Consistent reading time for comic pages?

I had a link sent to me recently asking about this blog post that claims all comic pages regardless of content are read at 3.75 seconds per page. A relevant section:

I read in one of Frederik Schodt’s excellent books on manga that a study concluded that readers spend an average of 3.75 seconds on a comic page. My own observations of myself and others has led me to believe that time frame to be fairly consistent, by which I mean not dependent on the contents of a page. Unless a writer really creates an absolutely confounding monologue or an artist completely botches an integral sequence, readers do not seem to change their flipping speed for “difficult,” wordy, nor beautiful pages. This yields somewhat counter-intuitive results, in my estimation. Single panel pages, which should ostensibly be flown through, allow one image to be lingered on or “drunk in” because that one drawing is granted the full 3.75 seconds. Pages with many panels, taken to the extreme above, should require a slower, more contemplative pace. But they do not. They seem to clock at the same 3.75, meaning the eyes need to whip through these images to make it in time.

He then goes on to advocate different strategies of layout based on the idea that readers will go through it at this magic time of 3.75 seconds. Since I wrote a lengthy counter-rebuttal to this claim, I figured I ought to post it here too.

According to the science I've seen, this does not seem to be the case. The amount of time people spend on each individual panel varies based on how much information is in it, it's order in the sequence, as well as possibly size of both panel and page, and a whole page time varies definitely the way the page layout is organized.

From a very general study of my own relating to times it takes 4 panel comic strips to be read, I found each panel at an average of 1.5 seconds per panel when readers press a button to advance through panels. But, it does vary per position and narrative structure — first panels are consistently slower, panels after major events much slower. However, if you just take that average and multiply it by 4, that gives you 6 seconds for one 4 panel Peanuts strip that has no words in it.

In my last study, I found reading times varying between .6 and 1.8 seconds per panel (small times for panels that had very little information, such as blank panels or those with just action stars), with the full 6 panel strips clocking in around 6 ±2.5 seconds.

Plus, the *uncited* study that was mentioned in the blog is for manga (and if I recall correctly, Schodt also doesn't cite the actual study), which consistently 1) use slightly less panels per page (my corpus study — "Cross Cultural Space" — showed both American and Japanese books to have 5 panels per page, but manga had a lower standard deviation), and 2) use less balloons per page. Furthermore, eye-tracking studies show that fluent readers skip over far more balloons than non-fluent readers — so, less balloons means less reading time, especially for fluent readers.

The poster here then says that he finds this time to be consistent to his own experience — but you can't know such a thing from anecdotal evidence. You would have to have measures to substantiate it.

And, even if it were true that on average pages are read at a pace of 3.75 seconds — which, I imagine there is some average time out there if one were to crunch all the numbers — there is no way that we would feel the need to allot different time to different panels based on some intuitive feeling that we "want" to read each page in a specified amount of time.

Rather, the time it takes to read a page all depends on its content and the fluency of the reader.

Elements of Page Layouts

Belgian graphic design student Chris Vosters sends along this 7-page pdf graphic essay that expands on my paper, Navigating Comics, on how people move through page layouts.

Chris does a great job of categorizing both explicit and implicit ways in which an author can direct the flow of movement across a page layout without relying on the rule system my paper describes.

I highly recommend checking it out!

Navigating page layouts = defining "comics"?

One of the topics I debated closing out my new essay on page layout (pdf) with was its relationship to McCloud's definition of "comics."

As most know, McCloud's definition is that "comics" are "juxtaposed sequential images in deliberate sequence." Yet, he never places any constraints on that. He means all sequential images are "comics" — regardless of the characteristics of content.

On the one hand, you can say that his notion of closure demands that there is some content to those panels, and that closure is the underlying force behind his definition. However, there are places where this appeal to content is a bit slim. This is most apparent in his panel transitions, where "non-sequitur" as a catch-all for anything his other transitions don't cover. Or, in claiming that empty panels that represent "time" still maintain the essence of "comics." In other words, it doesn't matter what's in them, as long as they're in sequence.

If really all McCloud means by "comics" is that two graphic units are place next to each other, is he really just talking about the system I've proposed for layouts? This system just tells people how to navigate through a comic page — how to read from one panel to another. And, since my experiment used blank comic pages, it has nothing to do with content — just like McCloud's definition.

So, if navigation between panels is really all he's talking about for "comics," isn't that a little... I don't know... unremarkable?

It's also interesting in this interpretation, since (as I note in the paper) McCloud's notion of the Infinite Canvas essentially desires to simplify this navigational structure. Food for thought...

Navigating Comics Plus

Since various concerns have popped up here and there about my latest essay on page layouts (pdf), I figured I should take the time to reiterate responses to some of them here...

First off, the types of navigation I talk about here are absolutely intended to be part of a broader network of how people move through layouts. Certainly, panel locations aren't the only influence on people's movement through layouts. Among the other things potentially are color, content, etc.

What I was trying to get at is that people do have idealized preferences for reading directions given various conditions and that those preferences emerge *even in the absence of content* in non-left-to-right ways. My suspicion is that people use these sorts of preferences from panels as their first influence for navigational choice, which can then be further influenced by content and maybe color. However, that's something that would need to be empirically tested.

Another reason for creating this study that I didn't mention in my last post was that the year before I did the study, John Barber came out with his own paper about layout (unfortunately now taken offline). While he had some great ideas and observations, I disagreed with his basic claim that layout and meaning were expressly tied. Layout and content most definitely can be connected in important ways, but I think that this experiment nicely shows that they are governed by separate (yet interfacing) systems.

Finally, several people have been curious about instances where panels do not have borders at all. I do mention borderless panels, but only in a footnote, where I basically say "more testing needed"! My guess is that certain permutations follow the same principles that I outline in the paper, but others lead to greater violations because of the ambiguities they create (ahem, testing needed).

As with many of my papers, this should just serve to lay the groundwork for future work (by me or others). I'm still convinced that the really cool stuff will come far down the road!

Essay origins

So far I've been very pleased at the response to my latest essay, "Navigating Comics", on how people navigate through page layouts (pdf). As several responses have been rolling in via email and elsewhere, I intend to do a post soon addressing concerns in that feedback. However, I think it'd be informative to first talk about the origins of this paper.

Back in 2003 when I was drawing our political book We the People, every now and then my editors would tell me they had trouble knowing exactly where to go in the sequence. Often this happened in consistent situations (like what I call "blockage" in the paper).

Most of the times I'd either simplify the layouts or make some graphic fix (like a trail) to indicate a clearer path. However, it got me thinking... My editors were quite a bit older than I was, and weren't all that experienced comic readers, so I wondered if this lack of experience mattered in their reading habits? (or if I was just needlessly making things difficult)

So, I designed this study to test that. I had a booth at ComicCon 2004 that year to promote the book and my other works, so I designed a simple pamphlet people could make responses in and tested people throughout the convention.

I could tell immediately that the results would be interesting, I just had to wait another three years to learn the statistics necessary to show them (d'oh!). The theory with the tree structures predated the experiment by at least a year, but it didn't really say much without knowing about people's actual preferences. It's exciting to see that my suspicions for creating the experiment were borne out in data.

Every now and then I get a response to my work along the lines of "Why do theory? Why not do something related to praxis?" While theory can be interesting, enlightening, and much of science is simply about discovery without practical applications in mind (ex: penicillin), another reason is that theory can sometimes wrap back around on praxis. I like to believe that this is one of those cases, especially given that it came from those origins.

New Essay: Navigating Comics

I'm very happy to announce that I have a new essay online: Navigating Comics: Reading Strategies of Page Layouts (pdf). This paper reports the findings of an experiment I conducted looking at how people navigate through comic pages. The big finding: people don't just mimic text going left-to-right and down.

The full abstract:

The spatial domain is often considered to be non-linear, given the analog nature of visual information. However, the visual language of comics defies this by siphoning images into a deliberate reading sequence. Most often this sequence is assumed to be read in an order that mimics text: left-to-right and down, a “z-path.” However, several scenarios can violate this order, such as Gestalt groupings of panels that deny a z-path of reading. To investigate these concerns, an experiment asked 145 participants to number empty page layouts in the order they would read them, and showed that readers use an alternate strategy extending beyond both the traditional “z-path” and Gestalt groupings to navigate through comic page layouts.

I should also say that this paper took a very long time to complete. The study was run in 2004 and I know I've been talking about releasing the results all the way since last summer. Thanks to all who participated in the study and to all for being patient as I finished it up!

Enjoy!

Podcast: The Functions of Panels

The last podcast I did with the VizThink folks was so fun I decided to do another. This one is about the various functional roles that panels play in the visual language used in comics. Among the topics I hit are:

• focusing information within panels
• navigating page layouts
• visual "storytelling"
• text-image relationships

It's a slightly pared down and also expanded (at the same time!) version of the talk I gave at the VizThink conference. Enjoy!

Graphic safety for airplane crashes

One of the reasons I've been a bit MIA from the blog lately is because I've been traveling a lot. On one flight, I noticed some interesting things in the safety manual. First, let's look at a section that isn't too bad:



This part shows a nice step-by-step for what to do in the event of increased cabin pressure: how to put on an oxygen mask. Note the "no smoking" sign to the side — smoking with one of those things would be bad. Here, the numbered panels may serve two purposes — it gives a reading order along the "z-path", left-to-right and down, and it also gives an order to the procedure. You need to follow these steps in order to put the mask on correctly.

Comparatively, this one from right below seems very strange:



If the numbers are meant to direct a stage-by-stage process (or even just an order for reading the panels), then perhaps I can paraphrase the overall meaning (and the signs to the side):

"In the event of either a land or water crash, first, African Americans need to cover their heads. Then, blond women need to grab their legs, followed by white men who should force their children down next to them into submission. Lastly, once all the other groups are safe, pregnant women should brace themselves. They're the bottom rung in our concerns."

Ok, so maybe I added a few minor embellishments, but my point should be clear: this sequence has no need of numbers for any purpose. Really, all people should brace themselves as quick as possible in one of these ways, not in any particular temporal order. And, despite that people probably would read it in a z-path anyhow, there isn't any need to read these in the numbered order here either.

I love this example especially because it provides great support for why sequential images are not always a sequence in time. Each of these panels simply shows a different viewpoint of a broader scene. It's a shift in Space but not in Time. Truly, you could rearrange these panels in any order and still maintain the same meaning — clearly a sign that no "time" passes across the panel boundaries.

Eye-movements reading comic pages

Omori, Takahide, Taku Ishii, and Keiko Kurata. 2004. Eye catchers in comics: Controlling eye movements in reading pictorial and textual media. In 28th International Congress of Psychology. Beijing, China.

A team of Japanese researchers perform two experiments examining eye-movements across comic pages to show that both page layout and balloon placement factor into how readily comic pages are read.

They found that, for an average of 8.5 panels per page, there are an average of 20.3 fixations. Most of their study focuses on panels that were skipped over for one reason or another, and examining modifications made to see whether they would still be skipped over.

There were two major changes that showed significant effects in decreasing the rates that they were skipped: balloon position and panel layout.

The first factor in skipping is if a panel is followed by another panel with dense text. They altered the "dense balloon panels" by distancing the balloon further away from the preceding panel. This change resulted in a significant reduction in the times that the preceding panel was skipped.

The other major factor was when panels were vertically stacked next to a long adjacent panel (what I call "blockage"). The lower panel was often skipped so the reading follows the horizontal path. When altering these layouts to make the panels horizontally arranged, the rate of skipping decreased. However, this phenomenon was only observed in a couple of scenarios (6 instances) and they don't mention how many of these skips lead to going back and rereading the skipped panel. They also don't state how many times "blockage" occured and didn't lead to skips.

Slight decreases in skipping were shown for moving characters' positions within a panel, though not to high percentages (significance is not shown).

Additionally, a recognition task asking whether various panels were or were not in the comic showed significant increases in accuracy for the modified versions. No differences were shown in accuracy of reading comprehension for the story.

While they state that their participants all had comic reading experience, I wonder the degree of "comic fluency" that they have. The desire to jump towards panels with dense text insinuates a focus more on text than on the visuals, which was characteristic of a naive comic reader's eye-movements compared with an expert reader in Nakazawa's eye-tracking study.

Further, this study supports an idea that "blockage" situations are harder to process (evidenced by the skipped panels). However, I have empirical evidence from my own experiments on page layout (to be posted soon hopefully) that following the vertical path of panels is the prefered reading path, and that preference for it does depend at least partially on expertise in comic reading. Also, their studies used only the results from 25 subjects (half seeing modified versions half not), whereas mine used 145, so looking at a broader populace would be good here.

Hypothetically, I could tackle this issue myself, since the lab next door to mine has an eye-tracker at my disposal. We'll see... I have a few other things on my plate right now.