Tinderbox Forum

Is this possible with TBX? 'Discourse' mapping

Q & A - getting started with Tinderbox
#1

Hi, I’ve had Tinderbox for a couple of years and have not really got beyond making the simplest of notes, so have hardly used it.

Recently, I’ve become interested in the ‘discourse graph’ for Roam research, and I think much of it would be possible on TBX, but I’m not sure about the query bit. I’ll have a HUGE learning curve ahead of me, so before I dive in, I’d be grateful to know what can be achieved. I’m not at all good with scripts / coding - I can modify existing code with little problem but some of the discussion on here goes over my head.

My studies requires fairly formal analysis of premises and arguments raised by different philosophers. So a rough outline of my requirements / use cases are:

  • Drag a Zotero entry for a paper into TBX and have it automatically named to the CiteKey
  • Take notes / excerpts for each paper and have them associated with the paper
  • Create notes for the premises and claims for each paper.
  • Create a link between each claim and premise that corresponds to ‘supports’, ‘opposes’, ‘informs’. Some will be a one to many relationship, of course, and will link between the notes from different papers. At the top of this hierarchy will be a ‘position’, which is set out to answer to answer the essay question - so every claim and premise ultimately either supports or opposes one or more positions.
    (I think all of those are possible, watching the awesome videos. But then…)
  • Do a search / query from a ‘position’ to find every related note for that position, so I end up with a map of alias notes that are just relevant to that position. Is that possible? It sounds a bit ‘recursive’ (probably the wrong term).
  • If it is possible, would those aliases arrive in the new map with the arrows between them still represented?

I really would appreciate some guidance here about what’s possible, then it might become a christmas project.

1 Like
#2

Everything in your list should be fairly straightforward, save for the final step. The aliases inside the agent would not be linked, but you could probably write a function to link them.

I assume from your description that you’re adopting an argumentation formalism derived from Toulmin?

Rosemary Simpson had a related paper at ACM Hypertext in 2020 on Augustine: https://dl.acm.org/doi/10.1145/3372923.3404814.

#3

Have you seen my training videos: Mastering Tinderbox: Training Videos (Complete List). You might find these useful, especially the latest linking training and the Zotero citation management videos.

To automated the positioning of the notes in map view we’d need to through the use of $XPOS and $POS Attributes against the category structure. Or, we some how find a way to use Hyperbolic view.

I’d love to learn more about the framework of this method: Question>Claim:Evidence:Oppose:Answer. Can someone explain it?

#4

You can check this topic for my partial implementation of Roam’s ‘discourse graph’ in Tinderbox.

#5

@eastgate Thank you, that’s useful.

I’m not sure I’m that formal, to be honest - argument mapping is a technique I’ve heard of but not in a terribly formal way (as yet), and then as I said my interest was piqued by the discourse graph extension for Roam Research - which I’ve been using on a trial of Roam, but I’m not keen to continue if I can avoid it.

@ShiJianhui oh fantastic, I’ll take a look, thank you. I’m sorry, I didn’t come across that before I posted.

#6

I have been watching them over the past couple of days, and they are great! Thank you. Some of them went over my head a bit and I’d need more time - e.g. your video on Zotero and how to extract the cite key from the note and make it the title, and how to reformat the extracts from Highlights to rename them. All of that is definitely something I’d like to do.

As @eastgage mentioned above, there’s some very formal argument mapping approaches, but I’m at the moment just looking to do something like what’s explained in this website (this is not the best explanation, by any means, but the one of the first that I came to. I can find better ones if it would be of interest)

and keep it simple for now, but with an aim of moving into more complicated analysis as my studies and my skills grow.

Edited to add: the whole thing is not dissimilar to this software, too:

https://sourceforge.net/p/hypernomicon/news/

1 Like