IMO it's underappreciated how these mistakes are not a bug of human intelligence, but a feature. They reveal a default non-literal interpretation of the world, which can be made less abstract when needed. However, it's the default abstractness that prevents us from making gross contextual errors based upon a literal interpretation.
This feature really is a bug for Chinese speakers. The same phenomenon as people being unable to accurately draw a bicycle, an object you're pretty familiar with, is called "character amnesia" when you're unable to write a Chinese character that you probably see every day and can read and understand with no problems. The meaning of the character is definitely abstracted away somewhere in your brain, but there might be hundreds or thousands of characters you're simply unable to reproduce on paper.
The Chinese characters are composed of a specific order of strokes, and sometimes it's like you can't bring a clear enough picture of it into your mind's eye to be able to reverse engineer it with a pen on paper. I probably experience this a lot more as a learner of Chinese as a foreign language, so I'm pretty familiar with the feeling of "Character amnesia". It happened to me the other day with 牙 (tooth) a pretty simple character that you'd think would be easy to remember. Once you get the first two to or so strokes down though muscle memory seems to take over and you finish it almost subconsciously.
From anecdotal examples, this is actually pretty common, and in mainland Chinese sources I've read they seem to put it down to people using pinyin (romanised pronunciation) input on phone and PC keyboards instead which gets people out of the habit of remembering stroke orders.
If you ask me to draw an ampersand and a G-Clef (𝄞), I'll probably render more or less the same thing.
Possibly because in childhood I've drawn several G-Clefs but never ampersands (that I recall), while the opposite has been true later on. I suppose I've just re-purposed whatever motor memory I had for "a kind of infinite-loop G".
Is your experience similar to spelling errors in written English?
I choose English because it's the primary (only?) language for this site, and while there are definite rules for the correct sequence of letters for a word, the rules for the language overall are not especially consistent. And it seems to me that English has a bazillion words in common use.
I have from time to time experienced that very weird sensation, as I look at an English word I've used in written language for fifty years, and it suddenly for a time it seems totally wrong and incomprehensible. I recall staring at the word "Our", thinking it could not possibly mean anything in English. This is rare, and doesn't last long. So far.
But it gives me some insight into aphasia, perhaps.
And we need spell-checkers for our writing, because in a stream of words we have written by pushing buttons, the brain elides the misplaced letters, it's almost impossible to spot them all, much less work to simply hand to someone else to read it fresh, then the errors seem obvious to them.
But it's the temporary, isolated dyslexia I find bizarre.
Yeah, you got it right, I'm a native English speaker. I wouldn't say the first-person experience is entirely the same as forgetting a spelling, when I misspell a word I usually get most of it right and mix one or two letters up like piece/peice.
With a Chinese character, it's like you can't get the first stroke down and you're unable to reproduce anything. But once you've worked out the first two or three strokes, maybe by getting a dictionary, for me at least muscle memory seems to just take over and you can kind of just finish it automatically.
It's interesting though, these little quirks definitely give some insight into how the brain's underlying OS works. Funnily enough talking about dyslexia I often wonder if Chinese dyslexia and English dyslexia is the same condition, or if they're both completely different. Would an English kid be as dyslexic in Chinese if they were bilingual, or would they find reading one language easier?
Having fixed my bike so many times over the years, I can easily sketch a correct bike - if not in terms of proportions (at least without any corrections), but in how the frame is constructed and how all the moving parts come together. I guess someone who doesn't need this knowledge will just use a more abstract model.
I, too, can correctly draw a bike. I think anyone who's spent any time fixing bikes will be able to do this; if you know that it's called a "diamond framed bike", and you know why it's called that, it'd be pretty hard to not know how to draw it.
I've always wondered -- is there something I can read/see that compares the drawings from savants like Stephen to actual photographs? I'd like to dig into the specifics.
When I tried drawing portraits I really had to relearn how to look. In my mind eyes were positioned at the top of the head while in reality they are in the middle.
But when you draw cartoons it looks a kind if silly when eyes are drawn at the accurate location. And maybe this is the feature you mentioned. It's fun to play with it.
This has been discussed at least a couple of times before but maybe I have a new angle:
Somewhere during the last three years or so I realized I'm one of those people who can hardly see mental images. Even my dreams seems to often be just feelings og having been somewhere and done something.
However I can draw bikes my house and everything else that I know well (and I often have too draw things since I cannot keep mental models in my head.)
Why is this, how can I draw something I cannot project in my head? Is it just because I understand it so well that if I try to draw it wrong my mind hurts?
PS: My drawings are not nice, but they are somewhat correct, the frame makes sense, the chain is in the correct place etc etc.
PPS: I keep wondering if I always was like this or if I lost it at some time (if I lost it my best guess is there's a mild form of something like PTSD because of a bunch of stuff that happened from I was 15 to 25).
Maybe I'm stating the obvious, but a bicycle has a more complicated (or specific) structure than most people realize that they find hard to reproduce when sketching it.
I think "mental modeling" is actually the problem in that case. Our mental model of a bicycle is "wheels; pedals; handlebar; (sometimes) chain". To a casual observer, "fork" and "frame" is background information that they don't think about or notice but those are obviously pretty fundamental.
So when you sit down to sketch it from memory you suddenly realize that you're not exactly sure how those parts fit together. Think about how often people attach the pedals directly to a wheel in a sketch, for example. That's 100% reflecting our mental model ("to turn the wheel I turn the pedals").
"Sketch a bike" is almost like a test of "can you invent the diamond-shaped frame off the top your head?" Because of our mental model we don't even notice that part.
> "Sketch a bike" is almost like a test of "can you invent the diamond-shaped frame off the top your head?" Because of our mental model we don't even notice that part.
Distracting from your point, but there some awesome reconstructed diamonds floating around today. The two below are the first examples that sprang to mind because of how they invert the "missing" structure.
Oof. The cantilevered rear wheel in the first link in particular makes me really uncomfortable for some reason.
I don't mind the "n-shaped" frame overall but something about the chunky frame combined with the rear wheel sticking off the back makes me think it's going to snap off - painfully for the rider - on the first pothole or hop.
I didn't realize that that the upper bar between the rear wheel and the saddle was so important to my psychological comfort, but I now I know. Even one of those skinny bars like the ones on old-school Schwinn road bikes would make me feel so much better about that bike.
Wait. Is there a bar between the bottom-bracket and the rear wheel on the non-drive-train side of the bike, or is it actually cantilevered in two dimensions?
> Is there a bar between the bottom-bracket and the rear wheel on the non-drive-train side of the bike, or is it actually cantilevered in two dimensions?
There is no seatstay on the unseen side, but the Ventum is at least symmetrical. My boss' TT bike is similar in design to Cervelo in the second link, but with an oversized chainstay on the non-drive side for claimed aerodynamic effect. It looks fine with a disc in, however with a 5-spoke it looks dangerously pre-broken to my eyes.
I understand feeling uncomfortable with the design, single-sided front "forks" are my personal cut-off for that queasy feeling.
Plus, if you've not seen TT bikes in person before, be prepared for weird buffeting sounds from the disc wheels the first time you do. They definitely won't make you feel more comfortable about them ;)
Read up materials for beginner artists. The thing they emphasis is that people by default dont know how things look like and dont remember that. When you are drawing from observation, you are supposed to look a lot and draw a little, because it is normal that your brain will throw away that information. The set of things you actually know how it looks like is called visual library and you are supposed to consciously working on that and making it larger. The general assumption seems to be that beginner has small to empty visual library.
The "seeing through artist eye" is basically about that - training to see things and tricks to remember how things look like. It can get better with practice.
By default, untrained people tend to draw symbols of things they see - how the thing is "supposed" to look like.
Quite likely, there is absolutely nothing wrong with you, you just don't have the particular training, just like most people, which is actually perfectly fine for most of us.
I only recently (last few years) discovered that aphantasia is a thing. Before that I thought that when people were talking about "mind's eye" they were using a figure of speech that was totally unrelated to being able to create images in your head. I am a decent artist. I can sketch out the layout of a building after having been in it a short time, including identifying exits, etc. (maybe I should thank my CIA dad for that one) But I still can't recall images on demand. It's very strange. I always wondered if it was related to whatever brain malformation that caused my epilepsy. I also wondered if the genetics that caused my (admittedly self diagnosed) aphantasia was related to my brother's ability synesthesia. The brain is remarkable.
This is very interesting, but I do think that there's a big distinction between drawing and imagining, so I don't think it's too surprising that a person may be better at one than the other.
Indeed, I suspect I am like you and am better at drawing a bicycle than imagining one. With drawing it, I simply know how to do it. There's a sequence of steps (draw a diamond shape, add the seat post, add the wheels and the front fork). I don't even need to keep the whole thing in my head while I'm drawing it -- the paper acts as the repository.
On the flip side, though, there are plenty of things I can imagine that I cannot for the life of me draw. Faces of family members. Animals.
But again, I think that artists who can draw life-like representations are ones who do not use their imagination, and, in fact, learning to draw from life is a process of learning not to use your imagination. When we draw from our imagination, we tend to use our mental models, which are abstract and functional. We put eyes at the top of the heads, because that's how we imagine them, because we focus on the face and all that wide-open forehead space is useless. It's a skill to not use those models when trying to draw accurately.
> On the flip side, though, there are plenty of things I can imagine that I cannot for the life of me draw. Faces of family members. Animals.
Lucky you :-)
Seems I can only recall photographs of my mom and dad when I try now (passport style photo og my mom and a photo of my dad making a funny face while carrying an oversize chocolate).
Same kind of goes for my wife: I immediately recognize her now but the first few months I was dating her I was afraid I would not recognize her :-D
Edit: and when I try to view my wife in my head I only see a 15 year old passport style photo :-D
> However I can draw bikes my house and everything else that I know well (and I often have too draw things since I cannot keep mental models in my head.)
This is quite similar to watching a movie and how you can instantly tell a CG human is fake. Because you know how humans look, walk, move, etc. Whereas if it's a CG plant or animal most of us might not notice it's fake because we don't live our whole lives observing them. An expert at elephants can probably see a CG elephant and notice the wrong things much faster because they know it well.
For myself, I can draw from my imagination but it's not that I see the image in my head. It's quite often being a little less present or concerned with the actual drawing while still having the muscle memory of drawing the common shapes that make up mass and form.
I would recommend starting off making marks on paper and seeing what comes out of it. A lot of cartoonists practice drawing an odd-shaped oval and then using that as the basis for the shape of the head.
I always wondered whether a good way to describe Aphantasia to people who don't have it is to ask them to "imagine" an odor. Most people can recognize specific oders effortlessly and remember them for decades (example: you smell some brand of aftershave, and you immediately know that this is how your great-grandfather always smelled, who has been dead for 20 years), but you cannot "think" of an odor and actually smell it the same way you can think of an apple and see and inspect its image inside your head. Yet you "know" what gasoline smells like, or fresh bread.
This leads to another interesting question: are there people who cannot imagine music and other sounds, like speech? Is this related to Aphantasia?
A question even more interesting: are there people who can imagine odors, or combinations of them? I would expect this to be very handy if you are designing perfume, for example.
Wow, I never thought about that, didn't realise that was a thing. I can only so very vaguely imagine what things would smell like that I wonder if I really am at all.
Music, sounds and voices have always been the most clear for me (I can think through some of my favourite orchestral pieces with multiple instruments at once), but I've never been as good at visual.
I can imagine odors decently. I didn't know this until I just tried it after reading your comment. But yeah, I just imagined what a jasmine bush smells like, my dog's poop, strawberry scent car freshener, baby powder, and distant brush fire.
My best guess is that I'm imagining them at something like 30% fidelity, and some are harder than others to synthesize in "my mind's nose"
It's not nearly as good as what I can visualize in my mind's eye, so your analogy is still helpful.
Yeah, I think I have a relatively poor visual imagination, but when I try to picture a specific object it's probably ~10% as vivid as actually glancing at it. When I try to recall a smell, it's probably more like 20% as vivid as smelling it.
I can definitely think of odors and smell them, certainly more vividly than I can see mental images.
Thinking of fresh bread nearly makes me wonder whether someone is baking it nearby. Sounds are similar for me. I can play most songs in my head with full orchestration, and pick out different instruments. After hearing my alarm and turning it off, I will sometimes think of it later and have to check to make sure it isn't playing again.
Images though, require sustained effort for me to maintain. I would never ever mistake a mental image for a real image - they are completely on completely different "screens" if you will.
Yes, I think so, it is so strong the first time I tried to take the test I was somewhat surprised that you were supposed to be able to do all these things.
I'll add the article to my reading list immediately :-)
Edit: yep, read a few lines, very similar and interesting. Haven't seen this article before, thanks.
There's a youtube channel called Psych of Play (by Daryl Talks Games) which has a video[0] about an ad for a Tetris game released in 2017, and the video goes on to explain some hypotheses about the ability to have stored knowledge being separate from the ability to recall it. Some links are made with sleep, too. There's also an example of a person with both retrograde and anterograde amnesia who has no idea who you are after a few seconds, and no idea how he knows the piano, but can play it anytime he wants.
There's a lot about how memories are formed and recalled that is still the subject of much research.
>Little I knew this is actually a test that psychologists use to demonstrate how our brain sometimes tricks us into thinking we know something even though we don’t.
They probably don't pose the question like this, though. I suspect you'd get different results if you asked a person, "Can you draw an accurate representation of a bicycle?" vs pestering them to draw a bicycle so you can make fun of them later. I know for myself, if someone handed me a sheet of paper and said draw a bike, I'd do it for fun, but it would not be correct.
I don't think "how to draw a bicycle" is the thing that your brain tricks into thinking that you know. Instead, it's "what does a bicycle look like?"
If you ask most people if they know what a bicycle looks like, I think they'd tell you, yes, definitely. But really they don't. They know that it has a couple wheels and some sort of metal tubes, and that's a workable mental model to convince yourself that you know what a bicycle looks like.
People can recognize bicycle from non-bicycle perfectly well. If you would make bicycle with unusually large wheels, people would recognize that perfectly. In particular, telling bike from random wheels and tubes is easy.
So yeah, people do know how bicycle looks like.
What people do not know is how to draw bicycle or any other thing. Unless artists, they don't know to abstract important details from unimportant and they are complete crap at proportions.
For fun, ask people to draw face. Everything will be at wrong place with wrong size, but I still would not say people don't know how human face looks like.
>For fun, ask people to draw face. Everything will be at wrong place with wrong size
That’s kind of an odd counter example to use. Most people will certainly draw an anatomically correct face (for all intents and purposes) with two eyes, a nose, and a mouth. And you’re unlikely to see anyone add three eyes, turn the nose sideways, or move the mouth above the nose.
In contrast, this is _exactly_ what people are doing to bikes, as illustrated by the ridiculously funny renders in the article.
> two eyes, a nose, and a mouth. And you’re unlikely to see anyone add three eyes, turn the nose sideways, or move the mouth above the nose.
That is not enough for realistic face. In the first place, people wont draw these on the right places and with the right relative size. While the eyes will be higher then mouth, they still wont be on the right place and it will be extremely visible. The nose can be sideways, it sticks out and thus is harder to draw.
Also, the beginners nose wont look like actual nose, eyes wont look like actual eyes and so on.
It will be kinda like those bikes. Note that none of those bikes have three wheels, two seats or four handles. All of them have seat around where back wheel is, all of them have handles. There is one bike that have back seat small and one that have pedals under seat. These are the worst proportions and relative position wise, but some people will do stuff like that with human face too.
When they teach you to draw faces, they teach you literal guidelines to get thing at the right places and sizes. You are then supposed to erase the guidelines. It takes effort to learn for most of us.
In experiments measuring perceptual thresholds (say of motion, color, or size), it is common that subjects say they can’t see a difference anymore between movement speeds/colors/object sizes, but, when pressed to answer the question, ‘guess’ right way above chance level.
I think it’s likely this is similar: they know, but not consciously. To test that, one could ask them whether the drawing they made represents a bicycle. I would guess many of them could say there was something funny with their own drawing, even though they can’t make a better one.
(One can have a philosophical discussion about whether that means that they know what a bicycle looks like, but they sure can act as if they know)
Yeah, if simply asked to draw a bicycle most people are probably happy to just sketch something that represents the concept well enough to, say, work as a Pictionary clue. Would be interesting to see the results of being asked to draw as accurate a picture as possible. And not just the results, but the drawing process itself—I'd wager that having to think about it would elicit more "wait a minute, this doesn't go here after all" type realizations.
On a normal bicycle when you lean off the vertical axis, a compensation is made by turning the front wheel towards the "inside" of the lean thus bringing the bicycle again closer to the vertical. i.e. the bicycle (with the help of your steering input) turns the front wheel to catch your fall. If you lean without steering you will simply fall over to the side immediately! Don't try to do this by the way =)
This is the same principle of balancing an upright broom vertically, handle first, on the palm of your hand. When the broom falls out of balance, you simply move your hand in the same direction to recover the center of balance. A bicycle is balanced by exactly the same principle.
The rear wheel would likely shear upwards along the seat tube when the rider sits on it. I suppose with enough strength it could work, would be quite a heavy bicycle.
Riding a bicycle presupposes movement and control. To accomplish the act of riding a bicycle you must be able to steer otherwise you will just fall over to the side and you will not be riding it anymore =)
Your link was about what's needed for bicycle to be self-stable. That steering is required to avoid falling over (as opposed to, say, keeping balance by shifting your weight slightly to one side or the other as needed) is an experimental result and not baked into definition of "not having a steering wheel", even if it is a physical consequence of not having it.
If you showed some people a sketch of something like a Ventum triathlon bike they would probably say it was wrong or fake. The shape of bikes has been held back for too long by tradition and outdated UCI race equipment rules.
Or the safety bicycle design is so perfect that any attempt to “improve” upon it winds up coming across more as an art project rather than a utilitarian upgrade.
In the case of standard racing bicycles, it really is rules rather than tech.
There are different formulae for road cycling (governed by Union Cycliste Internationale), triathlon bikes (International Triathlon Union), and "human powered vehicles" (International Human Powered Vehicle Association), typically recumbent and with fairings, and others.
Top speed (undrafted) for HPV over a timed one mile is 78 MPH.
If utilitarianism optimises for local travel, the biggest difference I've seen is a curved top-tube on omafiets. If it optimises for moving rider and luggage from A-to-B, touring bikes are pretty standard.
Fortunately, mainstream bikes have been decoupled from the UCI rules, probably since the introduction of the mountain bike. Most bikes sold today are things like cruisers, hybrids, city bikes, and low tech mountain bikes, that have no corresponding category in competitive cycling.
Designs are changing, but gradually. Most kids bikes now have non trapezoidal frame designs. But metal tubing is still a pretty compelling material to make a mainstream bike from.
Dunno, doesn't look that special to me. But I've had quite a bunch of different bikes (trials/mtb/bmx/cruiser/race/'normal' day to day) and am from a cylcing-heavy country so I'm used to seeing many different shapes.
I didn't claim it's special. Just that it differs significantly from what the majority of people think of as a "bicycle". If you ask them to sketch a bike it won't look like that.
Having worked in a bike shop for a bunch of years I'm really good at drawing a bike, but recently I tried to draw the outside of my house. The front of the house that I see nearly every day. My house that I've been living in for over a decade. I couldn't do it!
I'm the same.I can draw shoes,houses, spaceship and etc., however I can't draw familiar objects. I think this is partially because I wouldn't be able to describe familiar faces either,even though I see them on daily basis. It's weird how the brain works: I remember the email with some API specs that was sent to me 4 years ago but I struggle to remember what we spoke about over the dinner just a day before.. Would be interesting to see if anyone knows why it's like this.
What blew me away is that the author is using 2d vector graphics to draw all that. I think there might've been a follow up post where he shows the process.
I didn't know what part was missing from the first bike and it was a little frustrating the author didn't explain! I have an idea now from comparing to real bikes, but I am not confident that this bike would "immediately break." From forces on the rear wheel?
It's missing the horizontal struts between the pedals and the back axle. Those make strong triangle shapes, and without them the back fork is going to be quite flimsy and unstable. The person's weight on the seat would put a lot of torque on the welds connecting the back fork struts to the frame below the seat.
It's kind of hard to draw a chain and chainstays clearly. That seems to be one of the common missing pieces. They're in a similar region of the bicycle.
I wonder what else seems simple but is actually hard to draw from memory. My example is the simplest knot, although the difficulty seems a bit different from that of the bike.
I would guess the tricky thing with bikes are all the similar looking metal bars at different angles. It’s hard to perceive the overall shape of the bike and remember how the pieces are connected.
So maybe other similar things that are assembled from several parts would be hard: a folding chair, a pair of scissors, a suspension bridge, the Eiffel Tower, a catapult, etc.
Some of these are gorgeous! Hilariously impractical, but gorgeous. Nice job. Shame it's rendered; I was kinda hoping they actually made these bikes. Although I suspect they'd be too dangerous to try them out.
You are heavily punished for this comment but instead of giving you another flag or downvote I hope there is a learning opportunity here:
Think if you met someone who said any of these things while you were listening:
- There are adults who cannot fish!
- There are adults who cannot swim!
- There are adults who cannot milk a cow if their life depends on it!
- There are adults who cannot tell you 6 teams from premier league
- There are adults who cannot tell you the rules of baseball, basketball, football
- There are adults who cannot say even half of the American states or capitals
- There are adults who cannot tell you which countries are members of the European Union
- There are adults who cannot drive a car
- etc
It is frightening!
Or not. We are just different. And for the record, more than two of the bullet points above covers me, a highly successful software engineer :-)
Not everyone has the same skillset, the same background etc.
Once we learn to respect others for what they are not for what they are not, we can make progress.
I recommend you edit it and I'll try to delete or empty my comment afterwards :-)
Edit: I see two hours have passed now so it is too late anyway. My advice still stands for another time: People can be excellent in most areas of live and still struggle mightily in others, at least that is my understanding.
I am a professional artist; drawing stuff from out of my head and making it look plausible, if not actually correct, is literally my day job.
I have spent multiple decades practicing to be able to do this. Everything that I can draw out of my head and make look plausible rests on top of a solid foundation of spending time looking at real-world examples of the thing, or photos of the thing, and working from them.
There is a huge hump that artists almost inevitably have to get over somewhere in the transition from "amateur" to "pro": amateurs almost inevitably convince themselves that they need to be able to draw anything with zero reference, and if they can't somehow magically do it they are a failure; pros will happily draw a lot of stuff with no reference, but the instant they realize they don't know what a thing looks like, they will put down their pencil and go find something to work from.
Part of being a pro artist used to be maintaining a "morgue" - you would be constantly on the lookout for photos of things you thought you'd have to draw someday, and you'd clip and file them. The advent of image search has made this less necessary, but I certainly still have a few shelves full of coffee table books full of really nice photos of distinctive stuff - I have an entire book that is nothing but photographs of staircases.
How does the human body work? How do all its muscles sheath the bones and organs, how do you break it down into basic shapes that you can quickly put on the canvas and then flesh out into a realistic drawing? Which parts should you exaggerate when you want to caricature a particular pose? Which parts should you not bother with more than a basic rendering of to best serve the needs of the drawing at hand? That alone is a body of knowledge that artists have to spend years cramming into their head before they can be a pro. We regularly circulate little tip sheets[1] that talk about the ways we have found to simplify these things.
How much of your day-to-day work as a programmer comes out of your head? How much of it is looking through documentation and searching Stack Overflow?
A bicycle is a complicated object that has many forgettable parts. Everyone knows that there are two wheels (bi-cycle) and that there's a triangle formed from the down tube, top tube, and seat tube. Except, the triangle has a 4th side to hold the steerer and fork (the head tube). Then there are the other details. The handlebars attach to a stem, which attaches to a steerer tube, which goes through the head tube with bearings on each side, and finally tapers at a "rake angle" to actually hold the hub of the front wheel. There is also the rear triangle that consists of the seat stays and chainstays (and the seat tube). You have to put the saddle somewhere, that's attached to a seatpost that goes inside the seat tube and is secured by clamping it in there. With all those in place, the rest is easy -- where the wheels, pedals / bottom bracket, and chain end up are all very natural.
The problem is how often the bicycle is extolled as a simple triangle with some handwaving and two wheels. Turns out all that handwaving is quite complicated. The triangle becomes a shape with 4 edges. There are actually two triangles. And there are complicated details in joining the seat and fork to the bike.
So... I don't think these people are dumb. They have been mislead by "it's just a triangle" when it's actually nothing like a triangle. The devil is in the details.
