Webcomics’ Identity Crisis, Part III: Rethinking Reinventing Comics, Part One: The Finite Infinite Canvas (And a Brief History of Webcomics)

Read Part I and Part II of this series.


As it turned out, I have much more to say about two particular aspects of Reinventing Comics than any others (much like Scott McCloud), and as a result I've reframed my discussion from how I was going to frame it around Reinventing itself. For how boneheaded Scott McCloud can seem at times, he actually got more right than wrong in Reinventing when it came to digital comics, including the first of his three digital revolutions. Although it seems to be mostly used as an excuse for laziness at the moment, there's no doubt that digital production of comics is everywhere, and is evident in all sorts of strips from PartiallyClips to Dinosaur Comics to sprite comics. And there are even some comics, namely 3-D comics, that have used computer tools as more than a time-saving measure. (Oddly, Diesel Sweeties would also fall in this category, due to the amount of work it must have taken R. Stevens to make those characters.) So for all his faults, McCloud got a lot right.

Except for, well, the meat of his latter two revolutions.

In 2005, Scott McCloud gave a talk at TED, which, thanks to showing up in the StumbleUpon demo, was my very first Random Internet Discovery. That talk was only reposted last month, which naturally has prompted a whole new round of debate on McCloud and his ideas. At webcomics.com, one particular aspect of McCloud's talk attracted attention: his reference to and exploration of the infinite canvas, touching off a debate on how and why the infinite canvas hasn't achieved the penetration McCloud predicted. (IMPORTANT NOTE: As I posted yesterday, webcomics.com is currently down. If and when it returns in full, I will add a link to the post in question to this post. Unless it happens a week from today or later, I will NOT give any announcement when I do so; you will need to check this page on a regular basis, or else when webcomics.com comes back up. Which I sincerely hope is before I have to put up Part V...)

Although I read all of Reinventing, I always intended to focus all my attention on the last half (and skipping over some relevant elements in the first in the process). I'm probably going to cover ground many others have trod before; the Comics Worth Reading blog calls the second half "outdated and silly to today's readers" even as he endorses the first half. (McCloud, incidentially, seems to have forgotten what made Understanding so groundbreaking, and relies a lot on what the earlier book called "word-specific" and "duo-specific" combinations, such that I find myself wondering why he couldn't have just written it as an essay, other that then we'd be deprived of his sudden bank of icons for media and his revolutions, an obsession far more developed and incorporated than in Understanding.) My analysis will be drawn not only from Reinventing, but also from McCloud's later sequel webcomic, I Can't Stop Thinking! (which barely kept the thoughts of Reinventing a smidgen more up-to-date, only lasting into 2001 in six installments), subsequent essays, and the TED talk (although it starts out talking about innovation in general and turns into a retread of Understanding until about the ten-minute mark, and even then turns into a retread of Reinventing for the most part).

The concept of the infinite canvas is one of two major developments that McCloud is known for in Reinventing, and not only does he focus on it to the exclusion of all else that could be done with the medium in a web environment, he seems to disdain one of the others - multimedia comics - as either dolled-up print comics in an electronic environment, or trying to do something that film, and eventually virtual reality, could do far better - if not crossing over to one of those realms. (His words on hypertext comics seem to indicate he would moan at an idea I have that I think is truly groundbreaking, if a bit inspired by something that already exists.) That Comics Worth Reading review thinks it should be interpreted as a mere suggestion that McCloud got a little too enthusiastic about, one among many, but some of McCloud's actions since the publication of Reinventing suggest otherwise. McCloud's enthusiasm for the infinite canvas devolves from his philosophy of the core concept of comics, its Platonic "form" if you will, and ironically, in the process he arguably completely abandons his definition of comics from Understanding Comics, as we'll see.

(I actually agree with him on multimedia in that it starts turning comics into some other medium. It also abandons the one thing McCloud saw as late as Reinventing as the great thing about comics - that anyone can make them - and if it doesn't it makes it no longer unique, and opens more media to "everyone". But, again as we'll see, McCloud starts including multimedia comics in the notion of the infinite canvas.)

He starts his journey into the infinite canvas, however, with the evidence he presented in Understanding that comics is older than the written word, and that it can be traced to the oldest cave-paintings, through the ancient Egyptians and Romans, up through Mesoamerica and the European Middle Ages, and through numerous then-unrecognized "comics artists" to the present day. In Reinventing, McCloud notes that until the European invention of printing, comics were presented in a single line stretched out as long as possible using whatever means was possible to keep it going, whether it be zig-zagging along a huge wall, spiraling up a column or just having a really long scroll. From this McCloud devises the notion of a horizontal infinite canvas, a single unbroken line of panels stretching out as far as anyone can imagine, what one might call the "ideal" infinite canvas. The "horizontal" canvas implies the "vertical" canvas as well, although that's less conducive to creating a single, unbroken string of panels. (Still, McCloud did I Can't Stop Thinking! as a vertical infinite canvas with "trails" connecting panels that wander all over the page.) McCloud also suggested a "staircase" that's a combination of the two.

Plenty of people have pointed out the problem with the scrolling infinite canvas: that the problem is in the scrolling, which can be hasslesome to people without mouse wheels; even with a wheel, it's rarely well-adapted to horizontal scrolling. (A vertical canvas is more doable in this context.) One of those people is McCloud himself in an episode of ICST! (ICST itself, incidentially, also points to another issue, that it can send people going in nonintuitive directions.) But the issue McCloud saw involved technological limits, like "herky-jerky image redraws", that appear to have mostly faded, making vertical scrolling more viable now than it was in 2000. As for the problems with horizontal scrolling, they presume the only way to scroll is with the arrow keys, mouse wheel, or scroll bars. If the comic is a straight-up image, or a series of straight-up images, on a regular web page, then yes, those are the only ways to scroll. But that's not what McCloud was talking about (and if you've already listened to the TED talk, you know that).

This is what McCloud was talking about.

You can make infinite-canvas comics using PNG images on a regular web page, but Flash, Java, and similar applets open up the possibility of merely clicking to move to the next panel, and a more intuitive click-and-drag interface across the page. McCloud suggests a number of ideas that would never work without some kind of applet to handle it (given current technology), and going far beyond what most people think of when they hear of the "infinite canvas", such as packing comics into a "cube", arranging comics in three dimensions and forcing the reader to "turn", "circular narratives" (first proposed in Understanding but technically not directly tied in to the infinite canvas until TED), and instituting a high level of interactivity, even as a way of introducing sound and motion - indeed, when McCloud speaks of the "infinite canvas" it almost seems like he wants webcomics to become a completely separate medium. (Note: In ICST! #3 McCloud seems to separate sound, motion, interactivity, and "the various blessings of hypertext" from "an expanded canvas" but also notes "I think some applications of these features are more promising than others".)

Some of his ideas completely go against most notions of what "comics" are, not the least of which is McCloud's own Understanding definition, "juxtaposed pictorial and other images in deliberate sequence". For example, the notion of going from one panel to the next by "zooming in" on the previous panel (or other forms of viewing one panel at a time) may seem to violate "juxtaposed", which McCloud used to mean "side-by-side". McCloud introduced this word originally to distinguish comics from animation, which is also displayed one "panel" at a time, but isn't presented as discrete panels, but as a continuous experience. Unfortunately, the distinction McCloud presents isn't that, but rather, he now sees comics as a "temporal map" where moving through space is moving through time, which muddies the waters.

In truth, though, just about every medium has a shot at getting blurred with every other medium in the new medium of the Internet (and computers in general) that can display any and all of them. Ultimately, just about every medium can be distilled into a few core elements, and within their place in the resulting matrix, they can ultimately become the same medium. (McCloud, in fact, touches on this issue.) For example, movies and television have very few elements that they don't hold in common; they both share the combination within them of a visual aspect and a sonic aspect, of sound and images. Similarly, print books have no sound and no images, and comics (and paintings) lie at the intersection of images but no sound. As silent films have a problem in that they too can be considered a combination of images and no sound, we must add a third item to the visual axis, the concept of the moving image. Thus, movies and television share between them the combination of moving images and sound, silent movies have moving images and no sound, but comics have only static images and no sound. (Implied are axes for the other senses.)

Just about any medium through the first two millenia of the common era could be placed within the resulting matrix, thus explaining certain connections. For example, music and radio programs both fit where no images meet sound, thus explaining music's presence on radio. Even video games, with their emphasis on interactivity, combine moving images with sound and are thus often played on televisions. We also see other theoretical media that might not have been explored. For example, I wonder what the intersection of static images with sound would be? (Perhaps kids' books or greeting cards with built-in sound devices?) The theater is problematic, as it theoretically combines moving images with sound, but most theater afictionados would tell you that a movie or TV reproduction of a performance does not cover the whole performance - and not even necessarily because of a presence capturing the other senses, but a role in the "interactivity" scale as well. And about the only thing that can be reliably said about board and card games (and maybe even particpator sports) is that they include interactivity... but unless it's like Calvinball and the rules are made up by the players, does it even have that? (Answer: Yes, if video games do.)

But the Internet can play all of these to some extent or another, and even beyond the Web's current hypertext-based environment (which is best suited to static images with no sound, and even more, no images with no sound, thus establishing our first connection), programs installed on computers can have almost limitless possibilities. Text can institute images and sound as important parts, or even become part of an image. A comic can break into motion and sound and become more like a film. Video in the traditional sense can become interactive and 3-D, capable of moving on your command. All media could, indeed have to, become more interactive. The distinctions between media could become less important if they are ultimately part of one uber-medium.

But there is also a hint in Reinventing of part of why the infinite canvas hasn't caught on even with the idea of using Flash or Java: McCloud predicts that, even outside comics, "spatial models" would take their place alongside hypertext as a means of organizing information. And... they haven't, nearly a decade in, despite most of the bandwidth barriers falling away. This is partly because, for whatever virtues it may have, a "spatial model" requires enough work on the basic concept to become, essentially, art (regardless of your definition), and at that point the person's attention becomes focused on the model itself, as a piece of art, than on what it contains, making it too distracting for utilitarian purposes. It could be used for artistic purposes, but it's then subject to the same reasons infinite canvas hasn't caught on in comics. (The closest McCloud seems to come to any specifics on the use of "spatial models" for information is an icon of a book, which would be more gimmicky than anything else when compared to hypertext. Perhaps McCloud has a "game world" in mind, such as the modern-day Second Life? Certainly a part of ICST! #1 seems to imply such.)

But I think there's a bigger factor in play here: Webcomic revenue models work best on a periodical basis.

Read that paragraph again, because the rest of these two parts hinge on it.

When considering McCloud's views, it's important to remember that he was coming from a comic book background. It was the singular form of the comic book that McCloud imagined translating to the infinite canvas, no longer having to contort their stories to the confines of the page, but flowing in one continuous stream (or several) from beginning to end. But webcomics have evolved under a comic strip model, what McCloud called "a marriage of convenience between newspapers and comics", and that model has little use for the infinite canvas. If it's a gag-a-day strip, all you really need is to deliver your punchline and run, and dolling it up with weird shapes or stretching it out too long becomes too distracting to the message. If you're a strip with a lot of drama and continuity, and you start down the infinite canvas path, then unless you're delivering a complete story in a single strip, the place where you decide to cut it off and release the next part is mostly arbitrary, determined mostly by your desire to keep the audience in suspense. (Remember this distinction; it will become important again later.)

It's possible to break up a long story into "chapters" and publish it on the infinite canvas, breaking a story into bite-size digestible chunks while also making sure your audience has a reason to come back for the next part of the story, but either you create several separate "chapters" that are meant to be read as one story, yet aren't because they're arbitrarily separated (McCloud explores the potential ideas behind this in ICST! #4), or you have to keep dynamically updating the one canvas, and not only point people to where the new addition is, but - and this is an issue with any substantially long canvas - also provide a longer-term way for people to mark their place.

(I know the old "drama v. gag" conundrum is arbitrary, and as early as Understanding Comics McCloud suggested that comics were capable of a wide variety of genres. ICST! uses the infinite canvas well, for example. I'm using that conundrum, though, to illustrate that in any genre, it all breaks down to one of three schemes: short one-offs, installments of a larger story, and the same larger story completed. The first is unaffected by the infinite canvas, and apply the second to an infinite canvas and there's little reason for it not to become the third.)

Therefore the infinite canvas works best when it contains a complete story within itself – but how do you make money off of it? Forcing people to subscribe isn't the answer, because when something is released one time and that's it, there's nothing to subscribe to. Similarly, people have no incentive to donate to something once they've already read it and have no reason to believe their donation will result in anything more. Most self-sufficient webcomics today sell merchandising, but it's difficult to merchandise a single, solitary work - especially if it utilizes the infinite canvas (the whole point of which, as McCloud sees it, is to defeat the potential for selling a print version of the work) and especially if it's sufficiently artsy in its use of the infinite canvas. (In some forms you could sell a real-world version of the comic itself, if, say, a comic packed into a cube were reproducible in real life, but you're back to the defeating-the-point problem.) Advertising's doable but unless you can make a sufficient event out of the release, don't expect to make a living off it – and even then, no matter how big the launch is, you get that one-time infusion of money and that's it.

Perhaps we could modify the subscription idea, and put the comic behind a paywall? Since not everyone clicks on an advertisement (the vast majority don't, in fact), let alone buys what's advertised, and many actively resent even non-intrusive ads like I have, you're probably always going to make significantly more money from a paywall anyway. (The ad rate for the Sandsday-only ad box rarely gets above a cent a day; since I average five page views a day, if I could charge a single solitary cent for the strip I'd be quadrupling my income! McCloud himself covers these and other problems with advertising in ICST! #5.) But again, no matter how popular your initial comic may be, no matter if thousands or even millions of people enjoy it, if it's released one time and then you're done with it, you get that one-time infusion of money and that's it. Whatever income you get from it past that initial spurt is damn near zero. That's no way to make a living. Any medium, indeed any product, will ideally make you money and keep you alive best if your customers have to keep coming back to it. This is okay in film where people make a substantial amount of money on each picture and quite a bit of side money from endorsement deals. It's even okay in print books to an extent - but in both cases both industries are still fond of series of films and series of books.

It's possible to make sequels and trilogies, but there's a reason why webcomics have evolved under a comic strip model to this point, with very, very few emulating the comic book model.

Pete Abrams, creator of Sluggy Freelance, tried to get into the comic book industry at first without success. Eventually he took up a job at a web design company, and started Sluggy as a "creative outlet". Why did he make it as a strip instead of any sort of graphic novel form? In a 2002 interview, Abrams attributed it to, really, the problems I've just described: the full-length comic book format doesn't lend itself to repeat visitors on the Internet. (Abrams' solution to that problem was to hammer out something really quick and throw it up, which he decided lent itself better to a comic strip format.)

Abrams was the exception, not the rule, in the early days of webcomics, although his example is illustrative of just some of the obstacles to getting an actual original comic book on the web in those early days. New media, and new forms of that media, are rarely founded by people who set out to do so. The Scott McClouds of the world are few and far between. Most media evolve over time, and grow organically, as the proverbial million monkeys on a million typewriters so often tend to cluster accidentally, as though drawn by gravity. Someone invented film, but no one person invented the movie. Someone invented the television, but no one person invented the TV show. For someone to decide to put a full-length comic book on the web before Reinventing would have involved the conscious decision to explore the possibility of doing so. Even before Reinventing, there were such "comic books" littering the web and some even explored the infinite canvas, but as if to prove the point, most were done by these sort of artistes that explicitly set out to explore the new world.

The comic book format is rather rigidly defined by the ways and means of print. Change the size of the paper, no matter what the reason, and you change the shape and form of the resulting comic. Remove a sheet from the book (if doing traditional fold-in-half-and-staple-binding) and you've removed four pages of comics or ads, and that changes the shape of the story. By contrast, the comic strip format is far more organic than McCloud gives it credit for. It's really just a really short version of the ancient comics McCloud is so in love with, and it appeals to our desire for a good joke easily. When the Internet came along, the comic book could have slid right in to the Internet (and indeed, by 2001 Marvel was putting select numbers of its comics, including, for a time, the entire Ultimate line, on the Web, which played a role in getting me into comic books), but it would need to undergo one heck of a mutation to really stretch its legs in the new format, and those mutations rarely happen easily or quickly. The comic book was at home on the printed page and much less at home on the Web page. (Indeed, from the very beginning Marvel's online comics were in Flash, not hypertext or even PDF – though I suspect that was for security concerns.)

But comic strips could be picked up right out of the newspaper and plopped onto the Web page and look almost indistinguishable from their print counterparts. (Anyone who's been here since December knows that I thought User Friendly was a print comic at first.) Combine that with the Internet's tendency to draw itself to humor and it's easy to see why most of the early pioneers of webcomics were people like Illiad: they just wanted to make their friends laugh at their funny jokes. In fact, this is still a driving force behind webcomic, or even web site, creation. (See: Morgan-Mar, David; and outside webcomics but still within the community, Solomon, John.)

There's another factor at work here. Although it's always been difficult, it has always been at least possible to write, draw, publish, and distribute comic books yourself. (McCloud explains one really simple way to do so in Reinventing Comics.) While comic books are more suited for the page than the screen, when considered alone, comic strips are actually more suited for the screen than the page. It's in this sense that comic strips were "a marriage of convenience between newspapers and comics", because simply trying to distribute a strip as a single narrow sheet of paper... well, there are few paper makers who make paper that size. Comic books could adapt easily to any size paper; comic strips either were stuck in a special format or, in the case of Sunday strips, had to take on the comic book form. That made it a lot harder to distribute without the help of a newspaper, which meant comic strip artists were far more at the mercy of a newspaper or syndicate.

As webcomics caught on and as success stories like Scott Kurtz or Gabe and Tycho started springing up, aspiring comic strip artists, long chafing under heavy creative control and small panel sizes and all else that came with comic strips, became drawn to the web like moths to a flame. For a while there was an explosion of new comics following the comic strip format, but at first few comic book authors saw an opening. They still had a chance at either self-publishing or at least some measure of creative control from publishers of all sizes. What few dramatic strips existed were almost exclusively former gag-a-day strips that had undergone Cerebus Syndrome. Although I don't recall reading anything from him that would confirm this, I get the feeling that the reason Rich Burlew originally made The Order of the Stick a full-page comic in 2003, despite its initial gag-a-day nature, was to allow himself freedom to roam and engage in large panel layouts. (Look at the very first panel of the first strip.) As before, when it came to webcomics, the comic book format was the major application of the infinite canvas.

By 2005, the game changed; in the same month Gunnerkrigg Court launched and Girl Genius moved to the web. They and OOTS are the major examples of true comic book style storytelling I can think of on the web today. Comic books finally found an expression on the web, and it wasn't McCloud's infinite canvas (though in OOTS and a lesser extent Sluggy there is an element of infinite canvas in at least some installments), it was the hypertext-based system McCloud disdained, which turned out to work fairly well. That full-page hypertext system allowed creators to entice readers to keep coming back for the next installment (with the added bonus of ongoing, well-developing plot threads to help ensure it), and still allowed the creation of a later print version for ideal capitalization. And the game may finally be changing and forcing aspiring comic book writers and artists to consider a move to the web, for reasons I'll get to – along with an expansion on how comic books manifest themselves on the Web – in Part V.

Could it have happened differently? Could it still happen differently, as the developments discussed in Part V play out? Could the infinite canvas still work? Suppose there were people out there who decided to move the comic book format to the web and released full stories on an infinite canvas. Suppose they actually managed to do so in a way that was more than just experimental, and decided to make a series of these complete stories. Would the webcomics community find out about them and pimp them? My bet is, "Not unless Scott McCloud does." While there are some parts that may be open to it, by and large the webcomic community is woefully unprepared to handle comics that are released one time and then that's it, infinite canvas or no. Quite a few aspects of the webcomic community, like ranking sites like Topwebcomics or "tool" sites like Komix, seem to presume that your comic is a continuing comic that releases at least once a week, once a month at least, and not in a complete story all at once. (Da Blog, I freely admit, falls into this category, with my rule that I never review comics that have ended, and an all-at-once comic never really "begins" in the first place.)

I left open a loophole earlier when discussing revenue models. Suppose you decide you're going to release multiple completed comics in a series on an infinite canvas, maybe once a month or once a year, and put them behind a paywall – akin to the price you pay when you buy a comic book or graphic novel at the comic book store or bookstore. Theoretically, there's no reason why that can't allow you to live, at least on modest means, in the span between releases, or even, with a lot of hard work, become the comic equivalent of JK Rowling with the infinite canvas on your side. This is the allure of "micropayments", the other one of McCloud's major innovations that Reinventing is remembered for, and the one presented first – and it, like the more minor innovations, is actually pretty close to fruition. Micropayment systems allow transfer of money to be cost-effective even if all you paid was a penny or two, and McCloud proposed that they would work more like cash where the seller doesn't get nailed with an additional flat-rate charge on the transaction.

(The idea is alluring enough that a recent wave of speculation that micropayments could be used to save newspapers, and a very recent appearance by McCloud's fellow micropayment advocate Walter Isaacson on The Daily Show, has crept it into the news on its own right, in a far broader sense than the relatively small world of webcomics, far more recently. In one sense, I actually lucked out with my delays, if only because Part IV can get the "my comments on the news" label!)

Although it sells what McCloud calls "eyeballs" rather than "bits", and it allows deposits of $5 or more and withdrawls of $10 or more (outside of the buying and selling going on within the system itself), Project Wonderful (which runs the advertising on this site) runs much like this, and it's telling that it has its origins in the webcomic field. (Micropayments are distinguished from subscription services in that subscriptions charge based on time, and micropayments based on each time you access the product. Arguably, that disqualifies PW from technically being a micropayment service - and qualifies most other advertising services.) Several other businesses, indeed, have sprung up to carry out micropayments for actual content as well; it's pretty much the norm in MMORPGs to some extent, and it's arguably the bread and butter of services such as iTunes, which allow song purchases for about a dollar each. PayPal is rarely used as an actual paywall, and I'm not certain how much like "cash" it really acts, but it still comes pretty close to having the mechanism for micropayments, if not the reality (and many webcomics do use it as a donation scheme). So to some extent or another, micropayments have arrived.

So why aren't webcomics using them? Especially considering McCloud, so much of an eminent and dominant thinker in the webcomics field, predicted they would?

I'll answer that question – and that answer's implications for journalism's consideration of micropayments – in Part IV tomorrow.