Catala

27 views
Skip to first unread message

Connor McCormick

unread,
Mar 31, 2021, 3:06:29 PM3/31/21
to Legalese talk
Interested to hear general thoughts on Catala: A Programming Language for the Law and, in particular, the merits or demerits of Default Logic based systems. 


Jason Morris

unread,
May 30, 2021, 4:16:31 AM5/30/21
to Legalese talk, cnnr.m...@gmail.com
I haven't used it, yet, though I want to, so I don't know how much of the following applies to Catala specifically, yet.

I worry that "correct by construction" encodings actually make it harder to encode existing legislation with errors, and then use testing and analysis to demonstrate the existence of the error. If you are forced to rely only on compiler errors, you lose access to information. Strict languages may also force you to make certain assertions that are not included in the law, and may not be legally true. I haven't yet done any encodings in strict languages, but Catala will likely be the first one I try.

Isomorphism between the law and the encoding is critically important to ease of use and maintainability for legal knowledge engineers. Because of how law uses defeasibility, you can't get isomorphism without default logic (or another non-monotonic alternative). But default logic is not also not enough on its own. You need syntax that eliminates the need to name the justifications inside the code for the default. Because in real laws, the information about what justifications apply to a default is sometimes included in the segment of the legislation that expresses the default ("default a is true, subject to section 2"), and sometimes it is included in the segment of the legislation that expresses the exception ("notwithstanding section 1, exception b is true").

There lots of other nice-to-have features for getting defeasibility and isomorphism at the same time, but the upshot is that for that purpose default logic is necessary and insufficient.

As far as demerits are concerned, checking for the presence of exceptions, particularly when they are implied to potentially exist, is going to be slower; it has multiple semantics, so you have to pick one or let the user pick one; and the more isomorphic the code, the less clear the semantics become, which makes some things implict in the encoding that are neither implicit nor explicit in the legislation.

Jason

Roland Turner

unread,
Jun 10, 2021, 8:55:21 PM6/10/21
to ta...@lists.legalese.com
On 30/5/21 4:16 pm, Jason Morris wrote:

> I worry that "correct by construction" encodings actually make it
> harder to encode existing legislation with errors, and then use
> testing and analysis to demonstrate the existence of the error. If you
> are forced to rely only on compiler errors, you lose access to
> information. Strict languages may also force you to make certain
> assertions that are not included in the law, and may not be legally
> true. I haven't yet done any encodings in strict languages, but Catala
> will likely be the first one I try.

Surely exact encoding is a fool's errand? Even natural language doesn't
perfectly encode legislation (e.g. it's always open to the judiciary to
look at the path of a particular piece of legislation in order to
understand its context); post facto attempts at perfect translation
would seem pointless, or harmful.

Presumably the value of logical formalism in legislation is its use
during the original encoding in order to facilitate discovery and
correction of errors and omissions prior to adoption?

- Roland


Christy Sweet

unread,
Jun 13, 2021, 12:16:42 AM6/13/21
to ta...@lists.legalese.com
Hi  I have no idea how I was included in this chat but please take me out- it certainly sounds interesting IF I could understand more than about 10%  of it 
  Cheers 

--
You received this message because you are subscribed to the Google Groups "Legalese talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to talk+uns...@lists.legalese.com.
To view this discussion on the web visit https://groups.google.com/a/lists.legalese.com/d/msgid/talk/6e7fdb19-e6a7-a9c5-0bf5-ad4fd7066d64%40rolandturner.com.


--

Christy Sweet
 (66)   (Inside Thailand, dial 0...)  94 807 0376 
 
Please note, I'm not always online, so the best way to contact me quickly is through mobile calls or text messages to the phone above.  

Dane christian Neilson

unread,
Mar 18, 2024, 3:52:39 AM3/18/24
to Legalese talk, Christy Sweet
MT103/202 DIRECT WIRE TRANSFER
PAYPAL TRANSFER
CASHAPP TRANSFER
ZELLE TRANSFER
LOAN DEAL

TRANSFER WISE
WESTERN UNION TRANSFER
BITCOIN FLASHING
BANK ACCOUNT LOADING/FLASHING
IBAN TO IBAN TRANSFER
MONEYGRAM TRANSFER
IPIP/DTC
SLBC PROVIDER
CREDIT CARD TOP UP
DUMPS/ PINS
SEPA TRANSFER
WIRE TRANSFER
BITCOIN TOP UP
GLOBALPAY INC US
SKRILL USA
UNIONPAY RECEIVER

Thanks.



NOTE; ONLY SERIOUS / RELIABLE RECEIVERS CAN CONTACT.

DM ME ON WHATSAPP
+13855324926
Reply all
Reply to author
Forward
0 new messages