***ROUGH DRAFT WORKING COPY***
With a tool such as QuantCollab, it is possible to begin quantifying the interaction dynamics of the co-creative process. Using the conventions of the tool, we have developed some theoretical models of co-creation in a hypothetical co-creative scenario. These figures can be easily implemented in the system given the description of the values and computations of each chart.
The cognitive dynamics have a y-axis of clamped or unclamped cognition, with the zero point being clamped. The positive y-axis is player 1, while the negative y-axis is player 2. In the beginning of the collaboration, each player makes new idea offers. Neither player accepts the offer of the other, and they keep generating new ideas. The more ideas that are suggested that are unrelated to each other, the further unclamped the participants are from a shared mental model of the artwork. 5 turns pass as the partners finally agree to an idea to work on together and coupled interaction begins for 3 turns. Then, player 1 has a new idea, which unclamps the session and begins a negotiation period where each player offers ideas that do not necessarily agree with what their partner just added. Finally, a new idea is established, and both participants work on the idea together to end the collaboration.
The creativity scale is based on the novelty of the ideas and their relative value as they are added during the drawing. The novelty of the idea is gauged by whether it was related to the partner’s last turn. If it is not related, then the idea is determined as novel. The value is determined by whether the partner actually adds onto the idea in the next turn. It is a retroactive value that is applied after the partner reacts to the turn. The creativity value for a turn will be highest when it is a new idea that is added upon by the partner. Creativity is evaluated as lower for when a player adds onto existing ideas in the drawing. This is viewed as an act of extending existing creative content. However, it does help with the collaboration dynamics, as calculated in the next section.
During time periods of multiple reciprocal offers, or disagreements, during the drawing, the creativity reading goes up as there are multiple ideas being generated in the drawing. However, the value goes up the most when there are consistent new ideas that each get added upon by the partner. During periods of coupled interaction, the participants are not generating new ideas, so that particular feature of the collaborative turn is low, while their partner adds upon their idea, so the turn has a collaborative value, which means interaction couplings will generally be coded as low on the creativity scale.
The creativity chart can be generated by taking each new idea generated as adding 10 creativity points to the scale. When a new idea is accepted, 20 creativity points are generated. During coupled interaction where one turn is based upon the other, the creativity value is 5.
The y-axis of the collaboration dynamics chart is whether the individuals are coupled or not. Coupling is a linear function of time, and the longer one spends coupled, the higher the coupled value becomes. This value can be calculated in two ways: 1) it could be a one-to-one mapping of the amount of time in seconds the users spend coupled and the value on the scale, or 2) it could be generated based on the number of turns spent in a coupling. The value could also be a composite score of both the amount of time spent in a coupling as well as the number of turns a coupling lasted.
In Figure XXX, in beginning of the collaborative drawing, the individuals negotiated content, so they did not have any coupled turns. Then, about halfway through a coupling began and lasted approximately 3 turns. During that three turn period, the coupling value consistently raised on the collaboration dynamic chart. Then, another period of negotiation comes, and the value decreases. Finally, a new coupling begins, and the value increases accordingly.
To code the cognitive state of the interaction dynamics, it is possible to use the sense-making curve generated by the QuantCollab program. This sense-making curve is based upon the question asked after every turn, i.e. Did you add to your partner’s contribution? With that single piece of information, we can determine a variety of cognitive information in different states of the interaction. There are two basic states the system could take, 1) building onto their partner’s idea in an interaction coupling, or 2) rejecting their partner’s ideas in a non-coupling event. To show the domains of the partner’s diverging through time and becoming increasingly unclamped from a shared idea, the unclamped value gradually goes up .25 points for each idea rejection.
This coding scheme accounts for both divergent and convergent creativity. Divergent creativity occurs when there is great fluctuation around the x-axis during periods of sense-making by both participants. During this time, participants add new ideas during their turns until an idea is well established and accepted by both of them. The more time participants spend in this uncoupled state, the more divergent their respective domains become, which is why the fluctuation is shown to be increasing through the turns in the beginning of the diagram. Then, both participants converge on an idea and work collectively on that idea, which is represented in the coding scheme as 0.
This coding scheme is sensitive to the temporal elements of the interaction dynamics of a co-creative session, meaning that turns are reactive to the turns that came before. A coding scheme must be sensitive to whether that turn is related to the previous turn and series of turns in order to fully determine the impact a turn has on the overall creative trajectory.
© Nicholas Davis 2022