Improving online discourse: Encourage critical thinking with Argument Maps

Alignment Best practices
1

Most forms of online discourse is done via comment threads that are either linear (this forum) or hierarchical (e.g. Reddit). Comment threads are well-suited to general conversation, but are a rather poor tool to have real structured discussion especially when the threads get longer and veer off-topic.

What alternatives exist, are they used in actual implementation, and are they effective?

I think for humanetech and HTC it is interesting to collect these alternatives. I recently bumped into a free software project that uses Argument Maps to guide a discussion:

Wikipedia: Argument maps are commonly used in the context of teaching and applying critical thinking. The purpose of mapping is to uncover the logical structure of arguments, identify unstated assumptions, evaluate the support an argument offers for a conclusion, and aid understanding of debates. Argument maps are often designed to support deliberation of issues, ideas and arguments in wicked problems.

In our current post-truth era anything that encourages critical thinking online is a good thing, so this fits well with humane tech.

The software I found was Arguman:

arguman.org is an argument analysis platform. […]

Argument mapping improves our ability to articulate, comprehend and communicate reasoning, thereby promoting critical thinking. […]

Critical thinking is the intellectually disciplined process of actively and skillfully conceptualizing, applying, analyzing, synthesizing, and/or evaluating information gathered from, or generated by, observation, experience, reflection, reasoning, or communication, as a guide to belief and action.

And, as said, Arguman is free and open source software: Arguman · GitHub

Here is an basic argument map which demonstrates the concepts (some maps are way more elaborate):

arguman-example-argument-diagram
arguman-example-argument-diagram1359×831 106 KB

When I looked on their Reddit I found another argument platform (non-commercial, but not open source): Socratic Trees.

Another nice example is Debate Map, also open source. Here is the argument map for Climate Change.

Last but not least I want to add a link to Argdown. Argdown is an extension to Markdown (the same text-based formatting language you use in this forum when posting). With Argdown this text:

[Argdown is the best]: Argdown is the best
tool for analyzing complex argumentation
and creating argument maps.
  - <Editors easier>: Argument map editors
    are way easier to use. #pro-editor
    + <WYSIWYG>: In argument map editors what
      you see during editing is what you get
      at the end: an argument map. #pro-editor
  + <Pure Data>: With Argdown no user interface
    gets in your way. You can focus on writing
    without getting distracted.

Is transformed to this map:

argdown-argument-map-example

4 Likes
2

These arguments maps are definitely interesting, but I get why forums don’t structure threads this way by default. Human IRL debates/discussions/conversations are sequential too (with all ensuing problems :smirk:), it’s more natural to copy that.

Missing from this list is the for-profit platform Kialo, which offers a simplified version of the above argument maps. You can only map a tree of “pro” and “con” arguments, no additional logical relationships such as “because” and “but” exist explicitly. The platform has many ongoing debates, including “Modern technology is a disadvantage to society”. I’ve been browsing some of these for a while, but I feel that they’re so extensive that they become yet another internet rabbit hole.

In the end I think it’s easier to deal with a contained article or discussion that elaborates on a few pro and con arguments in a story form. Then you might get only a part of the puzzle (or rather, a particular frame), but reading a few articles “from the other side” could help with construing a more complete picture.

1 Like
3

You are right. Another possibility is to construct the argument maps in parallel and in the background of an ongoing discussion. Either automatically or manually triggered. It could be used later to get an overview of pros and cons and to ‘jump into’ the appropriate part of the discussion.

Edit: Kialo looks like a nice implementation of argument maps.

1 Like
4

That’s a good point, that might also help with not repeating what was already said in giant discussion threads. (“Want to add something to this discussion? Check this map first!”). I believe Kialo has comment threads on every argument, a bit like the Wikipedia “Talk” page.

And of course, not every thread has the intention of being a debate. It would be cool as an optional add-on for those that are intended to be!

1 Like
5

Something interesting found in this toot:

This is a very rough implementation that shows how to display discussions as semantic graphs: DiscDAG discussion. Linking here, because it provides a slightly different approach and may be combined with argument maps functionality (and a great UX on top of it).

1 Like
6

Hey everyone! I’m the one who posted “that toot”, really interesting conversations going on here!

I can only really talk for myself here, but I feel like what I want to express often has other structures than just a linear series of concepts. It’s rather that the linearity of our spoken language constrains us. So to enable expression of the associative modes that we think in to me is a very “natural” thing too.

I say this with some conviction because, having used the DiscDAG system linked above, I was surprised by how it compared to conversations on Mastodon.

I think in reply-tree conversations (like Mastodon, Twitter, …) I tend to try and minimize sub-threads consciously and unconsciously because the conversation would become unwieldy otherwise, due to the UI provided. For example I might try to quickly reply to two sub-points of a comment, then extend one other point that I want to draw attention to out over longer comments; or reply to a series of questions in separate comments in one longer reply underneath the last one.

The graph visualization on the other hand lets me change context as I want and track down different threads very effectively. I was quite surprised how good it felt, especially given the very simplistic and rough UI.

8

I think I disagree with that. When I have a lengthy discussion we often “pin” things, ideas or partial conclusions, and then return to them later to weave them back into the conversation. For me, genuinely IRL, a discussion is a directed graph, usually (but not always!) acyclic, which was the inspiration behind DiscDAG (linked elsewhere).

I wrote that to explore having a platform that mimics discussions as I have them, rather than as they are forced to be by (most) existing on-line forums.

Here’s the example:

https://www.solipsys.co.uk/cgi-bin/DiscDAG.py?DiscussionID=DiscDAGsFuture

It’s daunting because it’s an experiment that was just made up as I went along, and the discussion there is already quite large. It serves to show that we need tools to collapse parts of the diagram, and tools to navigate, but it reflects the conversation/discussion that we’re actually having.

9

Welcome @s-ol and @ColinTheMathmo happy to see you here :slight_smile:

Before I respond I’ll first paste the contents of your latest toot here:

Is anyone interested in doing some open-source design work for this graph-fedi-thing I’m working on?

Anything from color choices and themeing to thinking and experimenting with UX would be very much appreciated.

Current (static) demo is here: https://dag.s-ol.nu/

You can join the ongoing discussion here:
https://www.solipsys.co.uk/cgi-bin/DiscDAG.py?DiscussionID=DiscDAGsFuture

Code will probably be MIT licensed, DM or reply for access :slight_smile:

Example-Discussion-Graph

Mastodon’s seemingly sequential comment threads can be very confusing, especially for newcomers who do not realize they are flattened trees. Also, because many people come into a thread on a sub-branch (not seeing the whole tree), any discussion has the tendency to quickly veer off-topic. It would be great if Mastodon (and Pleroma and others) had the capability to switch between ‘timeline view’ and ‘tree view’.

Indeed. The quoting is what you do here, on Discourse, for example, plus replied comments get a link list below it to its replies. But it gets unwieldy for long threads (as you can see on Discourse Meta with many 500+ long comment threads… entire discussion becomes TL;DR).

Yes, I can see that too, especially on the revamped static example (image above). Yet there should be some real clever UX work to make it truly intuitive and easy to work with. The graph is quite overwhelming to take in and to glean / grasp the useful information from across the entire thread takes care and attention.

I think in such UX much of the graph complexity should be once again hidden from view, but steering dynamic views that form by clicking interactive UI elements. Maybe some animation occurs with that and/or there’s a small graph viewbox in the corner of the screen that shows your position. I don’t know (am no UX / Interaction designer either, ha ha).

That is accurate for IRL discussion, I think, yes. So marking these things / ideas / conclusions should be part of the UX capabilities, where they are collected in a ‘pick list’ or something to be touched upon again further on in the discussion.

Here it gets exciting to me (going wild brainstorm from here on), as in the above I hear “semantics” e.g. ‘topic’, ‘idea’, ‘conclusion’, etc. and the fact that the Fediverse is based on Linked Data standards. Note that I am explicitly NOT thinking “Semantic Web”, but having some closed vocabularies to describe discussion structures. Then “Argument Maps” - the topic of this discussion - also come into the picture.

I think you have different Discussion Types, some way more structured and analytical than others, which may be just casual chatter, so there may be Templates for the semantics / vocabulary to use e.g. for Scientific discussion, or political decision-making, or conflict resolution or what-have-you.

Finally I will bring in some random unrelated-but-close ideas into the mix, and make some cross-links here and there (inform people), that may help the brainstorm and entice others to join the fray :slight_smile:

  • Majority of knowledge / wisdom / insights in online discussion gets lost + forgotten over time. At the moment you type or read it, is the best time to gauge if it is valuable for reuse later on. I wrote an idea to help people automate this knowledge aggregation (which I’ll update with a link to this thread).
  • There’s great work underway on the Fediverse to use Linked Data for knowledge representation, most notably openEngiadina.
  • At openEngiadina they follow with much interest the Underlay open distributed knowledge graph project.
  • Along a similar vein there’s the Anagora project (see also Hacker News)

I’ll inform the Anagora and openEngiadina projects about this thread, just FTI (for their information).