James Grimmelmann has an interesting pointer to an article purporting to describe the difference between how computer scientists and lawyers think. The core of the article is that legal data has “color”, or provenance.
It's a fun essay, but as someone who often straddles this divide, I think it's missing something important. But darned if I can put my finger on what it is. Maybe that law is often about shades of gray, when computer logic is binary?
The Paranoia metaphor in the essay is slightly off, since Colour is explicit in Paranoia.
Saying that computer science is entirely “binary” is a little too reductive, I think. Hackers know that there are things out there called ‘real numbers’ that can split any difference. My sense is that law is computer science and arguments, with the two halves weighted about equally.
Pingback: Letters of Marque
Pingback: Letters of Marque
Pingback: Letters of Marque
I wasn’t entirely impressed. The author’s definition of “bits” is shaky. Basically, in many examples he’s looking at only the current state of an object (to use CS speak) in situations where the object history is important. But there isn’t any reason that the object’s history and provenance can’t be encoded in the object. By choosing only some of the fields to be the “bits”, he ignores others that provide the Colour.
As far as not taking tide times from a website, I assume the legal disclaimer is there to absolve the site of responsibility for error, not to make plaintiffs seek out an expert witness. (Non-lawyer question: wouldn’t a court take judicial notice of the official tide tables of Government authorities without recourse to an expert witness??)
Lazarus – I think you’re missing a main point of the essay. The fact that an object can encode history and provenance doesn’t matter. There is always metadata that cannot be encoded. Example: If I take a Britney Spears track, modify the history and provenance metadata, and claim it as my own, the law will likely not respect the object’s version of the story.
In other words, objects don’t inherently have audit trails. But is that so different from real life? Isn’t it possible that Cage recorded his silence precisely by the /dev/null technique, in which case that would be the more authentic method of replication? Colour isn’t safe from fraud or corruption either. Indeed, I would say that the “bit” world with checksums and public keys is potentially more reliable.
The difference between law and computer science, in my view, is that lawyers don’t have to think things through rigorously. They can get away with “close enough”, and sometimes they can even get away with “I would like this to be true, and I’m smarter than you”. Computer scientists, on the other hand, test their “arguments” against an unforgiving reality, and so they have to be more concerned with being right than with sounding right.
—–