like representation of a number and a picture-like depiction of a color collage? I have no idea. Quite likely, synaesthesia operates in its own distinctive way, unlike either language or pictures.
Also, does it turn out that my concept of Marilyn Monroe is grue-ish, if this blank canvas thing works. Because my concept just is my capacity to reidentify a thing as being this thing – and that capacity is instantiated in a machine that in fact is calibrated to reidentify Marilyn-or-this-weird-canvas as being one thing?
I think it’s a pity Schier didn’t realize his Marilyn canvas case had been anticipated by Edgar Allen Poe, in his Ligeia:
There is no point, among the many incomprehensible anomalies of the science of mind, more thrillingly exciting than the fact – never, I believe, noticed in the schools – that, in our endeavors to recall to memory something long forgotten, we often find ourselves upon the very verge of remembrance, without being able, in the end, to remember. And thus how frequently, in my intense scrutiny of Ligeia’s eyes, have I felt approaching the full knowledge of their expression – felt it approaching – yet not quite be mine – and so at length entirely depart! And (strange, oh strangest mystery of all!) I found, in the commonest objects of the universe, a circle of analogies to that expression. I mean to say that, subsequently to the period when Ligeia’s beauty passed into my spirit, there dwelling as in a shrine, I derived, from many existences in the material world, a sentiment such as I felt always aroused within me by her large and luminous orbs. Yet not the more could I define that sentiment, or analyze, or even steadily view it. I recognized it, let me repeat, sometimes in the survey of a rapidly-growing vine – in the contemplation of a moth, a butterfly, a chrysalis, a stream of running water. I have felt it in the ocean; in the falling of a meteor. I have felt it in the glances of unusually aged people. And there are one or two stars in heaven – (one especially, a star of the sixth magnitude, double and changeable, to be found near the large star in Lyra) in a telescopic scrutiny of which I have been made aware of the feeling. I have been filled with it by certain sounds from stringed instruments, and not unfrequently by passages from books.
I love the absurd specificity of the star of the sixth magnitude. But, philosophically, I like the fact that it’s unclear whether the narrator is supposed to be seeing these things as ‘pictures’ of Ligeia, or experiencing some vivid synaesthesia. And we also get the related thought that, if indeed these diverse things are naturally generative recognition-inducers of something, then the something they are inducing re-cognition of isn’t Ligeia herself but something else that she expresses, i.e. is, likewise, an obscure recognition inducer of.
Namely: some kinda vampire! (WARNING: plotspoilers!) I don’t think this solution will generalize, however. The Marilyn canvas is of Marilyn because she and it are … of? – a vampire!
Instead, I think it turns out – skipping, like, 100 steps – that we should understand language itself in a more Schierian way. So say I. We take the standard sort of ‘pictures should be thought of as language-like’ line and don’t just deny it. We stand it on its head. Language should be thought of as picture-like. Not in an early Wittgenstein way. In a Flint ‘natural generativity’ Schier meets Donald ‘nice derangements’ Davidson with some Ruth ‘proper function’ Millikan added in. Too much for one post.
Please discuss: how naive is Holbo for not knowing that talking about ‘natural generativity’, as an explanation for pictoriality, is manifestly hopeless? Of course Holbo doesn’t understand Derrida. (Probably he only pretended even to try to read Derrida.) But he could at least have read Goodman a bit more carefully. Or even Gombrich? ‘Myth of the innocent eye’ anyone? Doesn’t Holbo pretend he likes Nietzsche? Sheesh. Yet here we are.
UPDATE: Before anyone points it out, I know she isn’t exactly a vampire. I don’t want to spoil the plot.
{ 35 comments }
Z 06.16.11 at 6:02 am
A few thoughts.
1) We are all synesthete. For instance, repeated experiments (the google friendly term here would be Stanislas Dehaene) have shown that human beings who can count invariably arrange numbers in space (the bigger they are, the further on the right they are).
2) Provided I understood well your recap of natural generativity, there is a definite natural generativity of language, and I have my son to prove it (you can’t argue with that now, can you?). Being able to recognize dogs, and knowing they are called ‘dogs’, in English, does not provide any leg up regarding what the English word is for other things you can recognize. True. But it does provide a pretty huge leg up regarding the fact that other thing you recognize “have” words. Now you “just” have to learn them. Or in true life: one-year-old points a pig on his book. That is a pig, I say. Does it again. That is a pig, I say. Sudden flash in the eyes, mouth opens, eyes widen. Does it again. That is a pig. Points to something else. That is a curtain. Points to something else. That is a wall. Points to the pig. That is a pig. Mouths open even more. Points to another pig. That is a pig.
Your one and the only word you recognize is Maman but already, thanks to a pig on a book, you surmise that each and every possible particular form of the universe is attached to a universal word in a potential lexicon. Pretty remarkable, that. And of course the same is true about pictures. I mean, whether or not universals and forms exist (and whether or not this even means something), the fact remains that infants certainly have the capacity to recognize that they potentially exist. Which brings me to…
3) Edgar Allan Poe rocks. And Ligeia contains a severe criticism of the platonist theory of form in the guise of The Conqueror Worm. The poem is usually read as a meditation on the absurdity of human life, but come on, the music of the spheres, three instances of form (one being form of God)… And the meaning? Well, I think Poe is saying that all this jargon about universal forms and what-not is well and good, but when nostalgia (the worm) hits you, it will come from no definite place in your mind and with no specific universal form (but rather from the scenic solitude of your subconscious) and will feed upon each and any sensation, word, idea, memory, feeling you might have, even though no direct connection can be made between it and what you are directly thinking or experiencing until it has conquered your mind. And now, try as you might, you might see a star, know it is called a star but you will feel the eyes of Ligeia. And all my days are trances etc…
4) Oh, by the way. The star is epsilon Lyra. It is a quadruple star system (two larger star orbiting around each other and each having a smaller satellite star), and it is frequently used to test the resolution of optical telescopes. Cue some comments about language (telescopes) being unable to separate reality from hallucination (a star from its double).
John Holbo 06.16.11 at 6:10 am
“the google friendly term here would be Stanislas Dehaenez’
Yes. I just now finished “Reading In The Brain”. Which is very good and interesting and to be recommended to anyone interested in all this.
“Cue some comments about language”
Or perhaps comments on Nelson Goodman’s “Star-Making” paper, and responses thereto.
“and will feed upon each and any sensation, word, idea, memory, feeling you might have, even though no direct connection can be made between it and what you are directly thinking”
I love The Hellboy “Conqueror Worm” volume, especially the scene where the head-in-a-jar is wailing ‘I want to see the worm!’ Although otherwise Mignola’s adaptive employment of Ligeia is rather tangential to Poe’s text.
ben w 06.16.11 at 7:40 am
“A system of representation is iconic just if once someone has interpreted any arbitrary member of it, they can proceed to interpret any other member of the system, provided only that they are able to recognize the object represented.†(44) And pictures are icons, in this sense.
Surely “recognize the obejct represented” must mean “recognize actual instances of the object represented”, because otherwise the recognition involved is going to look a lot like the interpretation for which recognition is a necessary condition. (I can interpret any other member of the system of representation provided only that I can interpret the representation in question. Sure, I buy that.) And presumably the recognition in question is not only visual, but a visual recognition based on, well, not a still image, but anyway not based on movement, either the object’s or one’s own: so that if I recognize such-and-such an animal from its close cousins the thus-and-sos by its gait, or peculiar head-bobbing, or whatever, I need not be able to interpret a still image/photograph/painting/etc. of a such-and-such as being of a such-and-such rather than of a thus-and-so. And obviously a quite specific sense of “interpret” is at issue; it is presumably a recognitive sense.
Questions:
(1) the condition states not, once I can interpret any arbitrary member, but once I have interpreted any arbitrary member. Of course I can only actually interpret specific members; I can’t be presented with (e.g.) the arbitrary member. Suppose then that we have some kind of color-reversing photography—presumably a system of representation—and I have been presented with, and accurately interpreted, an instance of such photography in which the thing (/scene) depicted is entirely black and white, so that what once was white now is black and vice versa. I could probably figure out—recognize, even—what it was, if it was simple enough. But things that I ordinarily can recognize with their ordinary colors might be quite unrecognizable to me.
And, in general, it could be true that whatever I’m presented with might be one of the few, or the only, instance of this system of representation which I can actually interpret; I can’t interpret any of the others. (Perhaps I even mistake the system of representation to which the instance with which I’m presented belongs—since it could belong to arbitrarily many.) It doesn’t seem fair to object that I haven’t, in that case, interpreted an arbitrary member of the class. How not?
Perhaps these are just not iconic. But what is? I mean—what are “pictures”, here? Surely I can fail to be able to interpret photographs (even those that haven’t been subject to darkroom effects, or outlandish lighting) that are of things that I can in fact recognize (perhaps not from that exact perspective, but perhaps so).
(2) Here a system of representation for representing letters:
‘a’ is represented by ‘n’
‘b’ by ‘o’
…
‘z’ by ‘m’
I think it’s pretty plausible that (modulo previously mentioned concerns), once you interpret one of these guys, you’ll be able to interpret any other one, provided that the object represented is one you’ll recognize (which you always will, since we’re just talking about sequences of letters here and the same sequences are employed by representer and represented).
vpbavp? be abg?
(3) there were some other things but I’m tired.
Chris Bertram 06.16.11 at 7:51 am
I enjoyed that post John, and it is the second recommendation of Schier’s work that I’ve had recently (the first from a psychologist specialising in children and art).
For an example of a sign that is instantly recognizable as being of the thing it is of, without actually looking like that thing, see this example (scroll down to first featured comment).
http://theonlinephotographer.typepad.com/the_online_photographer/2008/08/pop-semiotics.html
maidhc 06.16.11 at 9:24 am
Z: When I view my number line, I view it as though I am suspended in space above it, and, if I want to, I can pivot around so that it extends forward in front of me, or turn around the other way to look towards the negative end. Furthermore, it’s not a straight line. For instance, it starts to curve to the left around 50, and around 60 it curves rightward until around 70 it’s straight again. I’ve read that this is not uncommon. So the part about arranging numbers in space is correct, but the big numbers being on the right is more dubious. I’m perfectly happy to come around the other side and have the big numbers on the left, which is actually my preferred viewing angle for the numbers 1 through 20.
I wonder if any of those research subjects were left-handed Arabic speakers.
Synaesthesia is common in LSD experiences, and LSD mimics natural brain chemicals that connect different parts of the brain, so synaesthesia may be the existence of a connection between different parts of the brain that only some people have.
My brain draws a connection between rock songs and places. Not necessarily the first time I heard the song, but maybe when I really paid attention to it. To me, Carl Douglas’s Kung Fu Fighting is forever associated with the place where Finch Ave. goes under the railroad line just east of Keele St. (in Toronto). I have hundreds of such associations stored away. But other genres of music, not so much.
The discussion reminds me of the Radiolab show about the man who learned language as an adult. If you haven’t heard it, this man was deaf and had never learned any kind of language. This woman was trying to teach him to sign, but since he had no concept of language, it didn’t go anywhere. She worked with him for a long time without getting anywhere, but suddenly one day the concept hit him and after that it was no time at all before he was signing fluently.
Then she tried to get him to describe what it was like to be without language, but he basically said that it was impossible to describe using language. He knew other people like he had been as well.
The podcast is online at http://www.radiolab.org/series/podcasts/, it’s called “Words” and right now it’s on the second page. For some reason it stuck in my mind much more so than most of the other shows.
My father had a stroke and lost the ability to speak, although he still tried to speak. We could never figure out if there were words in his head that wouldn’t come out, or if he was just performing the physical actions associated with conversing. But he certainly didn’t read, or show any understanding of pictures. He did respond to music though; he obviously enjoyed listening to the classical music station.
Before his stroke he had very intellectual interests, including opera and art, and he had an interest in photography since childhood. I tried to interest him in reading this site, but he never did quite get the concept of blogs.
It possibly suggests that an understanding of pictures is in a different part of the brain than language–which uses several different parts of the brain itself. And music is different too.
I recently saw Cave of Forgotten Dreams, and that raises the question of when humans started to create pictures. Those are the oldest known pictures, but they are already very sophisticated. They are 28000 years old. DNA evidence says that Australian aborigines separated from the rest of the human race about 50000 years ago, but they have art.
People sometimes say that Neanderthals created no pictures. We have no evidence that they did, but we have no evidence that homo sapiens did either, before the cave paintings. Maybe Neanderthals had a vibrant tradition of sand-painting or tattooing.
Sorry to go on at length, but it’s a subject that raises so many fascinating ideas. I didn’t get time to bring in Poe, but I agree with Z that he rocks. I have a vivid memory of reading The Premature Burial at age 8, and I think I hoovered up everything he wrote shortly afterward.
John Holbo 06.16.11 at 10:59 am
Thanks for the good comments. Responses later. (First, baths and bedtime for girls.)
Ajay 06.16.11 at 11:48 am
A synaesthete may be very strongly and systematically inclined to say that the number three is blue, hence the numeral three looks blue? Is that right? Synaesthetes think the numeral gets its color from the thing it is of.
It’s not blue, it’s green! For me at least ;-) This is a chicken-and-egg problem for synaesthetes as I for one don’t remember having an idea of a number sans its colour, so it’s hard to say if threeness is green, or if the shape of the numeral 3 is green. While ‘my’ triangles are also green (but a lighter, yellower shade), so are the letter B and numeral 8 (once again, different shades)! Drawing nonsense diagrams which resemble 3/8/B is surprisingly unhelpful; my impression is that if the symbol has no meaning to me, it doesn’t have a colour either :-/ I am living in Germany now, and feel the esszet symbol ß (= ‘ss’) is blue, as is the letter S, so that would seem to be more sound or meaning influenced, though it too is rather sinuous…
So a series of numerals (or letters) may function as a ‘picture’ of a color collage.
In my experience the meaning of the numerals in the moment of comprehension far outweighs their colour – they don’t depict the colours, they just happen to have colours.
The system is not for me ‘reversible’, i.e., while each letter and number is an extremely specific shade, if I were to see that shade in another context I would not think to myself, ‘Oh, there’s a very 3-ish shade of green’. And while I have a (rather weak) sense of colour and shape for music, too, I don’t hear music when I see colours either. Equally, some colours are shared – both numeral 1 and letter I are black. So this seems a problem for the ‘synaesthetic colours of numbers = picture’ idea, because while one can move from a picture of a pig to a real pig and recognise one from the other, one could not move from a given colour to a specific numeral/letter/sound with certainty.
The one thing ‘looks like’ the other, reliably. But in a picturing sort of way? Do synaesthetes think (feel?) a series of numerals is both a word-like representation of a number and a picture-like depiction of a color collage?
My feeling is that the colour of a series of numerals is not what it represents – it represents its numerical value, and happens to have colour only in the same way that ‘3’ has two curves or ‘2’ rhymes with ‘roo’ – those things are true of the individual numerals but are not how they are initially comprehended when one first reads them.
I emphasise the above is merely my experience as I believe synaesthesia varies quite a bit from person to person in terms of range and intensity. I have also always felt FWIW that everyone is synaesthetic, if only they could see it in their mind’s eye – try drifting off to sleep as you listen to music; something in your brain ‘lets go’ and the swirling, pulsing, vibrating colours are much more vivid. And no drugs are required ;-)
Tim Wilkinson 06.16.11 at 1:39 pm
Not that anyone ‘naturally’ recognizes Marilyn Monroe without having ever seeing her. – I recognise pictures (including movies) of her without having seen her.
What here is meant by a ‘system of representation’? Does everything that could be called a picture belong to a system of representation called ‘depiction’? Or are we talking about a certain style of depiction? Is this something like recognising a new typeface, where (perhaps, I dunno) we can rapidly adapt to a (sufficiently consistent) style of deviations from (let’s say) some set of stereotypical forms of the characters? The stereotypical form of pictures perhaps being what the depicted object actually looks like seen (through a window) from a standard angle and distance, or what a trompe l’oeil would look like?
Trompe l’oeil surely is for humans a naturally recognisable (too recognisable) depiction (my dog recognises and distinguishes moving images of dogs, cats, birds on the telly. Not sure about stills – he certainly doesn’t recognise pictures on a postcard if you pick one up and show it to him. Presumably he just, correctly, sees some idiot waving a coloured/shaded bit of card in his face. Whether it might be possible to fool him with a big still mounted in a frame, I’m not sure. I think there are eyesight/reliance on motion issues there.)
Are we/S saying something roughly like: other forms of depiction than a perfectly set-up trompe l’o involve adding/learning systematic transformations of that basic form (or cumulatively, to already extant forms)?
Oddly enough, icons in computing in some cases seem (abrupt intro of analogy alert) more like a code than a cypher – i.e. you learn each one rather than learning a transformation method. I’m sure I’ve got to know some icons without knowing what object they are supposed to represent – though generally, curiosity makes one want to work out wtf this shape is supposed to be (a psychological-interpretative project?) so data is unclear.
John Holbo 06.16.11 at 1:59 pm
ben w, the bits I quote from Schier are preliminary stages. I say the rest is twiddling, but there’s quite a bit of it – over 200 pages. So I’m not really managing to do him justice. (I myself am still in the digestion phase.) But here goes:
“Surely “recognize the obejct represented†must mean “recognize actual instances of the object represented—
Yes, pretty much.
“And presumably the recognition in question is not only visual, but a visual recognition based on, well, not a still image, but anyway not based on movement, either the object’s or one’s own: so that if I recognize such-and-such an animal from its close cousins the thus-and-sos by its gait, or peculiar head-bobbing, or whatever, I need not be able to interpret a still image/photograph/painting/etc. of a such-and-such as being of a such-and-such rather than of a thus-and-so.”
Yes, there are going to be lots of qualifications here. There are so many variables to do with how we recognize things in life and in pictures that there is no hope of nailing this down neat. The main thing is the contrast with natural language, I suppose. Imagine if, once you learned the word ‘cat’ for cats you generally ‘got it’ and could go on to call dogs ‘dog’ and so forth. Obviously it doesn’t work that way at all, but pictures work that way. Maybe some styles are weird, or some objects hard to recognize drawn at funky angles, or standing still. Maybe you are from a tribe that has never seen anything like it. Still, once you get it – if you do, and most people do – you can ‘go on in the same way’. Picturing bootstraps up from a general capacity visually to recognize things as being the kinds (or the specific individuals) that they are. Exactly how that works remains to be sorted, but this in itself is important, because it is enough to establish that a lot of ‘picturing is a kind of language or code’ talk is misleading at best.
As to the stuff about ‘any arbitrary member’. He doesn’t want to swear up and down that a simple figure drawing of a smiley face will always work, or that an up-the-nostril, crazy foreshortened view of the face that makes it look like a mountain-range wouldn’t work. He’s just saying: pretty much people get it if you give them a ‘normal’ sort of picture, and then can go on at least to other normal sorts of pictures. Past this point, he does try to deal with twiddly counter-examples, such as you suggest. But I won’t try to go into all that here.
“Suppose then that we have some kind of color-reversing photography—presumably a system of representation—and I have been presented with, and accurately interpreted, an instance of such photography in which the thing (/scene) depicted is entirely black and white, so that what once was white now is black and vice versa. I could probably figure out—recognize, even—what it was, if it was simple enough. But things that I ordinarily can recognize with their ordinary colors might be quite unrecognizable to me.”
Schier spends a lot of time talking about a hypothetical code in which a red tree stands for a green tree. A color-reversed code. He argues, against Goodman, that it’s not possible for red to be iconic for green. That is, it’s not just that we have two different conventions. Green for green, or red for green. The former is just natural in a sense that the latter is not. I was somewhat surprised that he didn’t consider the case of photographic negatives, which are obviously informationally equivalent to developed black and white photos. Yet it is almost impossible to recognize anyone – even very familiar and famous people – in a negative. And, interestingly, scientists who study this can’t really figure out why we have so darn much trouble. It’s a bit of a mystery, I gather, although there are some hypotheses.
Re: some of your later points. The code point. Yes, he has to handle that one. Basically he does it by saying that your ability to work out any further instance of the code is not a more or less direct function of your ability to recognize the thing that the further instance of the code is about. As you say, it isn’t iconic.
Thanks for all the colorful stuff, madhc, Ajay and Chris.
John Holbo 06.16.11 at 2:20 pm
“Not that anyone ‘naturally’ recognizes Marilyn Monroe without having ever seeing her. – I recognise pictures (including movies) of her without having seen her.”
Yes, I obviously flubbed that one in the post. Schier gets it right. It obviously supports his point that sometimes we can recognize unfamiliar people from their pictures, as well as pictures of familiar people. And we can recognize people we’ve never met in one picture on the basis of having seen them in another picture. Again, a disanalogy with language. No one can recognize dogs from ‘dog’ – the word.
“What here is meant by a ‘system of representation’? Does everything that could be called a picture belong to a system of representation called ‘depiction’? Or are we talking about a certain style of depiction?”
Schier thinks once you get one style of picturing, i.e. can go on with it, having your capacity to recognize things triggered, thereby informing you what the picture is of, you are quite likely to get other styles of picturing, but this is obviously no sort of necessary truth. System of representation is mostly an attempt to avoid backing into that obviously wrong implication. It could be a matter of style. It could a lot of things.
“other forms of depiction than a perfectly set-up trompe l’o involve adding/learning systematic transformations of that basic form (or cumulatively, to already extant forms)?”
No, Schier doesn’t say that and it wouldn’t be plausible, I think. The reason we can recognize that a picture of Charlie Brown is a picture of a person – even before we know it is Charlie – is not that we are running some dog-leg from our ability to recognize people in real life, through some sort of implicitly known truths about Albertian perspective, then to Charlie Brown. It’s just that Charlie Brown’s face sets off our facial recognition software. How it does that is an interesting question. Maybe something about Albertian perspective, but that’s highly doubtful. But anyway: it does it. That’s the point. The same thing that causes us to see a face on the person standing in front of us, and in the moon, and in the wallpaper, is the same basic software module that causes us to see one on Charlie Brown. That’s natural generativity. Picturing is a convention that simply exploits it. That’s quite a simple view, in this bare-bones form, but it implies that picturing works differently than many philosophers and theorists (especially literary theorists) have supposed. So it’s worth thinking about.
John Holbo 06.16.11 at 2:29 pm
“Maybe something about Albertian perspective, but that’s highly doubtful.”
In Charlie Brown’s case, anyway. We have highly specialized modules for recognizing faces that have to do with being on a hair-trigger for really basic patterns of light/dark contrast. Nothing Albertian. In other cases, for all I know, our brain is an implicit Albertian. Maybe it’s a technical question for brain scientists. Maybe it’s a bad question because somehow it doesn’t make sense. I don’t really have an opinion.
Tim Wilkinson 06.16.11 at 3:00 pm
Well, I’d guess (rankly) that faces are a special case, and we reconstruct them from minimal cues in a firaly ‘primitive’ way; cf. various wierd things animals try to mate with/attack/get fed by/etc. I’d be very wary of extrapolating from that to pictures of other stuff for which recognitional capacities are not so firmly-wired.
Also the naturalistic visual field + transformations idea doesn’t require that all tranformations – including cumulative progressive stylisations – in conventional depiction are done ‘on-the-fly’ – as one becomes used to/learns a pictorial convention (which may be a very fast process), one might ‘cache’ in increasingly broad vocab of conventional forms (of objects, constituents, mere shapes) linking directly to the concept involved without translating into typical-visual-field-normal-form first. A decyphering mechanism might be used for novel forms, or to resolve ambiguities etc. Not that any of this is likely to be supraliminal(?) of course.
Maybe there’s an analogy (or a closer relation) with a debate about concepts – very roughly, are concepts based on stereotypes and family resemblance, or on definitions and compositional rules in mentalese (or something) – or, more plausibly I’d suppose, a combination and complex interplay of the two.
Tim Wilkinson 06.16.11 at 3:01 pm
Didn’t refresh page before posting
Random lurker 06.16.11 at 4:28 pm
I think that this idea of “natural generativity” is misleading, beacause it trat “pictures” as sorts of units, whereas they are more sort of aggregates.
For example, the picture of a dog, if realistic enough, is composed by the picture of the leg of the dog, the picture of the other leg, the tail etc., all of wich carry a recognizable meaning. On the opposite, the letters “d”, “o” and “g” carry no meaning by themselves.
In this sense, a picture of a dog is more similar to a narrative description that to a word; narrative description are not “arbitrary” in the same sense that words are arbitrary, in fact are more on the “iconic” side.
But if this is true, “pictures” are not more iconic than “language”, we just make the wrong confrontation between complex iconic pictorial discourses and simple “base level” linguistic signs, we could as well confront a biography of Marilyn Monroe to a stick figure and say that the relationship of the biography to the subject is iconic, whereas the relationship of the picture to a human is (almost) purely arbitrary.
Henri Vieuxtemps 06.16.11 at 4:51 pm
I read that it’s either Stalin or Orwell – not clear which one, but one of them is living inside my brain and is messing things up big time.
phosphorious 06.16.11 at 5:41 pm
A point of clarification: is the Holbo/Schier opposed diametrically to Goodman? My dimly remembered reading of “Languages of Art” has me believing that Goodman downplayed or even eliminated the role of ‘resemblance’ in visual art (claiming that even photos don’t resemble their objects in any strong, straightforward sense), and so ‘depiction’ (and ‘mimesis’) is explained in terms of. . . something else? Something more language-like, involving a notational system that has its meaning by convention.
Whereas Holbo/Schier also reject ‘resemblance’ in any simple way, and so explain depiction in terms of. . . something else again? A widely shared, brute faculty of the human mind to connect pictures with objects?
Jed Harris 06.16.11 at 8:13 pm
Schier seems to be me to be describing families of representations have have structure preserving maps from the representations to the (category of) things represented. If you learn the map, then you can take any given instance of a thing or a representation and map it into its “image”.
We need such a map simply to recognize a dog as a dog. Of course the map from an iconic picture of a dog to a dog is different. More generally, different styles of iconic representations will generally have somewhat different maps, but they will share similarities — and will share similarities with the map that allows us to recognize dogs as dogs. (I guess formally these similarities would be defined by a map between the maps.)
This is a pretty deeply developed area in mathematics — it is central to category theory, topology, etc.
Also, as a practical matter there’s lots of technology for learning and using these maps. To pick a recently interesting example, Stanley the robot car (now further developed and replicated by Google) learned on the fly the map that defines “this road”, so it can generalize from local closely observed road, and find “the same kind of road” in more distant views.
So it seems strange to me to discuss the feasibility of non-conventional representations using primarily verbal / intuitive criteria.
John Holbo 06.17.11 at 1:14 am
Random Lurker: “I think that this idea of “natural generativity†is misleading, beacause it trat “pictures†as sorts of units, whereas they are more sort of aggregates.”
Actually, Schier doesn’t. But the preliminary gloss on the position I quoted simplifies the idea that way, by way of introducing it. Then he has to spend pages and chapters grappling with the complications of the cases you anticipate. I think, if you think about it, there’s not really any problem saying that only some elements of a canvas are ‘naturally generative’ of interpretations, in his sense.
Jed Harris: “Schier seems to be me to be describing families of representations have have structure preserving maps from the representations to the (category of) things represented. If you learn the map, then you can take any given instance of a thing or a representation and map it into its “imageâ€.”
Schier has quite a bit to say about maps. They are icons in his sense (mostly, although there are complications. It’s not quite right to say that I can go on to read a map of Russia, if I can read a map of France, if and only if I can perceptually recognize Russia.) One reason he wants to distinguish ‘systems’ of representation is that he doesn’t want to imply, erroneously, that just because you can recognize a smiley face as a face that therefore you can read a map.
But that is not to say that all pictures encode a map of what they are of. It’s a matter of triggering a recognition switch, not necessarily a matter of encoding sufficiency of information. Take facial recognition. It’s on a hair-trigger, as I’ve mentioned, so we get tons of false positives. The moon. Cracks in the walls. Light sockets. We see faces everywhere. All these things don’t encode face maps in any rich sense. They trigger the activation of face maps in the mind. Poverty of stimulus situation.
“We need such a map simply to recognize a dog as a dog.”
Not exactly. We need such a map simply to have the concept of dog. How it is that we visually recognize a dog, given that we have the concept, is a semi-distinct question. Something a lot less than anything map-like might activate the map, causing us to recognize whatever did the activation as a dog. Of course, if this is the case, we will be liable to mistake non-dogs for dogs. But, in fact, that is often the situation. This is important because picturing is a convention/institution that makes effective use of our error-proneness. We see faces everywhere. That’s a bug, a side-effect of our need never to miss an actual face. But picturing makes the bug a feature.
“So it seems strange to me to discuss the feasibility of non-conventional representations using primarily verbal / intuitive criteria.”
But pictures are completely conventional. It’s just that we shouldn’t infer from that that they work ‘by convention’, and that therefore we ought to analogize a system of pictures to a language. The ‘pictures work like language’ analogy is not totally useless but it’s very bad in a lot of ways.
“This is a pretty deeply developed area in mathematics— it is central to category theory, topology, etc.”
I don’t see how this is inconsistent with what I am suggesting.
Chris Bertram 06.17.11 at 7:56 am
bq. But pictures are completely conventional. It’s just that we shouldn’t infer from that that they work ‘by convention’, and that therefore we ought to analogize a system of pictures to a language. The ‘pictures work like language’ analogy is not totally useless but it’s very bad in a lot of ways.
Hmm, yes, but as I’m remarked in another context [property, possession], some aspects of language (assignment of meaning to words) are completely conventional, but language also rests on shared natural capacities and (possibly) universal grammar. In fact, I think there is a range of domains where people move very quickly from noticing the conventional aspects of what is going on to drawing a range of conclusions that they aren’t entitled to draw.
Random lurker 06.17.11 at 8:30 am
@John Holbo 18
I think I didn’t explain myself very well: in my opinion pictures can be “iconic” (wich seems to me more or less the same than “mimetic”) only because they are aggregates.
So for example a stick-man is iconic only because we can disaggregate it in a head, an arm, another arm etc.
The singular arm of the stick-figure is simply a stick, but it “represents” an arm more or less arbitrarily, like a word. It is true that nobody could recognize it as an “arm” out of the contest of the stick-man, but only leads to a polysemic interpretation of the line.
Also, “But pictures are completely conventional”. This is a common assertion but it seems misleading too to me: perspective, for example, is not conventional but an actual optical law, so that phtos respect perspective. In facts, most drawings resemble reality very few, but lookers are usually trained to recognize what traits are pertinent and what are not (in a culturally determined way), but this is a completely different thing that being conventional IMHO.
Z 06.17.11 at 9:18 am
r perhaps comments on Nelson Goodman’s “Star-Making†paper, and responses thereto.
And then, a transition to The Power of Speech perhaps, since we are talking about Star-Making. Damn, why is Poe everywhere in this thread?
Random lurker 06.17.11 at 9:23 am
phtos was “photos”.
I’ll add a better example that just came to my mind:
I can draw a picture of a dog with wings, that most people would recognize as “a dog with wings”.
The definition you gave for natural generativity is:
“If you can recognize a dog, when you see one, and a picture of a dog, when you see one, then (ceteris paribus) you can recognize a picture of a cat – or of a candle, a house, a tree – provided you can recognize those things themselves. Schier calls this ‘natural generativity’. ”
However, dogs with wings do not exist in reality, so that people cannot “recognize” a dog with wings in reality (though people can conceive a dog with wings,I take “recognize” as meaning “referring to an already estabilished meaning”). In this sense this would not be a case of “natural generativity”.
In my opinion, people can recognize a dog with wings only because they can recognize separately the “dog” and the “wings” in the picture and we can suppose that this process is repeated up to the smallest detail, at wich point we only have “lines” and “shapes” or “colours” (not symbols).
Tom Hurka 06.17.11 at 10:22 am
Flint Schier was a friend and grad school contemporary of mine from mid-1970s Oxford. He died, as many tragically did, in the 1980s. At the time he was writing an aesthetics book for Penguin that would have been terrific; its draft first chapter, published in a memorial volume for him called Virtue and Taste, is a delight.
Also wonderful, if you can find it, is a review for the NY Times Book Review of Roland Barthes’s book about clothes, The Fashion System. At once scathing and hilarious.
John Holbo 06.18.11 at 1:06 am
Thanks for that appreciation, Tom. I knew he died young, didn’t know anything about that. I have read the Barthes review and it is great.
Quick response to Random Lurker on ‘convention’. The trouble is, as I am sure you agree, that the term is elastic. There is a perfectly good sense in which Albertian perspective rules are conventional. They are human-made. You don’t have to make pictures according to them. You can make them some other way. If you want to explain why a given picture follows Albertian rules, it is perfectly reasonable to point to the artist’s culture and community and education and so forth. That said, there is a lot to be said for calling them natural, as you say. (You appreciate that, I can tell, so I won’t go into it.) My point is this. The best argumentative strategy, if you see the natural stuff, is not to deny that what the convention-minded theorist is saying is true. Read Nelson Goodman on Albertian rules and everything he will say is, strictly true (well, his foot might slip, but the things he tends to say tend to be narrowly defensible. He’s no fool or philistine about art.) But it just doesn’t follow that Albertian rules can’t be ‘natural’, in a perfectly good sense. You have to get to the point where you can say, as I do, ‘pictures are totally conventional’ and yet are prepared to acknowledge that the workings of the thing are ‘natural’. You have to see that those thoughts are no more incompatible than ‘the ball is round’ and ‘the ball is red all over’.
Martin Bento 06.18.11 at 6:16 am
“As Nelson Goodman always points out, pictures don’t much resemble the things they are of – they are flatter, often much smaller, etc.â€
I don’t know what’s hiding in the etc., but if “flatness†and “size†are seen as huge problems, it makes me think this man has not given much thought to how recognition works out in the world of real objects. Within huge parameters, size has to be disregarded because apparent size is so strongly a function of distance. The man who thinks “That can’t be my wife up the street. She is no taller than the length of my hand†is suffering some cognitive defect. As for flatness, stereoscopic vision is overrated. Close one eye and look at your dog. Look familiar? Any thought she might be your neighbor’s dog? Or a buffalo? Anything you think you would recognize with both eyes open that you wouldn’t if you closed one? Recognition does not rely on depth perception. Our vision is quite close to two-dimensional anyway, which is why one-eyed people have only a minor handicap. Most of our perception of depth is based on an intuitive understanding of how light interacts with three-dimensional objects, which is why two-dimensional images can get pretty close on the basis of modeling.
And if you want to know what is natural about understanding pictures, I think mirrors a good place to start, since they are more or less natural pictures. Images in a mirror are not a language, right (If they’re a language, who is speaking them)? There may be some issues with recognizing ourselves initially, since we don’t see our own faces directly, but otherwise the issue of recognition is straightforward, as the resemblance is so close, it is easy to get confused and mistake mirrored images for actual objects. Mirrors flatten and can distort size (especially if the reflecting surface is not flat) without impairing recognition, so we needn’t let those factors detain us, nor any others that do not prevent recognition in mirrors, including accidental mirrors – still lakes, store windows, cop glasses, etc. To the extent recognition of images in pictures operates the way it does in mirrors, it can be “cashed out†as resemblance, as what other currency could mirrors offer? That I think should be an important test, whatever you are asserting about recognition in pictures, would it apply to images in mirrors? If so, it is not fundamentally a convention. Ironically, the only thing I can think of that is difficult to recognize in a mirror is language; it is hard to read with left and right reversed, at least in English and in many other languages.
As for the blank Marilyn canvas, should we really consider such an example without looking at its (lack of) plausibility? If we do not observe people making such “recognitions†with no conceivable element of resemblance present (at least no more so of Marilyn than any other person, or any other “white†person, if you want to go there) and no convention either, is this not evidence that recognition is not in fact as arbitrary as this? If we try to find a case of recognition that is based neither on resemblance, nor on convention, do we find anything? We can stipulate one, as here, but it is implausible to us.
Martin Bento 06.18.11 at 7:29 am
Then there is the question of stylization. We see a bunch of cross-hatched lines on an ink drawing and interpret shadows. We can see the individual lines, which are not very shadow-like, but the denser the areas of lines, the deeper the shadow suggested. This gets into that area you mentioned of things simultaneously natural and conventional. Natural in the sense that the thicker areas of lines are darker, on average, as more deeply-shaded areas are darker, but conventional as we have to see this average density effect rather than just seeing the individual lines (which we always clearly see, and which provide a lot of the artistry of this approach). Even here though, it is worth keeping in mind that our natural recognitive apparatus has to be highly tolerant of variations: things look different in different lighting, from different angles, in different states, etc., and sometimes we can recognize them nonetheless. Even in forming a mental model of a specific entity, we have to generalize from specific impressions.
Jed Harris 06.18.11 at 6:49 pm
Thanks to John Holbo for the thoughtful response. Unfortunately we are largely talking past each other, but in respectful and perhaps somewhat illuminating ways.
I was using “map” as a mathematical term of art, and obviously very wrongly did not flag it as such or explicate it at all. Interestingly, there isn’t (as far as I can tell) a formal definition that neatly encompasses the meaning of this term of art. Wikipedia has a list of all the variant uses.
Trying to roughly get at the core concept, a (mathematical) map is a relationship between two or more “objects” (often many more) that picks out certain aspects of each object and “maps them into” the corresponding aspects of the other objects. The aspects it picks out define what makes each object “mappable” as one of that kind of object, and the map also defines how we can transform the aspects from one object to another so they are “equivalent”.
Let’s use this concept to reinterpret our exchange. John says “We need such a map simply to have the concept of dog. How it is that we visually recognize a dog, given that we have the concept, is a semi-distinct question.” A given concept of “dog” just is this sort of map, picking out the aspects that let us recognize it as a dog, and letting us transform those aspects between (potentially very different) dogs. Typically these maps are impossible to express in language, though we have gradually developed ways of capturing them.
A perhaps suggestive and accessible example of maps in this sense is D’Arcy Thompson’s transformations between biological forms in On Growth and Form.
Regarding pictures being “completely conventional” — I guess I don’t understand what this means — perhaps it is a phrase of art? Consider the example of medical X-ray photographs. People typically learn to interpret these after they are used to ordinary visual processing, recognizing pictures, etc. Initially they experience them as incomprehensible, but after repeated exposure they start to “see” tumors, occluded arteries, etc. However I think it would be an abuse of language to call these photographs “completely conventional” because they are produced by quite a simple physical process, in which the geometry and x-ray absorption of the body is simply projected onto a 2-D surface.
This process of learning to understand x-rays is a pretty good example of learning a map. How is it different from learning to understand a given genre of pictures?
Maybe more later.
Stephen Lathrop 06.18.11 at 9:24 pm
Not sure I have seen much that addresses the nature of pictures here. The nature of pictures is that if you are talking about them, you have changed the subject.
John Holbo 06.19.11 at 8:07 am
Stephen Lathrop: “The nature of pictures is that if you are talking about them, you have changed the subject.”
If so, then how can you be not sure there is much here addressing their nature, Stephen? Doesn’t the denial of that possibility follow necessarily from the fact that we are talking about them? (On your view?)
Jed Harris: “Regarding pictures being “completely conventionalâ€â€”I guess I don’t understand what this means—perhaps it is a phrase of art?”
Yes, of course. But I don’t think my use of it is any further from ordinary usage than any other philosophical usage is likely to be. Yours, for example. I am not opposed to a use of ‘convention’ on which x-ray photographs are not deemed conventional. But, somewhat for rhetorical purposes, I want just to grant the full conventionality even of X-rays, just to prove the very important point that the sorts of things that are generally thought to follow from this concession don’t necessarily follow from it. I think this point of view may be more instructive and compelling to the contructivist conventionalist than the one you propose. (But maybe I’m wrong about that.) The case of x-rays is complicated because in a lot of ways the institution of imaging piggy-backs on our conventions of picture making. We like to see the information presented on a nice two-dimensional surface. This way of presenting information suits us. It’s hard to sort out to what degree that suiting relation is a function of nature, or of convention – i.e. a kind of path dependence. Roughly I think things are conventional if they are 1) reproduced 2) by the weight of precedent. Now in a lot of cases precedent is mixed with a high degree of functionality. Yet it is functionality that could be performed some other way. (Think of the hammer case.) Think of a different case. Bird tracks. Are they ‘pictures’ of bird feet. Yes. And no. If we take them as such, we are in effect assimilating a natural phenomenon with our institutions of picture making. X-rays are a bit like bird tracks. So, to repeat, I don’t want to say your view is wrong, but it’s complicated.
Putting it another way, you say x-rays shouldn’t be regarded as conventional because they are “produced by quite a simple physical process”. But that’s confusing, because all painting is produced by a simple physical process. Namely, putting paint on canvas. And marriages are performed by simple vocal processes, which are characterizable in objective acoustic terms. Of course, a lot of human activity needs to be going on behind the scenes for this to come about, but that’s true of x-rays, too.
Random Lurker 06.19.11 at 11:38 am
@John Holbo 24&29
I appreciate your answer, however I disagree on your use of “convention”.
I believe that when I say “X has a conventional meaning”, I mean that there is a culturally estabilished meaning Y. In this sense, your use of the term “conventional” is so broad that it could mean “whatever the meaning Y that we infer from X, it is influenced by culture in some unspecified way”. But since we are more or less influenced by culture in everityng we do, this statement would be almost void of content.
Also, I took for granted that we were using the term “pictures” meaning someting like paintings, comics, or photos in a newspaper. I see that other commenters refer also to mirrors, and in fact also the word “dog” is a picture as long as it is written.
I think that we should differentiate the discourse in three branches:
– First there is a perceptive “skill” that makes us “see” pictures, wheter artificial or not
– Second, with regard to artificial pictures that we see as “texts”, we procede to an interpretation that uses a logic that would be better explained by concepts such as “model author” and “pertinence” (as in liguistic acts). This interpretation uses cultural parameters, for example when we read a black and white comic we take for granted that the world represented is not black and white, but that the “monocolourness” of the pictures is not pertinent to the interpretation of the pictures. At this level however pictures are not conventional, culture just direct interpretation away from “non-pertinent” aspects (it gives us the rules to interpret the “map” as explained by Harris@27).
– Third, as an historical process, the level of abstraction implied by the “de-pertinentization” of pictures can lead to the creation of actual symbols, as letters of the alphabet or traffic signs (but also very stylized representation of a eye), that are completely conventional, but could be used as parts of a more complex non-exclusively-conventional picture.
Stephen Lathrop 06.19.11 at 12:56 pm
John Holbo: “If so, then how can you be not sure there is much here addressing their nature, Stephen? Doesn’t the denial of that possibility follow necessarily from the fact that we are talking about them? (On your view?)”
Is there some way you could say that in other words? I’m not following your questions.
John Holbo 06.19.11 at 2:05 pm
Stephen, you take it to be that talk about pictures fails to be about pictures. (A standard view, to be sure. Good old dancing about architecture.) I was merely noting that it was somewhat odd, then, that you were ‘not sure’ I had said much about pictures. It’s like being not sure that 2 + 2 = 5. Nothing wrong with it. I’m not sure 2 + 2 = 5 myself. I just thought it was a bit funny.
Henri Vieuxtemps 06.19.11 at 7:48 pm
I read in a Dennett’s book that there is a theory, convincing theory, apparently, that the brain has evolved a mechanism to react to any kind of vertical axis symmetry. Because in nature, seeing vertical axis symmetry is likely to indicate that there’s another animal facing you, and you need to make a quick decision of which one of the four Fs (fighting, fleeing, feeding, or mating) you want to do now. That’s not conventional, that’s hardwired.
Martin Bento 06.20.11 at 12:03 am
I like “mating” as an “F” – that’s a bit sly.
John Holbo 06.20.11 at 3:20 am
“That’s not conventional, that’s hardwired.”
Yes, my personal view is that the way to think about picturing is as a conventional exploitation of the human propensity for visual recognitional false-positives. That’s the short version. Of course, most canonical pictures are somewhat more complicated cases: visual recognitional false-positives that do not result in actual perceptual/judgmental error. Our picturing institutional/conventions/practices depend on the fact that we often spontaneously recognize things ‘as Marilyn Monroe’ without actually being fooled into thinking Marilyn really is there. One part of the brain is saying ‘face!’ while a lot of other parts of the brain are simultaneously recognizing that it’s just a piece of paper, etc.
It’s not right to say ‘that’s not conventional’, however, just because it’s hardwired. It can be both. The rules of English grammar are conventional/arbitrary/paradigm case of path dependence. Nevertheless these conventions ride on a lot of hardwired human capacities.
Comments on this entry are closed.