Updates from the Canonical Debate Lab

It’s been a long time since my last article from the group. A LOT has happened over that time (both in the group, and obviously, the world). Many subjects have crossed my mind to write about here, but I haven’t found the time. I’d like to take those deep dives at some point, but I thought now would be a good opportunity to give some quick news as to the latest activities with the Canonical Debate Lab (CDL).

New Members

The CDL has continued to grow, slowly but steadily. We passed the 100-member mark in our Slack team a while back, but that sort of number isn’t very interesting. What is more interesting is that we have had a fairly constant flow of conversations and activity in the group, so much so that we long passed the mark of being able to get old messages on our free account. Maybe now’s the time to think about getting non-profit status…

It’s not just about quantity, of course. Very recently, we’ve been fortunate to have members with a much more academic background. This was our missing piece, that could help provide us with a much more solid foundation upon which to design and build. Many interesting conversations have already been started, well past the usual aligning of terminology that’s always necessary.

Software Development

The CDL came together from the very beginning as a group of people with similar ideas, most of whom were already trying to make their own version of a system like the CDL. We were able to talk through all the important points and resolve critical differences enough to put together our white paper. However, none of us are able to devote our full time towards the project, and we still haven’t been able to unify our projects into a single one. For the time being, this allows us to experiment with different ideas for features and for ways to represent debates.

Here are some quick updates on member projects:

• The Great American Debate — Our partner project has been going full speed ahead with collecting data and arguments related to the topic of Climate Change. They have already identified 220 sub-topics related to the debate, and are working on a web platform to display it all. In reaction to the COVID-19 crisis, they have begun focusing their sites on covering some of the information-rich topics related to the virus. They are hoping to have a working prototype of their user platform out this year.
• Debate Map — The most feature-rich version of CD-style debates continues to see steady development. It is now being used as the back end for The Great American Debate, and is hosting the debates they are analyzing.
• Gruff — This project was restarted from scratch in order to try out a graph DB (ArangoDB) as a back end for the CD debates. The project is focusing on providing full transparency features by keeping track of the full history of the debate, and allowing users to view each change. It is also planning to be the first to focus on testing out different scoring algorithms.
• Reason Score — The creator of Reason Score is taking time off from other projects to focus full-time on this platform and related ideas. Expect some interesting developments in the next months. Keep an eye out for some surveys and experiments on Facebook and Twitter, too.
• Socratrees — Developed by one of our newest members, this is another interesting take on a CD-style argumentation platform. Its creator, Steven Jeuris, has been focusing on researching the best designs for laypeople to interact this type of information, and has written an academic paper on the subject. We are looking forward to helping him with this research!
• Digital Peace Talks — DPT is a free and open discussion platform that enables anonymous one-on-one chats which both users can rate and publish in a 3D network model. DPT is currently hosting a global discussion on how to tackle the corona virus pandemic. Join at corona.dpt.world or search online for #discussCorona.

So Many Ideas

We have been discussing many ideas, many of which have already been considered in the white paper, but also many ideas that will potentially be featured in the paper and the system in the future. In the next few weeks, I will expand on some of these concepts. If a topic interests you, let me know, and I might write up a deep dive on everything we've discussed so far.

• Negation — This is an interesting feature that permits users to view a Claim as stated, or reverse the statement to view it from the opposite position. It seems simple, but there are some interesting twists to the idea. Some of our platforms, including Socratrees and Debate Map, already support this in some form.
• Belief, Truth, Acceptance, Agreement — In the CD white paper, we discuss the idea of a Truth Score for Claims. The term is a bit misleading, since we aren't actually measuring "the truth". As such, our members have been discussing whether or not it would be better to use a different term, like Belief, Acceptance, Confidence, or (my favorite) Agreement.
• Binary vs. Percent Scores — We have been discussing whether or not the best way to ask users to rate a Claim is via a percent score (0% to 100%), a smaller scale (like 1 to 5), or even a binary score (yes or no). The discussion considers not only the usability, but the basic philosophy behind what it means when we score a Claim.
• Uncertainty — We have been trying to keep the system as simple as possible for uninitiated users. However, in our experiments with mapping debates, it is becoming clear that there might be a need to introduce a new dimension to scoring: Certainty. This is the measure of how sure we are about the facts (or our own opinions).
• Curators or Crowd-sourced? — There has been hot discussion regarding the best approach to the Curation of debates. Should it be done by an elite group of skilled individuals, or should it be a process that is open to everyone?
• Crowd-sourced or Quality Controlled? — The CD is designed so that everyone can participate and contribute to the debates. However, the people from The Great American Debate have put a great amount of effort into creating a more centralized content analysis pipeline that is designed to ensure only the highest quality, bias-free information for each debate. Which approach is better?
• Map Equivalence — On a more nerdy note, we have been considering the idea that there is more than one way to represent the same debate. Given that argument maps can be represented mathematically, and that scoring mechanisms are a mathematical measurement of these maps, is it possible to create a definition that allows us to say two different maps are "equivalent"?
• Time-sensitive Debates— There is a large category of debates that are time-sensitive. That is, they only make sense within the "context" of the present (the time at which they were defined, and for some time after). In the Climate Change debate, there is the common example of "We only have N years until we reach the point of no return." Another example would be the Claim that "Candidate X will win the election in November." How should a platform designed to last for all time handle questions within a temporal context?
• Qualifying Qualifiers — The CD is designed to contemplate all types of debates, and at the same time eliminate duplicate Claims. So how should the CD handle variations like "Most successful rock stars are Sagittarians" and “Many successful rock stars are Sagittarians”. They obviously would share nearly all the same Arguments on both sides, and there would be as many variations on these Claims as adjectives would permit. How can the platform efficiently handle all the possible variations efficiently and usefully? Also, one variation would likely have more agreement (in terms of scoring) than the rest. How can the CD platform make the most of this phenomenon?
• Overlapping Arguments — The previous question shows that it's possible to have many versions of the same Claim with very slight variations. It would seem to be redundant and counterproductive to have every version of the related Claims used as an Argument in the same debate. Should the CD prevent more than one ("the best") of the same Claim be used in the same debate? How could that be enforced programmatically?

Conversations in the Canonical Debate Lab range from the hippie to the most erudite, sometimes in the same breath. We are trying to save the world by creating an information architecture that will allow us all easy access to the sum of human knowledge. It is no easy task, but it is fun and fulfilling. If you would like to join the conversation, come find us in our Slack Team!