The structure of reasoning about opinion-based things
I am hypothesizing that argumentation about opinion-based claims must generally follow a predictable structure: opinions near the top of the “supports” pyramid and truth-based claims at the bottom. At some point in between there must be a transition layer where oughts are dependent on is’s. For example, you might say 01 | We should ban GMO crops (opinion-based) and support it with the claim 02 | GMO crops can reduce biodiversity (fact-based). Notice this transition — in apparent violation of Hume’s law, an opinion is being supported with a fact. Whether this is convincing depends on other, unspecified things such as whether biodiversity is good, which are Toulmin warrants for the “supports” relation between 02 and 01. Much disagreement is probably caused by transitioning from ought to is too without a clear Toulmin warrant. Ideally, a “spiral” of warrants would extend out from the original “supports” relation, e.g., 03 | Biodiversity is good as the first warrant. Then 04 | Biodiversity makes the world more beautiful supports 03, which is warranted by 05 | Beauty is good. When you reach 05, you’ve pretty much gone as far as you can go; 05 cannot be justified in terms of anything else. It’s a statement of value at the deepest level. There is still a lot to explore here about why, even if a person agreed with these claims, they might not find them convincing. For example, even if you agree that biodiversity is good, what you really need is for biodiversity to be better than whatever you’d be giving up if you ban GMO crops. Evaluating tradeoffs is a pretty important part of modelling (and communicating) reasoning.
The role of Toulmin warrants
The structure of traditional argument maps consists of a root claim with branching supporting and objecting claims underneath it. Each relation is a binary structure; there is the support and there is the claim being supported. In contrast, Stephen Toulmin popularized a basic structure consisting of three parts: the claim, the support, and the warrant, whose purpose is to justify that the support does, in fact, lend validity to the claim. I have found this third element to be sorely missing from Kialo — a platform that supports only the traditional argument map structure — and is the reason that I wanted to introduce “relation” claims into Ameliorate.
// Under what conditions are toulmin warrants necessary? Why can you sometimes get away without them? Is there a way to make model them in a way that actually encourages people to think about this part of an argument (rather than, say, just having a “relation” claim in the middle, whose purpose may not always be clear)?
// Are there versions of Toulmin warrants that apply to relation types other than “supports”? Is there a warrant for “critiques”?