Show HN: International Legal Dictionary | Hacker News Reader

Show HN: International Legal Dictionary

public.law

105 pointsdogweather33 comments

I'm a pro bono attorney, and have been really interested in making the law easier to understand. I also see a ton of government resources online in varying degrees of usability. And for sure, there's no interoperability between them.

This is an international legal dictionary, an experiment in improving the situation: glossaries are scraped and parses from official sources: https://github.com/public-law/open-gov-crawlers. The results are saved as datasets in well formed JSON with Dublin Core metadata: https://github.com/public-law/datasets

I add Library of Congress subject headings to the sources, to enable filtering (still to come).

The web app is basically an old-school mashup, which I've always liked.

Another experiment is using the Dale-Chall readability formula to improve the reader's experience. Here's an example of it at work:

https://www.public.law/dictionary/entries/amicus-curiae

This is an experiment, using readability as a relative metric. I.e., not extracing an absolute grade-level score as its normaly used. Instead, using it to compare different definitions of the same phrase. My theory is, there's strong scientific validity for this use, even when applied to very short passages: All I simply want is to figure out, "Which is more readable? Passage A or B?" And then, my code sorts the definitions in order of readability to (theoretically) produce a newspaper-article-like effect: A reader can read the first couple of sentences to get an overview of the story.

Open articleView on HN

A somewhat similar project that I started at Google (this was US-only) was:

Common constructions of computer terms ("database", "processor", "thread", etc.) These words appear in patent claims, and the claims are what determine infringement (I use the word "construction" instead of "definition" deliberately).

You might well think there's a standard dictionary of them. There is not, at least as of 5-6 years ago. The parties in a lawsuit propose a construction, and the judge creates one. He's not obliged to follow precedent. Thus you find that "database," for example, has been construed to include a flat file. The construction is supposed to be what the patent writer meant at the time of filing OR what the general industry meaning was.

If that sounds murky, it is. With a chemical patent, there are standard meanings for words, and you won't get far arguing that "methylated" doesn't mean what it generally means. Not so in software.

Often after the judge's claim construction, one side will offer to settle, because they know they'll lose.

Sometimes you'll see a patent from some other country, and of course the meaning of the terms there may not be the same as in the US. So this would be a much, much bigger project than you've taken on.

Your experience reminds me of my own disappointment I experienced as an engineer learning the law: It's nowhere near as organized, structured, and rational as I assumed it would be.

There are some patent dictionaries based on how US courts defined terms. For example (over a decade old): http://www.pubpat.org/garrod-glossaries.htm

The USPTO also had a glossary pilot program at some point, but it ended and I guess was not considered useful: https://www.uspto.gov/patents/initiatives/glossary-initiativ...

I think patent attorneys and agents could avoid most of this problem if they wrote better. I'm a patent examiner (and this post only represents my own opinion, not that of the USPTO or US govt.), but it seems to me that patent applications are typically written quickly. Clearly defining terms is an afterthought, particularly given that many attorneys and agents see value in being vague.

Yes, maybe their explicit definition could cause them trouble later, but I think the lack of an explicit definition is more likely to cause trouble. I have examined applications with useful explicit definitions, or a reference to a preferred dictionary. I also examined one application where the applicant redefined a term contrary to its normal meaning in the middle of the specification, and that redefinition was critical to understanding the claims. I personally think all special definitions should go in an easy-to-find section.

Edit: I'll also use this opportunity to make a shameless plug for my patent claim analyzer, which will identify a lot of ambiguous language in patent claims.

Thanks, and I'm all about shameless plugs :)

Having been in Litigation: there's immense value in having vague terms. 15 years later when you're asserting the patent, you really need to be able to say your claims cover something you couldn't possibly have anticipated.

I know that's legal gamesmanship and I'm not applauding it.

I'm not sure how mixing legal terms from differing domains will pan out for your intended target audience, here in the Commonwealth hegemony we have issue with self representation in Court from those educated via central north american police and law shows .. quoting their US Constitutional Rights as a Free Citizen, etc.

That said, you might like, say, the New South Wales (Australian) guide to legal terms and self help:

https://www.lawaccess.nsw.gov.au/Pages/representing/lawassis...

which is government funded and assists in connecting people with legal aid and translating legal terms to plainspeak.

Best wishes with your endeavour.

Lol, thanks for the perspective from the across the water. I'm already surprised to learn how similar the concepts are (still) across all the common law countries. Even, say, between the U.S. and New Zealand. I was also surprised to find out how easy Canada's sites are to parse, which is why there are so many Canadian sources in the dataset. They're on top of their game with compliant and uniform web apps.

And yep, I found that page too, and it's on my list of Australian sources to add. This one's good, although not as readable:

https://fedcourt.gov.au/digital-law-library/glossary-of-lega...

> I'm already surprised to learn how similar the concepts are (still) across all the common law countries.

We can blame the Roman's and, in no small part, William Garrow for all that ..

[1] https://en.wikipedia.org/wiki/William_Garrow

Despite concepts in common there can still remain the complaint of Shaw (and later Wilde) that we are divided by a common language.

Readability is good, at some point in any legal procedure though there comes an importance of being precise about nomenclature.

In the field of mathematics, for example, such phrases as "if and only if" and "neccesary and sufficient" go deeper than casual reading might understand.

Now comes the twist in law that common phrases may have different interpretations between countries, or as may the case in the USofA, between different counties or states.

Still any move to increase understanding of the law in general with pointers to the law in specific locations is a good one.

Thanks! Interesting info. For sure, I agree that vocabulary is important, and specialized words are extremely valuable. On the other hand, I'm intrigued by professionally written dictionaries that use a simplified English of a few thousand words in the definitions. Then, of course, sentence length and structure also come into play.

BTW, the Dale-Chall formula uses two factors to score readability: (1) Average sentence length (longer is worse), and (2) percent of "difficult" words. They've done scientific validation of this small and easy set of metrics. (See The New Dale-Chall Formula.) To determine whether a word is "difficult", they created a list of 3,000 "easy" words that 80% of 4th graders know. (It was developed to help design textbooks for children.) It's an interesting algorithm which is an easy-to-implement proxy for deeper linguistic analysis.

I've found that this very simple algorithm does a decent job of putting the easiest-to-grok definitions at the top. Meanwhile, sentences with parentheticals, lots of prepositions and passive voice are rated less readable and pushed towards the bottom of the page.

Conversationally, in a beer over a barbie kind of way, no offence intended,

Dale-Chell lost me at "words that 80% of fourth-graders in American schools knew in 1948 (revised 1984)".

A cookie in 1984 is not the cookie of the 2020s web savvy teenager.

This particular article from the UK shares many of my own feelings about automated readability indexes:

https://www.effortmark.co.uk/readability-formulas-seven-reas...

I still admire your goal here, I'm simply suggesting to rely on a modicum of actual human feedback over automation where possible.

> beer over a barbie

I immediately pictured rude things being done to a Barbie Doll. Which is probably your point.

  > issue with self representation in Court from those educated via
  > central north american police and law shows

Interestingly, as one who does not watch television, I had defense attorney try to explain to me a concept as "like on television shows". This was a professional with a law degree, to whom I was paying a not-insignificant sum of my earnings, trying to describe some concept with pop-culture analogies assuming that everyone in the world would be familiar with American television. And we're not in the US, don't have the same legal system as the US, and don't even use the same language as the US.

In general, the documented precedent cases are used to develop a consensus by legal crowds. In many places "jury nullification" can theoretically supersede the courts judgement, but will likely get you kicked from the pool if you know such things exist. This is also why many punitive laws survive, simply by settling out of court via plea deals to avoid direct community/jury input. Citizens are still in charge, but decades of media often convinced people due process was somehow a bad thing (Socrates would be disappointed).

As a side note, if you are not using a legal rep, than expect people to make a bureaucracy more painful. Talk to a lawyer before you try to talk with anyone else, as you may be initially innocent.. but say/do something contradictory that still lands you in trouble. I also advise against doing your own dentistry as well. ;-)

It would be cool to have a law database effort which focuses on building something accessible that makes lawyers lives easier, and provides a tax free donation link to a nonprofit organization. Perhaps some sort of curated wiki style journal site with a validated legal comment section.

I like my lawyers and commercial accountants, as they were the best investment one could make for a business. =)

Impressive and really readable.

I feel this would benefit from a slightly different name such as International Common Law Dictionary or something similar instead of the general term used now.

I believe that many concepts found in common law just don’t translate to other law systems (let’s say Sweden, Spain, Germany) and a translation of the words (which would help in terms of understanding) actually doesn’t do the concepts justice because they just don’t exist in the other systems.

E.g. „Article III judge“. Probably won’t be found in other systems, but easily translatable or the Chapter entries (Chapter 11, Chapter 9 etc) - surely every law has some kind of similar law or code or whatever, but I believe it couldn’t be directly related to these code chapters.

> a slightly different name such as International Common Law Dictionary

That's a good idea. The generic name I'm using doesn't really describe it well.

And yes — it'll be really interesting to see what happens when I try to add Civil Law info. I mean, do they even explain the law over there? It's a whole other country. I kid, I kid. These will also be non-English, so that's another dimension of complexity.

Your thought about concepts not existing is a good one. Maybe a deft use of subject tags will help. I'm not convinced that LoC Subject Headings is the right choice, but they should handle anything under the sun. And they'll enable visitors to understand and filter the scope of the info.

Try Louisiana. Civil law, in English.

617 entries is nothing. Look at Black's Law Dictionary. The problem with Black's is that as much as a third is archaic. Maybe half. Uncommon words and phrases that lawyers use would be useful--"affiant" is missing, btw--but at best this is a minor glossary. "Process server" is also missing. I once saw this translated as "judicial notifier." It's a phrase "uncommon" lawyers have trouble with, since it requires an understanding of two concepts that have nothing to do with each other.

The idea of using the program for specialized vocabulary, or building your own glossary, is much more interesting.

Thanks, interesting critique. The app is an MVP - really a platform into which new definitions can be poured.

I did a little searching and found this source for affiant which will be good to add. https://www.sos.nd.gov/notaries-public/notary-terms-and-defi...

For sure, I agree, definitions of the lesser common words will be the most valuable to site visitors.

Great work! If relying only on government sources is not a pre-requisite, you could also consider this site for glossary of terms under Indian laws (https://nyaaya.org/glossary/). I am not associated with them, but they do incredible work in providing digital, accessible legal information to the general public. And they make it available under creative commons license.

Thank you! That's a really good glossary. Beautiful plain English. But yeah, so far I'm relying on government texts, and statutes (actual laws) if possible. Mostly to increase confidence in the resource. But also to not worry about copyright issues.

By the way, if you find any English-language Indian glossaries published by the government, I'd be very interested. I've done some searching.

Okay- so all of the legislations and government notifications by the central/federal government are now officially published only in digital form and are available on this website (https://egazette.nic.in/).

For your purposes, (a dictionary of key terms, unfortunately there is no government publication. However, one good hack you can consider is to look at the Section 3 of the General Clauses Act, 1897. The terms defined under this Act will govern the meaning in all the central/federal legislations and rules unless those legislations specifically define these terms. The terms defined under General Clauses Act include terms such as affidavit, immovable property, person etc. (essentially terms similar to the ones used in your dictionary). Here is the link; hope you find it useful (https://www.indiacode.nic.in/show-data?actid=AC_CEN_3_20_000...)

p.s.- https://www.indiacode.nic.in is also a government website where you can find copies of legislations and rules.

That Section 3 is fantastic. Thank you. I actually prefer definitions directly from statutes, because they're actual law and the copyright situation is better too. Are you generally familiar with Indian law? And a Python programmer? :-)

The main difference between this, versus using a web search engine, appears to be the readability sorting of definitions and the fact that the definitions are scraped from government resources.

The sorting is quite a bit of fun to see in action.

But, what advantage is there that the source is from a government resource? Versus, say, doing a Google search, or browsing Merriam-Webster's legal dictionary, or doing a search through Black's online edition, and so on.

There are several interesting advantages to government sources:

First, they're public-domain, or should be! In reality, there's crown copyright in the commonwealth countries. And every US State is its own fiefdom often with arcane publishing deals that keep the laws out of the public domain (or so they claim). In the US, only federally authored works are clearly public domain by statute. I got tangentially involved in the issue when I was in law school when I made my first app for Oregon: https://boingboing.net/2009/10/30/oregon-once-again-cl.html

Second, my thinking goes like this: For most people, interactions with their local government is a big part of the legal process. And so knowing how the government defines words is extremely important. IMO it's very persuasive to use the govt's own text in one's arguments. Some of the dictionary's definitions are from actual statutes, and so have the force of law. E.g. https://www.public.law/dictionary/sources/world.public.law__...

Black's is a fantastic book, and I'm a big fan of its editor, Bryan Garner. But as far legal arguments go, it's merely a somewhat persuasive "secondary source."

Ok, and finally, by me picking the sources, I'm filtering out tons of unattributed crap that's online.

Congratulations on the launch!

The content on public.law (great domain) is useful and looks like its headed in the right direction. A small suggestion would be to move the newsletter signup from home page (or) at-least from hero to footer? And get other USP like the legal dictionary there.

Since the site covers international law, Perhaps you might be interested in the need-gap: 'Compare law of the land in an easy way'[1] posted on my problem validation forum; Say like if someone travels from USA to India then what major laws should they know about.

[1] https://needgap.com/problems/255-compare-law-of-the-land-in-...

Interesting. I went to look for England&Wales entries and the first local (to UK) entry was from Ireland: https://www.public.law/dictionary/entries/emergency-care-ord....

It talks about "Tusla", which term was entirely opaque to me. Apparently it's a neonym for a child welfare agency. An otherwise very readable sentence became entirely opaque due to use of a proper noun.

The "ex parte" entry appears to say that it's a two party proceeding, in USA, is that true? In UK it's one-party before the court/tribunal. Multi-party proceedings where a party fails to respond are not ex parte in UK law (AFAIK) but are still inter partes.

This Singapore law glossary might be useful: https://www.judiciary.gov.sg/news-and-resources/glossary . Should be straightforward to scrape too.

Sweet! I've wanted a Singapore source.

Looks good, sending it over to legal team at my company

Well, my very first search yielded me:

Page not found

We're sorry, but we could not find the page you requested: www.public.law/search?term=perjury

https://www.public.law/dictionary/entries/perjury

Looks like the search function is broken.

Thanks - Not implemented yet. I put the form on the page before implementing the backend for it.

Ah - thank you - I'm removing the search form for now. Yep, I still need to implement search.

Very cool project. Wishing you good luck!

Reply on news.ycombinator.com