Immersion seems to be linked with proxemics - the sense of spatial surroundings.
According to proxemics an individual seeks the balance between spatial crowdedness and emptiness in his surroundings and interacts with his environment to maintain this balance. Often quoted example is the density of tables in a Mediterranean cafeteria vs. the density in a Scandinavian one. When a Finnish group sits down in a Mediterranean cafeteria they tend to move the extra chairs away and an Italian group in Finland crowds a single table, yet both groups are doing this to relieve proxemic stress. The principles seem to be global although calibrations are culture dependant.
above: proxemic spheres
I have played with the idea of four proxemic spheres that relate to physical distance
- intimate sphere - very few things are tolerated here
- personal sphere - this where most of the everyday actions take place
- social sphere - e.g. cafeterias
- global sphere - the culture you live in, rules, regulations, habits etc.
Immersion translates into the presentation's (display, installation etc.) ability to invade the personal and intimate proxemic spheres. This goes for theater, cinema, music, games, work tasks as well. Immersion here defined as maximal suspension of disbelief.
above: Attention in terms of proximity.
I contrasted this with an idea of circles of attention (in the image sphere of concentration):
- intimate circle of attention - imagination, thinking, immediate awareness of/to external stimuli
- personal circle of attention - operating things, dialogue, etc.
the whatever it is you need to be focussing on, knowing things and or people - social circle of attention - the group, the roles, landmarks, the place
- global circle of attention - basically the world perceived as a background or a setting
From the users point of view the display has to penetrate at least the social circle and the displayed content the personal sphere. Porn in almost any form seems to do this for the most of us, but is it learned and cultural like the proxemic feeling of crowdedness?
Both models above consist of four more or less concentric cocoons, each can be described in terms of distance -
span of attention vs. attention span :) - or as collections of things perceived as separate objects.
| sphere of | number of items | radius measured in | and something | else | this too |
| intimate proxemics | 1-2 | centimeters | and something | else | this too |
| personal proxemics | a handful, 6 | meters | and something | else | this too |
| social proxemics | some hundreds, occasionally thousands | usually tens or hundreds of meters | and something | else | this too |
| global proxemics | no object/background separation, 1 | kilometers, it has a cognitive outer limit | and something | else | this too |
| circle (sphere) of | number of items | radius measured in | and something | else | this too |
| intimate attention | 1-2 | centimeters | and something | else | this too |
| personal attention | 5-12 | meters | and something | else | this too |
| social attention | 5-12 | audiovisually in tens or hundreds of meters | and something | else | this too |
| global attention | again all sensory stimuli blends into one sensation of background | from some tens of meters on | and something | else | this too |
The table above is for us to complete and develop. The main purpose is to introduce some basic metrics for Kyperia.
The spheres and (circles) are concentric the inner carrying more precedence over outer ones. Again from the (human) user point of view, how many things stand out between the self-awareness and the background depends on the short term memory. The total ranges between 5 and 12 with a normal person and can peak sometimes at 20 for a very short period. Out of the twelve one goes for introspection (the me) and one for the awareness to the background.
Operating the short term memory is a consuming task, so it normally operates at half of the reserve.
Finally this leaves effectively only five items to operate with. If these things are seen for the first time they cannot be retrieved from the experience (the long term memory) and their short term memory has to be refreshed often enough for them to stick to the background.
The things, as items in memory are mainly conceptual, items that we are aware of are mainly representations and proxemics deals mainly with the space the things occupy. To coax and enable user participation the limit of handful things, that require user attention at any given time is good rule of thumb. In my experience this applies to maintenance tasks as well.
To overcome these limits a human observer uses two strategies: One, shifting the focus of proxemic or attention spheres, as in choosing where to sit in a new environment or observing a distant incident.
The other, to combine (with the help of long term memory) items with the help of patterns into complex items (black boxes), that way reducing the need for short term memory.
As an example of this: I am approaching a moderately sized public display, first it comes to my attention and proximity by jumping out of the background as a one thing occupying the social spheres, as I get interested I approach and start making out details that invade my personal sphere. At this point the long term memory is referred to: this arrangement of lines looks like a room, there's a boy -> boy's room. Finally something sparks up my imagination and I feel a satisfaction in my intimate attention.
The more of the observed stimuli are familiar the more they will be expected to be so - that is why in a truly artificial environment things are scarce and clustered.
above: Interaction and proxemics
Where I am trying to get with this is that I we should start experimenting with a world that is completely artificial and bears as little resemblance to our physical world as possible. In the beginning the user is left to discover that a) the display has active content, b) he/she can manipulate that image, then discover c) the first handful of primitives and finally gradually pick up d) features, gamerules, tool uses etc. - all this in nice handfuls.
I think that as a starting point we keep should limit the likeness to what is so well known from the real world to the minimum, we can gradually introduce environments that are better orchestrated for a variety of intended purposes - from entertainment to remote controlling a submerged robot.
In conclusion (for the moment)
Eventually some Kyperia agents will be complex in appearence and sophisticated in function, nevertheless they should maintain these spheres, since agents mainly represent human users. We can capasitate agents with larger spheres, but they should always conform to, respect and make use of these human limitations.
In agents the spheres are represented as collections of pointers to the ontology (concepts) they operate in (with).
... carried on in SemioticTriangle and KyperiaOntology
---
SpatialRelations? in cyc ontologies --Fri, 24 Sep 2004 19:21:41 +0300 reply
see http://www.cyc.com/cycdoc/vocab/spatial-vocab.html
- an old remark of non-necessity of addressing in a shared space
- reply
even though, quite interestingly, the example that the author had in mind there, even though being a space, is not spatial! http://an.org/tunnustelua/0054.html (well admittedly it does require other identifiers, then, 'cause there is no location) - GameDesign
- reply