The meaning of Int 2

[hymer]

What are animals capable or incapable of, combat-wise, on account of having Int 2?

Can they delay actions? Ready them? Use bull rush, disarm, trip? Know when it's good to grapple and when it isn't? What are your opinions? And feel free to ask and answer more questions along these lines.

[VoxRationis]

They can do all of those things except maybe disarm (the rules don't explicitly prohibit them from using disarm at Int 2, but I as a DM would rule that seeing a tool, identifying its purpose, and devising a plan to force the opponent to drop it would be outside the bounds of most animals' intelligences). Just keep in mind that the animals are going to be less adaptable in their strategies.

[torrasque666]

They can do all of those things except maybe disarm (the rules don't explicitly prohibit them from using disarm at Int 2, but I as a DM would rule that seeing a tool, identifying its purpose, and devising a plan to force the opponent to drop it would be outside the bounds of most animals' intelligences). Just keep in mind that the animals are going to be less adaptable in their strategies.

Seeing a tool and doing that all just from seeing it? yeah, too smart. But seeing them use that tool to inflict pain on a comrade/itself would then give it all the reason it needs to try and get rid of it. By, ya know, attacking the wrist.

[supermonkeyjoe]

The best way I find to run animal behaviour is to watch some documentaries, packs of wolves for example hunt in ways that naturally promote flanking and ambushes, solitary hunters would just run at the prey until it catches it or it gets away.

I'd say that creatures wouldn't use combat manoeuvres unless they were particularly built for it, wolves will trip but not grapple or bull rush for example, Bovines would bull rush and attack and maybe overrun but that's about it.

Depending on the circumstances the creature may flee as soon as it's hit or fight to the death, if it's on the hunt or if the PCs are approaching its young for example.

I usually give animals one or two simple objectives for the fight, maybe a simple tactic like attack the smallest PC and go from there, they can't innovate any new plans during the fight and without a good reason to stick around they will flee if their quarry seems stronger than them.

[Alefiend]

How about the difference between Int 2 and Int 1? In what ways is a snake dumber than an octopus?

[Eldan]

Well, an octopus understands fairly complex tool use and puzzle solving. Some even build things for themselves. Snakes don't. The related cuttlefish even use primitive language.

Really, that's the problem with Int 2. Some animals are far too intelligent. Like crows or chimps.

[HammeredWharf]

Int 2 animals tend to have a good grasp on certain types tactics (remember, they often have high Wis), but can't use other tactics effectively. For example, a pack of wolves is quite likely to surround their targets, trip them when necessary, avoid AoOs and go after the squishies, but they won't know what magic is unless it makes a big boom. They can evaluate the threat level of simple melee characters quite well by looking at them, but need to see the abilities of unusual threat to evaluate them. They're also very likely to value their buddies and retreat when hurt.

Int 1 animals only have a gimmick or two and are unable to use other tactics. They're often unable to adapt to situations. For example, a snake can sneak around and bite stuff, but can't work together with other snakes or use many of the complicated tactics described above.

Edit: or at least that's how I play them when I DM.

[The Oni]

Really, that's the problem with Int 2. Some animals are far too intelligent. Like crows or chimps.

^ This. A crow has got to be at least Int 4. Some dog breeds probably hit 5 or 6. If you optimize heavily, you've probably played characters that are dumber than a bonobo.

[hymer]

Well, an octopus understands fairly complex tool use and puzzle solving. Some even build things for themselves. Snakes don't.

Depending on what you mean by building things, king cobras build their own nests. Alligators and crocodiles put sticks on their faces and use them as lures for nest building birds. It's definitely not crow or moose level intelligence, but I'm not so sure the '1' is merited as a blanket score for reptiles.

[Arc_knight25]

Normally in the MM's they will have descriptions of how the creatures will act in combat, what abilities they have and how they will use them. When I DM (Which is rare mind you) I just use the descriptions to help with how they behave in combat.

Things get more interesting when Speak with Animal comes into play, for now you need a personality along with what information something with an Int 1 or 2 that doesn't normally talk. They have a very different description of landmarks and things seen.

[Segev]

A lot of the "animals should have Int 4-6" animals can have their tricks modeled as, well, tricks. Sure, that usually requires training by a character with Handle Animal, but that would still probably be the better mechanic than trying to claim their Int is above 2. Even the least intelligent, most mentally-handicapped people can do some surprisingly clever things if put in situations similar to those where we see "clever animals" doing "high intelligence" tasks. (Well, those who aren't vegetables.)

I'd say that Int 1 is a very, very rudimentary ability to adapt based on ingrained instincts. It doesn't entail much if any real learning ability, but it is sufficient to have stimulus-response behaviors develop into a short-hand that cuts out the middle-man. It's enough to learn a few tricks, but not enough to really perform even rudimentary puzzle-solving. At best, an Int 1 animal is going to solve a puzzle by trying different things until one works, with no necessarily rhyme nor reason nor adaptation shown in the different approaches. And they might well try the same thing that didn't work more times than is necessary to determine it doesn't work. By the same token, they could "solve" the problem by repeating the same activity until sheer stubbornness causes a change. (Beating on a door until it breaks, for instance.)

Int 2 would be where rudimentary reasoning becomes available. They learn faster and they adapt tools to their use through a combination of instinct and ability to actively learn. Still operating on instinct more than intellect, Int 2 animals nonetheless can have some highly complex learning behaviors within their instinctive bailiwick. Crows and parrots can adapt their sound-mimicry to language, monkeys can teach each other to use tools and can experiment, and dogs can read a situation to determine how best to carry out commands, even reasoning out new tricks in which they've yet to be trained.

Int 3, the first non-animal level of intellect, is where true understanding of language comes in. Abstract reasoning may be very difficult at this level, but it can be done. Symbols can represent things, rather than being rote learning or trick identification. A lot may elude the understanding of an Int 3 character, but an Int 3 creature has an ability to abstract concepts and apply them that is forever beyond the grasp of even the most clever problem-solving Int 2 animal.

[prufock]

^ This. A crow has got to be at least Int 4. Some dog breeds probably hit 5 or 6. If you optimize heavily, you've probably played characters that are dumber than a bonobo.

Based on the average human scale of 3d6 in D&D, this puts a crow at the .46th percentile. This equals an IQ of about 61. A human with IQ 61 would be expected to be able to harvest vegetables and repair furniture. Now, a crow could probably be trained to harvest vegetables, but I don't see it learning to repair furniture.

A dog, in your estimation, would be in the 1.85th percentile to the 4.63rd percentile. This would mean an equivalent IQ of about 69 to 75. A 75 IQ human is expected to have a 50% chance of finishing elementary school. How many dogs do you know that could finish elementary school?

I'd say 3 should be reserved for very unique animals - hand-signing monkeys, very well-trained dogs, and particularly smart birds.

[VoxRationis]

You would let someone with an IQ in the low sixties repair your furniture? My mother gets mad if I even TOUCH certain pieces of furniture, and I am quite intelligent.

[hymer]

@ prufock: Nice work. I wonder, though, whether you can reasonably conflate Int score and IQ. IQ is meant to measure human intelligence, Int is used for just about anything from dogs to gods.

[Ratatoskir]

You would let someone with an IQ in the low sixties repair your furniture? My mother gets mad if I even TOUCH certain pieces of furniture, and I am quite intelligent.

There's no great cognitive feat in repairing furniture. Some specialized knowledge on the higher end of it I'm sure, but wood nailed together isn't a highly complicated idea.

[Jeff the Green]

Depending on what you mean by building things, king cobras build their own nests. Alligators and crocodiles put sticks on their faces and use them as lures for nest building birds. It's definitely not crow or moose level intelligence, but I'm not so sure the '1' is merited as a blanket score for reptiles.

Wait, what?

[Arc_knight25]

Wait, what?

Maybe a somewhat botched Rocky and Bullwinkle reference?

[hymer]

Wait, what?

Well, I honestly don't know much of anything about moose intelligence, but it's such an amoosing creature I wanted to put it in. The closest I can get to a defence: I think Muldoon in Jurassic Park (the book) has an internal monologue about how some animals are really good at escaping from zoos (he's an expert, and is thinking about raptors and how good they are at getting out), and he notes that moose (or maybe it's caribou?) are surprisingly good at manipulating objects with their noses, so they have a knack for escaping. Now, Muldoon could be wrong, or Crichton could be wrong, or I could be wrong. But if we're all right, then moose are apparently not as dumb as their poor vision and blind angle right in front of them make them seem.


Maybe a somewhat botched Rocky and Bullwinkle reference?

I'm afraid you give me too much credit. All I know about Bullwinkle is that Sam from the West Wing won't eat moose because the cartoon character can hatch a plan.

[No brains]

To forewarn, this goes off, but mostly just beside your topic of just what actions an Int 2 animal may take, but I think it is worth going into why I came to my conclusions. I can't cover every animal, so the reasoning I used for some might still work for others.

I've wondered a bit about this myself and it's always important to remember that "intelligence" in D&D is "the ability to think, learn, and remember". It can help to try to think of times when an animal needs to do any of those things and make the call on whether it could reasonably do any of them. It can also help to look beyond animals and look at the tactics of other low-int creatures from the MM.

Before I really begin though, I should note that a creature's natural abilities should give a good idea of what it 'knows' to do. Wolves have a trip attack, so wolves must be smart enough to trip. Lions have improved grab, so they must know the benefits of grappling. Both of them have the same Int score, but they don't both know the advantages of each other's tactics, so it could be reasonable to assume Int 2 allows for the knowledge of one combat action. Moving and preparing allies will be covered later.

"Think" is not something many animals need to do in their daily lives, especially herbivores. Their primary goals in life are to get food and avoid being food Their entire morphologies attack problems better than their brain alone. Now Int 2 predators seem to know the advantage of setting up before charging in. Lions and wolves will try to maneuver into position before the attack to cut off escapes. Contrast this with Int 1 or - predators. They just move at prey (although sometimes stealthily) and then attack.

"Think" doesn't just apply to predators though. Most animals actually have Int 2 and Int 1 is the exception. The only creatures with Int 1 are actually solitary creatures that aren't domesticated. Either a social structure or domestication seems to give an Int of 2 so it can be assumed Int 2 knows that other creatures can help a creature survive.* There are obvious exceptions with mates and offspring, but even those exceptions have exceptions with cannibalistic animals.

*this could be the tl;dr of my post.

There is the exception of predatory birds, which are often solitary, but have Int 2. Maybe Int 2 is needed to usefully maneuver in three dimensions.

"Learn" is something no animal may ever do. Their bodies have a plan built in that they can stick to, and it takes a rare instance of success when the plan is defied for there to be a chance to "Learn". Learning seems to be built into a creature's capacity for a social structure. A creature can learn more when its sample size is increased. A solitary creature may never learn that something is instantly fatal, as it may never get the chance to observe something of similar ability falling prey to something. It obviously can't learn from watching itself die.

I could go on all day, but I'm starting to get a headache and "Remember" goes right along with "Learn". Expect me to come back to explain exactly what Int 2 creatures may never do.

[Nibbens]

But doesn't a wisdom score make up for a lack of intelligence in some way. My dog knows how to "disarm" me of a rope when we're playing tug, more or less - it might be a failed attempt most of the time, but then again he's not actively trying to kill me. Not to mention he "sunders" his toys and my pillows when he's mad at me, or "steals" the sandwich off my coffee table when I'm dumb enough to leave it there when I'm watching a movie. LOL.

[hymer]

@ No brains: It seems to me that the Animals in MM1 at least are simply assigned int based on whether or not they are warmblooded. Birds and mammals (including several solitary predators like leopards and bears) get int 2, everyone else gets int 1.

@ Nibbens: Maybe it does. Regardless, it seems you can ascribe a wide variety of actions to your (D&D Int 2) dog.

[Zubrowka74]

First off, D&D is not good at modeling reality. It's not meant for this.

That being said, a lot of "tricks" animals uses in the wild are innate. They are behaviors that have been recorded in the ADN. Remember, dogs do use disarm tactics. Most will attack the wrist right away, weapon or not.

So I would'nt play with INT scores for anumal, unless it's a special type, like a companion.

[hymer]

To sum up, it seems to me there is something like consensus that the limit imposed by int 2 is not so much in what actions and manoeuvres you can take, but rather in judging the right time to do them. And that suggestions in text or stat block would show the kinds of things animals are more likely to do, such as wolves mentioned trying to flank and bears having improved grab.
Thanks everyone! Any further thoughts are, of course, welcome.

[prufock]

You would let someone with an IQ in the low sixties repair your furniture? My mother gets mad if I even TOUCH certain pieces of furniture, and I am quite intelligent.
If they are trained to do so. Those benchmarks are based on studies correlating IQ with certain tasks or accomplishments. Note that this doesn't mean they're doing any intricate work, but they can deal with things like screwing/nailing pieces together, glueing, painting, etc.


@ prufock: Nice work. I wonder, though, whether you can reasonably conflate Int score and IQ. IQ is meant to measure human intelligence, Int is used for just about anything from dogs to gods.
Maybe not precisely, but I think we'd agree that Intelligence score is meant to model the same thing IQ measures (ability to learn, problem solving, etc). Animals can certainly be better at some specific tasks, but as far as general problem solving aren't on par with humans.

[hymer]

@ prufock: I don't really agree. IQ measures mostly things that you can train yourself or be educated in, which is why we get the Flynn effect (which means that the general public do measurably better at IQ tests today than we did ten years ago, meaning biologically almost nothing has changed, but the test results have). Intelligence is a much more raw, more purely biological stat. Putting it in D&D terms as best I can, IQ measures your skill points as well as raw Int, and while skill points are in part granted by Int, a first level rogue is likely to have many more of them than a first level wizard, despite the wizard having the higher int (assuming the same race and point buy).

[prufock]

@ prufock: I don't really agree. IQ measures mostly things that you can train yourself or be educated in, which is why we get the Flynn effect (which means that the general public do measurably better at IQ tests today than we did ten years ago, meaning biologically almost nothing has changed, but the test results have). Intelligence is a much more raw, more purely biological stat. Putting it in D&D terms as best I can, IQ measures your skill points as well as raw Int, and while skill points are in part granted by Int, a first level rogue is likely to have many more of them than a first level wizard, despite the wizard having the higher int (assuming the same race and point buy).

There's no reason to think that a "purely biological" trait can't be trainable. Strength can be trained with weights. Dexterity can be trained with yoga. Constitution can be trained with endurance running. Why would you think Intelligence can not be trained with testing practice?

Skill points measure your ability to perform specific tasks, but there is no "recognize pattern" skill, for instance. Skills are more similar to academic study or on-the-job training (since they differ based on class); your ability to learn certain jobs is a quality of your intelligence. So learning to pick a lock requires both the training (from your job/class) and your intelligence. Skill points are a reflection of both.

[hymer]

There's no reason to think that a "purely biological" trait can't be trainable. Strength can be trained with weights. Dexterity can be trained with yoga. Constitution can be trained with endurance running. Why would you think Intelligence can not be trained with testing practice?

No matter how many weights a character lifts or how much yoga they practice, their physical scores remain the same. They can be improved through experience and magic, or in one case a prestige class - perhaps there are other ways, but then they are pretty obscure. Most characters get no increase in any of their scores just because they train that ability more. This is another sign that you can't use a real-world attempt at measuring intelligence to say much of anything about how Intelligence works in the game.


Skill points measure your ability to perform specific tasks, but there is no "recognize pattern" skill, for instance. Skills are more similar to academic study or on-the-job training (since they differ based on class); your ability to learn certain jobs is a quality of your intelligence. So learning to pick a lock requires both the training (from your job/class) and your intelligence. Skill points are a reflection of both.

There not being a Take IQ-Test skill is a measure of how little the people making the game thought such a skill would contribute to the system and the game. If you took two people with the same Int, one with all his skill points in things like Jump, Balance, Swim and such, and the other with all his skill points in Knowledges and Profession (Teacher), the second one would score better, and probably much, much better.

[prufock]

No matter how many weights a character lifts or how much yoga they practice, their physical scores remain the same. They can be improved through experience and magic, or in one case a prestige class - perhaps there are other ways, but then they are pretty obscure. Most characters get no increase in any of their scores just because they train that ability more. This is another sign that you can't use a real-world attempt at measuring intelligence to say much of anything about how Intelligence works in the game.
Of course when we are trying to shoe-horn real life into D&D rules, there are bound to be inconsistencies. On the other hand, all of the racial paragon classes increase an ability score at level 3, and I'm sure there are others. And as you said yourself, gaining enough levels allows you to increase an ability score.


If you took two people with the same Int, one with all his skill points in things like Jump, Balance, Swim and such, and the other with all his skill points in Knowledges and Profession (Teacher), the second one would score better, and probably much, much better.
I'd need to see some stats on this to comment, and I can't find any IQ stats for average athletes. A +2 circumstance bonus for a teacher wouldn't be out of the question.

[Vogonjeltz]

For D&D purposes Int 1-2 only reflects the inability to read/write/speak languages and determines the trick limit.

Nothing else. So animals can be just as crafty as anything else. This is represented by wisdom (tool use is fairly common in animals including insects, Int doesn't model this, wisdom does)

[Coidzor]

What are animals capable or incapable of, combat-wise, on account of having Int 2?

Can they delay actions? Ready them? Use bull rush, disarm, trip? Know when it's good to grapple and when it isn't? What are your opinions? And feel free to ask and answer more questions along these lines.

I think they can ready and delay them, though they wouldn't always do so. I can't really think of any good examples for situations where they'd ready or situations where an intelligent creature would ready and an animal might or would not.

Bull Rush? Some of them can certainly. Like Moose or Bears or Boars or Bovine-Types. Maybe Elephants.

Disarm? I think several of them are capable of it, but I'm not quite sure how often it would occur to animals that aren't specifically trained to disarm. I can see training an animal to disarm as part of its default approach to combat, I suppose.

Tripping? Definitely. Probably the most ubiquitous combat maneuver other than maybe grappling.

As far as knowing when it's good to grapple and when it isn't... I think their criteria aren't as fine-tuned as, say, an experienced, learned student of combat with 10+ Intelligence, but I believe that those which grapple have some ability to size up whether it seems like a good idea to grapple. Like a giant constrictor snake probably wouldn't try to grapple an even bigger constrictor.

[Jeff the Green]

Note that there are tricks to tell your animal to bull-rush, trip, etc.