Justin Hall's personal site growing & breaking down since 1994

watch overshare: the links.net story contact me

Date: Mon, 12 Feb 1996 11:59:58 -0800
From: Mark Petrakis <spoon@well.com>
To: Justin Hall <jahall@cyborgasmic.com>

Characters in the Interface

The following address was delivered by Mark "Spoonman" Petrakis at the ACM
Special Interest Group for Multimedia Conference in San Francisco on
November 9, 1995. The panel, chaired by Abbe Don, was entitled "What's that
Character doing in your Interface?" The other panelists were Jonathan
Steuer, Michael Powers, Marie Macaisa, and Tim Oren.
Mythology is born in the personification of nature, in the putting of a face onto raw experience. The sun becomes Apollo. The stars dance in constellations. The rivers are moved by water sprites. Zeus hurls thunderbolts, and Poseidon calls sailors to a watery grave.

Mythology is not a stylistic choice. Rather, it is human imagination reduced to its most common denominators, and as such, we cannot help but follow mythology's lead. We personify at every turn. Michael Jordan, God of the Nike Air, Michael Jackson, the Sybaritic Shapeshifter, Bill Gates, Father of Bob. We do the same thing on a personal level; Aunt Maria, the long suffering Saint, Little Georgie, King of the One-Liners, Cousin Steve, the Martini Man.

Such personifications may be as simple as the chubby little Letter O, or as complex as King Lear, or yourself, for that matter. From numbers to cars to housepets, there is nothing that we cannot personify. Trust me on this, I'm an expert.

In terms of the uses of characters, we find them used to drive all human ends; narrative, educational, entertainment, commercial, political, judicial. In short, characters take any and all imaginable forms for the purpose of achieving any and all imaginable ends. I hope that's a broad enough definition for you?

So, where does that leave us? Today, it leaves us staring into a computer monitor and wondering what our overactive imaginations are going to do with mega-bauds of bandwidth, body-mounted input devices, and ultra-sophisticated 3-D modelers. I'll tell you what we're going to do. We're going to use characters of any and all imaginable forms for the purposes of achieving any and all imaginable ends.

So with that in mind, I'd like to take the time I have to talk about self-representation, specifically as it might apply over networks. Avatar, I believe, is the word of choice here. An avatar stands in for you in a virtual world, similar to the way the little silver tokens mark your whereabouts on a Monopoly board. Add to that some kind of telepresent control, some limited mobility and functionality, and boom!- you've got yourself a boy, a digital doll, a virtual puppet, a doppelganger.

There is something to watch out for here and something to embrace as well. Having listened in over the years as my daughter gave voice and purpose to her legion of Barbie dolls, I can say that appearance is only one of many variables in determining how a character will be used. In my view, the quest for higher resolution is a path of seriously diminishing returns. Her Barbies remained part of the tribe, even with missing legs, chopped-off hair, and felt-tipped tattoos. In my view then, the iconic aspects of the character were more important in unlocking her imagination, than any sort of life-like-ness. WHAT IS FAR MORE SIGNIFICANT THAN APPEARANCE IN ENABLING CHARACTER IS THE OPENNESS OF THE CONTROL CHANNEL. So long as my daughter had full freedom to give voice and action to those dolls, there was no problem. But, to be perfectly honest, I have seen her perform this same enabling magic with a Reese's Peanut Butter Cup, so who can say? No doubt, she is the progeny of experts.

The situation would be very different though, if her Barbie's responses were all pre-determined, and all she could do would be to select one of several canned responses. She had those kinds of dolls early on. The ones that talked, with the strings or the buttons. They disappeared fast. But let's say, hypothetically, that through peer pressure or advertiser pressure, she wanted a doll that talked. Going one step further, let's say she was made to feel that any departure from that doll's list of pre-recorded responses was just "not done". What would we have then? We would have a problem, a problem of the very sort that I think we may well find ourselves facing in the next few years.

In my view, characters exist as a sort of symbolic shorthand for that which cannot be expressed in any better way. If a simple description can encompass what a character is about, then why go to the trouble of making a character? How could such a character help but be anything more than "cute"? It has no other purpose, than to be decorative.

I have been involved the past few years, as all of us on this panel have, in trying to integrate characters into the computational model. The intent as I saw it, was primarily to bring some of that "symbolic shorthand" to the task of navigating and authoring in virtual space. I have never had a problem with that. It seems an inevitable and ultimately interesting direction for programming to take.

The one potential problem that I have seen however, is the tendency among hard-focus types, to coerce characters into acting like humans. It is imagination's purpose to lead, I believe, not to follow. Characters are not people. They are inventions, and as such it is just plain silly to expect that all they should do is ape conventional human behavior. The French poet Apollinaire was speaking to a similar point early in this century, when he said that the theater was no more the life it represents, than a wheel was a leg. And the novelist Bruce Sterling recently echoed the same idea when he said that forcing the computer to conform to a life-like model is like judging an aircraft by how well it lays eggs. Personification is a magical sort of tool, and as such it should not be used ineptly or indescriminately. It is this lack of good judgement I believe that keeps the old anthropomorphic debate going round and round. I mean, it's silly enough that I have to carry a handful of presidents in my pocket. What more do I want `em to do - start talking to me? "Hi, I'm George. I'll be your quarter today." Clearly, not everything that can be personified should be personified. You don't need an expert to tell you that.

Good characters, like good avatars, enhance our abilities to communicate. In best case scenarios, they extend the reach of our personas just as the telescope and the microscope extended the reach of our sight. But I don't expect we'll reach that point without first having to wade through several laughable generations of software-based characters that will be one-dimensional, fairly undynamic, and way too cute. Like Bob, come to think of it.

As I see it, we have two choices. We can fall into line and follow marketing's lead, i.e. download the new Beavis and Butthead Avatar Set, oh boy!, or we can rebel, like the original Beavis and Butthead might have, make a stink, hack better alternatives, and basically insist on nothing less than full powers of control and personalization.

In dark moments, I sense a curious and rather insidious amorality as regards the future of personification in the virtual domain. To counter that, we must do with digital characters what my daughter did with her Barbies, namely instill them with something of our own spirit. The intent of true characters is to guide us through a process of self-discovery, a process more commonly called "Storytelling". It is not an obvious process, in that its underlying purposes can only be revealed over time. This process, which we have been engaged in from early infancy, asks us to inhabit the simulacrum of the character and to perhaps subsequently modify our own real-life actions based on experiences garnered inside that character-centered simulation. If we are to preserve the integrity of this process in the digital cross-over, we cannot do less than exercise full responsibility over our characters.

There is a significant social dilemma involved here that seems to me to necessitate our asking some direct questions. Who puts these characters there? What is the intent behind their appearance? What is the nature of the transformation that they ask us to undergo? Is it to relieve us of certain repetitive tasks? Fine. Is it to make certain computational processes more intuitive? Fair enough. How does it do that? By limiting the field of my responses? I question that. By reducing complexity to a few easily programmable choices? No, I don't think I want to do that. By giving me a mirror in which I have a hand in trying out new and previously untested options. I'll give that a try. I'll see how that feels. I'll take a step in that direction.

My point is that there is no absolute verdict that can be delivered on characters in the interface. The computer/human interface, as we all know, is growing ever more transparent, ever more resembling the world it presumes to reflect, and as such, the arrival at this juncture of entities, presuming to reflect ourselves, is inevitable. Our challenge then is to recognize the symbolic force in characters and to govern that force as wisely as we can. As Obie-Won said, and this expert can only reiterate, "Trust in the Force".

                       Mark "Spoonman" Petrakis
                        Mill Valley, CA  94941


web ethics course |

justin's links by justin hall: contact