We leave a trail besides that of the Journal's fork history. Here I illustrate working back from a page I constructed from a document George Mobus made available.
We pick up the trail with this lineup: wiki 
We start with Structural Assistant, a derivative page.
We click 'search' in the wiki page footer.
We scroll to 'Find pages sharing any of these items.'
We click to open 'The Problem Domain'.
We click the ☼ 'create' action in the Journal.
We double click 'Thu Sep 10, 2020 9:21:45 PM'
{ "type": "create", "date": 1599798105119, "source": "http://ward.dojo.fed.wiki/structured-discourse.html", "item": { "title": "The Problem Domain", "story": [ ... ] }
We follow the web link to another wiki page
We arrive at extraction details for Structured Discourse.
> We'll translate an html document into wiki export.json format by inserting a script tag for translate.js at the end of the original. Loading the original will run the translator and download the export file. Drag this to any wiki to retrieve contents. source
ConsensUs: Computer-moderated Structured Discourse
We've arrived at our source and the mechanical extraction we've applied to the source document, source code included. We've followed a twisty path but without any steps immune to automation.