How to Define Animal IQ?

[Pursuivant]

Originally Posted by Skarg

1. Assign a base-10 IQ to the individuals, which represents how smart each animal is relative to other animals of the same species, in the standards relevant to that animal's frame of reference. (i.e. an average bobcat has a bobcat-IQ of 10. A smart bobcat might have a 12 in thinking like a bobcat about things bobcats care about.) From human frames of reference, the standard GURPS IQ could still be used. I would research what expert species researchers know about the species' ways of thinking, preferring researchers who have spend lots of time observing the animals (q.v. the book Are We Smart Enough To Know How Smart Animals Are?)

This is how GURPS 3E Bunnies and Burrows handles it. It "re-norms" the GURPS attributes so that the rabbit is the baseline species.

That allows meaningful variations in ST and IQ among small animals.

As it stands, there just isn't enough "granularity" in the IQ 0-6 range to reflect subtle variations in intelligence, like the fact that dogs are generally smarter than cats or that chimpanzees are generally smarter than gorillas.

If you were to were to give partial levels of an attribute for ST and IQ, at the rate of +/-1 per partial level of ST and +/-2 per partial level of IQ, you get finer gradations between species.

For most purposes, fractional levels of attribute scores are ignored, but when dealing with another creature with the same level of ST or IQ, skill and attribute rolls are "renormed" so ST or IQ +/-0 = 10, +1 = 11, +2 = 12, and so forth. Treat a full +/-1 level of ST or IQ as a +10/-10 modifier.

That way you could have a "heroic" cat who's strong (ST 2.4) and smart (IQ 5.3) compared to other cats, but a run-of-the-cattery moggie with respect to humans.

The same concept might apply to HP and FP, but that also requires the concept of fractional points of damage, which would make things really complex.

Originally Posted by Skarg

2. Figure out what I am going to use the IQ-like ratings for (probably perception and skills, I would think, but maybe also certain situations), and then assign them whatever value seems appropriate for each of these. This is more or less what GURPS Bestiary does - animals get perception ratings and skill levels that are just what they are, and can vary a bit for each individual.

Given that traits like DX, HT, Per, and Will really don't scale (much) with Size or species, this is the way to go.

Given that a score of 12 is considered to be "professional" level in GURPS, it makes sense to give most animals an overall Per of 12, and then make whatever IQ-based skills they might have substitute to Per.

For example, a wolf with Tactics (Pack Hunting) uses Per rather than IQ.

Will should probably be set at 10, lower for noticeably distractable or tractable animals, lower for noticeably perseverent creatures.

[Skarg]

Originally Posted by Pursuivant

As it stands, there just isn't enough "granularity" in the IQ 0-6 range to reflect subtle variations in intelligence, like the fact that dogs are generally smarter than cats or that chimpanzees are generally smarter than gorillas.

It's both a low-grain scale, and a human-centric one. It also tends to favor rating how trainable an animal is to be gotten to do things humans think are interesting.

Meanwhile a cat may think both dogs and humans are less smart than they are, as judged from a cat's perspective. They both make too much noise, and dogs are so undignified... and chimps may think humans are deficient for not being able to swing on vines using their feet, etc.

Originally Posted by Pursuivant

That way you could have a "heroic" cat who's strong (ST 2.4) and smart (IQ 5.3) compared to other cats, but a run-of-the-cattery moggie with respect to humans.

The same concept might apply to HP and FP, but that also requires the concept of fractional points of damage, which would make things really complex.

Yeah... it depends on how interested you are in tracking the details of animal activities. Makes complete sense for Bunnies & Burrows or play that's really interested in what happens between animals, either because of caring, or because it's really important (messenger animal, pack dogs, human shapeshifted into a small animal, high-stakes animal betting, whatever).

Originally Posted by Pursuivant

Given that a score of 12 is considered to be "professional" level in GURPS, it makes sense to give most animals an overall Per of 12, and then make whatever IQ-based skills they might have substitute to Per.

For example, a wolf with Tactics (Pack Hunting) uses Per rather than IQ.

Will should probably be set at 10, lower for noticeably distractable or tractable animals, lower for noticeably perseverent creatures.

GURPS Bestiary did a pretty good job of listing appropriate base stats for perception and skills of various species.

[strange7person]

Originally Posted by Maz

Aren't the problems you mention here, like tactics covering crossfire, traps covering springloaded traps and so on, fixed by the fact that animals are TL:0 ?
Possibly introduce a TL:-1 (no tools at all, not even a stick) for most animals... but not the crow.

Nest-weaving birds, aphid-herding ants, anvil monkeys, and so on could then have a Cutting-Edge Training perk to represent their TL 0 techniques in a broader capacity than One-Task Wonder.

[Flyndaran]

None of them are anywhere near TL 0. Stone age people were MUCH more advanced than any non-human.

Negative TLs may not be canon, but cutting edge training to make some aspect TL-1 makes sense.

[David L Pulver]

Originally Posted by Flyndaran

None of them are anywhere near TL 0. Stone age people were MUCH more advanced than any non-human.

Negative TLs may not be canon, but cutting edge training to make some aspect TL-1 makes sense.

For tools, the distinction is often made as follows:

* A fair number of animals _use_ found objects or living things as tools.

* A very few smarter animals, e.g., corvids, directly modify found objects for use as tools.

* Humans modify objects to make tools and then use these tools to create other quite different tools, e.g., using a hammer stone as a tool to make a flint hand ax, and using the ax to carve something else, and so on.

Perhaps only the later is qualified as "sapient" and what TL 0 represents.

Some experiments seem to indicate that smart apes can be talk by example to proceed to using simple tools to make other tools, but don't do so on their own, or go on to use the tool they've been taught to make for different purposes than the one they were originally shown how to do.

[Flyndaran]

Cognitive bootstrapping using tools or language seem the one real human unique trait.

[Pursuivant]

Originally Posted by Skarg

It's both a low-grain scale, and a human-centric one. It also tends to favor rating how trainable an animal is to be gotten to do things humans think are interesting.

From the study I've done, it's clear to me that animals are really "smart" at doing the things they do every day.

For example, sheep are "smart" when it comes to remembering territory, fellow flock members, and ability to recognize different quantities.

Outside of their specialized skill sets, most animals, even humans, are "dumb." (After all, how well can you track by scent?)

Ethologists are still struggling to figure out how to measure animal IQ, especially for non-terrestrial or non-mammalian species.

For GURPS, we can disagree all day on what exactly IQ measures. For purposes of animal IQ, I took it to be a combination of neural complexity, learning speed with various tasks, historical and anecdotal reports of clever animal behavior in the wild, and animal trainers' reports of how easy various animals are to train.

Of course, animal trainers' reports are complicated by the fact that most animals have some form of the Greed advantage. Once a trainer figures out what an animal likes to eat, training becomes much easier and faster, limited only by the animal's attention span, distractibility, and patience.

Certain animals which have a reputation for being "stupid" or "stubborn" due to lack of trainability might actually be smart enough to recognize that what they're being asked to do is BS and refuse to go along with the program. (Cats, goats, and certain octopi might fit into this category.)

[Pursuivant]

Originally Posted by Flyndaran

Cognitive bootstrapping using tools or language seem the one real human unique trait.

Agreed about "cognitive bootstrapping", although some animals appear to have limited ability to imagine future scenarios and plan accordingly.

Tool-using animals are good at finding and modifying simple objects in their environment to achieve a single task, not so good at making Tool A to get the materials to make Tool B, and so forth.

Language is more of a debatable area. There's some very good evidence that certain animals have the capacity to recognize individual spoken words (gaining a vocabulary of hundreds to even thousands of words), and a few have actual, limited, ability to produce words in a single language and use them in a creative fashion. A rare few might have the ability to produce very simple sentences.

These animals are mostly highly-trained "living experiments" who easily qualify for an Unusual Background in GURPS terms, but there are a few cases where wild animals might actually be learning or using languages. For example, chimpanzee vocalizations vary slightly from troop to troop, and who knows what cephalopods are communicating via their chromatophores.