UXperts Needed to Design the Future of Debate
The Canonical Debate Lab is already starting development on our joint project to create the "one debate to rule them all": the Canonical Debate. One of our major design goals, quite possibly the one factor that could mean the difference between success and failure of this critical project, is to Make it effortless.
While this may seem like an easy task — Google made search effortless, Twitter made public harassment and humiliation effortless, and Facebook made convincing millions of people to believe in fantasies effortless — debate is a deceptively complex subject.
Basic Types of Debate Visualization
Debates come in many forms. The most focused of these either circle around a specific problem (e.g. "What should we do about climate change?"), or a dialectic (e.g. "We should adopt Public Law 132/A"). To some degree, a debate regarding possible solutions to a problem is just an umbrella for various dialectics, with each proposed solution being a sub-debate about whether or not that solution should be adopted.
Dialectic debates are usually represented as an argument map — a tree of pro- and con-arguments that try to refute or support the main point, and each sub-argument. These maps are often displayed in one of the following forms:
Argument Tree
Perhaps the most common approach is to represent a proposition and a set of arguments in a tree format, with some way to indicate whether an argument supports or attacks its subject.
Outline View
Another way to approach what is essentially a hierarchical (although potentially very deep) structure is to represent the debate as an outline, with some way to differentiate the supporting arguments from those that are attacking the previous statement.
Columnar View
The previous approaches attempt to show several levels in a single view. In order to improve the readability of a single statement or argument, some platforms prefer to display only one statement at a time, with the support arguments on one side and the attacking arguments on the other.
Graph View
The Canonical Debate may permit arguments to be made in a circular fashion, since we do not naively expect humans to adhere to machine-like thinking. In such circumstances, a tree-style view would be insufficient. Whether this is a requirement or not, some platforms prefer to display debates as a graph, with nodes and edges representing the arguments and their connections.
Multi-Dimensional
A more radical approach to graph-like visualization is to offer navigation through a three-dimensional space of interconnecting topics and arguments. This can be mesmerizing, entertaining, productive and intimidating all at once.
Hybrid
This isn't a real category, but there are other creative ways of combining the need for being user-friendly and readable with a complicated multi-dimensional structure.
Although certainly not a requirement, problem/solution debates tend to be displayed in a manner similar to dialectic argument maps, often by just adding the problem as a level, layer or node above each proposed solution.
Our Own Approaches
The Canonical Debate Lab is a coalition of developers, activists and debate enthusiasts who have come together in part because we realized we were all trying to solve the same problem: how to create a Wikipedia-liken centralized, definitive place to hold debates so that information would not have to be lost to the Internet ether and repeated ad-infinitum. Rather than work separately, we decided to join forces, but even in our small group it is evident that we each have our own preference for representing debates.
Debate Map
Stephen Wicklund has done some fantastic work on creating Debate Map, a mind map-like graph view, incorporating some basic elements of a problem/solution space into an argument map.
Gruff
Timothy High (ok, me) has created and re-created several times (with the help of Bruno Sato in this latest edition) the debate site Gruff, similar in style to the "columnar" Kialo mentioned above.
Reason Score
With Reason Score, Bentley Davis has taken a unique approach to debates, in what could best be classified as "hybrid" (because that's the non-category). He has applied the text messaging paradigm to hierarchical debates in order to make them more accessible and feel somewhat conversational.
Digital Peace Talks
Although not strictly a project of The Canonical Debate Lab, our member Iwan Ittermann is working on a separate project to improve online discussion in order to promote better understanding and consensus. The concept includes a three-dimensional representation of the arguments related to specific topics.
Cruel and Unusual Requirements
Most argument maps and debate sites follow the strategy of keeping it simple for the users, which may be average citizens, or may be experts focusing on a single aspect of the map structure. The Canonical Debate is a very audacious attempt to capture the many complexities of human argumentation into a single structure, and yet keep it as intuitive and easy to use for the uninitiated as possible. Here are just some of the requirements that we need to support while maintaining usability.
Argument Ranking
In order to keep debates readable, it is important to rank arguments by how well-vetted they are and how big an impact they carry. This is easier to do in a "columnar view" like Kialo.com than in a tree- or graph-like display.
Curation
On the other hand, there are use cases in which it is more important to see the entire structure of the debate at once. Such is the case of curation, the process of reorganizing and maintaining the debate in a pristine state. Graph-style layouts are much more appropriate for this.
Scoring
A major topic of concern is the support for providing a score for each claim and argument regarding its truth or quality. However, as described in our white paper, we believe that no single scoring system is sufficient, as each has its own inherent biases. Instead, we seek to support multiple forms of scoring, each of which may be viewed separately, or perhaps simultaneously.
Some of the current sites may represent scoring as a simple count of likes, and perhaps dislikes, which can work as a popularity-based filter (it seems to be a good approach for Reddit, for example).
A somewhat more sophisticated approach would be to ask users to rate each argument according to how effective they believe the argument to be.
Debate Map provides a much more detailed and visual approach to scoring, allowing the user to define even a range of certainty for their vote. While not explicitly included in our white paper, it would be an interesting topic to explore.
Current solutions to scoring don't appear to support what we call roll-up scoring, which is an important feature for understanding how arguments can impact the outcome of a debate. It is also a major focus for academics in the field of argumentation. Thus, we need to support yet another dimension of complexity in terms of scoring visualization.
Knowledge Graph
Debates and claims in the Canonical Debate are meant to be related to one or more elements in an external knowledge graph. Knowledge graphs are themselves a totally separate entity which may require some form of representation within the Canonical Debate.
Size
The Canonical Debate is designed to encompass every possible topic that humankind is willing to discuss, all within an interrelated map of claims. This can become very difficult to visualize and navigate, especially if one tries to take it all in at once.
Kialo.com provides a very creative and aesthetic solution to the problem by offering a topological graph with hover-over dialogs to provide summary information. This works well in a tree-style structure, but is harder or impossible to apply to a possibly circular graph of arguments with unpredictable depth.
How to convey the number of possibilities, give a 10,000-foot view of a subject, and still provide reasonable navigability is a serious challenge. Trending debates, text-based searches, subject-based searches and knowledge graph-based navigation are all options that need to be considered and possibly supported.
Trustworthiness
There are two separate aspects to this question. First, it is important for the platform to convey that it is itself as trustworthy as possible: there must be a way to show that the platform is fully transparent, incorruptible and censorship-free.
Secondly, it is important for each argument or claim to show to what degree the proposition has been vetted by the community.
Versions
Part of the effort for transparency involves the ability for anyone to view and audit the changes that have occurred to a debate over time. Wikipedia.org provides a very basic listing of the linear history of the evolution of its articles.
The dearly departed Google Wave web app offered a fairly creative slider-based approach for allowing users to move back and forth across the "timeline" of a conversation.
User-Friendliness
We have already mentioned this, but it is critical to reiterate that we wish to produce a site that is feature-rich while still maintaining an inviting atmosphere that makes it fun to use. Details must be available, but perhaps only to those that seek them.
Kialo.com offers a magazine-style introduction to each major debate topic, which may be difficult to provide and maintain in a fully community-supported ecosystem.
Multiple Interface Support
Perhaps it is not possible to support all these goals simultaneously via a single interface. Fortunately, our plans include providing access for any number of interfaces, each of which may be optimized for their own target task and user base. In terms of UI, this may not be one project, but many.
Help!
UX/UI Experts
The Canonical Debate Lab has (perhaps unfortunately) begun development without the expertise of a UX/UI specialist! We consider usability to be a critical factor in the success of our project, which we truly believe will be a fundamental tool in building a better future for everyone. This is a challenge that we hope you will find to be worthy of your skills!
Graphic Designers
We are in dire need of help in this area as well, from our logos to our colors, fonts, buttons, everything!
Product Owners, Requirements Specialists, Users, etc.
We are creating a list of user stories right now to drive the interface design, in the general format "As an X, I would like to be able to do Y so that I can Z." If you have something you would like to see, please add your comments below this article, or directly in our spreadsheet.
Concerned Citizens
We are gathering quotes from people on the question "What should be done about climate change?". The actual goal for this project is not to try to answer this question (yet!), but rather as data to develop machine learning algorithms to analyze and group human input. If you'd like to contribute, you can quickly add your thoughts on our dedicated form.
Everyone Above, and Anyone Else!
Even if you aren’t an expert, you can definitely help out! We are looking for volunteers to do user interaction testing on some of our own sites, listed above. All you have to do is let us record you trying out one of our test sites.
Contact Us
If you would like to help us, please contact us via Twitter, in our Slack channel, or in the message section below.
