There’s been something of an explosion recently in discussion over the merits of design in the community of blogs I read - it was precipitated by a post by Robert Scoble that was commented upon by Scrivs; since then, the thread has been taken up in one way or another by Zeldman, Design by Fire, and any number of others in comments and blogs all over. Frankly, the whole thing makes my head hurt; while Scoble is generally a fine author (I read him occasionally, but his focus is not my own), his argument here stinks and it really isn’t worth the time people’ve been spending on it - especially when the rebuttals they produce stink just as badly. I’d rather we all just get along with our lives doing visual design or not as our individual preferences guide us - but since that doesn’t appear to be a viable possibility for the immediate future, I thought I’d give a quick breakdown of the primary argument from a (somewhat informed) philosopher’s perspective.
I’ll go through Scoble’s original post, and I’ll include comments made by him elsewhere when they add to the argument. Here we go!
[be warned, this is a long post]
Lars Holst is doing a series on “well designed weblogs.” Except his is not well designed.
This is the opening of Scoble’s argument - hmmm, looks like it’s going to be an ad hominem, like saying “Lars says X, but Lars is ugly, so pfft.” Beyond that, does it matter what Lars’s blog looks like? Has he put his own on the list of well-designed weblogs? Are people only allowed to judge that in which they have a publicly demonstrated competency? Should the judge in the Gaiman-McFarlane lawsuits (via Neil’s blog) have been an artist or a writer?
Anyways, continuing on….
Why do I say that? He’s using a gray font over a white background. That makes it a lot harder to read.
OK, so we’ve moved on from the ad hominem to a slightly more substantive point - Scoble appears to be identifying ‘well-designed’ with ‘easy-to-read’. Of course, we don’t have enough information here to evaluate Scoble’s claim; what gray is Lars using? #333 isn’t significantly more difficult to read over white than black.
Beyond that, however, Scoble appears to miss the point that readability is only one aspect of design. Granted, for a blog it’s highly likely to be a very important aspect, but it’s still only a part of the whole. What about giving the user a feeling of control? (This will come up again in a few moments.) What about aesthetic appeal? What about legibility, as opposed to readability (see Bringhurst’s The Elements of Typographic Style for the difference? All of these can conflict with Scoble’s apparent conception of readability as the highest aim of design.
Well, at least we know what Scoble thinks good design is… or do we?
Personally, design in weblogs is over rated. Make the darn thing easy to read. Forget making it pretty.
Wait! Just a few sentences ago Scoble claimed that Lars’s blog was not well-designed because it was hard to read, which implies that good design just is (at least mostly) readability. Here, it seems that Scoble is identifying ‘design’ solely with its aesthetic aspect. Of course, this is a common error, but it’s still an error - design is not just about making things look pretty. Design is about solving problems, and in many cases those problems include getting people to read something. In those particular cases, readability is an important aspect of the overall design, but even then it is rarely the only aspect of the overall design. It seems that Scoble’s view of design is a bit confused.
If you try to make it pretty we’ll just read it in our RSS News Aggregator which doesn’t pass along design info anyway.
As others have pointed out, the vast majority of the public has no clue what an aggregator is; could it be a bad idea to make major design decisions solely on the basis of a vocal minority?
Personally, my weblog might look ugly when compared to those that Lars points to, but do some analysis of how hard it is to read.
OK, here we’ve got Scoble trying to set himself up as an authority - I may not have a pretty blog, but by my definition of good design (readability) I’m the tops!
This bears some scrutiny. Why does Scoble think his site is so easy to read? He gives some details in one of the comments to his original post (in describing why another site is more difficult to read):
Look at seriocomic’s blog. He uses justified text. That’s harder to read.
He uses column widths much wider than mine (per character). That’s harder to read.
He uses a font that’s much smaller. That’s harder to read.
His font size isn’t overrideable in IE. Mine is. That’s frustrating.
He sets the font for me, where I give you control of the font (I let you change mine by changing the default font in the browser). That makes my site easier to read.
So we’ve got: ragged-right is good, columns wider than mine (Scoble’s) are bad, much smaller fonts are bad, adjustable font sizes are good (though it’s not clear if Scoble thinks this makes his site easier to read or just more enjoyable to use), and user-controllable fonts are good.
This is a mish-mash of fact and fiction; taking each claim individually:
- Ragged-right is in fact more readable than ragged-left or justified text for extended passages. See, for example: Trollip, S. and G. Sales. “Readability of Computer-Generated Full-Justified Text.” “Human Factors”, 28, 1986 pp. 159-164.
- Line length (column width) is indeed important for readability, but the solution is not to fix column width in absolute terms (like Scoble does); instead, line length should depend on the size of the font. For an alternative, improved method, check out Max Design’s em-based method. Interestingly, the ideal line length for readability and the ideal user-preferred line length differ - which goes to show you that the user isn’t always right about what’s best for her (a point which will come up repeatedly in the following analysis).
- Font size is important to readability, but it’s not at all clear that adjustability adds anything if you set it correctly in the first place; see this study in Usability News for a breakdown of reading speed (note that is not all there is to readability) by font face and size - turns out 12pt is generally read faster than other sizes, so shouldn’t Scoble just set his font size there and user-font-size preferences be damned? Ah, but continue reading that study and you’ll see that there were significant differences in perceived legibility and attractiveness that had no relationship to reading speed (or to reading efficiency, for that matter). It seems, then, that Scoble’s adjustable font size ought to be seen as more of a nod to user enjoyment than to readability. (As I mentioned above, however, it’s not completely clear that he really thinks this feature increases readability.)
- Font choice is also important to readability - clearly some fonts are more readable than others (*cough* Vivaldi). But this is where Scoble really drops the ball; he leaves the choice of what font his site appears in to the user. Does he expect his users to understand the finer points of typography and the psychology of perception? Does he expect them to understand that the relatively low resolution of a computer monitor renders serifs (so useful in 600+dpi print) something of an eyesore and an impediment to reading online? Does he expect them to even know how to change their default font to one highly suited to reading? This is where the rubber of user preference really meets the road of empirical data, and I’m afraid Scoble is left in the dust.
So where does the argument go from here? Continuing on with the original post:
Also, look at how long it takes for the average “well designed” blog to come up in the browser. Mine is fast. Some of theirs? Well, not so fast.
OK, here’s an empirical claim. Where’s the evidence to back it up? Here’s some data for various blogs’ front page weights (minus images) for 26 Feb 2004 0839 +0500:
- Zeldman: 19.43kb
- SimpleBits: 24.9kb
- Kottke: 16.71kb
- hicksdesign: 23.81kb
- Stopdesign: 26.71kb
- Dive into Mark: 3.94kb (wow!)
- Scoble: 65.93kb
Whose blog do you think will take the longest to load?
My site is #49 on Technorati’s list (out of 1.4 million blogs). Why do I have readership while many of these “well designed” weblogs don’t? Cause I focused on making my blog readable. If you want a pretty looking site, go somewhere else. (Most of the top 100 are “ugly” blogs — but they all are easy to read).
Others have jumped on this point well enough; suffice it to say that I think Scoble does himself a disservice here by implying that he has readership merely because his blog is readable. He consistently provides content that is fresh, useful, and interesting to a large number of people - that’s why he has a large readership (and why Zeldman and Mark Pilgrim and all the other A-listers have the readership they do). It’s just too bad that he missed the mark with this post.
Before I wrap this up, there is one other piece of Scoble’s argument I’d like to address. It came out in another comment on the original thread:
And, don’t assume I don’t know a bit about design and aesthetics. In college I won several newspaper design awards.
A couple of things here: I really, really hate the argument from authority that Scoble is trying to press with that last sentence. So what if he won awards - I won a He-Man phone for designing an action figure when I was a kid (I really did), but that doesn’t make a lousy argument better. It’s generally a good idea to provide empirical data - not memories of one’s halcyon college days - to back up empirical claims. Heck, Scoble even mention his “research” in one of the first comments on the thread - where is it? How does it address the well-known findings I cite above? That’s the way to respond to critiques, not by pointing to a set of trophies on the mantle.
PS. Hey Scoble, your comments page uses black text on a gray background. That’s not very readable.