I've been talking about my plans to create The Wikipedia for Debates, and I think it's time I pause to illustrate some of the concepts I've been discussing with more detail. There are many sites out there already with the goal of providing an easy digest of the main debates we are facing, and/or an opportunity for the public to chime in. However, none so far have all the elements I believe are necessary to have wide-spread impact in reversing the trends seen in recent elections of a divided and misinformed public.
As the subject of this illustration, I have chosen the U.S. Electoral College. It is both an intellectual/philosophical issue that has spanned the history of the United States, and an incredibly relevant topic now that Senator Barbara Boxer has put out a proposal to end it. It is also currently a very emotional and divisive issue, giving rise to accusations of personal interest and lack of critical thinking. But please keep in mind that my objective here is not to discuss the issue itself. I'm just using it as an example (if you are interested in the actual debate, you can try zooming in on my mind map above).
Obviously, you can't have a debate if no one can find it. There are two main ways someone would come to a site like this: either they're looking to search for a specific debate, or they are following a link someone else posted (I'm assuming a browser-based interface, although there will be more ways than that). Links are easy enough, but searching has to be done right. This should be trivial, but there are a couple of issues that need to be resolved:
- The debate actually needs to be there
- Duplicate versions of the same debate need to be merged into one, canonical debate
- Debates that SEEM to be similar, but are really different, need to be kept separate while still respecting their commonalities
The first issue is obvious, but critical. In some of the debate sites I researched, there was no result for a debate on the electoral college. You can't have an informed public without information. And, if you can't be sure the site will have the information you need, you are less likely to use it at all.
The remaining issues are more complicated. If a debate is being held in more than one place, the information will not accumulate as needed. This is one of the main points of having a "Wikipedia for Debates". But the debate also must not be repeated more than once on the same site. Not only does it mean information isn't accumulating, it also creates a bit of a "broken windows" effect, such that people will lose interest in maintaining the debate and apathy can set in. Note that Wikipedia, for example, does not allow multiple entries for a single subject.
My solution to this, and many of the problems we will see, is to have official curators. As with Wikipedia, there is a class of people out there that care more about maintaining the quality of our debates than they care about the individual debates themselves. Just as the ACLU stands for freedom of speech in all its forms, there are people (myself included) who will want to dedicate their time to cleaning up and improving what are otherwise free-flying open debates. Part of this work will include identifying when duplicate debates have been created, and having the tools to merge them into a single item. It will also entail refining the context of the debate, so that a question like "Should the U.S. accept Syrian refugees?" can be kept related to, but distinct from "Should the EU accept Syrian refugees?", as well as defining temporal contexts (are we talking 2016 or 2022?).
I'll write more about the curator role in future articles.
At a Glance
It may seem simple, but the most important part of the debate is the first look. In order for the project to be a success, it MUST be a place that everyone can go and get what they’re looking for with zero effort. I break this down into a few essential factors:
- It must be easy to read and understand the top 5 arguments (give or take) from each side in under a minute
- The reader needs to be able to trust that these arguments are complete, based on fact, relevant, and ordered by the impact they have on the debate
- It must be possible to "drill down" for more information if desired
Many of the debate-focused sites I have seen fulfill the first requirement. Most arguments can be summarized in a tweet, if properly written (again, curators can help with this). However, the requirement that we see the "top 5" requires itself some method for choosing which are the "best". Some sites choose to do it by hand, others by popular vote. Those methods work pretty well, given the proper care, and given enough eyes on the issue to trust the voting. However, I think there are many ways to improve on this, based in part on the second item on my list.
Before going on about those ideas, take a moment and compare the list from Debatewise.org with the mind map version I scraped together. I find it interesting that a number of the Debatewise.org arguments are distinctly partisan in nature. My list only contains arguments "for the general good", because my personal lens is one of finding the best solution for the system as a whole. Although I personally think making an argument based on personal benefits (I'm using this loosely — you can, of course, argue that the Republican party is bad for everyone), there is no one way to look at a debate. You can't (or shouldn't) force people in a democracy to make choices based on someone else's criteria. This is another deep subject that I will try to tackle in a separate post, but let me summarize by saying: the tool SHOULD be able to adapt to each reader's preferences. That is, what are considered the "top 5 arguments" may need to change based on who is viewing. Not automatically, because that would be contrary to the goal of combating the internet filter bubble, but I do believe we need tools to interact with the debate based on perspectives.
Getting back to my list, the second item is a list of several attributes that must work together in order to give an argument the proper import: it must be BELIEVABLE. People will lose interest in coming to the debate site if they feel it is a waste of time. Users must trust that this list is the best we can do:
2a. It must be complete: if the debate is a ghost town, then you can't trust that we've found all the best arguments. Users will have to come back later, or not at all. True, when a debate is fresh, it won't be complete (and in fact, a debate may never be truly complete), but all it takes is a few interested experts to come in and flesh out the top arguments. Looking again at the Debatewise.org list, you can see that some of my top arguments, the one about one vote counting more than another, and the possibility of electing someone without winning the popular vote, are missing. Worse, there are no "No" arguments whatsoever in the debate. This is not a problem with their platform so much as the problem of not having enough participants, nor curators.
2b. It must be factual: we are going through a period of coming to terms with the notion that much that we see in the media is not completely, if at all, true. I'll go more into how this platform can ensure that we are dealing with facts here and in other articles, but suffice it to say it's not an easy task. It requires crowd sourcing, debating, fact-checking, algorithms, and all we can throw at it to make sure we have done our best. I haven't yet found a debate site that does this.
2c. It must be relevant: just because something is true doesn't mean we care. But relevance is a sliding scale, and something that can't necessarily be easily determined. For example, the statement that "illegal immigrants have been guilty of assault, rape, robbery and drug trafficking" is definitely true. However, is it relevant to the debate on deportation of illegals in the U.S. right now? That depends on how they compare, as a group, to the general population, and to other populations with similar traits (socio-economic situation, geographic location, and so on). It may be that the problem isn't as bad as those in favor of the argument would have you believe. I consider this to be one of the largest problems with our current state of debate: giving our arguments the proper context and proportion. But how should we decide how relevant an argument is? Relevance, as with truthfulness, is debatable.
2d. It must be ordered by its impact: I'll admit here that I don't yet have a working definition for what I mean by this. But there is something else to debating and decision-making that goes beyond just facts and relevance. I can't describe it, but I know it when I see it. It's what matters most when the time comes to actually make a decision. "What was the one argument that really convinced you?" The thing is, I think it is more of a personal X-factor. It is a question of how much each argument resonates with your own personal values. It's most important in decisions that require making trade-offs (which is more important to you? The right to privacy, or your personal security?). So, perhaps this element can only come into play as a user begins to interact with the debate. For those keeping track, this is one more topic I promise to tackle in the future.
The system should show these factors attached to the arguments themselves. You can see in the image to the left that I tagged some of the "No" arguments with the names of some well-known logical fallacies. I also marked one as "Irrelevant". That's not exactly how I picture this working, but is a good start. You can imagine something like a mini scoreboard: "probably true", "low relevance", plus tags for any fallacies identified.
Get Down and Dirty
Finally, and very importantly, users must be able to go beyond the 140-character description of an argument. This is another place where most debate sites, as well as most news articles and other internet sources, fall flat. There are two structures commonly employed: the long-form debate, wherein one side explains all its points then another responds in kind, and the crowd-sourced "comment thread" method, where average people state brief opinions, and are voted up or down, or presented side by side. There are other ways to have a debate as well: live discussion (moderated or unmoderated), question and answer sites (Quora), unthreaded comment sections (e.g. on news sites), threaded discussions (Reddit), and so on.
Each presents a different mix of two opposing factors: detail vs. "usability". Long-form debates are very readable, emotive and convincing, because they present multiple arguments and supporting evidence in a single go. However, if the opposing side wishes to take issue with a single argument or piece of evidence, it distracts from their effort to present a comprehensive view. Live debates and conversations have this issue as well: you can ignore a point you disagree with to focus on the big picture, or stop everything to go down the rabbit hole of a single point. Neither is ideal, nor gives a complete treatment.
Threaded comments like Reddit, on the other hand, give you all the time you need to go down all the rabbit holes exhaustively, even multiple times. But this comes at the cost of readability. There are the duplication issues mentioned before, as well as a general lack of organization for the debate. It is very hard to get a quick overview, and then drill down without a discouraging amount of effort. In the Reddit example here, good effort was made to curate the conversation, but Reddit provides no guarantees this will always be the case (nor the tools needed to do it well).
Here's a quick, and admittedly superficial example of a deep dive into an argument:
One of the first things to note here is that the argument has two halves: "agree" and "disagree". Any statement in favor of or against a proposition can itself be supported or contested. There is theoretically no end to the possible arguments, but in practice, it won't go on forever. Argument "depth" is not the same thing as going back and forth in disagreement over the same topic.
But note how diving into deeper levels of the argument changes the nature of a discussion. At the top of this "branch", the argument is about how the rules of the Electoral College influence the manifestation of our politics. On the branch farthest down, there is an argument about whether or not the Civil War in anyway reflects the success or failure of the system. It would require very different people to argue each point, with different levels and types of expertise. The bottom argument may not even affect the outcome of this debate, but would certainly be an interesting place for history buffs to add their piece. This richness of information exists, but almost never finds its way into our current debates. This is an example of how this model, like Wikipedia, can help knowledge to accumulate, bit by bit.
More observant readers may have noticed that some of the arguments are tagged as "examples". It would be useful to support tagging an argument as an "anecdote", for example, to separate it from something that has true statistical significance. But there's one more thing I'd like to point out here: Some arguments attack the logic of the statement, but others are just evidence to pile against or in favor of another argument. This is an important distinction that should be made clear, supported and encouraged. It is a critical part of promoting fact-based discussion. Of course, even "evidence" is debatable, on the same merits as any other argument: truthfulness, relevance, and so on. You heard me right: a piece of evidence should be treated as its own debate.
I also tagged one argument informally as "conjecture". This is really shorthand for "you have absolutely no way to prove that." In an actual debate, this would be readily apparent, as it would have a low "truth" score, and no supporting arguments. That is, this statement has no evidence to back it up (at least, not yet).
Now look a the argument below, on the "Yes" side of the debate:
This is an incredibly easy proposition to prove: just point to the results in 2000, 2016, or any of the other cases. I skipped that evidence in this example to focus on some other arguments I encountered. These arguments are not about the validity of the statement itself, but rather the relevance (or impact?) it has on the subject. As I mentioned, relevance is debatable, and that means it should be debated, right there, attached to the argument itself (although separated from the arguments about validity).
I won’t go over these here, but I would like to jot down some attributes that the arguments should have for those that care to read more:
- Detailed description — there should be a place to explain in detail the meaning of the argument and try to convey some of the emotion behind it. This could be picked from a crowd-sourced selection.
- External references — when discussing evidence, providing links to external resources is essential. But it also would be helpful to link to external articles that provide a more complete and prosaic (or poetic) description of an argument.
- Related debates — other arguments or debates related to the current one, but perhaps related to a different context, should be readily available to help identify related arguments and so on.
- Comments and discussion — there should be a place to state opinions (beyond supporting arguments) on the current debate.
- Version history — all versions of the debate should be readily viewable in order to ensure that information is not lost, and to ensure with full transparency that curators are not favoring one side or the other.
- Tags, meta data, etc. — the more people know about the arguments, the easier it is to find, categorize and relate them, subscribe to changes, and so on (without sacrificing usability!), the better.
Reduce, Recycle, Reuse Reason
One last concept to toss out there: I have pointed out the need to prevent a duplication of debates, and yet to provide space for related debates with slightly different contexts. It may have crossed your mind that some arguments might apply to both debates. Consider the example I gave with illegal immigration: there can be many specific debates related to this topic (illegal immigration to the U.S., to Europe, from Europe, to Canada, and so on). And yet, if someone proposes as the argument "it is a basic human right to go freely wherever one wishes", there's no doubt this argument should apply equally to all these debates.
We need to avoid repeating the same debate twice, and that's exactly what my example argument is: it's own debate. This implies two things: first, any debate can be used as an argument in another debate. Second, it does not belong to that debate. Rather, it should be possible to reuse it as an argument in as many debates as can be considered appropriate. Or, rather, as is relevant. The validity of an argument stays constant, but its relevance may change from debate to debate.
All the debates (and arguments and supporting evidence) I have shown so far bear one trait in common: they are all binary propositions that can be answered by a "yes" or "no", a "true" or "false", or "agree" or "disagree". These follow the lines of Oxford-style debates, which focus on a single positive proposition around which participants debate (an Oxford-style rephrasing of my example might be "The United States Electoral College should be abolished."). But it turns out that in the real world, things aren't always so black and white.
You can see hints of this coming through in some of the sub-arguments that are proposed in my example:
The argument that "We should fix the balance… not abolish the Electoral College" is in fact raising a bigger issue: "What can we do to improve the U.S. electoral system?" In this much larger context, "abolish the Electoral College" is but one of many options available to resolve what some consider to be a problem:
This goes beyond debates to, perhaps, actual decision making. I will have to cover how this type of debate can work in a future article. Suffice it to say that this type of approach is sorely needed, and I have yet to see a public site that tackles it adequately.
I'm sorry if this was a long read. The fact is that debates are very complex entities. It is no wonder that we have up to now dealt with them in a somewhat haphazard, inconsistent and superficial fashion. But this is an idea long past due, a gaping hole in our knowledge infrastructure, and must be tackled. It may require a genius in Information Architecture, or in UX design the likes of Apple, to tame this complexity and produce something deceptively simple (looking for volunteers!). But I really do believe it can be, and must be done.