Types of Legal Argument

The article explains why it's simplistic to think of legal argument as pure maths-style logical deduction. This goes back at least to Oliver Wendell Holmes (US supreme court justice from 1902), who made fun of those who treated a dissenting judgement "as if it meant simply that one side or the other were not doing their sums right, and if they would take more trouble, agreement inevitably would come."

Holmes said that law should be seen as induction, not deduction. Induction involves pattern matching, analogizing, and so on. I think that at least the first 4 of the 5 types of argument described in the article can be thought of as types of induction.

Holmes also said that the law consists of nothing more than prediction of what courts will do. And he described the development of the law in language that would make any Machine Learning person think of nearest neighbour classification: "The growth of the law is very apt to take place in this way: two widely different cases suggest a general distinction, which is a clear one when stated broadly. But as new cases cluster around the opposite poles, and begin to approach each other [...] at last a mathematical line is arrived at by the contact of contrary decisions."

It seems to me that the legal theory of Holmes, and the modern practice of machine learning, are remarkably similar. They're both about prediction based on experience / datasets. Shameless plug: I've written a (free!) book about it, "On the path to AI: Law’s prophecies and the conceptual foundations of the machine learning age".

https://link.springer.com/book/10.1007/978-3-030-43582-0

I've been dabbling in expert systems (aka rules engines) for the last few years. See CLIPs and Rete algo in your language of choice.

This, combined with exposure to KNN over TF-IDF (either live indexes like Elasticsearch under the hood or GiST over trigrams in Postgres or custom trained models), allowed me to discover a kind of "ratcheting" of knowledge.

If the I/O of the program is like an interview with a client, then it's possible to have a "topic stack" where each frame represents a volley in that conversation. I think chatbots also go deep on "dialog engines" like this (see RavenClaw).

For each fact being input, it's possible to check all the policies (rules) that might apply in a declarative way where the developer doesn't need to care about order. Rete takes care of performance concerns, but it's not horizontally scalable (which doesn't really matter IMHO).

Then, for the same stack frame, checking what all other facts are by doing that light KNN to see which other fact patterns were similar, and computing the rule matching on those selectively. This is like making analogies to case law.

Rinse and repeat that flow until some kind of solution is found/suggested/validated by the user, etc. See SHYSTER paper which covered this approach in the 90s.

The article does address the two points you mention, and also addresses contractual arguments and examples.

> Argue your point as best you can and hope!

This sounds like a recipe for distress. Isn't the utility of the idea that whatever argument you face, whatever tort or contract or criminal allegation you are fighting for, your best will be better if you understand the logical premise of the argument and the way it is carried forward?

What do you suggest the students learn in addition to or instead of this?

The legal concept you're describing is "comparative responsibility"[1]:

> [A] doctrine of tort law that compares the fault of each party in a lawsuit for a single injury.

> Comparative responsibility divides the fault among parties by percentages, and then accordingly divides the money awarded to the plaintiff. The plaintiff may only recover the percentage of the damages he is not at fault for. If a plaintiff is found to be 25% at fault, he can recover only 75% of his damages.

The answer to your hypothetical is, of course, it depends! The "United States" section of the wikipedia article has a good outline of the different ways states implement the doctrine. It boils down to: sometimes it can be a complete bar to recovery if the victim is at all responsible, sometimes there's a threshold level of responsibility where it will bar recovery, and sometimes it's a percentage reduction as described in the example above.

[1]https://en.wikipedia.org/wiki/Comparative_responsibility