williamb@milton.u.washington.edu (William Bricken) (08/21/90)
Two comments, intended to serve as reminders of the domain we are exploring: #1: NOT PHYSICAL: VR is not necessarily a simulation of physical reality (PR). To implement physical simulation in VR, we add *constraints*. That is, VR is more general than PR. So those who argue that we cannot travel in time in VR are confusing domains. In the HITL architecture, TIME, like space, is a parameterized dimension. Just like time-lapse photography and temporal cuts in films (remember Slaughter-house Five?), we have time-travel tools available in VR. This is a central practical issue for telepresence, where signals from Mars robots are delayed 40 minutes. How do we tweek time to coordinate actions that occur at astronomical distance? Are we relativistic yet? The Varieties of Space is a reminder that we are not constrained to mass-like assumptions of continuity, connectedness, or measurability. Recall that Newtonian physics coevolved with the calculus. Our concept of PR and the language we use to describe it are incestuous. Also note that *humanly-perceived time* is very much a function of our frontal lobes. Have a look at J. T. Fraser's Time as Conflict for definitions of times that are atemporal, prototemporal, eotemporal, biotemporal, nootemperal, and sociotemporal. The central idea is Jakob von Uexkull's notion of Umwelt (species-specific universe): an animal's receptors and effectors determine its "conceptualization" of the world. Let's remember that *progress* (time as change) is only about 600 years old. Science as truth is even younger. And truth is (quite obviously) Umwelt-specific. There are scientists who have spent their lives studying time, let's look at their results. In case my position is unclear to physical scientists: PR has little to do with VR. We are exploring a new science with a new basis. *Immaterial realism*: VR is constructed, not given. It is completely representational, but not a priori rational, empirical, or verifiable. Not mass, mind. #2 PHYSICAL AS BODY: There is a connection between PR and VR, it is expressed by our bodies. Our body is our interface to VR. Bodies expect some PR cues, and good VR design respects these cues. For instance, we are about 3 5/6 dimensional: for translation, X and Y but not Z; for head-centered orientation, full yaw but 1/2 pitch and 1/3 roll. Aside: I'd sure like to see those avionic terms disappear for human orientation. What about TURN, BEND and LEAN? So, two distinctions. 1) Native vs learned: Kids bring one set of physical responses to VR, adults bring another, depending on training. In the Autodesk VR demos, most people got disoriented when given 6df control devices. Pilots got annoyed when not given 6df. VR design must first respect native bodies, then make explicit what needs to be learned. The physical scientists need to ask: what do people bring to natural laws? 2) Mass vs information: We live in two overlayed worlds, that of mass and that of information. VR is the place to transact with information, its "physics" is the cognitive organization of bits. The issue of SPACE is not physical measurement, it is cognitive comprehension. William Bricken william@hitl.vrnet.washington,edu "If it's not both accurate and paradoxical, it's not worth saying."