On synthetic casuistry

The main idea: limit judicial fiat and legal bloat by enforcing sentencing through linking to a few hundred synthetic cases only.  

0. Posts on this blog are ranked in decreasing order of likeability to myself. This entry was originally posted on 19.06.2025, and the current version may have been updated several times from its original form. 

1.1 Consider the large de facto legislative freedom that being unduly harsh or lenient in sentencing gives judges nowadays.

1.2 Further consider the severe bloat that precedent tracking enforces, the freedom it allows to cherry pick, and the degree to which it makes law unknowable.

1.3 So let’s step back and design a simpler system from scratch, in round two of my first attempt. 

1.4 You create a set number of synthetic cases, likely in the low hundreds. These must be such as to account for reasonable variations of relevant real-life factors, such as type of crime/tort, degree of uncertainty, strength of evidence, extenuating circumstances, etc. Note how these cover both civil and criminal cases.

1.5 Each of these synthetic cases is linked to a specific sentence, going from nothing, to slap on the wrist, to death. I of course would like to stick to the “can’t have any party profit from the proceeds” rule and only have punishments and no damages ever being awarded, but this is not strictly required for the system to work.

1.6 The set of synthetic cases and sentences would be adopted by the legislature, and could of course be updated at any time with new cases, (as long as others are dropped for a strict limit on case number), changes in prescribed sentences to existing cases, etc. 

1.7 These are now the only precedent that are allowed to be brought in court, no real-world cases can be referenced at all.

1.8 After hearing the case, a judge or jury is required to decide which of these synthetic precedents most closely fits the matter at hand, not in terms of facts, but in terms of how guilty the defendant is in their entirely subjective view. 

1.9 Defence and prosecution now be free to play in a new dimension, not just arguing the facts but also the subjective likeness to this set.

1.10 It’s important to reiterate that we don’t require the case itself to be similar to the synthetic case. It’s just the subjective degree to which the defendant is guilty must match. For example, a jury could link the case of a guy alleged to have purposefully burned his friend’s car down with the synthetic case of John having put Jack’s life at risk through negligence. 

1.11 Of course, a judge or jury can still link to a grossly mismatched synthetic case on purpose in order to be lenient or harsh, but this would be far easier to observe and argue for in higher courts.  If the case at hand was of Andrew raping Julie, and the synthetic case of John raping Anne was right there but was skipped in favour of John hosting a loud party, you know what went down. 

1.12 Note how, though numerous, a few hundred synthetic cases are feasible to know of, and very easy to research, in stark contrast to the practically unending repertoire of real-world casuistry we rely on. Even a jury could easily be made to narrow down the few cases of interest through A-B search.

1.13 Legislative changes would be just as easy to judge as judicial abuse, as laymen would have a far better grasp of the severity of a few hundred synthetic cases compared to the knowledge of legal minutia in this or that Code. 

1.14 Further note how shonky cases would not automatically become part of the precedent for other shonkier cases to be decided on, but remain isolated incidents with no impact on the casuistry.

1.15 Finaly, note how feasible it is for a few hundred curated synthetic cases to be devoid of any internal contradictions (at least in the eyes of the Legislature), with no instances of more severe crimes receiving more lenient sentences. An internally consistent body of law, how about that?

1.16 Is it feasible to distil all relevant legal knowledge down to a few hundred synthetic cases? Using modern understanding of factor analysis and other statistical techniques, it wouldn’t surprise me at all if it would. A few hundred is a lot if you are free to mix and match relevant factors at will.

1.17 Moreover, is moral intuition so reliably standardised that various parties would select reasonably close cases when presented the same set of facts? Probably not, but this issue affects all legal systems, and is just being laid bare here.  There is an argument to be made that nations should be formed of those who judge these issues in the same way.