Omni Chat:
Eileen Gunn and
Steven Johnson

Eileen Gunn: Welcome to OMNI's E-Media. Our guest tonight is Steven Johnson, co-founder and editor-in-chief of the lively insider Webzine Feed. Steven's new book, Interface Culture, is a cogent examination of the history and aspirations of the "graphical user interface" – the window into the virtual space behind your computer screen.

In his book, Steven gives us a peek at the Primal Scene of GUI (Doug Englebart's first mouse demo), and a tour of its adolescence – from the Mac to the Web, with a nod to the importance of game developers – then moves into the McLuhanesque territory of how the medium influences how we think. Then – and this is the part I'd like to focus on in the discussion – he jumps off into the future of the medium. Where is it going? How much will interface-design be influenced by visionary individuals, and how much by venture capitalists and entertainment interests?

I'm particularly interested in having a lively conversation during the discussion part of the show (the latter half ). This is a chance to bounce ideas off one of the most influential people in cyberspace, and I think we can have a discussion here that will reward re-reading. I've invited a few inquiring minds with an interest in cyberspace design to join us and participate, and I hope all of you in the audience will join in too.

Steven Johnson: I'm here in the "room", whenever you want to begin, Eileen.    
Eileen Gunn: Great. Hi, Steven. Hi, everyone. Steven is here, and I think we're ready to get started. I have some questions, to get the ball rolling. In about half an hour, we will open the cybertalk flood gates, and I hope that a lot of you will also have questions and comments at that time.    
Steven Johnson: Sounds great!    
Eileen Gunn: Hi, Steven. Your book is wide ranging one, covering not only the history of computer interface design, and its possibilities in the future, but also the changes in human imagination since the dawn of the Industrial Revolution. We could do an entire interview on just a chapter, or even a page. I'm going to do a little selective surgery here, but please excuse me for slicing it up into infobites. One of your concerns is the actions by which a society tells itself who it is. Novels, movies, architecture - these are all ways in which, through the artist, we examine who we are and where we are going. Does the computer interface - cyberspace specifically – tell us anything about who we are?    
Steven Johnson: Yes. I think it tells us a great deal. First off, the *need* for the computer interface – all those visual metaphors and icons and windows – tells us something about the information overload that most of us face these days. Secondly – different types of interfaces emphasize different qualities in ourselves. This interface, for instance, gathers together a bunch of strangers into a virtual space, and lets them communicate with one another... That's a rare thing in the history of recent technology, which has been largely about keeping strangers separate.    
Eileen Gunn: You pre-empted my next question – which was, What on earth does =this= interface say about us?    
Steven Johnson: Well, let me then build on my answer – the other thing that's worth noting about this interface is that it's text-based. It's one of the little-noted – but crucial – elements of the 'net revolution: the renaissance of literacy that it has ushered in, that people are writing more via chat and e-mail, and reading more because the web is still largely a textual experience. Of course, some of that text is *hypertext*, which gives it a new dimension, and some of the e-mail and chat conversations are just that, *conversation*, less formal than old-fashioned letters...    
Eileen Gunn: I think of media as being extensions, in a McLuhanesque way, of people and to other people. How does that translate to your keeping people separate?    
Steven Johnson: I think of media in the same McLuhanesque way, but there's no real communication between strangers in media like television or the radio. In fact, strangers are discouraged from interacting because 20th-century technology has pushed people to the suburbs and their living rooms, and out of the traditional urban cores, where strangers used to interact more regularly.    
Eileen Gunn: Good point. Do you think there's much future for text-based communication?    
Steven Johnson: I certainly hope so. In my darker moments, I think all the powers that be in Silicon Valley and Redmond are conspiring to turn the Web into another version of television, which seems silly to me, given that we already have television. But in my optimistic moments, I like to think that the textual element of the Web will remain essential... ... particularly if the channels remain two-way. In other words, as long as people can write back to what they're reading, they'll keep reading. (It also helps if what they're reading uses hypertext well.)    
Eileen Gunn: But text is so much more work! And not everybody types well. (I'm sure this interview is much more work for you than it would be if you were just talking.) Do you envision a graphical interface that would provide this same ability to exchange ideas, to converse?    
Steven Johnson: For real-time conversations, yes. there are already a number of software packages out there – like the palace or comixchat – that add a level of visual and gestural information that enriches the whole experience. but for asynchronous communities - like the WELL or most bulletin boards – I think text is still the best solution, and will continue to be. You know, the one thing that occurs to me is that we're uncomfortable using so much text because we're still trained on television!    
Eileen Gunn: Your book actually focuses more on the graphical than on the text-based interface. And that seems to me a more likely direction for the web to go in. How do we give people the vocabulary communicate in graphical ways? The comixchat is a start, I suppose. It seems that it wold be necessary to teach people to communicate in completely different ways. Or for them to teach themselves. Like comixchat, we could give people pictures that they could employ to convey whatever the pictures mean to them. That might be something that would grow and change. Especially if they were non-representational pictures.    
Steven Johnson: Certainly giving people a visual way to represent themselves would be a major improvement, as in the onscreen "avatars" that everyone's talking about. But also, it would be wonderful if you could represent clusters of communities visually, almost like neighborhoods in a city. The BBS software we used to use at FEED gave you the option of attaching a little icon to your post – basically a thumbs-up, happyface, question mark, angry face, etc... People seemed to use them, but they were a little too cartoonish for my tastes.    
Eileen Gunn: I was very interested in your description and in what I've seen myself of the magic-lens approach to interface design. Would this be a way of enabling different communities to relate to one another?    
Steven Johnson: Interesting question about the lens. It's been traditionally used to filter out information – drag the lens over the map, and you'll only see the roads you're looking for... I'll have to think more about that – maybe you could use the lens to highlight thoughts posted by people you were interested in...    
Eileen Gunn:



(The magic lens approach to interface design adds classic Gibsonian 3-dimensionality to the interface: you move toward data that interests you, and it gains in detail as you get closer to it.)    
Steven Johnson: Actually, what you describe there is a little more like Apple's HotSauce browser, although the two are related. In HotSauce, you "fly" through the 3-d space of your hard drive (or the web); with the lens you drag it over data and it changes depending on the criteria you've chosen to accentuate... The classic magic lens implementation is in Corbis' Leonardo CD-ROM, where dragging the lens over Da Vinci's backwards handwriting, translates it into English and inverts it...    
Eileen Gunn: HotSauce is what I was thinking of. Duh. The problem with icons is that the meaning is *intended* to be unambiguous. What we need are more ambiguous icons.    
Steven Johnson: Excellent point about ambiguous icons. It's a problem with interface design in general right now. All it knows how to do is make things more simple. But that's not always what we want.    
Eileen Gunn: I think we're ready to start the interactive part of the evening. Please feel free to start throwing out questions, and Steven can field them as they come. In order to keep things moving, Steven and I will keep talking, while you are preparing your questions. Please don't be shy!! This is a chance to talk about the future of cyberspace with one of the guys who's imagining it. Feel free to "interrupt." The interface is slow and linear, so nobody will actually perceive it as an interruption.

Let's move away from the conceptual to the actual. What site out there on the web now embody a design that is truly 3-dimensional in the Web sense, rather than simply being linked pieces of paper?

Steven Johnson: Depends what you mean by 3-d. There are definitely some intriguing VRML-based sites out there, though they're still pretty slow and blocky. You can get to some of the best ones through Mark Pesce's home page, though I don't have the URL on me right now.    
Guest: Isn't worldschat a 3-D based chat? With pretty good representational figures?    
Steven Johnson: But there are also some very intriguing hypertext projects that are 3-dimension in a looser sense, and that are far more complicated than just "linked pieces of paper." (Our Documents at FEED for instance, where commentators and readers annotate a text document with their own comments...) Worldschat is another good example. The other example are the online versions of games like Quake or Myth...    
Eileen Gunn: The VRML-based sites I've seen have been too slow on my 486 machine to capture my imagination. Are there sites that share FEED's aspirations to nonlinearity of idea flow?    
Steven Johnson: From what I've seen of future chat rooms to be released, they are primarily using mics and cams rather than text or representational figures (I agree with you that they are a little silly at times..) I like text, but my main reason for liking it that is clear (mics tend to be choppy) but I think they can overcome that transmission problem eventually – do you think they can make it better, and therefore more viable?    
Eileen Gunn: I admire the density of FEED's overwritten documents.    
Steven Johnson: There aren't as many sites as I would have expected, Eileen, though people seem to be coming around to the importance of nonlinear structures. (The Atlantic monthly Online has started a roundtable series that uses more branching paths, somewhat like our Dialogs or Documents.)    
John D. Berry: The odd thing about text-based communication in the electronic world is that "text" has turned into something completely divorced from what it looks like. People talk about how they "just want it as text," when in fact the text that they see is a visual lowest-common-denominator that isn't very easy to read or very flexible to write in or design visual information in. (You can probably tell that I work with type.)    
Steven Johnson: As for guest's comments about using mics: voice-mail based BBSs are probably around the corner. Is that something you'd like, do you think? I can't make up my mind.    
David Levine: The problem I see with VRML-based and other 3-D, nonlinear user interfaces is Difficulty-of-Use. When our input and output devices are only 2-D (admittedly an improvement over the 1-D text interfaces common just a few years ago), any attempt at 3-D manipulation feels like random, uncoordinated groping. Is there really any point to implementing a 3-D interface with the current generation of I/O devices?    
Eileen Gunn: Right now, an enormous amount of technical knowledge is required to take advantage of what it's possible to do, such as to create the overwritten text on FEED. Does the complexity of interface design mean that a single-author work cannot be done in this new medium?    
Steven Johnson: Good point about the quality of text, John. Doesn't it amaze you that my onscreen fonts haven't improved in readibily one pixel since the Mac came out in '84! That's thirteen years! Doesn't that just drive you crazy?    
Eileen Gunn: I think David and I are asking the same question.    
Steven Johnson: David's right about the i-o problems today, though when I think about this issue, what comes to mind are the kids playing Mario 64, who've spent months roaming around in 3-D spaces, and feel supremely confident learning the rules of new environments...    
John D. Berry: There are more possibilities available now – things like anti-aliased type and more typefaces actually designed to be readable on-screen – but the original screen fonts still stand out above most of the later ones. It's a scary thought.    
Steven Johnson: I'm actually in the middle of playing Riven, the sequel to Myst right now, which is related to this question I think... Anyone else playing it?    
Audience member: (From previous to Steven:) I think the reason I can't make up my mind either is because the desire to be more removed from humans by text, and the computer interface itself, comes in conflict with the human need to be closer – which drives us to want images and sound? It's still removed... but less so    
Eileen Gunn: I'm afraid to take that first step on the slippery slope of Myst.    
Steven Johnson: The thing that strikes me playing it is that the game is a wonderful example of a virtual world – an interface, in other words – that is *deliberately* difficult to use. It's hard to figure out how all the machines work, and how to get from one space to another. And the fact that it's hard is what makes it fun!! But to think of it as fun, you have to be more or less fearless about "cracking the code" of complicated interfaces...    
Audience member: It's not just a Communication Application, it's an adventure.    
Eileen Gunn: My site, The Difference Dictionary, is based on the fact that making connections and figuring things out is about as much fun as you can have. I wish I knew enough to include the kind of worlds that Myst is an example of. But my VRML skills are sadly lacking.    
Steven Johnson: I may be wrong person to ask about voice in BBSs, because I *hate* the phone! Actually, my problem with the telephone is directly related to my love for e-mail. I started screening my calls 100% of the time, a few months after I got hooked on e-mail. And now I'm terrified to answer the phone cold.    
David Levine: I haven't actually played any of the Nintendo 64 games, but my experience with previous generations of 3-D games hints that their 3-D ease-of-use may derive from giving the player an extremely limited range of motion (controlled by a fairly simple joystick or steering wheel), then jazzing it up to make it feel more 3-D than it really is. By "jazzing it up" I mean tricks like being able to move on top of a tall object (actually a simple X-Y transform with a fixed Z component) or a zooming point-of-view. A general 3-D GUI couldn't use these tricks. Or could it?    
Steven Johnson: What's the URL for your site, Eileen? I should add here that FEED's URL for anyone who cares is    
Eileen Gunn: But do you screen your e-mail?    
Audience member: I've never seen a BBS, these are web based, or private server chats.    
Eileen Gunn: Do you think that most people on the Web are ready to deal with a truly non-linear approach? What's holding things back – the designer or the market itself? (But isn't this always the case?)    
Steven Johnson: You're right about most 3d games, David, but it's not really true for the Nintendo 64. You should try it – it's so 3D, in fact, that the most intriguing games, like Mario, make manipulating the camera angle a signficant part of gameplay. (Since you can move in any direction, it's a lot more important...)    
Eileen Gunn: The Difference Dictionary, which is a nonlinear guide to Gibson & Sterling's "Difference Engine," is at    
Steven Johnson: I think, Eileen, that it's crucial for us to remember that the technology is now advancing faster than our perceptual or interpretative skills are, and that we sometimes have to be patient about these things. I think folks are just learning how to really navigate through complicated hypertext spaces – it's a genuinely new type of reading...    
Eileen Gunn: Re: David's question about 3-D GUI, I was working on the Fighter Ace game site due this winter on Microsoft's Computer Gaming Zone. And the charm of the game is the effect of motion rather than the complexity of the motion.    
Katherine Enos: Voice-based styles of web communication may be well-suited for certain things, among them BBS-style chat or sexual commerce areas, but i think that the text approach is more inclusive and easier on bandwidth. frankly, it's a joy to live in a cyberspace where the printed word is given some prominence. i never cared for the idea of comix chat, for example.    
David Levine: Navigating in hypertext is just as hard as navigating in 3-D. One of the main hills we have to climb is that, just as you must read broadly to become a good writer, you must navigate a lot of hypertext (good and bad) to become a good hypertext author. As a society, our hypertext navigation experience is very limited, so our hypertext creation skills are weak. Things will improve as the pile of existing hypertext gets deeper (compare today's online help, bad as it is, with that of ten years ago and you'll see what I mean). We stand on the shoulders of midgets.    
Eileen Gunn: What a retro-radical you are, Katherine!    
Steven Johnson: Yes, I think if there's a genuine art form coming into its own here, it has something to do with the pleasure, and richness, of moving through these virtual spaces, which is why the best achievements so far (Myst and Riven for instance) are really closer to architecture than, say, movies or novels...    
John D. Berry: "Higher and deeper," I believe, is what you meant to say, David.    
Steven Johnson: I'm with you Katherine! (kind of)    
Eileen Gunn: Does this mean that I should give in to my occasional urges to play Myst? Is this essential for my moral & spiritual well-being in cyberspace? In bed?    
Katherine Enos: Well, I adore text, I am a writer too... but the public seems to want sound and images (because they don't have text skills).    
Audience member: At least from what I've seen in the development of com apps presently in R & D    
David Levine: "Higher and deeper", John? Preferably, but there is much to be learned even from bad hypertext. In fact, the bad examples may drive innovation even faster than the good ones.    
Steven Johnson: Not so sure about the bed part, Eileen. But you should definitely check it out. You might as well start with Riven though.    
Katherine Enos: Forgive my ignorance, but in Comixchat, is the user allowed to "create" the character which will represent him or her? I guess I'm asking because it seems as though the ways in which you can represent yourself with Comixchat will be limited to what the corporation has allowed you.    
Eileen Gunn: We're running out of time. This has been great. Thank you very much, Steven, participants, and viewers for joining E-Media tonight.    
Steven Johnson: I had a great time. Thanks for having me!    
Eileen Gunn: Steven, we can run over if you have the time. Do you have any summary statements you want to make?    
Katherine Enos: I'm Game. I had a juicy question that this *urp* chat deleted on me.    
Steven Johnson: Will you believe me if I say that I'm on a deadline for my review of Riven? *grin* It's true, I swear.    
Eileen Gunn: I think Steven's outta here. Thanks a lot for you time, Steven. Katherine, can your question wait 'til next month?    
Katherine Enos: No problem.    
Eileen Gunn: Thank you all for being here!    
Steven Johnson: Bye all.    
Katherine Enos: It was really an honor to meet you Stephen. *smile* Good evening.    
David Levine: Time flies while you're having fun!    
This interview was conducted on Omni Online, November 17, 1997
Copyright 1997, Eileen Gunn