How could humans recognize another species as sentient / intelligent?

Question

There is this fantasy in fiction that humans meet another species and figure out that it's intelligent. This usually happens with dragons, dolphins, aliens, collectively intelligent plancton...

But sentience could theoretically be so different from what humans experience.

Let's put aside the fact that even recognizing life might be complicated. Say tomorrow a group of scientists encounter in a previously unexplored part of the world (mountain top, ocean bottom, innermost forest) a small animal. How would they recognize that it is an intelligent sentient being assuming that it actually is?

Follow-up question (or prequel to this question): What more than sentience do you need from a species for its intelligence to be recognizable by humans?

Answer 1

Besides all the moral relativism that permeates post-modernist philosophy shown on this topic, I will risk writing a more traditional, dialectics based answer.

What is intelligence?

Well, lets do a dialectical analysis of evolution of life in our planet.

A mineral/chemical world is one where all the observable phenomena is the result of physical or chemical interaction between substances. Energy tends to fall and the entropy of the universe tends to grow. Biological life starts out of this chemical world because it is a superior form of chemistry. We might say that biological life makes a better use of chemical energy, hitchhiking the energy and the general increase in entropy of the universe towards its own perpetuation. Once life sets in, competition starts and results in a trait: to preserve itself, life evolves from more concrete, hardwired logic, towards abstract reasoning.

If two lifeforms are intermingled in a pattern of predation, or other form of competition, both will evolve due to limited resources. From chemical warfare of single cells, to group defense of multicellular organisms. As soon as a limit is reached a very "creative" adaptation sets in.

For example, when a cheetah hunts a thompson's gazelle, both evolve. Soon we reach a point where evolution is not possible anymore, and the genetic pool stabilizes. If a cheetah was to evolve indefinitely (in a specific niche that does not change) it would soon reach a point where adding muscle mass would decrease its speed. In other words, certain traits (in this case, speed) are limited by the composition (the chemical makeup) of normal life forms. If a cheetah was going to be faster, he would need to be made from much different chemicals, like iron bones, or something like that, which is much more costly to be used. The result is that this "impasse" is solved dialectically by a lateral operation. In the case of the primates, that lateral evolution was the increase of intelligence. Primates are not the utmost predators of nature. They are not stronger than lions, nor faster than cheetahs, et cetera. But they use their intelligence to hunt in groups using lances and other instruments.

This means that as soon as a limit is reached at a level, a certain new trait differentiates one group of species from others. For one, the beasts are still stronger physically than humans, but we reached a technological level that allows us to dominate all beasts without too much problem (and we've done that).

But why do life tries to perpetuate itself? Because if you have a primitive world where there are life forms that don't try to preserve their own existence, this non-self preserving life will be killed and replaced by any life form that developed a trait of self preservation. As soon as the ambient is totally filled with life, competition will set in, and self-preservation will become the utmost objective.

Even if we are talking about another planet, and our curiosity and romanticism wants to see something different (to allow us to dream with a universe where men did not dominate almost everything, and there is nothing in ordinary life out of our domain) this another planet is in the same universe as ours. It's composed of atoms just like ours, and those atoms make molecules. Provided that some differences might come from different makeovers of such planets (silicon vs carbon, etc.), those differences pale in comparison to basic similarities, like the need for energy, the competition, etc.

So, whats the purpose of intelligence?

Animals, plants and similar beings evolve by changing their own genome to adapt to circumstances. Provided a fixed habitat like I said in the early part of this text, animals would stagnate in ideal forms without reason to change. Our habitats are not fixed, but changing continuously. This means that animals must adapt out of natural selection, which is slow, and while, usually, able to save species from extinction, natural selection cannot save specific individuals from death out of inadequacy.

As soon as an animal is able to adapt the environment to his own needs, on a much larger scale than most animals, this single specific animal can save not only his species but himself from death out of changes in the ambient. The intelligence allows a much more flexible approach to competition with other animals. Instead of death of the weak and survival of the fittest, we get power over the ambient to change it as needed, and the more we evolve intellectually, the more we can adapt the ambient to our tastes.

So, regarding your question about how to detect intelligent/sentient life, we might answer the following:

1. Sentience can be detected out of the animal capability to react quickly and unambiguously to outside stimuli. A plant can react to its external environment, but not quickly or unambiguously. A sentient animal will have a basic reaction called fear. The need to save his own life at any cost.
2. Intelligence is a superior form of sentience. The animal at hand will show capability to change its environment in a non-spontaneous way in order to fit it to its own needs. We call that adaptation of the environment into human made forms "anthropized environment", or the geographical space unambiguously changed from natural to man made.

TL;DR

Life evolves intelligence to allow better adaptation and survival of the fittest. Intelligence allows changing the environment to something that suits the intelligent life-form better. This means that a intelligent/sentient life form will show a trait of environment manipulation that can range from simple tools and hunting tactics to terraforming and exploration of extreme energy forms.

Answer 2

Defining intelligence is tricky. Here's my take on it:

Assumption 1: Intelligence is a label applied by humans based on observation, not an a-priori trait of an entity.

I choose this assumption because I want a science bound answer. Religious answers are free to use a different definition, based on whatever deity derived divinations they please. Intelligence seems less to define the creature, and more defining how we treat that creature. We give more respect to intelligent things.

Assumption 2: Highly predictable things are not intelligent

This is a carefully worded assumption. Obviously I want to come to the result that humans are intelligent and rocks are not. If I throw a rock, and it lands where I expected it, it clearly did not act unpredictably on the way.

I am carefully not assuming that all unpredictable things are intelligent, although I think there is a fine line of reasoning that goes down that path.

Assumption 3: Intelligent things stay around for a while

Highly unpredictable events like explosions are cool, but they soon flare out. Their action eventually extinguishes that which they need to live, and they stop exploding.

Because this is a human defined trait (assumption 1), this means their lifespan needs to be on an order where humans can observe their behaviors long enough to decide they are intelligent. It also means that very long lived intelligences might appear unintelligent because we lack the patience to see them.

---

Results of these assumptions

It is clear from the predictable thing that intelligent things must have some "internal state," meaning information that is not easy to acquire from them. If you could know everything a person was going to say or do for the rest of their life, they would not seem very intelligent to you.

To be long lasting, that internal state must not be a limited source. Consider a firecracker fountain, with lots of surprises, but when it's over, you know everything about the hull that remains. I think Chaos theory offers great potential for small internal perturbations to have tremendous outward outcomes. It also has the advantage of making it very hard to measure those internal perturbations without killing the creature.... meaning its internal state could be, indeed, unknowable through violence.

I would expect a model of this creature-under-test to appear to be very broadband. This is an artifact of the definition. It is unlikely that the creature's lifespan and daily life matches pace with ours, so we will have trouble observing all of the creature's actions. Accordingly, any creature that actually seems to have internal state on our timescales is likely doing everything from long slow life goals to ultra-fast snap decisions.

The test

The big test I see for testing intelligence is interaction. If we observe the creature silently, without interacting, we can build up a model of how they interact without us. If we then visibly observe it, interacting freely, if they seem to change around us, there is a good chance they should be treated as intelligent.

The neat thing about this test: it also forces us to allow them to decide if we are intelligent, by their definition.

Answer 3

From the wikipedia page on "Animal consciousness"

In 2004, eight neuroscientists felt it was still too soon for a definition. They wrote an apology in "Human Brain Function":[38]

"We have no idea how consciousness emerges from the physical activity of the brain and we do not know whether consciousness can emerge from non-biological systems, such as computers... At this point the reader will expect to find a careful and precise definition of consciousness. You will be disappointed. Consciousness has not yet become a scientific term that can be defined in this way. Currently we all use the term consciousness in many different and often ambiguous ways. Precise definitions of different aspects of consciousness will emerge ... but to make precise definitions at this stage is premature."

As far as I know, this is still true today.

I would personally define the type of consciousness humans have (vs animals) as:

• ability to recognize oneself and others
• ability to comprehend abstract ideas
• ability to have a "thought-process" - resulting in the ability to "predict" outcomes
• ability to learn through forms of communication rather than only personally-experienced stimuli
  • This one is difficult for me to define, communication could be interpreted as form of stimulus. What is really meant is that dogs would not pass this; one dog cannot share his knowledge of tricks with another dog. Each dog must individually learn through his own training.
• ability to analyze the results of ones actions and choose different actions in the future

Many of those could be applied to some animals, I think the biggest things that make humans unique is the ability to comprehend abstract ideas and being able to learn through communication.

---

Most of the time, it seems to me that when you would want to define it in this way, you would want to be able to speak to the other creature and have it understand you. You would want to be able to tell it things, and have it learn from your experiences, and you would want it to be able to tell you things, so that you can learn from it.

As for How would they recognize that it is an intelligent sentient being assuming that it actually is

Assuming the above, the scientists could observe behavior changes (learning) being passed from one of the animals to others. They could recognize the animals attempting to communicate with them and convey ideas. Likewise, trying to convey ideas to the animals should be met with an attempt to understand.

You wouldn't be able to tell right away. You'd have to start small and slowly build signs of matching the above criteria. It would take time and effort from both sides.

Answer 4

It depends on how you define sentient.

Dolphins, for example, are sentient according to some definitions of the term, as are chimpanzees. However, we don't interact with those kinds of animals in the same way that people interact with elves in a fantasy setting. If what we're after is the level of sentience that something like a dolphin or a monkey exhibits, there's a ton of tests for various forms of animal behavior that have been developed. Some of these tests, such as the mirror test, have been put forth in other answers as a test for sense of self.

By many definitions, this defines sentience. However, some intelligent birds, like magpies, can pass this test, and they aren't what we typically want to define as a sentient species.

According to the test you choose, different animals that exist on earth could be defined as sentient. Dogs, magpies, dolphins, crows, and elephants can all be defined as such by various tests. Some scientists would argue in favor of each of these being sentient. If a certain group of these scientists were to discover a new marmot on a far off mountain top, it would not have to be discernibly different than other marmots for them to describe it as sentient.

Of course, just because a small group of scientists says a marmot is sentient doesn't mean that everyone else will look at it as a dumb animal. A better question to ask might be:

What traits of intelligence must a creature possess to not be widely viewed as a dumb animal?

This, of course, won't apply to all people, just most of them. There have been, at many points in human history, lots of groups of people that view other groups of people as dumb animals. We can still look to see what traits a creature would need to have in order for most people to think of them as being intelligent.

In my opinion, the main (and perhaps only) question that must be answered is this: can the other life form force us to regard them as (at least) intellectual equals? There are two main ways that a creature could go about doing this: intelligible speech and development of technology. Of these, the latter is far more important.

Intelligible speech

If an organism has a language that we can understand which uses abstract representation of concepts, we're likely to regard them as being intelligent. This, however, is no guarantee. Dolphins and whales, for example, may have the capacity for abstract speech, but we can't understand them. Because of this, we have no real need to treat them as equals. A dolphin won't swim up to a fishing boat and tell it to stop killing dolphins in a way the fishing boat will understand, so the fishers can keep thinking of dolphins and whales as big dumb fish and spearing them for meat.

Development of technology

Of course, if those same dolphins were to swim up to the fishing boat with machine guns and open fire, we'd probably have to think differently. Tool use is one of the primary ways that we as humans differentiate ourselves from animals, so it's likely that an other creature that we encountered using and developing tools would be viewed by most people as intelligent, even if those tools were at the level of bows and arrows or primitive stone knives.

The important thing, though, is the development and progression of technology, not merely the use of a rock or stick as a tool. A crow that uses a rock to smash a crab is a smart bird. A dolphin with a rocket launcher, on the other hand, demands respect as an equal.

Answer 5

To determine if something is intelligent, and especially how intelligent, you may consider running tests and making a lot of observations.

The Mirror Test

This is, perhaps, one of the more famous ways to determine if an animal is intelligent. It basically amounts to putting a mirror in front of the creature (or within its habitat) and seeing what it does to its own reflection.

The simple idea is that a creature who "passes" the mirror test will show behaviors that it recognizes itself in a mirror. There are various derivations of this test, but it helps set a high standard for "intelligent."

Learned Behavior

Is this thing capable of learning from its experiences or from observation? Some people even use this to claim that plants (and slime molds) are intelligent. Even so, you can look at what Joshua Klein did with crows in his TED talk.

This may set the standard pretty low, though, as you can get a robot to do these things, but not actually have a self-aware robot. That being said, testing to see if there are any learned behaviors can apply to a very wide range of creatures, from many different kingdoms.

Tool Use

This indicated at least some level of intelligence. Does the creature in question use tools? Use of tools shows that the creature at least can recognize beneficial items and figure out which ones are better than others.

Culture

Does this creature have culture? That's a tough one to figure out, too, but it's an indication of intelligence. You're looking for art, social hierarchies, music, dance, etc.

Answer 6

Only one test needs to be administered. Is it a living thing?

Anything else assumes we are able to determine all forms of communication and or know what anything is thinking at a given time.

If we meet a bipedal that communicates with pheromones we are unable to detect make it not intelligent/sentient or does the mere physiological fact it is bipedal make it so?

If we place a life form in a room with a mirror and it does nothing mean anything or does it know it is confined and waits to see what develops. What if the same life form breaks the mirror, does it think it is attacking another creature or does it think the mirror is the weak link of its prison and can escape.

Does something that is able to grow in size or replicate make it intelligent or sentient? A crystal accumulating mass that shatters and makes smaller versions of itself alive?

The use of tools a valid test of intelligence or sentient life? An otter would fit this criteria by using rocks to open mussels but wolves who hunt in packs and drive prey to other wolves waiting to ambush not?

The ability to adapt to its environment and propagate? So if a catastrophe kills all land based life forms on earth and only sea life exist are humans no longer sentient or intelligent?

The only thing these test and restrictions do is feed the human ego to make it feel superior and special.

The only test needed is is it living.

Answer 7

I think the other suggestions here are making this way more complicated than it needs to be. I would use a simple mathematical test. Something like 3 + 2 = 5. So for example I could use small pebbles or twigs to make a stack of three objects and a stack of two objects, then draw arrows in the dirt from each the two stacks to another spot, where I then place five objects. Then do the same with 1+6 = 7 or something. Finally I would set up stacks with 3 + 3, draw the arrows, and let the creature fill in the stack of 6. If it does so correctly, it clearly has substantial intelligence; not only does it understand basic math, but it can correctly appraise a non-natural scenario and reason out a rudimentary form of communication.

Obviously, while 'passing' this test would indicate intelligence, failing it does not automatically imply the reverse. I can think of three possible reasons to this wouldn't work:

a) The creature doesn't understand mathematical logic. In this case, I would argue that it is not intelligent

b) The creature understands logic, but does not understand the communication aspect, i.e. that I want it to fill in the final stack. I really don't think this is likely. The only symbol I use is an arrow to indicate direction, which I imagine is pretty clear to any reasonable being, especially if it watches me draw it. With appropriate body language (i.e. backing off to indicate that it is the creatures turn to finish the third problem) I can't really see the communication aspect failing. Even if it does, if the creature is sitting there watching me work, then trying things of its own, mimicking etc, that is clearly not normal 'natural' behavior.

c) The creature understands logic and communication, but for whatever reason is not interested in demonstrating its intelligence to me. I can't really imagine an intelligent creature not being interested in meeting a different intelligent species, especially if they have come to our planet (or we have gone to theirs). But if it is completely disinterested in communicating with me, I suppose all I could do to determine its cognizance is observe it for a while, like the other responses suggest.

Answer 8

This question was the core subject of H. Beam Piper's "Fuzzy" novels, the first two of which (Little Fuzzy and Fuzzy Sapiens) are now public domain and fairly easy to find in e-book format. The universe has its fair share of anachronisms; humans have interstellar travel, antigravity, video communication, the ultimate in lie detection, and highly reusable ranged stunning weapons, but the characters smoke like chimneys, drink like Europeans, still use chemical film, magnetic audio tape and room-sized computers with the processing power of a desk calculator, and powder firearms are still the predominant weapons. Still, the series is a cult classic in science-fiction.

The plot of the first book centers around Jack Holloway, a freelance prospector on the newly-colonized planet of Zarathustra, and a bipedal quasi-humanoid creature about two feet tall, about twenty pounds, and covered in golden fur, who wanders into Jack's camp one fateful day. The creature seems highly intelligent, in fact much more intelligent than other animals on the planet; "Little Fuzzy" is able to problem-solve, figuring out in a matter of seconds how to unscrew a jar lid and then screw it back on. That indicates reasoning ability, and also the ability to conceive of right and left, concepts not directly relating to an object, and that's abstract thinking. Little Fuzzy also demonstrates the ability not only to use tools, but to make them, and to make tools that make tools, again demonstrating a degree of abstraction of thought. However, he seems unable to speak, instead producing high-pitched "yeek" sounds, and he seems to have no innate knowledge of how to produce fire, two qualities that are, as of the beginning of the story, the only definitive and legal proof of sapience.

The existence of Little Fuzzy (and as it turns out a whole population of "Fuzzies") on Zarathustra, along with the possibility they may be sentient beings and not just intelligent animals, is very disconcerting to the leadership of the Chartered Zarathustra Company, who was granted more or less complete ownership of Zarathustra as a "Class III uninhabited planet" to spearhead colonization and economic development. If the Fuzzies are a race of sentient beings, then Zarathustra is a "Class IV inhabited planet" and the company's charter isn't worth the parchment it's printed on. The company's CEO, it turns out, will do anything to prevent that.

Without going too much deeper into the story or spoiling it too much, the final definition of sapience arrived at by the characters in the book is multi-part and breaks down to the following:

• Conscious Thought - A sentient/sapient being has the ability to think as all animals do, and it is "conscious", as it is aware of itself and its surroundings, but more than that, a sapient being is aware of the fact that it is thinking, and knows what it is thinking about. While an animal is normally limited to dealing with the immediate, based on memory and instinct, a sentient being can think about things that are not occurring or directly related to the present situation. A high level of conscious decision-making can be observed in some animals when dealing with novel situations, which makes them readily able to learn how to deal with new situations they haven't seen before based on things they know from past experience. We ascribe to these animals a high level of intelligence, even if we wouldn't call them sentient. Sentient beings are contrasted by having more or less sustained conscious thought.

• Abstraction, Generalization and Classification - The awareness of "thought" itself is a concept that doesn't have a material component; humans know that we think with our brains, but a thought, as a concept, is classically nebulous and immaterial. It is "abstract". A non-sentient intelligence might encounter an object and make a simple, one-dimensional judgment about it; "food", "nesting material", "mate", "predator". This is all a non-sentient intelligence typically needs, or in many circumstances all it has time for.

  The sentient being has the ability, through sustained conscious thought, to begin with an object, say an apple, and methodically assign to it a quality or set of same that is not in itself any one material thing, but that might apply to many examples of objects, such as that they are red, round, edible, sweet and grow on trees. These many qualities form a "class" that can be used to identify an object as belonging to the class of apples because it has all the qualities, and to apply the qualities of an "apple" to an object known to belong to that class without having first-hand experience of that specific object's sense-qualities.

  This extends further to applying more or fewer of these qualities to larger or smaller groups of things, or to other abstract ideas; the class of apples, and all their examples, are also part of the larger superclass of "fruit", which grow on trees and are edible but aren't necessarily red (oranges), round (bananas) or sweet (lemons). The class of fruit in turn belongs to the larger class of "food", which is simply any edible, nutritious thing. This abstraction and classification occurs in the sapient mind almost without limit.

• Symbolization - Having formed these abstract concepts, it becomes necessary to represent them in some way. The representation of an abstract concept is a "symbol". These symbols have many forms, but of particular note are visual and audible symbols, which can be used to communicate abstract concepts to other sentient beings (of your race and others). The sentient being is a symbolizer, and a symbol communicator, able to use visual or audible cues as more than mere signals, but as material (if transient) representations of an abstract concept. Humans, as sentient beings, are so inured to thinking of things in terms of verbal symbols that we literally talk to ourselves in our own heads using a symbolic language, and it is difficult or impossible for us to think in any other way.

• Imagination - This last thing is a "more than the sum of its parts" combination of the three previous tenets. A sentient being, able to consciously think about abstract concepts in the form of symbols, can mentally conceive of situations the being has not experienced, by combining elements of things he has experienced with additional concepts symbolically communicated to him, to create a new situation which does not exist. If this situation is desirable, the sentient being is able to act consciously to change his actual environment to more closely or exactly match the conceptual one. If the situation is not desirable, the sentient being can take conscious action to ensure he never experiences it in reality. The sentient being, through his symbolization, can also communicate this hypothetical reality to other sentient beings, transferring the idea of the possible situation to them and allowing them to decide to help or to hinder his own efforts to realize or prevent it.

Answer 9

There are really three keys we could use:

Communication

We recognize communication as being an artifact of intelligence. If we can recognize how something communicates, and can manage a translation/conversation, then we know it's intelligent and can measure that intelligence in terms we can understand.

Tool Use

This is the other big one - we create and use ever more complex tools (this includes things as diverse as hammers, corn, and computers). If we see a species using tools, we can recognize it as intelligent based on those even if we can't necessarily communicate with it.

Fighting

Whether it's battles, warfare, or a hunt, these are all scenarios where we could recognize intelligence from adaptive tactics. If a creature shows it can out-think a human tactically, by changing how it approaches the problem, then we'd see it as intelligent even if it doesn't use tools and we can't communicate with it.

Answer 10

This question has actually already been answered in the domain of artificial intelligence. The Turing Test was proposed as a way to determine if an artificial system was intelligent. The modern analog of the test is basically, "Can anyone reliably tell which is a human and which is a computer while chatting over text messages (or any other chat system)?" At this point, no artificial system has passed an unrestricted Turing test but some have passed more restricted versions of it (bounding the domain or not telling the human that they're trying to tell if this is a computer).

As I said, this is focused on the domain of artificial intelligence created by humans, and one can imagine life forms that could not pass this test but would be significantly more intelligent than humans. However, I think it focuses on the most basic aspect of intelligence from the human perspective - the ability to communicate thought. Obviously, accommodations would need to be made if they communicated via hand gestures, etc., but that doesn't affect the basic premise of the test.

Language and communication is the most fundamental invention ever. It allows us to exist beyond ourselves and transfer knowledge between individuals. For a human to consider a species intelligent, I think we'd need to be able to communicate with them. We have been able to communicate with some individuals of animal species such as Koko the gorilla who has a relatively large vocabulary of sign language. While most people would consider gorilla's more intelligent than many species, we don't interact with them the way we do humans. I think it's safe to say that Koko could probably not pass an unrestricted Turing test even with accommodations for sign-to-text translation.