3 ms·
> There's a large cohort of smart people who've read a contract and think this means they grasp the law; but without understanding jurisprudence or equity, comm
by burrows 10y ago
> There's a large cohort of smart people who've read a contract and think this means they grasp the law; but without understanding jurisprudence or equity, common/case law, the myriad forms of legislation, actual legal proceedings &c &c.
There's a large cohort of smart people who've read a program and think this means they grasp its semantics; but without understanding the language specification, compiler bugs, undefined behavior implementation, posix specification, operating system bugs and configuration of deployment environment.
- smallnamespace 10y agoI don't see why you're getting downvoted here, because the analogy is pretty apt. Much like programming languages, law language has evolved to be written in a way that's unambiguous as possible. Like programming languages, contracts and the like don't succeed in having 100% predictable semantics. One source of unpredictability of results comes from the lack of a stable, 100% predictable execution platform (namely, judges, juries, lawyers, and current case law and legislation), much like compilers and platforms do not 100% specify their behavior, or even when they do, the run-time behavior is not perfectly predictable (e.g. when does the GC run). Another source of unpredictability is that law at the end of the day must interface with the non-legal world ('facts' vs. law) and so touches global context which isn't known when the law or contract was drafted, which one might compare to dealing with unforeseen interactions with the OS, deployment environments, user input, etc. So while laws and programs try to accurately and precisely encode their own internal semantics, they sometimes fail because they depend on other things that don't have perfectly defined semantics.