IETF defines may, must, shall, etc. in every document but doesn't use the same format as law.
1) I remember being in a Requirements Review meeting and spending an entire hour debating the meaning of a word. Don't remember the word, but it was something like "includes".
2) I separately remember being in a Requirements review for a feature that concluded successfully. Then later in a Design Review for that feature. And after that, Code and Unit Test reviews of the same feature. Then, months later, the feature was presented to the person who asked for it in the first place and then said, "that's not what I wanted. I wanted xxx instead."
Yes, this person was at the Requirements and Design reviews!
To do it, lawyers would have to learn some category theory, but I think many of them would enjoy it.
Many probably discovered it already, without realizing it has a name.
When I was I law school they taught us to write as clearly as possible and avoid legalese. Hopefully that catches on, I don't think legalese is good or necessary. Clarity is better
There’s a famous story of Warren Buffet buying Nebraska Furnitur Mart with no paperwork whatsoever, just some verbal assurances, a handshake, and a TV interview with the esteemed owner.
They certainly did.
coliveira is saying a change in legal language opens up a contract to a reasonably different interpretation in the future. And we know it's reasonable (even likely) because there are myriad historical precedents where such ambiguities were used in the past to win cases against, say, GrantMoyer's Optimized Non-Standard Legal Language Law Offices. By using standard language that specific class of ambiguity issues goes away.
On the other hand, any good faith definition of "superstitious" is a pejorative that holds the belief to be irrational. So when GrantMoyer's Optimized Non-Standard Legal Language Law Offices uses its patent-pending optimized, non-standard language in their contracts, they are throwing out all the superstitious terminology on the basis that nothing bad will actually happen as a result of doing this. (At least, aside from things like random chance or any of the other true reasons that just happen coincide with superstitious falsehoods.) That's a bad idea-- it isn't supported by historical precedent, and it is likelihood to cause predictably bad outcomes in the future.
The only mitigating factor is this weasel sentence:
"Except sometimes the superstitions are right."
But to be generous we must take that to mean only the truism that superstition sometimes coincides with things like random chance. (Otherwise it'd be like saying no one goes to a particular restaurant anymore because it's so crowded, which is funny but incoherent.)
Edit: clarifications. Also, in closing, Your Honor, I call for one of those "bad court thingies." :)
Sure, some "superstitions" are the result of random folklore, and can be ignored without consequence. (Knock on wood!) But others are the result of meme-ized facts, passed on via something that's folklorish and easy to remember to share.
Thus, it is fair to use the word "superstition" to describe something you think might be harmless, but may actually have serious consequences, if ignored, after all.
It's irrational to think the awkward phrasing is actually better, but when you're being judged by other people sometimes they're going to enforce the bias.
Everybody learns partially through pattern recognition, partially from trial and error, partially from armchair reasoning. And functions mostly by repeating what they already know.
But it's emotionally satisfying to put all the negative parts onto others (cargo cult sheeple reciting magic spells), and save all the positive identifiers for yourself (brave scientist uncovering truths about the universe)
Hackers (in the HN sense) also excel with programming language nuances.
Here's an example. Patent lawyers absolutely love the word "plurality", a very uncommon word in normal English. They mean "more than one", but why don't they say that? It's not only clearer but it's also less ambiguous (check the many definitions of "plurality").
It's because they are copying a rule without thinking about it. Other examples of that:
* "an historical occasion" - this is wrong, it's "a historical occasion", but people hear it and think it's some special rule
* Not using "I" in research papers, even with a single author! There's no rule that requires this but people think there is. I read a paper once that used "I". It was great.
But, in any case, for those who pronounce the h, “a historical” is easier to say.
In the case of lawyers I’d suspect tradition first and fear of typos reversing meaning (“ahistorical”) second.
You're only doing it because you've heard other people do it and you think there's some special rule for it. Maybe subconsciously even.
And don't say "you can't say something is wrong in language because prescriptivism Vs descriptivism etc."
Yes generally it is the case that language is defined by how it is used. This is an exception, where something has a very very clear purpose and usage, and you're doing it wrong.
I actually can't think of any clearer cases where the prescriptivists are right. It would have to be something so obvious you don't even think about it, like using spaces between words. Yet for some reason everyone messes up "a historical". It's as bizarre as if everyone omitted spaces before "historical" but not "history" or any other words.
The grammar police have yet to pull me over. I'll continue speaking as I do, thank you.
There is nothing wrong with the use of ‘an’ before ‘history’ and the forms of that lemma. It is unusual, and perhaps pretentious or otherwise silly, but only in the same way as ‘connexion’ is an unusual (&c) but perfectly correct spelling.
The point of saying "an" instead of "a" is to make it easier to say a following word that starts with a soft sound (there's probably a technical term).
"A igloo" is hard to say, so we change it to "An igloo" which is a lot easier.
"A history lesson" is not hard to say so there's no need for "an".
Have you ever said "an history lesson" or even heard anyone say that?
No. Of course not.
Not arguing correctness, just intuition.
Yes, often.
For that matter, the first day I was in Great Britain, I noticed that the other Americans I met had British accents, and I noticed in the few hours talking with them, I was already picking up on the accent myself.
It's only natural to pick up the language and idioms of the cultures you happen to be immersed in!
(And this applies to Law and Engineering as much as anything else, too.)
I would say maths has its own issues with language accessibility, but they're a bit different from law. Probably the biggest problem is naming everything after people instead of coming up with meaningful names.
I'd prefer it if previous cases had a bearing on rewriting/amending laws, rather than influencing a current case, and if the feeling is that a current judgement was not correct, rewrite/refine the law taking the case into account.
That way a judgement does not require previous case research, this does mean that law needs to become a more well specified, and unambiguous language. We have decades of classification and computer language theory to draw from, and LLMs to stub-translate existing law, with exceptions going to a dead-letter-queue for individual appraisal.
Law should be understandable by the layman, else how does someone know if they have broken it?
This is especially important in a world where precious few scenarios allow ignorance as a legal defense.
your process seems reasonable though.
> What matters is all the relevant case law
Not in all jurisdictions. And even in those places, the laws are often difficult to decipher.I think that the problem is trying to encode both intent and formal instructions in the same language. If a law were to state clearly "our intent is to prevent people from falling off tall buildings" and then specifying building code, then any ambiguities in implementation or new materials or new methods could be addressed by referencing the intent. This would be especially helpful in e.g. criminal law, where ambiguities or oversights in language (e.g. with regard to women raping men) often lead to both innocent people sitting in jail, and guilty people going free.
This is how American laws are usually written. Courts even venture into the debate records to discern intent.
Similar to the way it's brittle when there are money numbers that don't include inflation (either on purpose or not), or (rarely) use a hilariously wrong inflation number instead of an index (say for a law passed during a time of very high inflation). Then the numbers change (or don't), and everything gets weird.
Maybe with a new nation that had time to adapt to the scenario it could form a functional system, but if implemented in America today I'd give it a not insignificant chance of actually crashing the entire government like a computer attempting to open a zip bomb
reminds me of this: https://themedialine.org/by-region/irelands-push-to-alter-ic...
As something of a litmus test, if a crime is so minor that the police wouldn't do anything if you brought them multiple eyewitnesses, video evidence, and an address to knock on, then having that law on the books probably does more harm than good (many petty crimes like mild speeding would fit those criteria).
I see you are not American, and evidently are not really familiar with the second amendment.
Careful past that Chesterton fence
I read some of the argument transcripts and usually it is a bunch of lawyers trying to sort out what the law should be given it works this way in this case and that way in that case and this one goes up the middle or the circuits disagree. But sometimes it is much more nakedly political.
It's all people's decisions, there's no magic about it.
If you pay a repairman to come to your house and fix something, even if he does fix it in five seconds, you're going to feel a bit cheated because you paid him assuming it would take hours.
Same with a lawyer - if you pay one to write something for you, or review a contract, they WILL write something or find something that makes you feel like you got your money's worth.
The target audience is both lawyers and laymen. Every lawyer and politician who claims that it is not worth optimizing this also claims that they don't want ordinary citizens to be capable of obeying the law.
Tell this to every politican that if they don't work brutally hard on changing this fact they don't want ordinary citizens to be capable of obeying the law, and that they are thus (in the view of the ordinary citizen) working on to establish a arbitrary justice. Thus, they work on destroying (in the view of the ordinary constitutional state) on actively destroying the constitutional state. So, what makes these politicians and lawyers than differ from a despicable high traitor?
It’s always an y in under-emphasis. It’s like two generations of reviewers ago, there were a bunch of defamation lawsuits and every reviewer has mimicked this style since.
(Mostly what I’m saying is that legalese is essentially natural(ish) language ‘software’ that has been developed and maintained over centuries by non-coders[1] and that most of the seemingly pointless or obtuse stuff is probably actually significant in weird edge cases.)
[1] Not by any means saying that systems code is written by non-coders! But some of it is written by wizards and becomes Deep Magic that gets repeated by less eldrich contributors because if it isn’t then the authors sometimes meet a fiery doom.
I know this might be difficult for those of us who have specialized in computer science and have found a lot of things very easy, but sometimes, other professions are difficult and complicated too! Not everything is gatekeeping, and needing to consult an expert otherwise you'll footgun yourself, isn't a personal blow to you, as a person!
Some things are just complex! Programming as a field is "complex" but compared to any actual scientific field it probably the simplest to learn, pick up and understand. There's a reason why Doctors and Lawyers are highly paid and take forever to train.
Techbros have already tried to reinvent the wheel in other fields — crypto payments and metamed come to mind. Metamed learned that "actually, medicine is really absurdly difficult and you can't just instantly automate the job of a doctor with Technology" the hard way, and crypto payments are learning "why financial infrastructure and regulation exists in the first place", the hard way (through losing billions of dollars over and over again through exploits and ponzi schemes). I suspect that soon techbros like you will try to invent the legal system from scratch and learn why laws and the legal system are that way, hopefully it won't be the hard way! (Through a revolution)
Computing became cheaper because it has layer upon layer of economies of scale.
Law doesn’t because every case is unique and is literally decided upon minutiae and exceptional details.
Working in software breeds some truly wild perspectives on the world.
And smart software engineers never over-complicate their code, right?
“Wild” is the most generous word I could come up with, sorry!
Is software engineering arcane for similar gatekeepy reasons or is that for more legitimate reasons?
My personal suspicion is that software is one of the easiest fields in the world to learn, and it has the sneaky benefit that you don't have to brush up against many constraints of the real world, but then — this doesn't explain the apparent god complex many programmers have.
And the answer to your question about the legal system being faster or more efficient is that it has gotten much slower but this is due to the massively increased caseload across the system (that has not been matched by an increase in court resources) which is, ironically, likely due to an increasing ease of access to the courts caused by everyday lawsuits becoming cheaper and easier to file.
And yes to be fair, yes some stuff in a field is simply complicated and we need to use complicated tools, specifications, processes and other such things in order to describe it accurately. But lets face it - most lawyer stuff is not complicated, the field is just swamped with complexity, "technical debt" and they are relying on processes and fluid human-driven interpretations in order to make it all chug along. What makes the whole thing complicated are the unspoken, verbally and experience-driven pieces of arcane knowledge. Nothing is documented in simple flows, nothing can be automated according to them, and they require actual human actors to drive the process forward from one step to the next.
This is why they are so against automation and simplification of their field. They know that if they automate the processes, that their field becomes very focused on knowledge, and at that point we'll all just realize that most of what they know can be codified in some sort of unambiguous specification format.
A machine cannot be held accountable, and therefore should not be in charge of making decisions.
In theory, yes. In practice, not “everyone”, and very much not equally true of everyone to whom the law practically applies at all, even in the places which tout “equal protection of the laws” as a bedrock principle.
You can learn a lot by watching or asking others.
Your fear seems unfounded.
In medicine if you get things very wrong, you die, your patient dies, or everyone in the room dies.
In programming if you get things wrong, you give someone radiation poisoning and they die, a plane crashes, or someone loses days of work.
Surely, those are bigger injustices.
In that context, does it really seem so absurd that the rules are "obtuse"? Many people feel that way when reading someone else's code, but that doesn't mean all the weird conditionals are unnecessary.
Realistically, you’re probably not going to go to jail immediately but you can have a bad time if you’re caught doing some things you didn’t even know were illegal.
I’ve done software (a lot) and worked in a legislative office (a little), they are extremely different.
> Laws are programs.
Laws are not programs, or even meaningfully analogous to anything so deterministic. They are more like, if we must make a computing analogy, components of prompt templates (other portions of which will be filled by evidence and…well, a bunch of other stuff you don’t control when writing the laws) that will be used by an extremely high temperature LLM. Which you can’t actually test freely against, though you can observe past behavior – but, even then, you are stymied by the fact that the LLM itself is constantly updated with retrained, and sometimes rearchitected, versions, without useful release notes.
It's just that the programming language is mushy-- 400 definitions of the word "set" is the example my AI offers-- and the interpreters vary. In a pinch, the code writers know that nearly anything they write can be rescued by the chief interpreter, which has led to a model of legislatures intentionally writing ambiguous code and top-level courts interpreting with an emphasis on public good.
But, when you go down some arcane rabbit hole that will normally be interpreted by functionaries -- bankruptcy comes to mind-- suddenly everything is prescribed with remarkable precision and it becomes obvious again, law is code.
If this were really true, there would be zero appellate cases (and probably not even trial cases, as the clear and only correct resolution would be apparent to all parties without a trial which would only add expense) in these areas, and yet... That's very much not the case.
If law, even in these areas, is “code”, it is so more in the sense of the Pirates Code in Pirates of the Carribean, not computer software.
[1] https://www.law.cornell.edu/uscode/text/28/part-II/chapter-3...
If you look at code as in, only syntax and its meaning for the language, I can see how folks might reason along the lines of 'law is similar enough to code. You are saying what can and can't be done'
However, code is rarely only its syntax. Its meaning is contextual, often to the business of which it applies, and would have no real meaning outside of that context, and context may change at any given time, which means reinterpreting the code, replacing the code etc.
That is what law in practice, is what code is in practice. Its syntax is very technical, but its meaning is entirely dependent upon its applied context.
I been thinking alot about this lately, and understanding the sociological implications of code in an organization (or in more extreme cases, society) has really changed my point of view of how to build software
If I were to translate this into code, it would mean that variables that were constant will sometimes still change value. It would mean that operations on data would differ depending on the scope in which the operations were applied. And it would mean that certain variables that were closer to the base address (more money) would operate differently than variables stored in distant memory offsets [2].
[1] https://en.m.wikipedia.org/wiki/Plaxico_Burress#:~:text=On%2...)
[2] https://en.m.wikipedia.org/wiki/Trial_and_sentencing_of_Robe...
But on the other hand, you shouldn't need a law degree and bar license to sue an insurance company to pay up. Their client hit you, you're injured, now pay up or have a judge order you to. But the lawmen have carved out their own moat that puts ordinary people in a position to rely on them.
Plenty of jobs employ around that number: house maids, bookkeepers, groundskeepers, and a host of professions the average person rarely or never employs.
I suspect many of them may wonder why we have twice as many programmers :)
If the American legal system is any indication, the vague, bad stuff is the desired outcome.
It's amazing the regulations that the EU has where it says "no, do not do this", and not only do people generally not do it, the people who do get caught doing it are fined into oblivion for it. There's no culture of hiring teams of lawyers to poke holes in the law to get around something society clearly desired, because the court is going to do what society desired.
Why, yes, I am talking about data privacy, but there are other examples.
Wat. I’ve literally seen tax laws in Germany and Sweden which could only possibly apply to one family. European law is in sans serif; that doesn’t mean it’s less convoluted than American law.
You’re seriously arguing that the EU bureaucracy is streamlined and comprehensible without lawyers? Why do you think every rich person and powerful firm in Europe has fleets of lawyers?
Are we watching the same Supreme Court?
That, to me, sounds a lot like a judiciary that wants to have things decided via the interpretation of legalese by a lawyer instead of by facts as determined by the present will of the people.
… there are consequences for lawyers if they fail to do this translation effectively…
There can be no lawyer at every time you do almost anything in life, be in traffic, raise children, participate in education, purchase something, travel, anything! Do you need an engineer sitting right next to you when you turn on the light or operate a vechicle? No! You need them for advanced things.
If you argue that ALL law have to be understood only by lawyers then you are very very lost! What you have to obey, you have to understand! At least the relevant part (nuclear waste storage is perhaps not for you, I am just guessing), but even knowing which part is relevant you need to understand! Not enough lawyers to sit next to everyone guiding through everyday life that is regulated by law!
It is difficult to understand because it made to sound ominous (fucked up for everyone), that's it, see the article. Form over essence. Also evolving without effort for restructuring so it becomes overcomplicated (see article). All formality over practicality and common sense, that's why it is so convoluted, or "it’s an accidental property", mimicing for centuries.
Also don't mix up everyday language here used with friends over a beer that could be vague for outsiders - or themselves - but there ARE ways to be specific and understood with ordinary people too using the generic communication method law uses too: language. That could been used, that is being used in other areas in communicating towards layman (like doctors, I mean the smart ones, not engulfing you with latin explaining your condition and prognosis, right? Also instruction manuals for complex gadgets are not in ancient greek just for the sake of it, right, but ordinary and specific or even formal language that are still being understood by the ordinary consumer, as that is important.)
Additionally: "Lawyers tended to prefer plain English versions of documents, and they rated those versions to be just as enforceable as traditional legal documents.". I rest my case.
Everyone knows (sans lawyer) you can’t kill people, steal things, run red lights, commit fraud, lie in contracts, and so on.
Regarding your closing argument, what a wonderful argument against your point. Of course plain language contracts are just as enforceable as traditional ones. That is, you cannot argue, “your honor, this contract was written in plain language therefore it is not valid.”
But enforceability is a distinct concept from (really a subset of) defensibility. The problem with plain language is it tends to be less exhaustive and more ambiguous, which while enforceable (a valid contract with force of law) will be easier to wiggle out of in court.
To be clear, I understand your point and share your desire for this to be the case. Just having worked in a complex regulatory environment (health data), I don’t see it as even close to realistic, and it’s unrealistic for very good reasons.
It's "diff" but for legal text, using natural language (well as much as legalese is natural).
I don't know in english but in french you can definitely have a legal text (say for your company statuses) and then patches when you did modify things. Or you can have the "consolidated" text, which is the text with the patches already applied (it's definitely more of an easier read as there's no need to jump back and forth between the patches).
And of course the patches can reference another legalese text, which itself can reference another legalese text. And all of these can have... Patches.
These patches/revisions going out of their way to introduce a patching logic using natural language is plain weird.
Besides these patches/revisions, I'd say the second most WTF legalese thing is that somehow, at some point, where there were already rules and laws governing what a person could and could not do, someone, somewhere (in Rome probably), decided that we could create new persons, but not physical persons: entities dealt with as if they were individuals. But virtual.
I did set up a company: I created a person. And we do shareholders meetings and vote stuff and go in front of a notary and that notary comes up with revisions and we patch the person I created by sending the patches to the authorities.
It's completely mindboggling.
For example, every german has a right to control pictures of their own self, and the way the law was worded, it already applied an idea on the legality of sharing/distributing deepfakes of regular citizens.
However, this kind of generalization requires very careful wording. Every single word and every single statement will be examined for inclusion and exclusion of situations across many court cases. This means it's hard to formulate these laws.
And then you rely on age-old, court- and battle-proven formulations... even if they are clunky and not intuitive.
Thus, an artificial eye recording -- and even recording to an SD card -- would be no problem.
And that is how a good law should be written: It should be written with the rights, duties and protection of citizens involved so it allows the judges to make new technology fit into the framework of the rights of the people.
And sure, it may fail at times. But that is what appeals and adjustments are for.
Well to add to that, representing yourself is a really bad optics and you can't fervently defend yourselves. That's why even lawyers when put on trial will use other lawyers.
I also think lawyers don't find it worth optimising the language much, because the target audience is other lawyers who know all the incantations anyway. The target audience is lawyers instead of laypeople, because the language isn't what matters in the first place. What matters is all the relevant case law, which can vary so much from the "obvious" or even "right" interpretaion of the written language, that a layperson not using a lawyer for all but the most trivial legal needs is a recipe for disaster.