We have size categories, why not intelligence categories?

[Greywander]

I was thinking about this. A lot of systems are designed under the assumption that you'll be facing foes of a similar size as you, then add Size as a game statistic with accompanying rules to govern how things change when you and your opponent aren't the same size. Not every system does this; some systems just make appropriate modifications to a monster's stats to account for it being larger or smaller. But a fair number do.

But I don't think I've seen a system that does the same thing with intelligence. And this can lead to some strange cases where a human PC has less intelligence than some animals. This also tends to restrict the intelligence values that animals can have; one point more and they're too close to human-level intelligence, one point less and they're virtually nonsentient. This makes it difficult e.g. to have one wolf that's smarter than another wolf. If we introduced a system for mental capacity that worked as an analogue to physical size, then these problems could be alleviated.

There's a lot of variance on how many categories we could have, or what the specific criteria would be for different levels, but I think we can broadly divide intelligence into four groups.

Nonsentient

This generally includes anything without a brain. I can see having two possible subcategories here. One could be Inanimate, and refers to anything that's not even alive, like a rock. Then again, such a category might be redundant, and it's not clear what mental stats would even represent in such a case.

The other category could be called Unthinking. As mentioned above, it includes anything without a brain: plants, fungus, bacteria, jellyfish, and so on. These are things that entirely lack the ability to make decisions, and instead merely respond to stimuli. Higher mental stats would symbolize more complex responses to a wider and more detailed list of stimuli. In some cases, entities in this category can even "learn" and change their response to a stimulus. However, this isn't true learning, merely a complex response to a complex set of stimuli; it is predictable and can be easily reproduced with the right stimuli.

Sentient

This covers anything that has a brain and is below human-level intelligence. It could be quite a broad category, and it can be difficult to divide it up into subcategories since there seems to be a lot of variance between different types of animals. There isn't necessarily a single metric for intelligence, as some animals are capable of certain intelligent behavior but not others, while a different animal is capable of the other intelligent behavior, but not the first. Instead of splitting intelligence up into different levels, it might make sense to create a checklist of intelligent behavior, e.g. using tools, being self-aware, using language, etc.

If we did break it down by level of intellect, however, the lowest level might be called Instinctual. These are animals that are capable of thinking and making decisions, but those decisions are almost entirely informed by instinct. They're also capable of learning new information, such as where to find food or what things or places might be dangerous. Not sure exactly what kinds of animals this might apply to, but one thing you could use it for is mindless undead (although those might be classed as nonsentient). Higher mental stats signify more complex and complete instincts that allow the creature to respond appropriately to a wider variety of situations. Higher mental stats can also indicate the creature has a greater capacity to learn and understand, though they'll always be primarily instinctual.

As you move up in intelligence, you'd start moving away from instinct and more towards learned behavior and logical thinking. You'd also likely start seeing social animals that opperate in packs and can coordinate with one another. These creatures are often ideal for domesticating since they're smart enough to learn but not so smart that are difficult to control. There might be more than one middling category, as this is a pretty broad set. Higher mental stats would generally correspond to learning more quickly or exhibiting more complex behavior (e.g. social behavior).

At the top would be Subsapient. These are creatures that fall just short of human-level intelligence, such as dolphins or apes. They can generally recognize their reflection in a mirror (i.e. they are self-aware), use tools and solve puzzles, do very basic math, and learn and use language. They fail at more abstact thinking that humans are capable of. Higher mental stats would have similar implications as it would for sapient creatures.

Sapient

Sapience is a tricky thing to define, as IRL only humans have it. Basically, this just refers to anything with human-level intelligence. In a fantasy world, this would doubtlessly apply to creatures like elves and dwarves, and most other humanoid races. Likewise, in a sci-fi setting, aliens would span all levels of intelligence, including this one. After all, it's easier for us to understand if an alien world is like our own, with it's own aliens animals and alien people.

Transcient (I basically made this word up; is there a more appropriate word?)

For beings of higher intelligence than humans. This could be gods, angels or devils, aliens, ancient precursors, family pets, etc. By its very nature, it's difficult for us to properly comprehend what this might look like. That said, there are a couple things we could do to roleplay an entity such as this. One would be to give them a greater capacity for abstract thinking. This could be represented by a higher ability with mathematics, possibly including thinking in higher dimensions (e.g. 4D or 5D).

Another would be heightened awareness. Two ways this could manifest would be in an apparent ability to see the future or read minds. In reality, they're doing neither of those things, it's just that they have a greater ability to take their knowledge of the current situation and predict potential outcomes, including the anticipation of unexpected complications. Likewise, they'd be more aware of things like body language and have a greater capacity for empathy, helping them to see things from the perspective of another person. If someone is rude to them, they might pick up on a number of cues that indicate that person might be severely stressed and anxious, rather than them just being a mean person.

These kinds of entities might also have an easier time overcoming their own feelings and doing the things they know to be best. If they know they should work out, then they will, and the temporary pain is less bothersome to them since they're more able to see ahead to the rewards. They're more able to resist eating food that isn't good for them. And so on and so forth. It's not that they don't have similar feelings to humans, they're simply more aware of the benefits of doing the things that they should that it makes it easier for them to overcome those feelings. This can also extend to things like feelings of revenge or jealousy or envy. It doesn't mean that they won't seek revenge, only that they won't allow themselves to be consumed by that desire and let it destroy them along with their enemy.

Anyway, thoughts on this? The effect would depend on the system you were putting it into, and it might work best with an original system. There would need to be some kind of accompanying rules for how creatures with different levels of intelligence interact with one another differently. It could also influence things like the ability to learn magic (if your dog can learn wizardry, and can also have a high Intelligence stat, then they might be a better wizard than you, even though you're technically smarter; on the other hand, wizardry would likely not even be available to anything below sapient or perhaps subsapient).

[rel]

This style of design has also been applied to strength, with power based systems often having a super strength ability and scale which is similar to size but a more pronounced change. Generally entities of different scale can't meaningfully interact so you can't blow up a battleship by stomping around on the deck for a few hours.

Your idea seems novel, I haven't seen categorisation applied to intelligence or body plan (the fundamental differences between say a human and a cow or the similarities between a human and an animated human skeleton) in a game yet.

[RedSand]

No offense but I don't really see why you would need to systemize this like that.

Size categories exist because you frequently need to quantify a thing's size relative to another thing so you know how they interact, in a different way then stats do. A gelatinous cube is bigger then a flesh golem but the flesh golem is built to be insanely strong and tough-you couldn't just say "oh this gelatinous cube is bigger so it must be given a higher strength score then the flesh golem". That doesn't realistically quantify how the cube interacts with the world. Intelligence, on the other hand, is an already extremely murky thing to quantify in real life, so it seems alright to me to just abstract it via a number.

[Breccia]

We had that in First Edition, but it was just a name given with their Int score.

[Bohandas]

Transcient (I basically made this word up; is there a more appropriate word?)

I think you mean "transcendent" or "transcendental". A transient would be a drifter or hobo.

[Composer99]

I'm inclined to think that size categories exist as an abstraction that provides some sort of game-mechanical utility.

So I guess the question is, what game-mechanical utility are you getting out of intelligence categories? For the purpose of game-mechanical utility, is the potential difference in intelligence between any two wolves significant enough that one of them might belong in a different category than the other?

In my "OGL fantasy heartbreaker", I don't really bother with categories, as such, but I do have two, you might say, "keywords" that are applicable: mindless and thinking.
- A mindless creature is, for instance, a gelatinous cube or a zombie.
- A thinking creature is any creature with an Intelligence of 3 or higher that understands at least one language.

(This means there are creatures that are neither mindless nor thinking.)

My goal isn't necessarily to have a sensible set of categories, but mechanics that are easy to hook other mechanics to - for instance, mindless creatures might be immune to charm or fear effects, or some spells might require that you target thinking creatures.

---

I think you mean "transcendent" or "transcendental". A transient would be a drifter or hobo.

If I'm not mistaken, Greywander's thinking of a step down from "omniscient" there, hence the neologism.

[LibraryOgre]

So, as mentioned, 1e and 2e divided Intelligence into categories, but they just corresponded to a range of numbers... a Low intelligence meant something between 5 and 7, inclusive. Very intelligent was 11 or 12. That's more or less what the size categories are, though they're a bit more nebulous... in some editions, dwarves are medium, in others they're small, and sometimes, it depends on their rolled height.

Rifts, on the other hand, has a few different kinds of strength, that run on roughly the same scale, despite being absolutely divorced from each other in numbers.
There's normal strength.
There's exceptional strength, which is normal strength, just with numbers at the theoretically higher end (16+, I think; but you can roll up to 30, and there's a ton of way to get bonuses).
There's Augmented (formerly bionic) Strength.
There's Robotic Strength.
There's Supernatural Strength.

Now, in theory, these run off the same number scale, but a 20 (normal/exceptional) strength and a 20 supernatural strength are WILDLY different.
Normal 20: Carry 400 pounds, punch does d4+5 damage
Augmented 20: Carry 400 pounds, punch does d4+5. Frequently have special modifiers to the carry weight, and if their strength is 24, they can do a power punch that will do 100 damage (or 1 mega-damage)
Robotic 20: Carry 500 pounds (25*PS), can do 2d6+5 with a normal punch, or 100 damage (1 mega damage) with a normal punch.
Supernatural 20: Carry 1000 pounds, punch does 3d6+5 on a restrained punch, 100-600 (1d6 mega-damage) for a normal punch, or 2d6 mega-damage for a power punch.

You will note: In these few ranges, absolutely none of it is related except for the number. Knowing someone has a 20 PS is meaningless unless you know what kind of PS they have, and someone with a 20 PS might be WILDLY stronger than someone else with a 20 PS.

[Greywander]

I think you mean "transcendent" or "transcendental". A transient would be a drifter or hobo.

If I'm not mistaken, Greywander's thinking of a step down from "omniscient" there, hence the neologism.

Correct, it was supposed to sit between sapient and omniscient (which probably should have been listed as the highest category of intelligence). "Scient" comes from the same root as "scientist", and "trans" means "across or beyond". So it was meant to mean something like "beyond human understanding" while being less than "omniscient" (which is "all-knowing").

I'm inclined to think that size categories exist as an abstraction that provides some sort of game-mechanical utility.

So I guess the question is, what game-mechanical utility are you getting out of intelligence categories? For the purpose of game-mechanical utility, is the potential difference in intelligence between any two wolves significant enough that one of them might belong in a different category than the other?

Just as an example, let's say we want to implement a mechanic that works in a similar fashion to grappling, but involving mental stats. As things are, all wolves would be equally capable, but under this system you could have some wolves with higher or lower mental stats than others, making them more or less competent with this new mechanic. But, as with grappling and size, this mechanic might not be as effective against creatures in higher intelligence categories. Or maybe it would be.

Right now, using D&D 5e as an example, you can't really introduce any kind of mechanics that allow for some wolves to be smarter than others, spanning the typical range of ability scores. I don't have a specific mechanic in mind that I'd want to introduce, but intelligence categories would be a prerequisite before I could even consider such a thing. My point is mostly that introducing something like this would open up the possibility of adding such mechanics. That said, there are some things it would affect immediately, such as making it so that not every animal absolutely sucks at INT saves and checks. There may be some situations where you want creatures of lesser intellect to perform worse, even if they have the same INT score as a higher intelligence creature, but you can specifically write that into the ability. This could also pave the way for things like mind control or manipulation abilities that only work on creature of lesser intelligence, e.g. things like Dominate Beast could be based off of intelligence level rather than creature type, or the Necromancer's Command Undead might only work on up to sentient undead, but not sapient undead.

So, as mentioned, 1e and 2e divided Intelligence into categories, but they just corresponded to a range of numbers... a Low intelligence meant something between 5 and 7, inclusive. Very intelligent was 11 or 12. That's more or less what the size categories are, though they're a bit more nebulous... in some editions, dwarves are medium, in others they're small, and sometimes, it depends on their rolled height.

But those are just fancy names given to specific ranges. You'll never get a wolf with 16 INT, for example, and if you did, they would be smarter than a 15 INT elf. With a system like this, a 15 INT wolf would be quite intelligent for a wolf, but still less intelligent than an 8 INT human. Well, maybe. It's possible, for example, for a smaller creature to have a higher carrying capacity than a larger creature if their STR score is high enough, and the same could be true here as well. The 8 INT human would still be of a higher intellect, but the 16 INT wolf might be, say, more "efficient". They have less understanding than their human companion, but they're more adaptive; they don't know what a book is, but they might be able to solve a simple puzzle more quickly.

One place this could manifest is with learning time. In D&D 5e (again, just using this as an example), the amount of downtime required to learn a new tool is decreased by having higher INT. So, for example, a high INT animal who is capable of tool use might learn just as quickly as a human would. Being of a lesser intellect might not affect the learn times, only whether or not that creature can learn to use tools at all.

[sandmote]

You could maybe reasonably tie particular skills to intelligence categories: for instance, a creature must be at least subsapient (although I don't like the implications of this term) to make checks involving technical knowledge like that gained by formally studying a subject.

I do prefer Inanimate and Unthinking to Nonsentient, partially because I think it could clarify why some effect doesn't work: exceptions that allow you to "trick" a jellyfish wouldn't work on a rock. Instinctual I also like as a category, but I'm worried about the implication of a specific term for creatures directly below humans. For categories above humans, I also can't think of an existing term. Although I do feel there should be at least three categories above sapient. I'm having trouble putting them into words, but I would expect illithids to be above sapient and below the category the elder brain falls into, but the elder brain to still fall below the "near-omniscient" category I'd expect of deific figures. I'm using illithids as an example case, but there's probably some similar sorting that could be done between outsiders.

Ending suggested categories:

• Inanimate
• Unthinking (able to process states)
• Instinctual (able to react differently to the same stimulus based on other conditions such as past outcomes)
• Sentient (capable of learning, at least by doing it first)
• Sapient (able to abstractly test and experiment)
• [Term pending] (sapient creature can follow along)
• Transcient (beyond what sapients can follow but must be manually exposed to information)
• Near Omniscient (deific levels of knowledge, where the being can simply be aware of information there's no real explanation of them knowing)

[Greywander]

I think I like the term "Mindless" better than "Unthinking". It just sounds a bit more natural.

As for creatures like apes, I'm fine with the term Subsapient as it's an accurate description. But depending on how many levels of intellect there are, it might make more sense to roll these very smart animals into Sapient and simply give them a lower INT score (this doesn't have to be D&D, but it's an easy example). If, for example, there's only one level between Mindless and Sapient, it really wouldn't make sense to lump apes, dolphins, and parrots in with, say, beetles and koalas. The main reason not to do this is if there's some kind of mechanical effect that results from being Sapient; it's basically classifying these animals as "people" rather than "animals", which raises a whole lot of moral/ethical/philosophical questions and creates the possibility for e.g. clerics and wizards. Though the idea of an orangutan wizard is pretty entertaining.

TL;DR, fantasy creatures aside, it's probably best that real life animals remain unambiguously animals, and not people. Otherwise, it gets wierd.

How about the following as levels of intellect?

• Mindless
• Instinctual
• Crafty (trainable animals)
• Cognizant (can use language and tools)
• Sapient
• Sublime
• Transcendent
• Omniscient

Everything between Mindless and Sapient (noninclusive) would fall under a broader label of Sentient, while everything above Sapient would fall under a broader label of Transcient. Some Cognizant creatures will use language or tools naturally, while others are capable of using them but must be trained first (e.g. a non-social creature who has no need for language). For levels above Sapient, it's difficult to define what they are or what special rules they might use. I'd imagine a lot would depend on the system. Based on what I stated in the OP, we could imagine that Transcient creatures might get advantage on Insight checks against a creature of lesser intellect (able to read them better and see things from their point of view more readily).

Hmm, we might need to define some kind of secondary statistic, similar to carrying capacity for Strength, that would scale with your INT score but also be affected by your level of intellect. The closest analogue I can think of is Memory, or your "carrying capacity" for learning things. Sounds like kind of a pain to integrate into an existing system (e.g. D&D), but if you built your system around it from the ground up it might work well. That said, I feel like intelligence is less about "how much" you can learn and more about "what" you can learn, e.g. intelligence requirements for certain types of abilities, and as long as you meet those requirements then you can qualify. As with carrying capacity, the requirement could be some combination of your actual INT score as well as your level of intelligence, so that a lesser intelligence with a high INT score might still be able to qualify.

You know, this could work really nicely for a system built around psychic powers. Maybe you could reframe some of the everyday things humans do as an expression of psychic ability, which is why animals can't do those things. Basically taking something mundane and reframing it as magical, then extrapolating that to higher intellect creatures who would have the same abilities, but stronger. For example, language could be a psychic ability, which is why higher intelligence is associated with telepathy. It's just taking language to the next level. Meanwhile, animals have a much more primitive psychic ability, so their forms of language (barking, calls, etc.) are more primitive as well.

Crafting could be another example of a "psychic ability". After all, crafting is essentially a creative act, and creation is the realm of the gods. Not even all gods, but generally an overdeity, that is, if any god at all in your setting is capable of such an act. An Omniscient entity using a crafting ability is basically how universes get created. Then again, this is sounding less like psionics and more like divine power. Humans are more divine than animals, and gods are more divine than humans. Hmm, maybe some potential for an in-setting conflict where you have the religious on one side who believe in divine power and atheists on the other side who believe that they've "unmasked" the false gods with their understanding of psionics.

[Herbert_W]

The closest analogue I can think of is Memory, or your "carrying capacity" for learning things.

I've held the opinion for a while that the so-called intelligence score in DnD ought to instead be called memory becasue "memory" is a more accurate representation of what the score does mechanically. A character with a higher int score cannot think, reason, or make decisions more adeptly because it is the player who does all of those things. What a character with a higher int can do is make more successful skill checks to remember information, be more effective with forms of magic that rely on memory, and (in 3e) learn more skills. That's all memory!

(There's also the fact that "orcs are stupid" is reminiscent of real-world nastiness in ways that have been discussed to death; we hopefully won't derail the thread by rehashing that here. I'll just say that "orcs have poor memories" sounds a lot nicer and would hopefully make for a more inviting and therefore potentially larger game.)

If you replace "intelligence" scores with "memory," which could very well remain mechanically exactly the same thing, that creates room for intelligence categories to be their own thing which has a clear purpose that's distinctly separate from what are now memory scores.

So, what purpose might I have in mind, I ask rhetorically? Why, it's to do that other thing which the current so-called intelligence score purports to (and does not) do: to represent a creature's ability to reason. This means, practically, how real-world humans could model that creature as players or as a DM. We would need at least four categories:

1. Nonthinking: Creatures in this category never make decisions. They either react on instinct, follow a preset routine, or do nothing at all. These creatures cannot be PCs - a player makes decisions for their character, and there are no decisions to be made here.
2. Limited thinking: Creatures in this category have some of the cognitive toolkit of humans, but not all of it. Players playing these creatures must remain mindful of these limitations, and may need to justify to the DM how their character could reach any given decision without the player "cheating" by using facilities which their character does not have.
3. Human-like thinking: Creatures in this category are capable of learning, abstracting, reasoning, etc. as humans are. A real-world human such as a player or DM can make decisions for them. In a normal game, all of the PCs are in this category.
4. Inhuman thinking: These creatures reason in ways that are beyond human comprehension. They may have a "higher" form of reasoning, or just a different one. Modelling the decision-making of these creatures in a game played by humans necessitates some shortcuts. Humans playing these creatures need help - they may get to peek at the DM's notes, to fudge rolls, to enjoy the effects of divination spells at-will without needing to cast the spell, or to discuss things with other players out-of-game even if their characters would normally not be able to communicate.

Categories two and four might be subdivided to represent which cognitive facilities are missing or added.

This has given me an idea for something that's a bit of a weird tangent from the current conversation:

Spoiler: Weird tangent

Show

We could also expand this system into one where, instead of an intelligence category, each creature has a cognitive faculty list. A creature could have some inhuman extra faculties while also missing some humans ones.

How would you like to play as something that doesn't understand the concept of groups, sets, or numbers - who literally cannot comprehend the forest for the trees - but which can see the future? You might foresee that a house will be destroyed by a man wearing furs, that another house will be destroyed by another man wearing furs, etc. - and when you tell your human wizard friend and his fighter friend about it, they get very upset and start working to protect a "village" (whatever that is; you've never seen one) from the "viking raid" (whatever that is; you've never seen one).

You know, this could work really nicely for a system built around psychic powers.

I can see a thematic connection here, but I don't see a game-design benefit to having intelligence categories be linked to a creatures psychic capability. Having these things run along the same scale is limiting. Having a creature with a low capability to understand the world but a high capacity to change it is a common horror trope. ("Oh no, he doesn't realize that the monsters he's killing are actually people!") A creature with a low ability to directly affect the world but a high intelligence is a common protagonist trope and a real worry for AI safety researchers. Tying intelligence categories to psychic power levels makes these archetypes impossible to represent in your system.

[Rynjin]

But I don't think I've seen a system that does the same thing with intelligence. And this can lead to some strange cases where a human PC has less intelligence than some animals.

When? Unless you're feebleminded (or the animal is supernaturally intelligent) this will never happen.

[Bohandas]

When? Unless you're feebleminded (or the animal is supernaturally intelligent) this will never happen.

Int damaging debuffs, diseases, and poisons can do it too

[LibraryOgre]

But those are just fancy names given to specific ranges. You'll never get a wolf with 16 INT, for example, and if you did, they would be smarter than a 15 INT elf. With a system like this, a 15 INT wolf would be quite intelligent for a wolf, but still less intelligent than an 8 INT human. Well, maybe.

Which is why I started talking about the charlie foxtrot that is Rifts strength categories.

The main advantage I see in this would come in 3.x and 5e, where ability damage is more prevalent. That 16 intelligence wolf would resist Int damage better than his 8 Int counterpart, of any species. But what that 16 would mean outside of that is the big question.

[Rilmani]

When I try to imagine homebrew (which arent’t spells) I like to remove the spells Bless and Bane from existence. And then I like to imagine one or two positive conditions (like Invisible) and negative effects which remove positive conditions or remove resources from them.

I think that the Sapience intelligence category could be interesting for player characters because sapient creatures can do amazing things when they cooperate. In that vein we should look at the Help action (in and out of combat), the Improvise action, the spell Aid, Luck points, and the spell Resistance. We all know that not all actions are created equal- Attack and Cast a Spell are quite dominant. So I reckon we should have a resource which lets a character take the Help action as a Reaction, then we augment the Help action from there. Pack creatures with Sentience tier intelligence can have this resource, but they have restrictions when using it, while a Sapient creature can use it to Help an enemy against a greater threat.

Let’s call this resource Synergy Points. When you take the Help action, you can transfer up to half of your maximum Synergy points to your ally. Your maximum Synergy is equal to a number related to your mental size category plus your lowest mental ability score. If you have the Tuned condition, at the start of your turn (if you are not Incapacitated) you recover 1d4 Synergy points. You can use your Reaction to take the Help action; you cannot do this again until you complete a Long Rest or until you spend 2 Synergy Points to refresh it as a Bonus Action.

Synergy Actions:
1. Stroke of Ingenuity (5 base cost): Make an untrained attack roll using Strength or Dexterity against a creature or object within 25 feet. On a hit, you deal 0 damage and must declare one damage type. The first time the target takes damage matching that type before the start of your turn, their Armor Class is reduced by 1 for each damage of the selected type they take (to a minimum AC of 10). This penalty lasts until the end of your next turn.

2. Terrain Clash (5 base cost): This can be take as part of the Attack Action. After a successful weapon attack against a creature in Difficult Terrain, roll a Strength (Survival) or Dexterity (Survival) check If your check exceeds the target’s Armor Class, they fall prone. If your attack was made with an improvised weapon, this action costs 2 fewer synergy points.

3. Defy Reason (3 base cost): As a reaction when you suffer massive damage or are reduced to 0 hit points, spend any number of hit dice and roll them. You regain 1 hit point per hit die and gain temporary hit points equal to the highest hit die’s result. Until you complete a short rest, you must use a bonus action to get up from Prone and you must roll a dexterity save whenever you are hit by an attack. On a roll of 1 or lower your armor (if any) is destroyed. If it is magical armor, its magical qualities are inactive for 1d4 days.

4. (Don’t) Break a Leg (2 base cost): As an Action choose one ally who can hear you. They gain a Synergy die, 1d4, which lasts until used or one hour elapses. Synergy dice can be used for ability checks which do not benefit from a proficiency bonus, saving throws against madness, and Honor saving throws.

So once again: Mental size category determines your maximum Synergy points and how many you can give to an ally through the Help action. Some attacks and conditions probably chip away at one’s synergy points.

[Anonymouswizard]

I've seen it done, generally either in two or three levels going downwards (Unthinking->Animal->Sapient) or in supers games like Aberrant going upwards from human (but Aberrant 2e lets you gain Scale on pretty much anything). It works, mostly because it avoids awkwardness like 'dogs suck at searching because they have minimum INT'.

I'd argue you also need an Algorithmic category for entities than can calculate, but can't 'think'.

[Greywander]

It does seem like it might be better to move away from a linear scale and more towards a checklist of traits. That said, an intelligence ladder could still be useful for some things, but it could potentially be compressed, say Mindless -> Sentient -> Sapient -> Transcient. For alien creatures, we could include an Aberrant category, but if it's just functionally equivalent to Sapient then maybe there isn't a point. Those with alien minds might just have radically different items checked off on the checklist. Likewise for AI/robots.

As for what the different intelligence levels mean, it might come down to getting (dis)advantage on contested INT checks, or getting (dis)advantage on certain types of INT checks, and so on. Or whatever the system equivalent is, if it's not D&D. If there isn't a useful application for these levels of intelligence, then there's no reason to keep them around.

For the checklist, we'd probably want a mix of crunchy and fluffy options. Can they use tools? Can they use language? Can they learn magic? (Might only apply to wizardry, as sorcerery is innate.) But then some fluffy options, like: Can they do abstract math (algebra, calculus, etc.)? Are they self-aware? And so on. It would probably take some work to come up with a good checklist.

When? Unless you're feebleminded (or the animal is supernaturally intelligent) this will never happen.

If you roll for stats you can start with as low as a 3. An ape has an INT score of 6; you have about a 2.8% chance of rolling a 6 or less using the standard of 4d6 drop the lowest. Which isn't a lot, but it still happens.

If anything, I can see value in the potential to have a "higher" intelligence who dumped their INT stat competing with a "lower" intelligence who has a high INT stat. It makes it less clear cut as to which creature is actually "smarter", instead highlighting how the two are different. Characteristics of a higher intelligence, such as the ability to use tools, read, and so on, give a clear advantage, but don't mean you can't still be outsmarted by an animal. In a way, it's similar to the comparison of martials and casters: a highly intelligent animal is like a martial who can rip things to shreds with weapon attacks, while a dumb human is like a caster who is much weaker with weapon attacks but has the capability to cast spells. It's not a straight comparison because one isn't clearly better than the other in every situation, and while the caster might come out on top the majority of the time, it's still going to lose often enough that it isn't a one-sided contest.

[Herbert_W]

For the checklist, we'd probably want a mix of crunchy and fluffy options. . . . It would probably take some work to come up with a good checklist.

There's a shortcut that could be used in order to avoid needing to come up with a good complete shortlist. Most people have a fairly good intuitive idea of how a person thinks, how a dog thinks, how a computer doesn't think, and how a person would appear to a dog. That could be used to establish a set of baselines which are then modified by (incomplete, growing) lists of more specific faculties.

This would have the advantage of simplifying presentation. Instead of each creature having a long list of faculties to read, remember, and apply, each would have a single easy-to-remember description plus a (usually small) number of modifiers.

Using the four categories of your compressed intelligence ladder, we'd have:

1. A mindless creature is similar to a modern computer. Mindless creatures act in accordance with preset rules. They cannot think or exercise creativity; if they appear clever, it was becasue they were cleverly instructed in anticipation of their present circumstance.
2. A sentient creature is similar to a dog. They are aware of their environment and themselves, and they can make decisions. However, they cannot engage in abstract reasoning or construct plans with multiple steps. For example, the notion of turning sideways so that they can fit through a hole while carrying a long stick in their mouth is at the very edge of their comprehension. Many sentient creatures have instinctual behavior which mimics a multistep plan, such as ambush predation.
3. A sapient creature is similar to an adult human. They are aware of their environment and themselves, can reason abstractly, plan, consider hypotheticals, be creative, etc. - basically everything that you can do.
4. A transcient creature can reason in ways that are beyond human understanding. A transcient creature may seem to humans as humans would seem to a dog: they do things and then stuff happens as they will, but with no comprehensible connection between the two. A transcient creature may appear to have supernatural powers even if it does not.

The more specific faculties could potentially be very very specific. For example, let's suppose that zombies are mindless in your campaign setting. That means that they can't understand language - but you want necromancers who create zombies to be able to control those zombies. The 5e solution to this problem is to allow the caster of the create undead spell to issue commands mentally, but granting zombies a specific mental faculty could make for some more interesting solutions. For example, zombies created by a certain ancient ritual could understand commands issued in a dead language (no pun intended), but not any other language (and only commands). Learning that language would prove extremely useful to necromancers who know that ritual. Heck, even just learning a few useful words and phrases from it would be of great value. This provides a reward for players that isn't just more gold or magic items with bigger numbers, a motivation for villainous necromancers to kidnap scholars, and an opportunity for a dramatic tables-turning when a BBEG marches up to the king's castle with his army of the dead, only to find out that the players know a few words of the dead language too!

[Rilmani]

I think that the Sapience intelligence category could be interesting for player characters because sapient creatures can do amazing things when they cooperate. In that vein we should look at the Help action (in and out of combat), the Improvise action, the spell Aid, Luck points, and the spell Resistance. We all know that not all actions are created equal- Attack and Cast a Spell are quite dominant. So I reckon we should have a resource which lets a character take the Help action as a Reaction, then we augment the Help action from there. Pack creatures with Sentience tier intelligence can have this resource, but they have restrictions when using it, while a Sapient creature can use it to Help an enemy against a greater threat.

Any rebuttal for the italicized text? If we want to talk strategic battlefield competence outside of training and experience (feats, maneuvers/superiority dice, proficiencies), I think that is the core around which intelligence “sizes” will matter.

What other angle, if any, should this quality affect? Is there any reason intelligence categories should not connect to cooperation at all?

[Herbert_W]

Any rebuttal for the italicized text?

You're not wrong, but I do think that you're suggesting adding something that DnD already has.

DnD is already a cooperative game where parties that work together do much better than parties which do not. Basic synergies like casters buffing martials, martials door blocking to protect casters, characters using the help action to ensure that a big limited-resource attack lands, etc. are bread and butter for this game. Characters plan to exploit those synergies as a natural extension of players planning to exploit those synergies, becasue both (most) characters and players are sapient. Likewise, players and characters plan to avoid anti-synergies such as devil's sightlocks screwing over the rest of the party with darkness.

I think that the italicized text is more appropriate as a description of a high-level consequence of the rules of a game than something that the game should directly model with a dedicated mechanic - and it's a consequence that DnD already models quite well.

Where the concept if intelligence categories is needed IMO is when a player and their character aren't in the same category, or when a DM and an NPC aren't in the same category. Players and DMs may need or appreciate extra guidance (meaning advice and/or mechanics) when roleplaying something whose mode of thinking is different from their own.

I do see some interaction between intelligence categories and cooperation being appropriate to model with rules, but those occur above and below the sapient level. Above the sapient level, creatures can influence the world in ways that humans can't understand - and there's no reason why that can't extend to cooperation. I would see it as appropriate for some above-sapient creatures to be able to e.g. use the help action as a free action once per round, at any range, by setting up favorable 'butterfly effect' chains. I can also see it as being appropriate to forbid certain forms of cooperation at the subsapient level. Outside of certain specific instinctual behaviors (such as swarming), cooperation requires planning and a rudimentary theory of mind, which subsapient creatures tend to not have.

[sandmote]

(There's also the fact that "orcs are stupid" is reminiscent of real-world nastiness in ways that have been discussed to death; we hopefully won't derail the thread by rehashing that here. I'll just say that "orcs have poor memories" sounds a lot nicer and would hopefully make for a more inviting and therefore potentially larger game.)

I was trying to avoid bringing up the example, but this is why I've been opposed to a category including creatures very slightly below "sapient." I'd rather include orangutans in the sapient category than leave it unclear where other humanoids fall.

Though the idea of an orangutan wizard is pretty entertaining.

So's Terry Pratchett's execution on the idea.

Any rebuttal for the italicized text? If we want to talk strategic battlefield competence outside of training and experience (feats, maneuvers/superiority dice, proficiencies), I think that is the core around which intelligence “sizes” will matter.

What other angle, if any, should this quality affect? Is there any reason intelligence categories should not connect to cooperation at all?

I think the problem you're going to run into is that cooperation can be done between even unthinking creatures. It isn't necessarily the sort of cooperation that sapient creatures are likely to think of, but there are things like lichens and corals that include cooperation between species in different phyla for their mutual benefit. Many creatures considered sentient but not sapient can also work together, including various species that hunt in packs or form complicated family structures.

I vastly prefer Herbert's basic concepts for the categories, although I'd want plants, spiders, and dogs to fall into separate categories. We can make some sense of what dogs are thinking, despite them having less reasoning ability, but need a massive amount of research to understand anything a spider does; and the lines along which a plant processes and responds to information is limited to recording their response to stimuli as we observe it (or, I guess, figuring out which chemicals send the message).

For categories above humans, I also can't think of an existing term. Although I do feel there should be at least three categories above sapient. I'm having trouble putting them into words

As another attempt on this, I suppose I'd expect it to be helpful to categorize creatures of increased understanding in a similar manner. There are those below us capable of having basic principles explained to them, and so should be creatures whose thinking can have the basics explained to us. There are creatures capable of some sort of reasoning or method for differentiating responses that's too different for us to follow, and think there should be such a category in both directions.

[Rynjin]

If you roll for stats you can start with as low as a 3. An ape has an INT score of 6; you have about a 2.8% chance of rolling a 6 or less using the standard of 4d6 drop the lowest. Which isn't a lot, but it still happens.

Is this a 5e thing? It wasn't labeled as 5e homebrew so I wasn't sure.

Previous editions did not have this issue; no animal had an intelligence greater than 2. Int 3 is the cutoff for sapience.

[Greywander]

Yup, the PHB offers a couple different ways of generating your ability scores. There are stat arrays and point buy (neither of which go any lower than 8), but you can also roll. 4d6, drop the lowest. Roll your own array, then assign as you please. It's more forgiving that some other methods of rolling stats (e.g. 3d6, in order), but it still has a chance of getting something less than 8, and all the way down to 3. Apes and dolphins, which are quite smart for a beast, have an INT score of 6. Giant eagles and giant owls go as high as 8, but they speak a language and should probably be considered sapient. There's a few more beasts in the 3 to 5 range, but most beasts have an INT of 1 or 2.

[awa]

2nd edition was worse there dolphins had an int of 11-12 most whales were dumber but still smarter than ogres. Though in looking that up I was surprised how much smart ogres and orcs were back than, I mean I suppose 8 to 6 isn't that big a difference but it feels like it to me.

[olskool]

Older edition sourcebooks (I can't seem to remember which one right now) had a thing called FIXED INT for animals like horses, dogs, and cats. They weren't "intelligent" as humans see Intelligence but they could follow certain directions based on how complex those directions were. Whenever the animal was given a command, you would try to roll UNDER the animal's INT score to have the animal execute that command (the older Proficiency system was "roll under." Thus, the higher the animal's INT was, the easier it was to train!

[Rynjin]

Yup, the PHB offers a couple different ways of generating your ability scores. There are stat arrays and point buy (neither of which go any lower than 8), but you can also roll. 4d6, drop the lowest. Roll your own array, then assign as you please. It's more forgiving that some other methods of rolling stats (e.g. 3d6, in order), but it still has a chance of getting something less than 8, and all the way down to 3. Apes and dolphins, which are quite smart for a beast, have an INT score of 6. Giant eagles and giant owls go as high as 8, but they speak a language and should probably be considered sapient. There's a few more beasts in the 3 to 5 range, but most beasts have an INT of 1 or 2.

Gotcha. I can see how it might be useful in some senses, but I don't think this needs to be made as complex as the OP in that case.

There really only needs to be two designations: sapient and non-sapient.

So while an Ape and someone with a low stat roll might both have Int 8, their intelligence is "different".

Int 16 (non-sapient) has strong problem solving skills and the ability to learn more than the average animal. Int 8 (sapient) is still orders of magnitude more intelligent in many ways, just due to simply processing and acting on information differently.

There is no need for "non-sentient" as a designator, because Int -- covers that. (I think; 5e does still have Int -- for vermin, constructs, plants, etc. right?) and due to the hard stat-caps for player characters present in 5e, "Transcient" or "super-sapient" characters are really just...super-intelligent. Int scores higher than 20, or in the case of gods, effectively infinite. Int -- in a different way.

[Herbert_W]

Gotcha. I can see how it might be useful in some senses, but I don't think this needs to be made as complex as the OP in that case. There really only needs to be two designations: sapient and non-sapient.

Well, that depends on what you expect intelligence categories to do.

Two categories would be perfectly adequate if you just want to avoid weirdness where player characters can end up dumber than animals, or effectively expand the range of intelligence scores without actually giving the int scale bigger or fractional numbers. (I'd still consider "int --" and "int effectively infinite" to be categories in the same way as NaN is a value, but maybe that's too nitpicky.)

If you want to provide guidance to people who are playing characters whose cognitive capabilities are dramatically different from their own, then more categories (and/or further modifiers to those categories) could be helpful.

. . . I've been opposed to a category including creatures very slightly below "sapient." I'd rather include orangutans in the sapient category than leave it unclear where other humanoids fall.

I'll preface this by saying that this might be a moot point since we might not want more categories anyways, and I don't know whether raising this question will help or make matters worse.

With that being said, what about human children? If we're going to portray them realistically, then infants will be in the same category or lower than adult apes and children will gradually climb the categories as they grow. Would you be comfortable with putting apes into a "slightly subsapient" category if it is explicitly acknowledged that they share this category with human children?

I think some people might be unconformable with the existence of a "slightly subsapient" category because of the fear that placing a creature in it would be seen as endorsing mistreatment of those creatures. Yet, as the example of human children shows, diminished mental capacity does not justify mistreatment. If anything, it implies a greater duty of care.

On the other hand, likening adult apes to human children still carries uncomfortable implications - just of a different sort - in a game where they are likely to be opponents in a combat encounter.

I'm beginning to suspect that modelling the thoughts and feelings of combat opponents in too much detail will always make people uncomfortable, becasue it draws attention to the fact that they have thoughts and feelings. Different models may provoke different types of discomfort, but the discomfort is always there - except when the model is abstract and impersonal.

I think the problem you're going to run into is that cooperation can be done between even unthinking creatures. It isn't necessarily the sort of cooperation that sapient creatures are likely to think of . . .

I agree, but I think that this is a reason to modify Rilmani's idea rather than to discard it. Intelligence category could very reasonably affect the type of cooperation that creatures exhibit (as you said, the type of cooperation that subsapient creatures exhibit isn't necessarily the same as what sapient creatures think of).

One point that I'd like the emphasize is that we don't need new mechanics to represent cooperation at the sapient level, becasue DnD characters already do that becasue players do that. (It'd be nice to bring back flanking bonuses, but that's neither strictly necessary nor really new.) Adding more complexity to model something that the game already models is just cost with no benefit. It's above and below that level that new mechanics could become useful.

I vastly prefer Herbert's basic concepts for the categories, although I'd want plants, spiders, and dogs to fall into separate categories. We can make some sense of what dogs are thinking, despite them having less reasoning ability, but need a massive amount of research to understand anything a spider does.

There's certainly room for a category in between "mindless/computer" and "sentient/dog" - I just can't think of one where both (a) the category itself is a helpful one to include in a game and (b) there's a commonly understood example that could be used to explain that category in a way that most people would understand.

That second concern is why I avoided having a "spider analogue" category: most people don't have a good idea of how a spider thinks. If you hand a DM a monster and tell them that it thinks like a spider, they won't be any the wiser about how to run it. (Or, worse, different people will have different ideas about how the spider-like creature should behave, leading to arguments.) Another complicating factor is that different species of spiders will get different lists of functions. If you hand a monster to a DM who does know about spiders and tell them that it thinks like one, they'll probably say "OK, but what species?"

I suspect that the best way to explain a spider-like mind to the average person is to start with a foundation of "mindless" and then add specific cognitive functions to that.

[Devils_Advocate]

I've held the opinion for a while that the so-called intelligence score in DnD ought to instead be called memory becasue "memory" is a more accurate representation of what the score does mechanically. A character with a higher int score cannot think, reason, or make decisions more adeptly because it is the player who does all of those things. What a character with a higher int can do is make more successful skill checks to remember information, be more effective with forms of magic that rely on memory, and (in 3e) learn more skills. That's all memory!

There's Investigation / Search / whatever, but that is kind of an edge case. And it ain't like Intelligence can represent general mental competence and at the same time be only one of three equally important mental stats! Honestly, if you just want more nuance to "how smart something is" in D&D than just a single score, just taking into account that there are two more to consider should help a bit.

Mostly, I like how "memory" is clearer than "intelligence", which is a lot more nebulous.

However, this isn't true learning, merely a complex response to a complex set of stimuli; it is predictable and can be easily reproduced with the right stimuli.

Well, "predictable" is relative. A sufficiently advanced mind might regard us as quite predicable. But saying that nonsentient learning is predictable by sapients gives some helpful guidance on how the categories interact with each other.

(You could decide that sapient beings are inherently unpredictable, and that's a valid setting decision, but I wouldn't include nondeterminism as part of any level of intelligence; that's something different.)

Maybe you could reframe some of the everyday things humans do as an expression of psychic ability, which is why animals can't do those things.

Isn't "Humans have mental capabilities that other species lack" kind of the going explanation?

(Yes, I know, you didn't intend for "psychic" to mean "mental" or "of the mind" in this context. Still, funny.)

it's basically classifying these animals as "people" rather than "animals", which raises a whole lot of moral/ethical/philosophical questions

As Herbert's post already touched on, you're not really avoiding those questions.

Humans don't all have the same level of intelligence. That's... a big part of why games like Dungeons & Dragons have Intelligence scores. To model those differences. You can define an upper bound as "the smartest human who ever lived" or "the smartest a human can be without artificially modifying or augmenting the brain", but the realistic lower bound is the lower bound of the rating system as a whole. And that's pretty much the degree of lack of intelligence sufficient for someone to be categorized as "not a person". Humans with all the mental activity of broccoli get called "vegetables". A human only as smart as the smartest dog is not going to be called "an animal" by most people, especially not in a sense of the word that doesn't include humans in general.

So, just because there are lower categories, "human-level intelligence" obviously isn't supposed to cover the full range of human intelligence. It's meant to cover everything from the smartest human to ?????, where ????? is rather conspicuously undefined. But leaving it vague where that line is drawn doesn't avoid the issues inherent in saying that that line exists.

There are basically three possibilities here:

1. Everyone below some level of intelligence is not a person, and thus is unworthy of the consideration given to persons, whether human or not. Why shouldn't the poor sell their babies to the rich to eat? Yum yum, delicious!

One can't fault the internal consistency of this position on its own, but one may feel moved to ask how it is that the less intelligent are less worthy of consideration.

2. Humans deserve greater consideration than other species because of something special about us other than intelligence.

What would that be? Is this difference demonstrable rather than purely hypothetical? How does this make humans deserve greater consideration?

3. The only thing that can make someone less deserving of consideration than someone else is their freely chosen actions, if that.

Bluntly: This is the ethical option. Every alternative is always gonna be some flavor of evil, regardless of which subcategory of sentient being you arbitrarily treat as being okay, or even more okay, to crap on. Devaluing others for things that aren't in their control is morally wrong. Which, let's be clear, does not mean that it's okay to devalue others for whatever choice you, uh, choose.

Note that I say "consideration". As I would hope would go without saying, I am not saying that everyone should be treated the same.

[sandmote]

There's certainly room for a category in between "mindless/computer" and "sentient/dog" - I just can't think of one where both (a) the category itself is a helpful one to include in a game and (b) there's a commonly understood example that could be used to explain that category in a way that most people would understand.

I was thinking of how the DM should adjudicate illusory and "mind effecting" abilities. In this case I'd actually put most computers in the category between "mindless" and dog-level; a creature with that level of intelligence can fall for and recognize illusions, whereas a "mindless" creature cannot. While the edges between the "mindless" category and spider-level category would be fuzzy (as all these categories are) I kind of assumed this would be a helpful situation to quickly include and relatively common in-game (although I admit is wouldn't be as commonly understood in real life).

That second concern is why I avoided having a "spider analogue" category: most people don't have a good idea of how a spider thinks. If you hand a DM a monster and tell them that it thinks like a spider, they won't be any the wiser about how to run it. (Or, worse, different people will have different ideas about how the spider-like creature should behave, leading to arguments.) Another complicating factor is that different species of spiders will get different lists of functions. If you hand a monster to a DM who does know about spiders and tell them that it thinks like one, they'll probably say "OK, but what species?"

I'd give it a more generic name, and was using plants, spiders, and dogs as examples. For instance, I'd put cats, horses, giant striders, varguilles, pegasi, and basic air/earth/fire/water elementals into the same intelligence category as dogs. While all these placements are debatable, my point is that it doesn't mean everything I would put in this category should be run the same way as the dog.

I suspect that the best way to explain a spider-like mind to the average person is to start with a foundation of "mindless" and then add specific cognitive functions to that.

Arguably too detailed to really be helpful, but I do think making a checklist of traits would make it easier to avoid the uncomfortable things that can happen when attempting to make this sort of scale.

Then check boxes like "object permanence," "visual individual recognition," "olfactory individual recognition," and "self awareness," then let you mark off what a creature can or can't do. Then humans pick up the "object permanence," category at 8 months while never qualifying for "olfactory individual recognition" the way dogs do.

[brian 333]

My question is, how would this be useful in a mechanical sense?

Size=Strength is not always true, but larger things tend to also be stronger. Elephants are stronger than mice, for example. But a giant pillow golem would not be stronger than flesh golem half its size. And strength has nothing to do with how difficult a target is to hit or how small a door the monster can chase you through. Size has mechanical utility in the game.

It seems to me that the premise here is as a role-playing aid. Players will not be playing animals, so it isn't a useful tool for them. A DM already has discretion over how NPCs are played, so it is not a necessary tool for a DM.

I admire the creativity and consideration, and it may form the basis of a good training tool for DMs, but how would it be useful in a game? Adding another stat to track adds complexity to an already complex system. There needs to be a payoff that makes the complexity worthwhile.