On the other hand when they want hard on the game-language in 4e, people hated that too.
PF2e, I think, benefited from the experience of others failures in striking a middle ground.
D&D 5E Dying From Exhaustion While Petrified
PF2e, I think, benefited from the experience of others failures in striking a middle ground.
EzekielRaiden
Follower of the Way
I mean, I don't think anyone has ever meaningfully defended the presentation failures of 4e.
But a significant number of the places where 5e falls down...are the ones where the designers went 'eeeeew, 4e cooties' and assiduously avoided doing anything that 4e did. They promised the world with "natural language," and then it ended up being really not that great. To the point that 5e itself actually does (falteringly) strike a middle-ground, because going "hard on the [natural]-language," to steal your phrase, would have been an actual disaster.
It turns out, when you're playing a game that has rules, being clear about what you mean is really important. Who knew?! How could designers and playtesters have ever discovered this deep wisdom?!
Rabulias
the Incomparably Shrewd and Clever
I think one can make clear and comprehensive rules with natural language. But if one wants rules to be complete and cover everything, it can get a bit dense and wordy.
Unless the game is very simple, no ruleset will cover every situation. Even less open-ended rules like boardgame rules written with "game language" will miss some things. This is more errata than a failure of natural language.
The creators of 5e chose to leave some rules open to interpretation, which I think has contributed some to its popularity. Think about the various debates over rule interpretations we have just here on ENworld. Despite sometimes very strong differences of opinion, all participants are playing 5e. If the rules had a definitive answer for all those, some people would feel the rule goes against balance, fairness, worldbuilding, and/or the fiction of their game.
I think in this particular case (exhaustion and petrification), the 5e authors probably did not think they needed to clarify further, and common sense and familiarity with petrification from myths and stories would serve. I think natural language could cover this issue with another sentence or two, building on some of the Pathfinder text:
While petrified, your mind and body are in stasis, so you do not age or notice the passage of time, and you do not need to eat, drink, breathe, or sleep. Your petrified body is an inanimate object.
billd91
Not your screen monkey (he/him)
Yeah, I know common sense is sometimes not so common. But sometimes you have to use it or at least appeal to it.
Paul Farquhar
Legend
The point of using standard English is to discourage people from legalistic interpretation of the rules. The rules don’t need to spell everything out because game world logic is the more important determinator.
EzekielRaiden
Follower of the Way
I am skeptical of the first claim, and outright reject the latter claim. The really severe problem with "natural language" is precisely the same as the really beautiful strength thereof: different people understand common words differently. A poet can do so much, express so much, with such a severe economy of words, specifically because natural language is vague and subtle, because natural language is a many-to-many mapping. But many-to-many is exactly what a rule should not be. It can be many-to-one (e.g. an extensible framework, which I'll get to in a sec, or mere basic abstraction, e.g. HP abstracts such that losing 20 hit points has no specific meaning or anchor, while AC abstracts such that "14 AC" could come from many different origins, high dex, natural armor, plate, etc.), but it should never be one-to-many (a single situation which correctly points to multiple results).
I fully expect to be told "but that's what GM judgment is for!" and to that I reply: Yes, because you are using the GM judgment to convert a one-to-many case into a one-to-one case. That's precisely why folks talk about how GMs need to be "consistent" and (usually) "fair" and "reasonable" etc. As soon as that consistency is broken--as soon as the GM rules on functionally identical situations in contrary ways--the rules have broken down and a problem occurs. The one-to-one or many-to-one connection still remains: play must produce one, clear, cognizable result or else problems arise.
Now, as for the second claim, the aforementioned "extensible frameworks" are the disproof thereof. See, there's a missing premise in the argument, which (generally) goes as follows:
1. There are infinitely many potential cases that rules could apply to.
2. Making infinitely many rules to cover those cases is impossible.
3. Therefore, having complete coverage is impossible.
But there's a hidden premise here required to reach the conclusion: each rule must be singular and only cover finitely many cases. Once we include that premise, the argument is of course airtight! In any system which only permits singular, finite rules, it is definitionally impossible to achieve complete coverage. You will hear absolutely no argument from me on that count.
You will, however, hear me dispute vigorously the idea that singular, finite rules are the only possible rules that can exist. If you can construct an extensible framework--one that leverages abstraction in order to cover not just individual cases, but whole classes of cases, then it is no longer guaranteed that complete coverage is impossible. The trivial, bad-design version of this would be to replace all rules evaluations with "flip a coin, if heads, the player decides what happens, if tails, the GM decides what happens." This, of course, abstracts out absolutely all of the details completely, and (as stated) would be very bad game design in the vast majority of contexts--certainly, it makes for rather dull gameplay unless you just really really love coin flips. But the point stands: comprehensive coverage is possible. It leaves open the question of whether comprehensive coverage exists that is effective, that achieves the experience for which the game was designed, actually exists. But that question is open, not closed; we do not know for sure either way, unless and until we actually start designing.
Again: No. Extensible frameworks and other forms of abstraction can get you there. As stated, that doesn't guarantee that these results are good--but there's also nothing saying they'll all be bad either. Which means the ball is now in the critic's court, to show that abstraction-in-general is unacceptable, in order to pitch the ball back the other way so the proponent of abstraction has to defend why this abstraction is good. However, that's something the critic will usually struggle to do, because D&D (every version, doesn't matter which one you've played, every single one) contains abstractions. Since very few are willing to throw out absolutely all abstractions (mostly because there would be hardly any game left if they did!), it thus becomes a matter of the critic needing to show how the proposed abstraction must be bad in general...which is an extremely tall order, one I've never seen happen.
Even designing under a maxim like "minimize abstraction whenever possible" doesn't achieve that goal--because there are too many grandfathered-in exceptions for classic D&D abstractions we've all grown so used to, we don't even think of them as abstractions anymore.
In which case, they can (and should) override the rules. As they have always been able to do. Nothing has ever taken away that power. Ever.
But that's precisely my point. "Natural language" was explicitly sold to us as avoiding having to talk about edge cases. As preventing rules debates in their entirety, because folks would already know what all the words meant and wouldn't ever be confused about what was intended. That's why I said they promised the world; they essentially claimed that "natural language" would let them speak both briefly and precisely. That was quite openly their intent.
We have seen that this is manifestly not true. If you need extra sentences to explain what "petrified" actually means, then you already aren't using natural language. You're defining a jargon term, and specifying how it works in edge cases.
EzekielRaiden
Follower of the Way
Does this actually happen?
Because I don't actually think this happens. I think rather the reverse. Vague terminology encourages rules-lawyers to hound for exceptions and to force favorable interpretations of unclear statements. Having clear, precise rules that mean exactly what they say and say exactly what they mean leaves the legalist with no defense; you must simply accept the rules as they are, or dispute the system itself, which...is not likely to be effective with very many DMs, who IME take a dim view of "but...but...that's stupid and I should get what I want instead!"
Frankly? This argument sounds like presuming the strong Sapir-Whorf hypothesis is true, when...well, it just isn't. You can't control someone's thoughts, preventing them from thinking legalistically, by controlling the words they have available. And whenever adjudication and benefits are in play, folks are going to push for the best result that achieves their aims.
It seems so funny to me, honestly, that folks are so ready to claim that rules cannot make bad GMs good, but (somehow) rules can make bad players good.
CleverNickName
Limit Break Dancing
Wow, I...hadn't ever thought of that. I always figured that if you were turned to stone, you were slain and could be Resurrected or Reincarnated as if your body had been destroyed. And if you had access to a Stone to Flesh spell, you could be restored instantly without all that rigmarole.
Bacon Bits
Hero
I don't really think we can blame natural language for the failure or incompleteness of a condition/keyword. Conditions and keywords are precisely what synthetic language is. The problem is already in the antithesis of natural language. You can't just point to any incomplete description and blame "natural language." That isn't what that means.
If anything, the problem is that the condition and keyword aren't property defined, not that they used natural language.
This is exactly like the problem in 3e where "dead" doesn't say you're incapacitated, and "dying" ends at -9. So, RAW the only drawbacks to being dead are that you can't benefit from healing and that your soul has left your body. But nothing stops you from taking actions, because they didn't tie the synthetic conditions together correctly. The problem isn't natural language. It's the synthetic language.
It's also the same problem why, RAW in 5e, you can't wake up from a long rest. The game says you go to sleep, and that means you're unconscious. But the unconscious condition says you can't take actions while unconscious. So you can't wake up, because waking up is doing something! The problem is in the keyword's synthetic definition is incomplete not the natural language. Just because it's written in English doesn't mean it's natural language.
Like, I'm sorry, but that's what they didn't like about synthetic language. You had to define everything down to the last little detail, and it was extremely easy to miss some normally self-evident assumption when you do that. That's why too many keywords doesn't work. It's difficult to untangle and difficult to imagine all uses for every keyword. Magic The Gathering says, "forget verisimilitude, I want determinism." TTRPGs don't do that. They say, "verisimilitude is king and the rules can't cover everything you could imagine, so we put a referee at the table at all times precisely to deal with what happens when mechanics and verisimilitude conflict."
Natural language is, "Light works like light works. You can't try to hide when someone can see you. You can see what's in your line of sight. You can use an Arcana check to see if your character knows something about magical lore. You can use a Survival check to find food and fresh water in the wilderness." That is natural language. The only synthetic thing there is the word "check" and the names of the skills.
Similar Threads
