- From: Wayne Dick <wayneedick@gmail.com>
- Date: Sat, 29 Mar 2025 22:09:06 -0700
- To: kerscher@montana.com
- Cc: W3C WAI ig <w3c-wai-ig@w3.org>
- Message-ID: <CAJeQ8SCtpaRgVA_hY6S2RuwAR8pWf7UTubueXmd7jOBgcVdjLg@mail.gmail.com>
There are several special difficulties with mathematical language, beyond notation. Notation is nothing trivial but it is the first model of programming language. We can parse it. But the language style is stilted. It is deeply understated: *nontrivial *means really hard, etc. There is also, passive voice: This drives grammar parsers crazy, but if we substitute "the universe" as the subject for "it satisfies..." Mathematical sentences make sense actively. That is what 99% of mathematical passive voice means. Different definitions: Words can be mis-parsed. For example: "Irrational" has two meanings: Mathematical and common. In mathematics it can be an adjective or a noun. They are very different. Even a trained GPT engine can be confused. This setting of linguistics of expressions to specific structures is commonplace. So if one wants to use a word processor or AI bot to examine mathematical text there are two barriers. There are the usual barriers, a babel of formats, there is a semantic barrier. I present this analysis, because I believe it is universal. When WAI speaks of "jargon" is really is trying to align standard language with discipline specific vocabulary and notation. If I succeed in mapping PDF with Mathematics into LaTeX, I could map the Latex into HTML, or someone could. I'm gonna pay $200/month to get a dedicated Chat to do this. I'll report back in May. Best
Received on Sunday, 30 March 2025 05:09:48 UTC