See also: Implementing Things

Thing

Things can be manipulated, and we experience them having some kind of essence and wholeness. They aren’t experienced as a separate input and output like a Mirror might be, nor do they merge innumerably like matter (eg. water droplets).

The design of a Thing is useful because:

How is it different than the Mirror?

Example 1. Three things, each directly manipulable and with their own size and mass

The goal when making Things is:

Avoid accidentally making an Agent that has a liveliness of its own and is merely influenced by external factors. Thing behaviours should always have a basis in simple matter, what we might expect of non-living physical objects in the world.

Defining

We might perceive the Mirror as being made up of an input thing and output thing linked together. Pressing a button here does something over there. In comparison, the Thing has a unity: integrity and essence. We don’t typically perceive its components or composite materials, although that may become apparent under scrutiny. A Thing is separate from other Things and the environment, even though it might not always ‘stand out’ in experience.

Things are inert, with no liveliness or agency of their own. They simply get pushed, pulled and carried along by external forces. We might expect to be able to perform basic manipulations of the thing: place, orient, destroy, stretch, pile-up, hold, point-at, and so on.

Examples of physical Things abound:

In the digital realm, examples are:

There is nuance in how we are able to act on, with or through things and how they are perceived. We can position them ‘just so’, for example.

Perhaps it is only in manipulating the thing that we begin to understand it and more nuance is revealed. For example, in the demos on this page, it is only in manipulating the squares that we get a feel for how each behaviour is different. For a given demo, all the Things have the same behaviour, so we can begin to figure out their logic and the world they reside in.

Below, we find Things that look like the previous ones, but the behaviour is different, and they are less nuanced in how they can be placed.

Example 2. Less nuanced. Can only drag on one axis, coordinates are quantised.

The dynamics of the Thing are enacted and sustained by the forces acting upon them. For example, consider the marble. It can roll when we flick it, but we don’t experience it to have agency or self-determination of its own. If we (or some other entity) aren’t doing anything to it, it’s in a stable state.

We don’t experience the Thing as having agency, a kind of higher-level intelligence and reasoning we might expect of an animal or fellow human.

But in terms of how we implement and design a Thing, we do in fact encode some kind of very primitive reasoning, allowing it to ‘decide’ when and how to respond, based on events, state and logic. This logic is contingent on external factors - such as human input - but is decoupled enough that we can truly identify the Thing as something interactive and where the hand of the interaction designer is present in its behaviour.

In contrast, for Mirrors it may be less clear what ‘interaction’ the interaction designer is truly designing. After all, the mirror reflects what is happening. If a skilled dancer is performing in front of a Data Mirror that reflects movement, the expression will have a richness which is more a credit to dancer rather than the designer. It may be that the Mirror invites a certain kind of movement (eg. novelty distortion mirrors), and here it may make sense to think of it as a Dynamics Mirror. It’s still not an object which can be manipulated, and reaction is still directly concordant with ‘input’.

Things might be indirectly affected by the environment or external forces. The same marble might melt given a high enough temperature. The candle flame flickers as people move about a room, or if a draught comes in from a window.

It might even be that the designed interactivity of the Thing (in how it can be acted upon, how it responds, how it solicits) is actually not experienced as being interactive at all - because the interaction is implicit or the Thing is otherwise in the background.

Example 3. Things susceptible to explicit dragging and also implicitly by ambient pointer movement

Changes can take place at multiple time scales, and forces can affect the thing at different scales. The grains of sand dune move with the wind, and the whole dune slowly migrates over years. The leaves of a tree might flutter in the breeze, heavy wind might sway branches and the persistent wind might affect the angle of growth.

Disclaimer: This ‘Thing’ definition is only for the practical purpose of distinguishing these interactivity models. There is a deep philosophical discourse on things/objects/agents, which we gleefully gloss over. There are issues with this simple account when we start considering things and objects in a broader sense.

See also: Implementing Things