December 10, 2018 Beth Cardier

Final Report for the Form Project

Design in Information Flow investigated how principles of emergent design can overcome barriers of computational logic. The arts contain  information structures that are omitted by scientific frameworks; this project identified some that could extend the capabilities of computational modeling. We achieved this by developing new formal representations that could carry new, additional knowledge from neuroscience, virology and microbiology.

These new information structures concerned the interaction among whole systems. Due to the way knowledge is represented in computational systems, it is difficult to model information that is unexpected (outside its reference frameworks), that changes structure (it evolves), or differs in structure (moves among multiple disciplines). We drew from narrative and design to develop representational structures that would support transitional information. The goal was to newly manage the transfer of information across boundaries of context, scale, time and domain.

From both narrative and design, these new structures were: analogy (connecting similar structures across diverse systems) and abstraction (the aggregation and refinement of structure).

From narrative these new structures were: situations (contexts), governance (the influence of context over its components), transference (carrying structures from one system to another) and retroactive reinterpretation (re-casting information structures based on new conditions).

From design these new structures were: principles of flow (ease of affordance creates faster-flowing channels), entropy (the easiest affordances are preferred when transferring information) and hypersection (the optimal fit between systems in which each compromises least).

These principles were discovered reported in three peer-reviewed publications, which indicated how they could newly represent neurobiological processes. The first paper recast narrative in terms of the new project, the second linked this to neuroscience and computational modeling, and the third, an article in the journal Progress in Biophysics and Molecular Biology, brought in design, biology and the implementation. In order, these papers were:

  • 2017 Cardier, B.  The Moving Lens: Coherence Across Heterogeneous contexts in narrative and biology. Proceedings from Computational Context: Why It’s Important, What It Means, and Can It Be Computed?, Association for the Advancement of Artificial Intelligence: Palo, Alto, CA, p. 295-302.
  • 2017 Cardier, B., Sanford, L., Goranson, H. T., Devlin, K., Lundberg, P. S., Ciavarra, R., Casas, N., and Erioli, A. Modeling the Resituation of Memory in Neurobiology and Narrative. Proceedings from Computational Principles of Natural and Artificial Intelligence, Association for the Advancement of Artificial Intelligence: Palo, Alto, CA, p. 534-540.
  • 2017 Cardier, B., Goranson, H. T., Casas, N., Lundberg, P. S., Erioli, A., Takaki, R., Nagy, D., Ciavarra, R. & Sanford, L. Modeling the Peak of Emergence in Systems: Design and Katachi. Progress in Biophysics and Molecular Biology: Special issue on Integral Biomathics: The Necessary Conjunction of the Western and Eastern Thought Traditions for Exploring the Nature of Mind and Life, 131c, 213-241.

The above journal article represented the coming together of the entire team, plus two external collaborators who helped bridge our domains – Ryuji Takaki from the Tokyo University of Agriculture and Technology, Tokyo, Japan, and Denes Nagy, president of the International Symmetry Society, Budapest, Hungary. These two will lead their respective Symmetry and Katachi societies at the 11th Congress of the Society for the Interdisciplinary Study of Form in Japan, 2019, where Niccolo and Beth will present the final artistic results of this project. This will include images, rationale, design principles and an exhibition of the 3D printed sculpture at the congress’s associated art show.

There were a few recalibrations during this project. It took longer than expected to develop a genuine connection between fields, and we lost a key group member at 16 months. During this time, we also submitted 11 grant proposals to NIH, NSF, DARPA, the Virginia Biosciences Health Research Corporation and the Keck Foundation, and learned that more development would be needed to attract the next round of funding. Finally, Casas and I discovered that hand-rendering the complex networks of animated biological information was nearly impossible given our usual tools (Keynote and Maya). Together, these factors pointed us towards a different tack.

At this time, Old Dominion University’s Visualization Modeling Analytics and Simulation Center (VMASC) became a new collaborator. They contributed significant internal investment to build a prototype of the modeling platform. We recently reported on the development of this new modeling platform in a conference paper for the Association for the Advancement of Artificial Intelligence, to be presented at Stanford in March 2019 by Cardier:

  • 2019, Cardier, B., Goranson, T., Saikou, D., Shull, J., Casas, N., Nielsen, A., Lundberg, P., Sanford, L. D., Ciavarra, R., A Narrative Modeling Platform: Modeling Implicit Influence Among Contexts.  Computational Context: Autonomous machines and human awareness: User interventions, intuition and mutually constructed context, Association for the Advancement of Artificial Intelligence Spring Symposium, Stanford March 27-29.

It is also described in my blog. A presentation of the modeling platform prototype will occur at this AAAI conference, followed by an invited book chapter, also supported by the Navy Research Lab, which describes its computational approach.

This prototype will enable us to produce our final round of drawings and animations, to be published alongside our other artistic products on our online Gallery, which documents the different system-level principles we explored from an artistic perspective. The original seed grant has thus rolled into a new project and phase of development, preserving most of the original team.

Given the very broad spread of disciplines, new approaches to interdisciplinarity were developed during this NAKFI seed grant. These techniques were based on narrative devices (Cardier) and design principles (Casas). This new knowledge was shared in a number of ways: a book chapter funded by the Navy Research Lab, an online interview/continuing education course (EVMS), a faculty development presentation (EVMS) and a joint article by Casas and Cardier to be submitted to Leonardo magazine. These are listed below:

  • 2018 Cardier, B., The Shared Story: Narrative Principles for Innovative Collaboration, ed. Lawless, W., Akey, J., Computational context: The value, theory and application of context with AI, CRC Press, Taylor & Francis Group.
  • (in progress) Hypersection: Shared Spaces in Collaborative Communication, Casas, N. and Cardier, B. for Leonardo Magazine.
  • Online class/interview: Eastern Virginia Medical School Faculty Development, Building Successful Interdisciplinary Teams, September 2018. https://www.ustream.tv/recorded/118027705
  • Presentation: Building Collaborative Teams: Lessons from Storytellers, Burroughs Board Room, Andrews Hall, Room 526, September 19, 2018, 12:00 – 1:30pm.
  • Presentation: Narrative Architectures, Texas A & M University, Architecture class, Coordinators Gabriel Esquivel and Niccolo Casas, October 11, 9-10am.

In addition, three blog posts explored this ideas in this project and its approach, and another is planned:

On a personal note, it is still astonishing to me, three years later, to see such a diverse group of people working towards building something real by bringing their fields together. That buzz still energizes the project. One of our new VMASC collaborators recently tweeted:

  • Spent most of the working day discussing the next project, a true privilege. 10 minutes in I realized this was it. I’m listening to a neuroscientist, a rhetoric researcher, a researcher who builds computational models of narratives, a true m&s guru, and some other person who’s I believe a mathematical / 3D designer who might also be an architect? Just awesome conversations @vmasc_odu today. I love this stuff! Building crazy tools that brings all of these fascinating people together.#unity3d you’re about to be on fire.
    @TheJohnnyFuzz, twitter November 14, 2018.

My own enthusiastic take on this project can be found here:

Building a VR Narrative Modeling Tool

Here’s something unexpected that I learned about transdisciplinarity: We never mastered each other’s fields (well maybe Ted did but that is Ted). I thought we would and so this surprised me. We did become fluent in the new group language we developed, which was anchored by the target product itself. It enables neuroscientist Larry Sanford to connect genuinely with me about analogical structure, or designer Niccolo Casas to have insight into new paths for artificial intelligence. But these insights immediately require the partner whose learned expertise is in that field – they could not be executed alone.

So there are still regular moments when one person says what they plan to do, and the others need to ask for a careful explanation. The further we press into new territory, the more constant the translative effort required, even though we built a common foundation for the project years ago. Instead of becoming experts in a strange domain, instead we build patient communication that also becomes friendship. At a funding pitch, Ted jokes “I’d ask these guys to paint my house”. We bring chocolate pie and sometimes kids to work sessions, growing new communities as we move beyond our disciplinary boundaries.