Hacker News Reader

Comment by tc | Hacker News Reader

There's a legal scholar (and I'm blanking on the source at the moment) that makes a very convincing point that a small handful of laws (against violence, fraud [1], etc.) covers 95-98% of all cases. The other 2-5% of cases are necessarily judgement calls. What's interesting is that the percentage of cases requiring judgement calls doesn't decrease as you go from N laws to 10,000N laws.

[1] To your point: it's true that we have more things we can make laws about, but I suspect that it isn't true that we should make laws about every specific thing. Fraud is fraud, murder is murder. It shouldn't matter whether you use the internet, a telephone, or pencil and paper to defraud someone, in the same way that we don't need a separate murder statute for each choice of murder weapon.

View on HN

And the exact application of those few laws in nearly infinite situations would require first a caste of extremely powerful judges, and later an accumulation of precedents which would effectively become a massive and complex legal code in itself.

Some brilliant thinkers centuries ago developed an ingenious technology for constraining judges: juries.

Of course, juries work better when they are able to find on both the facts and the justice of the law. We've deviated from this at our peril.

Mens rea, criminal intent, has historically been the simplifying agent in law. That's why you don't see a terribly complicated set of case laws surrounding, say, murder. The problem with lots of current criminal statutes is that you can be guilty of violating them without any intent to harm.

And will your legal system include only such clear-cut cases as premeditated murder? What do you do when someone starts file-sharing your program? Apply a single, simple, almost certainly ambiguously worded law about theft? And let's pretend your law is perfectly worded so as to have a clear application in every single situation: do you deem file-sharing theft?

What about if, in self-defense, someone goes further than necessary? He's attacked and choked by someone bigger and stronger. He reasonably fears for his life. Through chance or skill, he gets the upper hand, and knocks his attacker out. Then, before caveman adrenaline has faded, he slams the guy's head into the pavement until he dies. Do you apply the single "thou shalt not kill" law?

I don't really want answers to these specific questions, but to the meta-question: how do you handle actions which are harmful to society, but which cannot be resolved by straightforward common-sense ethical reasoning?

I would refer back to my grand-parent post.

The hard, borderline cases that you mention are relatively rare. No matter how complicated your system of laws, these cases tend to require humans, either judges or juries, to make a careful decision.

If you have 10,000 laws rather than 10, you end up instead having to decide which of the many conflicting laws that could be applied should be applied to the situation. In marginal cases, human judgement is extremely difficult to replace.

Well, from this and another post, you seem to be concerned mainly with overlap between similar laws -- a natural result of a large and fragmented legal code. Why not prune it instead of scrapping it? Refactor instead of rewrite? Insert jwz quote.

What do you do when someone starts file-sharing your program? Apply a single, simple, almost certainly ambiguously worded law about theft?

One simple law about theft that can't neatly answer the question seems preferable to the current tens or hundreds of possibly applicable pedantically worded laws that can't neatly answer the question.

You've disregarded the second, stronger portion of that argument. Constrained though they may be, the trail of precedents judges would generate in the "1000 laws" scenario would simply constitute a complex shadow law.

I believe I addressed that with the note about criminal intent.

Common law has tended to focus obsessively on mens rea. Statute law has tended to disregard it. The difference probably has something to do with having to make a judgement about a particular case with the humans that are going to be affected standing there in front of you, as compared to dreaming up a bunch of laws in the abstract, or coming up with laws to address a politically attractive but wildly uncommon situation.

I believe this argument is empirically strong. You can see for yourself that there are (at least) hundreds of years of case law about murder. And yet the law surrounding murder seems quite a bit more straightforward than the laws surrounding, say, exporting electronics. This is because criminal case law has been sharpened by a focus on criminal intent.

As recently as this year, SCOTUS has made rulings that set new legal precedent for murder (Google "Tonya Harden"). I don't even have to argue that it's specious to compare murder to environmental destruction or occupational safety to rebut that argument.

That would be the West Virginia Supreme Court, not SCOTUS.

http://www.huntingtonnews.net/state/090605-rutherford-statet...

And in any case, arguing around the edges of self-defense is a far cry from the sort of felony paperwork violations that got us into this discussion.

So perhaps the most clarifying question I could ask is: how many examples can you find of case law criminalizing paperwork mistakes with no criminal intent or intent to harm others, as compared to how many examples of that are there in statute law?

You're right, my bad. Hopefully my point stands intact, which is that even for a cut-and-dry crime like murder, court precedent defines a fairly complex shadow law --- self defense, intent to kill, competence, etc.

Like (I think) most reasonable people, I'm not in favor of criminalizing paperwork mistakes with no criminal intent.

> an accumulation of precedents which would effectively become a massive and complex legal code in itself.

Common law? Worked out pretty well for the English for about 1000 years.

Reply on news.ycombinator.com