Regarding the evolution of emotions, consider this:
Imagine a group of life forms of the same species who compete for resources. Lets say that either they are fairly even in power level, and thus it is superior for them to cooperate with each other and divide resources fairly to avoid wasting energy fighting. Alternately, some (alphas) are superior in power level, but the game theoretically optimal outcome is for the more dominant to take a larger share of resources, but still allow the others to have some. (This is superior for them to fighting to the death to try to get everything).
If life forms cannot communicate with each other, then they suffer from a prisoner’s dilemma problem. They would like to cooperate in the prisoners dilemma, but without the ability to signal to each other, they cannot be sure the other will not defect. Thus they end up defecting against each other.
We can thus see that if the life forms evolved methods of signalling to each other, it would improve their chances of survival. Thus, complex life forms develop ways of signalling to each other. We can see many, many examples of this throughout a broad range of life.
The life forms also need to be able to mentally model each other now to predict each other’s actions. Thus they develop empathy (the ability to model each other), and emotions, which are related to signalling. If each of the members of the species have similar emotions, that is , they react in similar ways to a given situation, and they have developed ways of expressing these emotions to each other, then it greatly improves the abilities of the life forms to model each other correctly. For example, if one life form then tries to gain an unfair distribution of resources (defecting in the prisoners dilemma problem), the other will display the emotional response of anger. They are signalling ‘you cannot defect against me in such an unfair way. Because you are attempting to do so, I will fight you’.
Because the emotional responses occur automatically, they act similar to a precommitment. Essentially, the life form having the emotional response has been preprogrammed to have this response to the situation. This is a precommitment to a course of action, which can help the life form to achieve a better game-theoretic result.
(For example, if we are playing ‘chicken’, driving our cars toward each other on a road, the best strategy for me to win the game is to visibly precommit to not swerving no matter what. Thus, the optimal strategy would be to remove my steering wheel and throw it out of the car in a way that you can see. Since you now know that I CANNOT turn, you must turn to avoid crashing, and I win the game. This shows how a strong precommitment can be an advantage).
If we think of emotions as precommitments in this way, we can see that they can give us an advantage in our prisoner’s dilemma problem. The opponent then knows that they cannot defect against us too much, or else we will become angry and will fight, even though this gives a worse outcome for us as well, it is an emotional response and thus we will do it automatically.
We can see that emotions are thus an aid to fitness, and life forms that evolve them will have a genetic advantage.
Now imagine that the problems that the life forms need to signal about become much more complex. Rather than just signalling about food or mates, for example, they need to signal about group dynamics, concepts like loyalty to a tribe, their commitment to care for young, etc.
Under these circumstances, we can see the need to develop a much greater range of emotions to deal with various situations. We need to display more than just anger over an unequal distribution, or fear, or so on. We also need to have emotional responses of love, loyalty, and so on, and be able to demonstrate/signal these to each other.
This is my understanding of how emotions should evolve among complex life forms that are at least remotely close to humans. Perhaps for extremely different life forms, emotions would not evolve. Or perhaps in pretty much all complex forms of life which are communicating with each other, some form of emotions would evolve, I don’t know. I don’t see the need for a magical sky-father to have desired emotions to exist, who then guided reality to this point.
Also, I am not sure how the universe-creator would accomplish this. If the degrees of freedom they have are to create the underlying fundamental laws of physics, and the initial conditions of the universe, how would they be able to compute out what sets of laws/initial conditions would lead to the physics which would lead to the chemistry which would lead to the biology which would lead to the life forms which would have these emotions? And why would that be the goal of the universe-creator? In the absence of evidence which would distinguish this hypothesis from others, I don’t see why we should privilege this hypothesis to such an extent, when it is pretty clear that the real reason for “believing” in it is actually “I want it to be true”. (And also “A strong meme which many people believe in is threatening that I will be harmed if I do not “believe” that this is true, and promising to reward me if I do “believe” it is true).
If anyone can correct my thoughts regarding evolution and emotions, or can point to some studies or scientific theories which either support or refute this post, I would love to read them!
Thanks for the welcome, and thanks for sharing your thoughts! I love game theory, and all your connections look good to me.
The life forms also need to be able to mentally model each other now to predict each other’s actions. Thus they develop empathy (the ability to model each other), and emotions
This is something I still don’t understand very well about evolution. They need it, and therefore they develop it? Is there anything that leads them to develop it, or is this related to the “evolving to extinction” chapter? I should go back and re-read the chapters on evolution. Is this something you can somewhat-briefly summarize, or would understanding require a lot more reading?
They need it, therefore if it randomly happens, they will keep the outcome.
Imagine a game where you are given random cards, and you choose which of them to keep and which of them to discard. If you need e.g. cards with high numbers, you can “develop” a high-numbered hand by keeping cards with high numbers and discarding cards with low numbers. Yet you have no control over which cards you receive. For example, if you have bad luck and always get only low numbers, you cannot “develop” a high-numbered hand. But only a few high numbers are enough to complete your goal.
Analogically, species receive random mutations. If the mutation makes the survival and reproduction of the animal more likely, the species will keep this gene. If the mutation makes the survival and reproduction of the animal less likely, the species will discard this gene. -- This is a huge simplification, of course. Also the whole process is probabilistic; you may receive a very lucky mutation and yet die for some completely unrelated reason, which means your species cannot keep that gene. Also, which genes provide advantage, that depends on the environment, and the environment is changing. Etc.
But the idea at the core is that evolution = random mutation + natural selection. Random mutation gives you new cards; natural selection decides which cards to keep and which ones to discard.
Without mutations, there would be no new cards in the game; each species would evolve to some final form and remain such forever. Without natural selection, all changes would be random, and since most mutations are harmful, the species would go extinct (although this is a contradiction in terms, because if you can die as a result of your genes, then you already have some form of natural selection: selecting for survival of those who do not have the lethal genes).
Sometimes there are many possible solution for one problem. For example, if you need to pick fruit that is very high on the trees (or more precisely speaking: if there is a fruit very high on the trees that no one is picking yet, so anyone who could do so, would get a big advantage), here are things that could help: a longer neck, longer legs, ability to jump, ability to climb trees, ability to fly, maybe even ability to tumble the trees. When you randomly get any card in this set (and it doesn’t come with big disadvantages which would make it a net loss), you keep it. Some species went one way, other species went other way. -- A huge simplification again, since you cannot get an ability to fly in one step. You probably only get an ability to climb a little bit, or to jump a little bit. And in the next step, you can get ability to climb a little bit more, or to jump a little bit more, or to somehow stay in the air a little bit longer after you have jumped. Every single step must provide an additional advantage.
This is something I still don’t understand very well about evolution. They need it, and therefore they develop it? Is there anything that leads them to develop it, or is this related to the “evolving to extinction” chapter?
I’m not a biologist or anything, but I think I’m competent enough to answer this question.
You’ll often see biology framed in teleological terms. That is, you’ll often see it framed in terms that seem to indicate that natural selection is purposeful, like a person or God (agent) would be. I’ll try to reframe this explanation in non-teleological terms. Animal husbandry/selective breeding/artificial selection is a good way to get an idea of how traits become more frequent in populations, and it just seems less mysterious because it happens on shorter timescales and humans set the selection criteria.
Imagine you have some wolves. You want to breed them such that eventually you’ll have a generation of wolves that will do your bidding. Some of the wolves are nicer than others, and, rightly, wolves like these are the ones that you believe will be fit to do your bidding. You notice that nice wolves are more likely to have nice parents, and that nice wolves are more likely to give birth to nice pups. So, you prevent the mean wolves from reproducing by some means, and allow the nicest wolves to mate. You re-apply your selection criterion of Niceness to the next generation, allowing only the nicest wolves to mate. Before long, the only wolves you have around are nice wolves that you have since decided to call dogs.
In artificial selection, the selection criterion is Whatever Humans Want. In natural selection, the selection criterion is reproductive fitness; the environment ‘decides’ (see how easy it is to fall into teleology?) what reproduces. Non-teleologically, organisms with more adaptive traits are more likely to reproduce than organisms with less adaptive traits, and therefore the frequency of those traits in the population increases with time. Rather than thinking of natural selection as ‘a thing that magically develops stuff,’ imagine it as a process that selects the most adaptive traits among all possible traits. So, we’re not so much making traits out of thin air as we are picking the most adaptive traits out of a great many possibilities. You didn’t magically imbue the wolves with niceness; niceness was a possible trait among all possible traits, and you narrowed down the possibilities by only letting the nice wolves mate, until at one point you only had nice wolves left.
Like the things that we discussed earlier in the welcome thread, teleological explanations of biology are artifacts of natural language and human psychology. Well before humans spoke of biology, we spoke of agents, and this is often reflected in our language use. As a result, it’s also often much more concise to speak in teleological terms. Compare, “Life forms need to be able to mentally model each other, and thus develop modeling software,” with my explanations above. Teleological explanations are also often a product of the aforementioned mental-modeling software itself, just as we have historically anthropomorphized natural phenomena as deities. Very importantly, accurate biological explanations that are framed teleologically can be reframed in non-teleological terms, as opposed to explanations that are fundamentally teleological.
Please feel free to ask questions if I didn’t explain myself well. And to others, please correct me if I’ve made an error.
No, you explained that really well!! Everything is a lot less fuzzy now! Thank you :) I think with science, the first time I read it, it makes sense to me, but I had such a bad habit of filing scientific facts into the “things you learn in school but don’t really need to remember for real life” category of my brain that now, even when I actually care about learning new information, it still takes multiple explanations, and sometimes one really good one like yours, before it really start to sink in for me.
Welcome to LessWrong, and thanks for posting!
Regarding the evolution of emotions, consider this:
Imagine a group of life forms of the same species who compete for resources. Lets say that either they are fairly even in power level, and thus it is superior for them to cooperate with each other and divide resources fairly to avoid wasting energy fighting. Alternately, some (alphas) are superior in power level, but the game theoretically optimal outcome is for the more dominant to take a larger share of resources, but still allow the others to have some. (This is superior for them to fighting to the death to try to get everything).
If life forms cannot communicate with each other, then they suffer from a prisoner’s dilemma problem. They would like to cooperate in the prisoners dilemma, but without the ability to signal to each other, they cannot be sure the other will not defect. Thus they end up defecting against each other.
We can thus see that if the life forms evolved methods of signalling to each other, it would improve their chances of survival. Thus, complex life forms develop ways of signalling to each other. We can see many, many examples of this throughout a broad range of life.
The life forms also need to be able to mentally model each other now to predict each other’s actions. Thus they develop empathy (the ability to model each other), and emotions, which are related to signalling. If each of the members of the species have similar emotions, that is , they react in similar ways to a given situation, and they have developed ways of expressing these emotions to each other, then it greatly improves the abilities of the life forms to model each other correctly. For example, if one life form then tries to gain an unfair distribution of resources (defecting in the prisoners dilemma problem), the other will display the emotional response of anger. They are signalling ‘you cannot defect against me in such an unfair way. Because you are attempting to do so, I will fight you’.
Because the emotional responses occur automatically, they act similar to a precommitment. Essentially, the life form having the emotional response has been preprogrammed to have this response to the situation. This is a precommitment to a course of action, which can help the life form to achieve a better game-theoretic result.
(For example, if we are playing ‘chicken’, driving our cars toward each other on a road, the best strategy for me to win the game is to visibly precommit to not swerving no matter what. Thus, the optimal strategy would be to remove my steering wheel and throw it out of the car in a way that you can see. Since you now know that I CANNOT turn, you must turn to avoid crashing, and I win the game. This shows how a strong precommitment can be an advantage).
If we think of emotions as precommitments in this way, we can see that they can give us an advantage in our prisoner’s dilemma problem. The opponent then knows that they cannot defect against us too much, or else we will become angry and will fight, even though this gives a worse outcome for us as well, it is an emotional response and thus we will do it automatically.
We can see that emotions are thus an aid to fitness, and life forms that evolve them will have a genetic advantage.
Now imagine that the problems that the life forms need to signal about become much more complex. Rather than just signalling about food or mates, for example, they need to signal about group dynamics, concepts like loyalty to a tribe, their commitment to care for young, etc.
Under these circumstances, we can see the need to develop a much greater range of emotions to deal with various situations. We need to display more than just anger over an unequal distribution, or fear, or so on. We also need to have emotional responses of love, loyalty, and so on, and be able to demonstrate/signal these to each other.
This is my understanding of how emotions should evolve among complex life forms that are at least remotely close to humans. Perhaps for extremely different life forms, emotions would not evolve. Or perhaps in pretty much all complex forms of life which are communicating with each other, some form of emotions would evolve, I don’t know. I don’t see the need for a magical sky-father to have desired emotions to exist, who then guided reality to this point.
Also, I am not sure how the universe-creator would accomplish this. If the degrees of freedom they have are to create the underlying fundamental laws of physics, and the initial conditions of the universe, how would they be able to compute out what sets of laws/initial conditions would lead to the physics which would lead to the chemistry which would lead to the biology which would lead to the life forms which would have these emotions? And why would that be the goal of the universe-creator? In the absence of evidence which would distinguish this hypothesis from others, I don’t see why we should privilege this hypothesis to such an extent, when it is pretty clear that the real reason for “believing” in it is actually “I want it to be true”. (And also “A strong meme which many people believe in is threatening that I will be harmed if I do not “believe” that this is true, and promising to reward me if I do “believe” it is true).
If anyone can correct my thoughts regarding evolution and emotions, or can point to some studies or scientific theories which either support or refute this post, I would love to read them!
Thanks for the welcome, and thanks for sharing your thoughts! I love game theory, and all your connections look good to me.
This is something I still don’t understand very well about evolution. They need it, and therefore they develop it? Is there anything that leads them to develop it, or is this related to the “evolving to extinction” chapter? I should go back and re-read the chapters on evolution. Is this something you can somewhat-briefly summarize, or would understanding require a lot more reading?
They need it, therefore if it randomly happens, they will keep the outcome.
Imagine a game where you are given random cards, and you choose which of them to keep and which of them to discard. If you need e.g. cards with high numbers, you can “develop” a high-numbered hand by keeping cards with high numbers and discarding cards with low numbers. Yet you have no control over which cards you receive. For example, if you have bad luck and always get only low numbers, you cannot “develop” a high-numbered hand. But only a few high numbers are enough to complete your goal.
Analogically, species receive random mutations. If the mutation makes the survival and reproduction of the animal more likely, the species will keep this gene. If the mutation makes the survival and reproduction of the animal less likely, the species will discard this gene. -- This is a huge simplification, of course. Also the whole process is probabilistic; you may receive a very lucky mutation and yet die for some completely unrelated reason, which means your species cannot keep that gene. Also, which genes provide advantage, that depends on the environment, and the environment is changing. Etc.
But the idea at the core is that evolution = random mutation + natural selection. Random mutation gives you new cards; natural selection decides which cards to keep and which ones to discard.
Without mutations, there would be no new cards in the game; each species would evolve to some final form and remain such forever. Without natural selection, all changes would be random, and since most mutations are harmful, the species would go extinct (although this is a contradiction in terms, because if you can die as a result of your genes, then you already have some form of natural selection: selecting for survival of those who do not have the lethal genes).
Sometimes there are many possible solution for one problem. For example, if you need to pick fruit that is very high on the trees (or more precisely speaking: if there is a fruit very high on the trees that no one is picking yet, so anyone who could do so, would get a big advantage), here are things that could help: a longer neck, longer legs, ability to jump, ability to climb trees, ability to fly, maybe even ability to tumble the trees. When you randomly get any card in this set (and it doesn’t come with big disadvantages which would make it a net loss), you keep it. Some species went one way, other species went other way. -- A huge simplification again, since you cannot get an ability to fly in one step. You probably only get an ability to climb a little bit, or to jump a little bit. And in the next step, you can get ability to climb a little bit more, or to jump a little bit more, or to somehow stay in the air a little bit longer after you have jumped. Every single step must provide an additional advantage.
Yes this. Of course it is not a given that something that would be a useful adaptation will develop randomly.
Great analogies with the hand of cards.
Ditto to Ander’s comment—very nice summary and analogy, many thanks :)
I’m not a biologist or anything, but I think I’m competent enough to answer this question.
You’ll often see biology framed in teleological terms. That is, you’ll often see it framed in terms that seem to indicate that natural selection is purposeful, like a person or God (agent) would be. I’ll try to reframe this explanation in non-teleological terms. Animal husbandry/selective breeding/artificial selection is a good way to get an idea of how traits become more frequent in populations, and it just seems less mysterious because it happens on shorter timescales and humans set the selection criteria.
Imagine you have some wolves. You want to breed them such that eventually you’ll have a generation of wolves that will do your bidding. Some of the wolves are nicer than others, and, rightly, wolves like these are the ones that you believe will be fit to do your bidding. You notice that nice wolves are more likely to have nice parents, and that nice wolves are more likely to give birth to nice pups. So, you prevent the mean wolves from reproducing by some means, and allow the nicest wolves to mate. You re-apply your selection criterion of Niceness to the next generation, allowing only the nicest wolves to mate. Before long, the only wolves you have around are nice wolves that you have since decided to call dogs.
In artificial selection, the selection criterion is Whatever Humans Want. In natural selection, the selection criterion is reproductive fitness; the environment ‘decides’ (see how easy it is to fall into teleology?) what reproduces. Non-teleologically, organisms with more adaptive traits are more likely to reproduce than organisms with less adaptive traits, and therefore the frequency of those traits in the population increases with time. Rather than thinking of natural selection as ‘a thing that magically develops stuff,’ imagine it as a process that selects the most adaptive traits among all possible traits. So, we’re not so much making traits out of thin air as we are picking the most adaptive traits out of a great many possibilities. You didn’t magically imbue the wolves with niceness; niceness was a possible trait among all possible traits, and you narrowed down the possibilities by only letting the nice wolves mate, until at one point you only had nice wolves left.
Like the things that we discussed earlier in the welcome thread, teleological explanations of biology are artifacts of natural language and human psychology. Well before humans spoke of biology, we spoke of agents, and this is often reflected in our language use. As a result, it’s also often much more concise to speak in teleological terms. Compare, “Life forms need to be able to mentally model each other, and thus develop modeling software,” with my explanations above. Teleological explanations are also often a product of the aforementioned mental-modeling software itself, just as we have historically anthropomorphized natural phenomena as deities. Very importantly, accurate biological explanations that are framed teleologically can be reframed in non-teleological terms, as opposed to explanations that are fundamentally teleological.
Please feel free to ask questions if I didn’t explain myself well. And to others, please correct me if I’ve made an error.
No, you explained that really well!! Everything is a lot less fuzzy now! Thank you :) I think with science, the first time I read it, it makes sense to me, but I had such a bad habit of filing scientific facts into the “things you learn in school but don’t really need to remember for real life” category of my brain that now, even when I actually care about learning new information, it still takes multiple explanations, and sometimes one really good one like yours, before it really start to sink in for me.