Introducing Sequiturs

We can improve how we discuss online.

Since the Usenet days, we’ve been using threaded comments: a system in which blocks of text can be marked as replies to each other and displayed in a nested fashion.[0]

This system has many advantages. It is:

  • Intuitive, because replying approximates the back-and-forth nature of a conversation.
  • Powerful, because it helps the reader to make sense of what’s going on. It organizes an otherwise disorganized mass of information, separating threads of discussion that would be tangled if comments were presented simply chronologically.
  • Flexible, because in any comment, you can write as much or as little as you want about whatever you want.

But this last advantage, flexibility, is also a disadvantage. There are no guarantees about the content of a comment.

A comment that is a “reply” may not actually engage substantively with the parent. In the best case, the thread may stay on track and yield a worthwhile exchange of information. Frequently, however, participants talk past each other to no one’s apparent benefit. In the worst case, the conversation can devolve into something worse than useless. In all cases, it’s down to the individual people participating in the discussion to come up with words that are valued by the community. The format itself imposes no constraints on their expressiveness.

For online communities that value expressiveness for its own sake (or perhaps just eyeballs—think of comments on YouTube), this may be desirable. But for a community that values making sense, and values participants learning something from an exchange of views, the status quo is deeply flawed. Cogent views may sometimes get the attention they deserve, but they’re just as easily lost among the scattered thoughts of all the voices in the room.

A better tool can help with this.

What is Sequiturs?

At Sequiturs, we believe the key to better online discussion is (as in other domains of life) to impose some constraints on ourselves. We take what works about threaded comments—namely, the structure created between comments—and extend it. We extend it to the text content of the comment itself. The resulting format is an argument: a series of premises and conclusions. Instead of free-form text, the content is broken down into a very powerful series of steps.

This format is powerful because it demands coherence. The reader of a Sequiturs argument expects the conclusions to follow logically from the premises, as they do in any good argument. Since readers are expecting coherence, that’s what the author of an argument must deliver. That means arguments that can’t be expressed coherently will tend not to be published, so the quality of arguments asserted will increase relative to the status quo.

The argument format is also powerful because it keeps the discussion focused. If you disagree with an argument, you can identify precisely where you think the flaw is, by pointing to a specific step in the argument. Discussions about the different steps in an arguments can stay separate from each other. The conversation is less likely to become hopelessly entangled.

You might be having trouble imagining what this looks like. Why not use a Sequiturs argument to make this very case about why Sequiturs is great?

This is an example of embedding a Sequiturs argument in another page. Follow the link in the footer to see the full argument page on Sequiturs.

Sequiturs is built around this argument format. Every discussion begins with an argument. Sequiturs provides a few ways to engage with an argument:

  1. You can create a Challenge. A Challenge is an argument that is posed as a counterargument to another argument. If you disagreed with step 1, for example, you could write up an argument whose conclusion is that the Sequiturs format does not improve the quality of arguments asserted, and mark that argument as a challenge to step 1. To accommodate arguments that are incompatible with each other but whose steps don’t directly contradict each other, it’s also possible to pose a Challenge to an argument as a whole rather than a particular step. And coming soon, you’ll be able to upvote challenges that you think supporters of the original argument should address.

  2. You can create a Comment. That’s right—even though Sequiturs emphasizes the structured argument first and foremost, you have the option of leaving a standard comment on an argument. We felt it was crucial to allow users to converse less formally in this way, e.g. for asking questions and giving constructive feedback. We know that not all thoughts are a good fit for the argument format. We think it’s enough that all discussions must begin with an argument.

  3. You can revise an argument. Revisions allow arguments to be strengthened in light of the discussion. Think an argument needs an additional step, or could be worded better? Revise it!

We’re working on even more possibilities for tapping into the power of the argument format:

  • Sub-arguments. Arguments should be able to build on each other. Argument A’s conclusion could be used a premise in argument B. In this way, A would be a sub-argument of B. This would allow arguments to become arbitrarily complex.
  • Visualization. Arguments and their revisions form a tree data structure. This data structure is made even richer by the fact that steps can be reused across arguments (we refer to the underlying content of a premise or conclusion as a proposition). The possibilities for visualization and machine learning on the natural language content of arguments are huge.

Vision

Sequiturs can be many things.

In all cases, it is a tool for helping people understand each other. It accomplishes this by clarifying the reasons for believing something (namely, the conclusion of an argument).

It’s also a tool for speaking truth to power. Public discourse is awash in misinformation, rhetorical sleight of hand, and sophistry. Sequiturs’ argument format is acid to these. It eats through them: it highlights the premises on which an argument rests, so they are more susceptible to fact-checking, and it requires coherence among these premises. The argument format is so clear and compelling that, if enough voices support an argument, those in positions of power may be forced to address it.

It is a tool for better critical thinking. Breaking down an argument into a series of premises and conclusions is not always easy, but it makes you a better thinker. Sequiturs has enormous potential in the classroom.

It’s a resource. Sequiturs can become the way to inform yourself about the strongest arguments on any topic. In a public context, it can be a go-to reference for learning about all corners of politics, art, history, etc. In the private context of a team or an organization, Sequiturs can be a tool for reaching better decisions and providing a lasting resource to explain those decisions.

While Sequiturs has all the features necessary for these rigorous applications, it also supports more casual applications. Here’s a tweet by Bill Murray that’s a perfect fit for Sequiturs:

This tweet suggests a usefulness of Sequiturs at a more local level. Arguments like this one can be made to prove a point ironically. Friends could use Sequiturs amongst themselves, to follow up on debates from the other day.

I hope Sequiturs will grow to serve all these use cases. I believe it can. I believe it provides just the right amount of structural upgrade over the current way of discussing things: not so much that it feels inaccessible, but enough to unlock great new powers of understanding that we didn’t realize were there.

Thanks for reading, and thank you for using Sequiturs.

-Ian



[0] We also have publishing media like blogs and Twitter. To the extent that these tools facilitate discussion—that is, a two-way exchange rather than a one-way projection of views—they also reflect an approach of threading unstructured text.