Where are our lines?
March 24, 2005 8:59 PM   Subscribe

One of my bright fourth-grade students asked me something she said has been bugging her for a while: "Why don't people have lines? Like in the cartoons?" Not as in lines of dialogue; lines in the drawing, to mark our feature.

Every attempt at a reasonable explanation (we just imagine they're there when we make drawings; they help us organize visual information in our minds) didn't make sense to her. She wants to know where our lines are. What can I tell her? (Be polite; I may print this for her.)
posted by argybarg to Education (43 answers total) 4 users marked this as a favorite
I honestly don't understand what you're asking.
posted by bingo at 9:04 PM on March 24, 2005

Well, firstly, the end of my first paragraph should say "mark our features."

It took me a while to understand her question myself, as it seemed slightly crazy. What she meant is: When we see cartoons or drawings, the noses, eyes, edge of the face and everything else are outlined with lines. Yet when she looks around, no one has lines. Why are they in the drawings but not in real life?

Yes, I know the answer. But I couldn't explain it to her. Can you explain it in simple, calm terms a nine-year-old might understand?
posted by argybarg at 9:08 PM on March 24, 2005

But we do have lines...they are just invisible unless you look the right way. If you look at your arm for example, you can see where your arm ends and the air around it begins. Imagine that boundary as a line. We don't see that boundary as a line with people because our brains "know" that people don't have lines.
posted by schyler523 at 9:13 PM on March 24, 2005

They're invisible? Can you try "Lines are to organize the drawing and keep things inside. People don't have lines 'cause their skin does that." Not really scientific, but that's a really weird question. Cartoons are just moving drawings, when you draw you make lines so of course cartoons have lines . . . or perhaps that people are 3-D and our features are prominent, so we don't need lines to tell us where things are. But it sounds like you tried those.

Maybe explain shadows? Shadows help mark our features.
posted by schroedinger at 9:14 PM on March 24, 2005

because a drawing is a 2-D representation of a 3-D object. Isn't she old enough to read Flatland yet? :)
posted by j at 9:20 PM on March 24, 2005

Why don't lines exist? You can't really explain a negative like that. They simply don't. The question is similar to asking why I can't see the equator when I'm standing at 0 degrees latitude.

In drawings, the lines are a tool used to show boundaries. In real life, lighting and color form the boundaries. The brain is very good at recognizing shapes from the visual information that our eyes give it. It combines the information from both eyes to determine where objects are in the space around us.

Also, a drawing could just as easily be done without the lines, especially if it's in color. The borders between differing shades of color are obvious to our pattern-matching brains without the aid of an artificial black line.
posted by knave at 9:31 PM on March 24, 2005

Maybe if she understands that using line is just one way of drawing things?

Lines are used in cartoons and comics and some drawings as one way of showing where two things come together.
Another way is by changing how dark or light you draw areas.

Lips are a good example. Lots of people draw lines around the lips to show them. But actually, of course, there are no real lines. A different way to draw this is by using one shade for the bottom lip, another for the chin, and another for the top lip. These different shaded spaces come together and seem make a "line" where they meet.

A visual example will make more sense. This charcoal image by Kathe Kollwitz is good. Zoom in.

The thing with the lips is true here. Also take a look at the way the black of the background meets up with the grey of the arm. It's not a line, but it looks like one.
posted by Cuke at 9:32 PM on March 24, 2005

Have a look at Genndy Tartakovsky's Samurai Jack on the Cartoon Network; one of the striking things about the animation is that the characters don't have black outlines.

You might also talk about "Harold and the Purple Crayon", and discuss why everyone in Harold's "dream" world has purple outlines, while everyone in Harold's "real" world has black outlines.

Also, your student's question might be a good segue to a preliminary discussion of chiaroscuro in your class.
posted by Dr. Zira at 9:36 PM on March 24, 2005

Choose a simple but not blocky object (like a pair of scissors, for example, partly opened, or even her own hand). Place it on a piece of paper and ask her to trace it using a pencil to follow all it's contours. When she's done, she'll see a distinct representation of the object (made of lines) on the piece of paper, but point out that the item (scissors, hand, whatever) doesn't actually have real lines, though the lines on the paper can show us what the object looks like.
posted by taz at 9:37 PM on March 24, 2005

tangent: knave, what do you mean you can't see the equator? isn't the equator a location? thus wouldn't you just stand facing east or west and it'd be right there as far as you could see until the horizon?
posted by michaelkuznet at 9:40 PM on March 24, 2005

I was getting at the fact that there is no physical line on the earth labeled "equator", it's a man-made concept.
posted by knave at 9:48 PM on March 24, 2005

tell her that the picture is actually made of spaces and the lines define the spaces as objects ... then tell her in real life her mind defines the spaces as objects, from her experience of objects
posted by pyramid termite at 9:49 PM on March 24, 2005

Now I have a tangent: Dr. Zira, I can't believe you just mentioned that book. I was read that book frequently as a child, and, seemingly out of nowhere, I've been thinking about it a lot in the past week or two. Especially driving at night and watching the moon follow me. Wow, I just got a chill.
posted by knave at 9:52 PM on March 24, 2005

Neat question. I asked myself something similar the first time I tried to paint realistic portraits. In that context, the best answer I could give would be lines are changes in colour. The darker the line, the bigger the change. I would find her a 'learn to draw' book, as her question is really one of technique

On a deeper level, your young friend has touched on a cognitive science question. As I understand, our visual processing system is very good at detecting edges and splitting up our visual field into mental models of the objects we see. (This is such a complex process that an entire branch of artificial-intelligence research is dedicated to machine-vision.) I suspect that cartoon sketch-lines tap into the same brain pathways to give realism to the characters.
posted by Popular Ethics at 9:58 PM on March 24, 2005

I agree with Cuke. Show her a few drawings where there are no outlines, just variations in shading. Any beginning "how to draw" book probably has examples of this with cubes and spheres, which she might be able to copy. Then have her draw a cube and sphere cartoon style--just outlines. (You may have to give her examples to copy.) Explain how cartoons are abstractions that represent the idea of a thing, not realistic drawings that try to show how something looks. That's one reason cartoons are fun--a realistic face can only be that person while a smiley face could be anybody.

Scott McCloud's amazing book "Understanding Comics" explains this better than I can. See chapter two: he shows a series of faces that range from realistic to completely abstract. The book will be over her head, but you'll find it interesting. And you may find other ideas in the book you could show your students--like how if you add a simple eye (a circle with a dot inside) to any closed shape, our brains will find a way to see that shape as a face in profile.
posted by hydrophonic at 10:15 PM on March 24, 2005

I love questions that inspire me to learn more.

Just googling for "cognition vision edge detection" gives me a book by David Marr, who detailed his theory of mental image processing (posthumously) in a book called Vision: A Computational Investigation into the Human Representation and Processing of Visual Information. Summary here. According to Marr, the first 4 steps of human vision processing are:

1) convert to greyscale
2) sharpen the changes in intensity
3) *draw lines* at the highest intensities (vectorize)
4) join up the lines to form continuous edges defining bodies.

A sketch drawn with hard lines would play right into this mental image processing. Your fourth grade friend -is- smart for thinking about this.
posted by Popular Ethics at 10:18 PM on March 24, 2005

PE: From my perspective (cog. sci./computer vision), your second answer hits on it, and your first one is part of the maddening problem (though perhaps not in just the context of painting). Too often boundaries actually separate regions that are similar in appearance, which makes writing computer programs to detect them very difficult.

In some cases we perceive edges where there are absolutely no lines at all. Neurophysiological data suggests that these boundaries are in a sense "filled in" after the fact in primary visual cortex (sort-of the first way-station for visual data in the cortex), almost as if it were some kind of scratchpad.

So, here's how I might answer the student's question:

The lines in cartoons show boundaries between parts.

In real life and in pictures and movies, our brains can figure out where the boundaries are by looking at the very detailed images. You can prove this by tracing all the way around someone's nose in a magazine ad---it's pretty easy to tell roughly where it starts and stops.

But in cartoons, all of that richness is gone. So you have to draw lines to give hints about boundaries your brain would draw "automatically" in natural images.

Or, in a sentence---the lines are there in real life, but they're all in the student's head.


To follow up on PE's next point... Marr's book is a classic text in vision. What it doesn't allow for is for any kind of feedback effects in the visual system. For those who believe that a "scratchpad" in visual cortex is necessary to intuit contours where there aren't really any, Marr's model is limited.

I haven't gone into much detail on why some folks think you need a feedback scratchpad mechanism, but I can elaborate if you like.

Agreed, this kid is pretty sharp.
posted by tss at 10:23 PM on March 24, 2005

Actually, animals make a strong attempt to do the opposite.

If we had lines around us, we'd stand out to both predators and our own prey. Ever wonder why many animals have bellies that are whiter or lighter colored than their tops?

Since the sun shines from above an animal, and its underside is in shadow, an animal that was one color both top and bottom would appear to be lighter on top, because that's where the sun shines on it.

By having a darker top and a lighter bottom, the sun-lit darker top and the shadowed darker bottom both look closer to the same, and the animal doesn't stand out from its background as much.

Since it's the animal it's harder to see which gets eaten less and gets to eat other animals more, it tends to survive better than its unshaded conspecific which stands out, so evolution makes sure that shading is a trait that's conserved and increased in future generations. (Increased because once all the unshaded animals are eaten, the less shaded animal is next on the menu.)

A particularly good example of this is what appears at first to be a counter-example, the catfish Synodontis nigriventris. "Nigriventris" is Latin for "dark belly", and that's exactly what the fish has: a dark belly and a lighter dorsal (top or back) side. This doesn't undercut the explanation above, however, because nigriventris swims and feeds upside down; in that position, it is nigriventris's belly that the sun lightens and its lighter back that the sun does not illuminate.

This account is taken from Richard Dawkins, The Ancestor's Tale, which like all of Dawkins's works, is full of such fascinating explanations.
posted by orthogonality at 10:34 PM on March 24, 2005

If I can join the conversation myself...

What strikes me now, after reading these very fine answers, is that a nine-year-old has a lot of trouble conceiving of her own mind doing something to the world. She doesn't conceive of there being a concept of the world separate from the world itself. When I told her that we add lines to whatever we see, she said, verbatim: "I can't understand a word you're saying."

This is so alien to her mind that, in fact, she would rather locate the problem elsewhere. She sees lines in drawings and cartoons (and presumably in her mind), so they must be out there.

(I wonder how this reflects on teachers' continual nagging of children to record their responses to literature, as if their responses could be separated from what is out there.)
posted by argybarg at 10:39 PM on March 24, 2005

We don't have lines, but we do have high contrast changes, caused by color differences (red lips next to beige skin) or lighting differences (light and dark around our noses from shadows).

Lines are frequently used in drawings as an easy way to represent these high contrast changes.
posted by NortonDC at 10:47 PM on March 24, 2005

orthogonality's comment reminds me of the canonical image used to demonstrate how wonderfully your mind segments images:

Find the dalmatian.

Perhaps this image can be illustrative for your pupil.
posted by tss at 10:47 PM on March 24, 2005

NortonDC: I would suggest that under many circumstances, the lines that are used to denote boundaries between parts and objects in cartoons are often (not usually, perhaps, but often) not present, in any traditional sense (abrupt changes in color or intensity), in natural images.

This is why the "magic wand" select tool in Photoshop is so often lacking when it comes to picking objects out of images. It gets many boundaries right, but it also often bleeds off into the background or gets hung up on a spurious boundary. Of course it would also fail utterly on the dalmatian image, but that's perhaps a bit far from a natural image.

I maintain (and some would disagree) that our brains use a considerable amount of background knowledge about objects to bias our perception of where edges occur in an image. These biases manifest themselves even in the first stages of visual cortex, which is where edge detection seems to occur in any real sense in the first place.

This is why real image segmentation is still a specularity in the eye of the computer vision researcher---in order to get it to work, you have to know lots of things about the structure of objects and other complex visual entities in the world, then use that information to tell you which perceived lines are liars and which missing lines are actually there.
posted by tss at 11:04 PM on March 24, 2005

I'm revising my answer:

We don't have lines, but we do have edges or borders. These are caused by color differences (red lips next to beige skin) or lighting differences (light and dark around our noses from shadows).

Lines are frequently used in drawings as an easy way to represent these edges.
posted by NortonDC at 11:05 PM on March 24, 2005

That was weird.

You are right, tss, but so am I. I said that lines are frequently used for what I described, not exclusively. Lines have many other functions in cartoons, such as smell or action lines (American Splendor on the brain). I think that the use I described is the primary role of lines in cartoons.
posted by NortonDC at 11:10 PM on March 24, 2005

Sure, but, that's just the thing. I'm not sure that these basic percepts are necessarily the primary use of lines. I would say that the lines are more descriptive of our understanding of the segmentation of the image than what is actually there.

Naturally it's up to the artist. But consider (scrambles for picture) this image. OK, bad photo of me, but take a look at the tail fin of the poorly tied-down N738VL behind me.

There is virtually no contrast between the fin and the snow. There is some, but it's extremely small compared to the brown stripes on the fuselage and the white paint. You could run Photoshop's "edge detect" on that image and indeed the stripe boundaries would be among the strongest in the image, and the fin/snow boundaries would be among the weakest. Now, which one would an illustrator emphasize?

This isn't an isolated case, either---just about any image has a pathological boundary like this that I think 9 out of 10 illustrators would choose to make prominent.

(I'm sorry if I sound strident about this---I can get enthusiastic about sharing the woes of computer vision with the world.)
posted by tss at 11:22 PM on March 24, 2005

In art class, I had the opposite problem--everyone else seemed to be able to see where those lines were and reproduce them unerringly on paper. What were they seeing that I couldn't?? If she's interested in reading, Drawing on the Right Side of the Brain suggests some effective techniques for becoming away of the "the lines", starting by drawing the boundaries of spaces between objects instead of trying to see the objects themselves. Boy do those lines suddenly pop into distinct view.

This might also be a good time to start having conversations with her about media literacy. If she's expecting the human world to follow the conventions of one kind of media, she may have have accidentally adopted unrealistic expectations based on other forms of media.
posted by nakedcodemonkey at 12:08 AM on March 25, 2005

Even with the portions of the tail section that are the exception, I'd still say that your chosen image strengthens the case for the primary use of cartoon lines being to denote high contrast edges.

Something that's not primary can still be important, tss.
posted by NortonDC at 12:08 AM on March 25, 2005

I think a good way to explain this to her would be to introduce the difference between 2-dimensional animation and 3-dimensional animation, and you can show this to her by comparing and contrasting traditional flat cartoons with claymation/stop-motion animation/CGI. Or you could show her a movie where real life people meet cartoon people (Mary Poppins, Who Framed Roger Rabbit, Cool World).

The land of 2 dimensions is flat. The land of 3 dimensions has depth. 2D uses lines. 3D uses shapes. We live in the 3D world. Our world has millions and millions of colors. But the cartoon 2D world doesn't have as many colors. We do have lines, they're just very very thin and not black. In the flat world, they need to exaggerate those lines -- and that's all they are, an exaggeration -- in order to distinguish one object from the one next to it, and to tell whether one object is in front of the other.

Also point out that different cartoons have different ways of defining those lines, like The Simpsons vs. Disney movies vs. anime vs. Powerpuff Girls. Perhaps in this way she can see that those lines are truly only arbitrary, defined by the cartoonist who drew them.

You don't have to dispute the existence of lines, merely point out that they are a way of showing things you can't normally see in the real world but know are there. (This should cover all those stink lines and the spinning squiggles and stars toons get when they bump their heads, too.)

Just be sure to use a lot of examples that are within her grasp and consistently reiterate 2D on one hand and 3D on the other, and hopefully she'll be able to understand your point better. :)
posted by Lush at 2:26 AM on March 25, 2005

I don't think anyone has mentioned this and I thought I might chime in as a comic book artist and painter.

To clarify a hair - she means 'why don't people have outlines?' Someone mentioned that the cartoon Samurai Jack doesn't use lines. That's true in the sense of outlines, but from a painter's perspective the images are accomplished by a line structure that has filled in color and no outlines. Lines would be more thoroughly subverted in the cartoon if colors from the surrounding environments, or ground, leached into the figure past the assumed boundaries.

To me, the explanation is in large part that cartoonists have to make many, many, many, many drawings. It is quite a bit faster to use outlines rather than... you could call it shading for your student's benefit... shading to finish these images. I haven't quite figured out all the angles for a nine-year-old's explanation (I have to teach the exact same ideas to college students though, and we actually just covered this yesterday)...
posted by Slothrop at 4:08 AM on March 25, 2005

Artists use outlines in the same way that writers use letters to create words.

It is a "pattern language"...a sort of shorthand. Slothrop does a nice job of describing this.

The letters A, B, C and so on don't exist where we can see them as we speak. We have created these symbols and we use them to record what we hear. Just as we use outlines to record what we see.
posted by jeanmari at 4:33 AM on March 25, 2005

I would say that you should help her to turn the question around. Help her to understand that rather than asking why doesn't the real world have lines like cartoons have, she should ask why do cartoons have lines when the real world doesn't.

The cartoon world is just a representation of the real world, and not the other way around. So then the question becomes what purposes do the lines in cartoons serve.

I think coming at it from that angle might make some of the above about abstractions and demarcations of boundary spaces make more sense.
posted by willnot at 6:07 AM on March 25, 2005

This is a really great discussion. I just wanted to add that I think The Natural Way to Draw by Kimon Nicolaides has some inspiring language about seeing and "feeling the edge" when working on contour drawings. I would recommend checking it out from the library just to read the initial passages on it.
posted by safetyfork at 6:40 AM on March 25, 2005

OK, NortonDC, I guess we'll just have to disagree. I say that just because contrast edges and lines in cartoons seem strongly correlated, it doesn't mean that they share this primary relationship. I think lines represent something else, specifically the boundaries we perceive between objects.

Here is a more extreme example than the plane image:

Here are the famous Calvin & Hobbes snowman comics. Note the very crisp, constant edge between the snowmen and the (presumably) snowy ground. We see edges there that wouldn't substantially exist in real life. To prove it:

Here's where someone tried to make the snowmen in real life. (NSFW snow drawing at the bottom). Find the images taken during the daytime. As you can see, very weak edges.

Lest you think this is another special case, consider: To what extent would the illustrator go to respect the actual visible contrast? Suppose Bart Simpson walks in front of a yellow building, or Snoopy in front of a white one, or Ziggy in front of a slab of total lameness. In these low contrast situations, do the cartoonists stop drawing lines around the characters? Not unless they're making a joke. Lines like those that delineate objects are more permanent than reality, since they're part of the semantic information about the scene that has to be conveyed. I think that's what's really being drawn.
posted by tss at 6:45 AM on March 25, 2005

i don't know what fourth grade is, so this may be extremely stupid, but i could imagine some very smart schoolchildren enjoying gombrich's art and illusion, which must discuss this at some point. or maybe i really don't have a clue about kids.
posted by andrew cooke at 7:01 AM on March 25, 2005

incidentally, i once had concussion and lost my sight for a while. it returned gradually, and i could see "lines" before "colours". in other words, the part of my brain that decided what distinct things were came back online before the bit that gave me images to look at. it's impossible to describe - it wasn't seeing lines around things - but i could "see" "white" shapes that were distinct from each other, even though everything was "white" and otherwise indistinguishable.

it was fascinating, but i must have freaked out the kind stranger who was driving me to hospital, with blood dripping down my face, as i sat there in the passenger seat describing this strange world...
posted by andrew cooke at 7:05 AM on March 25, 2005

Great question, from both a childhood and adult perspective. I'm surprised no one brought up Impressionist art. So much of that entire movement is, IMHO, a response to that very question. Show her some Monet, or if she's more specifically curious about people, some Renoir. It's all color, but there are clearly shapes.
posted by mkultra at 7:20 AM on March 25, 2005

More advanced than a quick conversation with a 4th grader, but still interesting...the chapter on The Development of Symbolism in Drawing in this paper by Lev Vygotsky (The Prehistory of Writing).
posted by jeanmari at 9:05 AM on March 25, 2005

argybarg - a nine-year-old has a lot of trouble conceiving of her own mind doing something to the world. She doesn't conceive of there being a concept of the world separate from the world itself. When I told her that we add lines to whatever we see, she said, verbatim: "I can't understand a word you're saying."

That reminds me of a little girl I knew once who would say "I can do a magic trick - watch what I do to the lights," and then cross her eyes. She didn't realize that no one else could see the doubled images she was seeing when she crossed her eyes.

Even to an adult, the notion that vision is an inherently creative process, depending on the brain as much as the eyes, is a very non-intuitive one. When I was in college, I remember my dad explaining to me some of the issues of vision processing from a cognitive psychology perspective. He held up a book and pointed out that when he moved it around, the apparent shape and size changed completely, and yet we still perceived it as a constant entity. At the time, I really didn't understand what he was saying. To my eyes, the book was the same shape, regardless of its position. It took years for understanding of the concept to sink in all the way. If you come up with an explanation that the 4th grader can understand immediately, be sure to post your results.
posted by tdismukes at 9:20 AM on March 25, 2005

Everything you've learned in school as "obvious" becomes less and less obvious as you begin to study the universe. For example, there are no solids in the universe. There's not even a suggestion of a solid. There are no absolute continuums. There are no surfaces. There are no straight lines.

- R. Buckminster Fuller

However, this answer to a young person may result in a warped brain. Or someone who learns to think dynamically at a young age. Or both.
posted by loquacious at 12:56 PM on March 25, 2005

I acquired my first pair of eyeglasses when I was thirteen, long after my eyesight had degraded. When I first put the new lenses on, I perceived lines around everything. It was the only way my mind could process the sharp contrasts that previously had been fuzzy boundaries.

Interestingly, it has often been suggested that the early Impressionists developed their "fuzzy" technique which eschews clear boundaries because they were nearsignted.
posted by SPrintF at 2:55 PM on March 25, 2005

We *do* have lines, but you have to be trained to see them clearly. Artists spend a lot of time practising doing just that.
posted by flabdablet at 3:37 PM on March 25, 2005

SPrintF, I had the same experience, but at a slightly younger age. Maybe because of that, and my incessant cartoon watching, I had the same confusion about outlines that this girl is experiencing. I have a feeling that having her do a hands on experiment might help her better understand this.
Give her a photo that has an easily perceived subject in a fairly complicated background (I pulled this one off Google at random), and ask her to trace the outline of the subject in as detailed a way as she can. Chances are, she'll find that although she 'sees' the dog and understands where he is within the photo's space, she'll have problems defining a lot of the particulars (the interior edges of his nose, for example.) Then have her 'correct' the drawing by referencing the photo, instead of tracing it. She'll hopefully find that in looking at 2D images, we tend to fill in a lot of information ourselves, and that artists use outlines to clarify the image they wish us to see.

Once she's experienced that, I think she'll be more receptive to some of the other wonderful explanations here.

(I loved reading this thread, BTW. Great question!)
posted by maryh at 5:57 PM on March 25, 2005

I took one of those classes where you draw naked people with charcoal on newsprint.

One day my teacher took me aside and showed me a book of Matisse. She pointed to one of his many sketches.

"When you can draw a line like that," she said, "you can draw lines in my class. Until then - I want you to draw LIGHT."

That was an epiphany.
posted by ikkyu2 at 7:09 AM on March 26, 2005

« Older Seat height filter   |   Internet site blocking Newer »
This thread is closed to new comments.