crank wrote:Hey Mr Samsa,
Is that Morticia in your avatar? It sure looks like an Addams:

No - the artist/author is inspired by gothic works, but it's not an Addams. It's from a comic called Johnny the Homicidal Maniac by Jhonen Vasquez (the same guy who created the Invader Zim television series). It's a dark humour comic, but enjoyable:

crank wrote:Anywho, we have that many traits thought to be genetic, are really some other mechanism-"a product of more general cognitive processes". This makes me think it difficult to impossible to untangle this stuff. Much of it could be a kind of cultural evolution, we do things because they work, do things often enough that rewires the brain. How do you get to a point to distinguish a brain is genetically wired to do 'A' and a genetically-wired brained rewired itself to do 'A' and doing 'A' rewired the genetic wiring?

You're right that this stuff is, in some cases, impossible to figure out and generally just incredibly difficult. For classically innate behaviors it can be a bit easier, for example, innate behaviors have a number of necessary characteristics which can easily be distinguished from other behaviors: they are rigid, universal across all individuals, automatic, elicited by a specific stimulus, etc. This means that if we raise individuals from the same species in a controlled environment, then they should all behave in the exact same way to the same stimulus, independent of any learning or environmental variables.

crank wrote:And what if the genetic-wiring is what causes the brain to make the culturally universal decisions that are "a product of more general cognitive processes"? If some genetically run, general optimization routine leads inevitably [pan-culturally] to behaviour 'A', though the situation is entirely environmental, that behavior is genetic or learned?

This is where it gets difficult, when we move away from the "classical" innate behaviors. These days a number of researchers argue that certain behaviors are a product of genetic predispositions - that is, we are likely to learn these behaviors. This can be incredibly difficult to figure out, but again what we do is isolate individuals and see what happens. An example of some good research in this area is the tool-use of New Caledonian crows. Initially it was believed that they created "fishing" tools from Pandanus leaves (which have little "hooks" from the barbs on the side of the leaf) instinctually. However, when the crows are isolated something interesting happens - they still try to create tools from the leaves to fish out grubs from holes in trees, but their tools are nearly invariably flawed or faulty. For example, they cut out sections of the leaf in a step-wise fashion to get into small holes whilst retaining the strength of the leaf at the base where it is held in the beak, but when the crows are isolated they cut it out (often) from the wrong side. This means that the hooks are now facing the wrong way and can't catch anything. It turns out that there is a mixture of genetic predisposition here as well as cultural learning. Interestingly, on the island of New Caledonia where they live, there are different "cultures" on the island - that is, different groups have developed different tools to solve the same problem and they pass on this design to their children through teaching.

When it comes to humans though, it obviously becomes a lot more difficult due to ethical reasons (we can't isolate them from cultural influences). This really should mean that we need to be a lot more tentative with our conclusions, but some poorer evo psychologists use it as an excuse to make wilder claims (e.g. "It's true that we can't demonstrate this to be true but that's just because we can't perform those experiments - we think the evidence is enough to make this Big Claim".

One of the important things to keep in mind is what science actually is. We aren't attempting to explain what is actually going on, but rather we're trying to figure out which explanation is the most useful to us. In other words, even if (in reality) it turns out that a behavior is evolved, it wouldn't matter if our learning theories gave a better explanation for the behavior (and vice versa). This is because science is simply the process of generating successful predictions, not measuring reality.

zoon wrote:

Mr.Samsa wrote:This is partially true, in that it is of course difficult to distinguish between innate and learnt behaviors in people, however, if someone makes the claims that "there are moral truths that are known simply by virtue of being human" then this is a very specific claim that needs to be supported. It specifically claims that there are moral truths which are innate.

It’s true, I wasn’t paying attention to the bit about “moral truths” in the OP. That would imply normative morality, thinking something is essentially-in-itself right or wrong, which I don’t agree with. I’m discussing descriptive morality, the rules people live by in practice.

Fair point, that's a much easier claim to defend with the evidence we have.

zoon wrote:

Mr.Samsa wrote:The problem here would be to demonstrate that this "moral truth" is a domain-specific adaptation, rather than the product of a more general ability - for example, if you present someone with a comparable non-moral situation like "Would you rather throw away 5 apples or 1 apple?" then invariably, across cultures, people will choose to save the 5 apples at the cost of losing 1 apple. This isn't because of some evolved "apple-saving morality", it comes from the simple fact that we understand basic numbers and, all things being equal, more is better than less.

This is the part of the scientific methodology that causes evolutionary psychology to be mocked by scientists - they start off well by finding a cultural universal trait, but instead of finding evidence that this cultural universal trait is innate or evolved, they just stop there and assert it anyway. Very often the evo psychologists fail to demonstrate that an evolved explanation is preferable to it being a product of a species-specific constraint.

Your remarks are entirely correct, as applied only to the first part of the trolley problem. Yes, indeed, choosing to save 5 people at the expense of 1 makes numerical sense, and doesn’t say anything specifically about social decision-making. If that was the whole of the trolley problem, it would be wholly uninteresting, for the reasons you give.

The interesting part comes with the second question, is it right to push the fat guy off the bridge to save 5 people by killing him? At that point, for most people, numerical sense goes out of the window. Most of us suddenly have a powerful intuition that it is not at all right to push the man to his death, even though the upshot, choosing to save 5 people and kill one, is the same as in the first scenario, where most of us are happy to flick the fatal switch. The underlying logic of this second decision is harder to analyse. Actually laying hands on someone to harm them becomes punishable in a way that indirect harm does not (even when the indirect harm has clearly been caused by a person), and this looks more like a specifically social response.

Indeed, but again this is still explained by domain-general rules and the contingencies underlying our behavior change depending on the context. That is, the salient aspect in the first example is the absolute numbers, whereas the salient aspect in the second example is a specific act that you need to perform. We see this kind of change in response as a result of "responsibility" in a number of areas, and not just moral ones. For example, in self-control experiments people can correctly identify which is the more rational option of $10 today, or $15 next week. However, when we change the situation to actually offer the money to the individual (rather than just an hypothetical situation), we find that their response changes to the impulsive act.

(And, of course, it would still be incredibly difficult to separate from cultural effects since a fundamental moral across all societies is "Don't kill people", which would interfere with our interpretation of these results).

zoon wrote:

Mr.Samsa wrote: It's actually a situation that is almost perfectly comparable to the supposed finding that there is an innate predisposition towards sniffing out cheaters in a population, known as the "cheater detection module". It was supposed to be innate because all people (even across cultures) made the same cognitive mistake in the Wason selection task, but when presented with content relevant to detecting cheaters, the success rate increased dramatically.

For years evo psychologists believed that it must have been an adaptation because it was culturally universal, however, it was eventually shown that this is unlikely - instead the evidence strongly suggests that it's a product of more general cognitive processes.

It’s been suggested (by Joshua Greene and others) that the trolley problem may be the result of two different parts of the brain being activated in the two different scenarios. In the first, it’s all about calculation, 5 against 1, choose 5. In the second scenario, pushing the man off the bridge, the calculation is overwhelmed by a more emotional response, in another part of the brain, prohibiting the causing of direct harm to another person.

Well I don't think there is any "may" about it - it necessarily would be the result of two different parts of the brain since it is two drastically different scenarios.

zoon wrote:If that is the case, it could be argued that this is a collision of two more general cognitive processes, rather than a specifically evolved response, but I don’t think it could be argued successfully that the two processes did not evolve, or that the second is not social.

Well I think, at the very least, an explanation based on evolution would need to demonstrate how it is superior to general cognitive processes and attempt to demonstrate that the results could not be explained by learning.

zoon wrote:(The trolley problem is unlike the Wason selection task in that it’s not obviously adaptive, it’s an interesting observed quirk of moral thinking rather than part of an ideologically driven programme.)

I imagine this would be a problem for the theory though. Evo psych is necessarily limited to adaptive behaviors (since we currently have no methodology for assessing spandrels since behaviors don't leave fossils), so if no argument can be made for how the "fat man" example is adaptive, then that would be a bit of a blow to the researchers arguing for an evolutionary basis. However, I don't think it would be hard at all to dream up a few stories for why it'd be adaptive - these hypotheses would just need to be tested.

Mr.Samsa wrote:

zoon wrote:It’s been suggested (by Joshua Greene and others) that the trolley problem may be the result of two different parts of the brain being activated in the two different scenarios. In the first, it’s all about calculation, 5 against 1, choose 5. In the second scenario, pushing the man off the bridge, the calculation is overwhelmed by a more emotional response, in another part of the brain, prohibiting the causing of direct harm to another person.

Well I don't think there is any "may" about it - it necessarily would be the result of two different parts of the brain since it is two drastically different scenarios.

To the logical mind, using general purpose calculation, the scenarios are not drastically different. In both scenarios, the person who makes the choice is clearly, unequivocally, responsible for the death of one person, and for saving five. The only difference is that in one scenario the person who makes the choice merely flips a switch, whereas in the other scenario the chooser must actually push someone to his death. Yes, it’s intuitively obvious that this is a very different matter, but why is it so very obvious to intuition when it is not at all obvious to logic? (It’s not even entirely intuitively obvious in other contexts – for example, if the trolley didn’t have any passengers, and the person choosing flipped the switch, or pushed, purely to murder the other person for their money, would they be able to claim that it wasn’t really murder if they only flipped the switch?)

It seems to me that you are agreeing with evolutionary psychologists (not necessarily Evolutionary Psychologists) when you state firmly that different areas of the brain must necessarily be involved. Does a general purpose calculator suddenly start using different areas of itself for logically similar problems? Assuming, as you seem to be, that Joshua Greene’s hypothesis is correct, how did human brains come to have one area for dealing with numerical calculations, and another area, with an override, dealing with social harm by actual bodily contact? Wouldn’t this be an example of an interesting psychological structure created through evolution by natural selection?


A couple of links:

Massimo Pigliucci discussing Joshua Greene’s dual process findings.
[youtube]http://www.youtube.com/watch?v=NOfKyjyWiU0&feature=related[/youtube]

Joshua Greene and Fiery Cushman (in press 2011) [url=http://www.wjh.harvard.edu/~jgreene/GreeneWJH/Cushman-Greene-FindingFaults-HandbookSocNeuro.pdf
]Finding faults: how moral dilemmas illuminate cognitive structure[/url]

Another go at the links - I think IE9 has a bug or two:

[youtube]http://www.youtube.com/watch?v=NOfKyjyWiU0&feature=related[/utube]

Joshua Greene and Fiery Cushman (in press 2011) Finding faults: how moral dilemmas illuminate cognitive structure

zoon wrote:

Mr.Samsa wrote:

zoon wrote:It’s been suggested (by Joshua Greene and others) that the trolley problem may be the result of two different parts of the brain being activated in the two different scenarios. In the first, it’s all about calculation, 5 against 1, choose 5. In the second scenario, pushing the man off the bridge, the calculation is overwhelmed by a more emotional response, in another part of the brain, prohibiting the causing of direct harm to another person.

Well I don't think there is any "may" about it - it necessarily would be the result of two different parts of the brain since it is two drastically different scenarios.

To the logical mind, using general purpose calculation, the scenarios are not drastically different. In both scenarios, the person who makes the choice is clearly, unequivocally, responsible for the death of one person, and for saving five. The only difference is that in one scenario the person who makes the choice merely flips a switch, whereas in the other scenario the chooser must actually push someone to his death. Yes, it’s intuitively obvious that this is a very different matter, but why is it so very obvious to intuition when it is not at all obvious to logic? (It’s not even entirely intuitively obvious in other contexts – for example, if the trolley didn’t have any passengers, and the person choosing flipped the switch, or pushed, purely to murder the other person for their money, would they be able to claim that it wasn’t really murder if they only flipped the switch?)

It is obvious to logic though, they are vastly different questions, with different contexts and subjects of concern. If we wanted to compare the two situations based on logic alone, then we'd need to balance out the confounding variables; specifically, in the second situation, we'd have to ask someone whether they would push a single person in front of the trolley to save five, or whether they would push five people in front of the trolley to save one. This way the involvement of the agent is equal in both options, and we'd get the same logical conclusion as the original trolley scenario.

In other words, there is the added "cost" associated with fat man scenario, and that is the agent having to make the decision as well as get involved. So of course it's a hard decision, and a fundamentally different one, because there are more variables to consider.

zoon wrote:It seems to me that you are agreeing with evolutionary psychologists (not necessarily Evolutionary Psychologists) when you state firmly that different areas of the brain must necessarily be involved. Does a general purpose calculator suddenly start using different areas of itself for logically similar problems? Assuming, as you seem to be, that Joshua Greene’s hypothesis is correct, how did human brains come to have one area for dealing with numerical calculations, and another area, with an override, dealing with social harm by actual bodily contact? Wouldn’t this be an example of an interesting psychological structure created through evolution by natural selection?

My position is not consistent with either of the evo psychs, since I'm proposing a non-evolutionary explanation for the same result. Basically, one situation involves making a decision, the other involves making a decision AND carrying out an action with various possible legal and social ramifications. It is necessarily true that the two situations will use different parts of the brain because they are two radically different situations - i.e. they are two different logical problems that required different analyses to solve.

Of course, it is possible that these two "areas" have evolved (or one of them is), but the point is that they need to demonstrate this. Pointing out that it uses two different areas of the brain is trivially true and doesn't support either position. Only a dualist would disagree with such an assessment.

Look at it another way: suppose that we lived in a world where the entirety of our behaviors were learnt, and there was no specific genetic influence (beyond a basic brain structure that allows us to learn and behave, etc). Now supposing this situation were possible, why would we expect the brain to solve a situation involving social harm using a part of the brain that solves numerical calculations? If we have learnt two different behaviors (i.e. how to count, and to care about other people) then wouldn't this perspective also argue that there should be two areas of the brain used here? The answer is that of course it does - to argue otherwise would suggest that it is either a) impossible to learn to care about social harm, or b) learning can only explain social harm in terms of numerical calculations.

Mr.Samsa wrote:
Look at it another way: suppose that we lived in a world where the entirety of our behaviors were learnt, and there was no specific genetic influence (beyond a basic brain structure that allows us to learn and behave, etc).

Busted! You're such a blank slatist!

rEvolutionist wrote:

Mr.Samsa wrote:
Look at it another way: suppose that we lived in a world where the entirety of our behaviors were learnt, and there was no specific genetic influence (beyond a basic brain structure that allows us to learn and behave, etc).

Busted! You're such a blank slatist!

Bah, I knew you'd get me one day. Your reinforcement history (without any impact at all from your genetic disposition) made such an outcome likely.

I feel so cool that i can make psychology jokes.

Mr.Samsa wrote:
~~~~~~~~~
One of the important things to keep in mind is what science actually is. We aren't attempting to explain what is actually going on, but rather we're trying to figure out which explanation is the most useful to us.
~~~~~~~~~~~~~~~~~~~~~~~~~~

jesus fucking christ Samsa, you just turned science into an ultra-right-wing, evangelical republican asshole!! Careful about how things out of context.... I will happily throw 5 of them off a cliff to save the trolley from getting messy.

Mr.Samsa wrote:
My position is not consistent with either of the evo psychs, since I'm proposing a non-evolutionary explanation for the same result. Basically, one situation involves making a decision, the other involves making a decision AND carrying out an action with various possible legal and social ramifications. It is necessarily true that the two situations will use different parts of the brain because they are two radically different situations - i.e. they are two different logical problems that required different analyses to solve.

Of course, it is possible that these two "areas" have evolved (or one of them is), but the point is that they need to demonstrate this. Pointing out that it uses two different areas of the brain is trivially true and doesn't support either position. Only a dualist would disagree with such an assessment.

Look at it another way: suppose that we lived in a world where the entirety of our behaviors were learnt, and there was no specific genetic influence (beyond a basic brain structure that allows us to learn and behave, etc). Now supposing this situation were possible, why would we expect the brain to solve a situation involving social harm using a part of the brain that solves numerical calculations? If we have learnt two different behaviors (i.e. how to count, and to care about other people) then wouldn't this perspective also argue that there should be two areas of the brain used here? The answer is that of course it does - to argue otherwise would suggest that it is either a) impossible to learn to care about social harm, or b) learning can only explain social harm in terms of numerical calculations.

Zoon:
Your post mischaracterises the two scenarios of the classic trolley problem. They are not by any means as different as you claim. In the first situation, the agent is required to decide between pulling, or not pulling, a lever. If the lever is pulled, the result will be that one person dies and five are saved. In the second situation, the agent is required to decide between pushing, or not pushing, a person. If the person is pushed, the result will be that one person dies and five are saved. In both situations, if the agent takes the active course and pulls the lever, or pushes the person, they will be carrying out an action with potentially very serious legal and social ramifications. If the trolley was empty, so that the agent was just deciding to kill someone for fun, it would be no defence at all to say that they only pulled a lever, they never laid hands on their victim.

The trolley problem is interesting and well known because the two situations are so similar, yet elicit strongly different responses. There is no obvious logical, or legal, or social difference between killing a person by pushing them, and killing a person by pulling a lever. So long as the killer’s agency is clear (as it is assumed to be in these scenarios), the killer runs a serious risk of being punished for murder, in both situations. On the other hand, if the killer is at the same time saving 5 people, the killing may be regarded as justified, and this is again true in both situations. Whether the killer actually lays hands on the victim is logically irrelevant. So it is odd that for most people considering the trolley problem, the laying on of hands changes the judgment of whether the action of killing is right or wrong.

I agree with you that the hypothesis of different brain areas being involved looks like a plausible explanation for the observed oddity of moral judgment. I do not agree with you that a purely logical general-purpose information processor would be likely to process two such logically similar situations in two wholly distinct areas. More to the point, it is claimed that neuroimaging has identified the areas concerned. I’m using the paper I linked to earlier by Fiery Cushman and Joshua Greene: Finding Faults: how moral dilemmas illuminate cognitive structure (in press). When someone is considering the causing of direct harm (as in pushing the fat man to his death) emotional regions such as the amygdala light up. By contrast, people judging that harm to one is acceptable for the greater good (as in pulling the lever to kill one and save five) are using areas of the brain associated with cognitive control and thinking guided by explicit rules (regions including the dorsolateral prefrontal cortex and the inferior parietal lobe). It seems to me that this may well be a demonstration that separately evolved structures are causing the odd moral judgments. The amygdala is an evolutionarily ancient structure, which in this case appears to be sounding off an emotional alarm when direct muscular action is likely to cause harm to another person.

It seems to me that this example, so far from being a parallel with the Wason selection task, is in some ways the opposite. The Wason selection task was claimed to show a specific adaptation for accurate moral judgment. The trolley problem, by contrast (assuming Greene et al are correct), points up a case in which evolution has caused the brain to produce two opposite moral judgments at the same time – pushing the fat man in front of the trolley is simultaneously good (the recent cortex calculates numbers) and bad (the ancient amygdala flags up an urgent alarm at the prospect of directly harming someone else). The make-do-and-mend of evolution has produced an outcome which is not entirely logical.

zoon wrote: Conversely, all learning has to be to some extent under genetic control (and evolved through natural selection), since the complex structures which make learning possible are built up by the proteins which are coded for by genes. For example, although nobody can speak a language without learning it, it’s clear that the ability to learn language is innate in humans and not in other species.

As others have noted, this gets so fuzzy as to be useless. Yes, humans do have a genetic predisposition to learning language (the ability to learn language does not appear to be unique to humans). But does that mean that ALL learning is to some extent under genetic control? It gets real fuzzy when we get to things like learning to use a microscope or playing raquetball or learning the equations for quantum mechanics. Fuzzy to the point that saying learning these is under genetic control is basically worthless. No, learning raquetball is not under genetic control.

The concept of "exaptation" is useful here. In exaptation a trait evolves for one thing, but the trait can do other things. A classic example used by Gould was that we evolved balance as part of bipedality. That sense of balance is useful for skateboarding. But we did not evolve to skateboard.

I think the OP was about whether some parts of morality are relatively invariant,

There's 2 parts to this:
1. How much of morality is due to evolution. Yes, it does appear that some of what we consider "moral" behavior is genetically programmed. One way to determine this is whether what is considered "moral" is invariant among widely different cultures. The trolley example seems to be one. It is unlikely that this type of thinking is "learned" because the example is so hypothetical. It's not something parents are going to sit down and say. "Now remember, if you are ever faced with the situtation of throwing a trolley switch so it kills one person but saves 5, throw the switch. But if you have to throw someone in front of the trolley, then that is bad." Instead, it seems to be genetically based. How would that particular idea of what is moral evolve by natural selection? So far I haven't heard a plausible scenario. The ability to detect cheating also appears to be genetically programmed into us. Along with that is the morality that cheating is immoral. For a species who lives in small social groups where individual survival is dependent upon the behavior of others in the group, it is easy to see how this morality evolved.

2. Is there some objective morality separate from culture? That is, whether morality is "natural law" that applies to all humans like gravity is a "natural law." And yes, many people (including religious ones) think that "morality" is an objective thing that can be determined by observation, deduction, and critical thinking.

zoon wrote:I agree with you that the hypothesis of different brain areas being involved looks like a plausible explanation for the observed oddity of moral judgment. I do not agree with you that a purely logical general-purpose information processor would be likely to process two such logically similar situations in two wholly distinct areas. More to the point, it is claimed that neuroimaging has identified the areas concerned. I’m using the paper I linked to earlier by Fiery Cushman and Joshua Greene: Finding Faults: how moral dilemmas illuminate cognitive structure (in press). When someone is considering the causing of direct harm (as in pushing the fat man to his death) emotional regions such as the amygdala light up. By contrast, people judging that harm to one is acceptable for the greater good (as in pulling the lever to kill one and save five) are using areas of the brain associated with cognitive control and thinking guided by explicit rules (regions including the dorsolateral prefrontal cortex and the inferior parietal lobe). It seems to me that this may well be a demonstration that separately evolved structures are causing the odd moral judgments. The amygdala is an evolutionarily ancient structure, which in this case appears to be sounding off an emotional alarm when direct muscular action is likely to cause harm to another person.

I think you are going too far with the "separately evolved structures". Instead, it is well known that thinking is distributed widely in the brain. Areas of the brain are only loosely associated with "cognitive control" or "emotion". As the paper notes, the authors consider BOTH moral dilemmas to be in the realm of the "cognitive". It's just that the areas of the brain involved with this cognitive problem can be distinguished. A simpler hypothesis is that we are looking at 2 separate "modules", each dealing with a slightly different situation.

zoon wrote:Your post mischaracterises the two scenarios of the classic trolley problem. They are not by any means as different as you claim. In the first situation, the agent is required to decide between pulling, or not pulling, a lever. If the lever is pulled, the result will be that one person dies and five are saved. In the second situation, the agent is required to decide between pushing, or not pushing, a person. If the person is pushed, the result will be that one person dies and five are saved. In both situations, if the agent takes the active course and pulls the lever, or pushes the person, they will be carrying out an action with potentially very serious legal and social ramifications. If the trolley was empty, so that the agent was just deciding to kill someone for fun, it would be no defence at all to say that they only pulled a lever, they never laid hands on their victim.

They are vastly different because in one scenario they are directly involved in the death of the people (pushing) and in the other they are indirectly involved (lever pulling). For example, look at the Milgram experiments (and replications thereof): subjects are asked to help "teach" a person how to spell by administering electric shocks whenever they spell a word incorrectly. When they are the ones to press the button, they become more and more anxious and stressed as the electric shock is increased over each trial. However, when they are asked to ask someone else to administer the shock, there is absolutely no hesitation - no discomfort, stress or anxiety.

This situation is not comparable to the Trolley hypothetical situation because in both cases there is a "lever" to be pulled that causes the harm to the individual, and all we add in the second situation is another layer of removed responsibility. The important variable, therefore, is this diffusion of responsibility in a purely behavioral sense. We can do the exact same thing in a non-moral situation to further illustrate the point: in a gameshow like Deal or No Deal, where the contestants have to either choose to take a guaranteed (lower scoring) deal, or gamble and try to win the million dollars or whatever. Now, when we ask the contestant what they should do, they'll show anxiety and stress, and (depending on the circumstances) choose the deal or go for the million. Ask somebody watching the show what they would tell the contestant should do, and they'll often make the same call but without the stress or anxiety. So we find the same behavioral difference that we see in the trolley example, but without a moral component.

Do we thus have an innate and evolved sense of when to gamble on game shows?

zoon wrote:The trolley problem is interesting and well known because the two situations are so similar, yet elicit strongly different responses. There is no obvious logical, or legal, or social difference between killing a person by pushing them, and killing a person by pulling a lever. So long as the killer’s agency is clear (as it is assumed to be in these scenarios), the killer runs a serious risk of being punished for murder, in both situations. On the other hand, if the killer is at the same time saving 5 people, the killing may be regarded as justified, and this is again true in both situations. Whether the killer actually lays hands on the victim is logically irrelevant. So it is odd that for most people considering the trolley problem, the laying on of hands changes the judgment of whether the action of killing is right or wrong.

They are vastly different scenarios. The lever removes the individual from responsibility. To illustrate why this is, change the lever to a more "direct" lever - the trigger of a gun. Ask subjects whether they would shoot a man in the chest (so that he falls on to the tracks) and saves 5, or whether they would push a man onto the tracks to save 5. You'll find that the "lever" and "push" examples will now gain equal responses.

zoon wrote:When someone is considering the causing of direct harm (as in pushing the fat man to his death) emotional regions such as the amygdala light up. By contrast, people judging that harm to one is acceptable for the greater good (as in pulling the lever to kill one and save five) are using areas of the brain associated with cognitive control and thinking guided by explicit rules (regions including the dorsolateral prefrontal cortex and the inferior parietal lobe). It seems to me that this may well be a demonstration that separately evolved structures are causing the odd moral judgments. The amygdala is an evolutionarily ancient structure, which in this case appears to be sounding off an emotional alarm when direct muscular action is likely to cause harm to another person.

Except of course the amygdala will also fire up due to learnt situations too, so we can't speculate as to whether it represents an evolutionary trait or not. The problem is that the "lever" example takes the responsibility away from the individual (due to how we've learnt what actions are "our fault") and thus there is less personal responsibility to trigger an emotional response from "direct harm" (because they don't view themselves as causing direct harm).

lucaspa wrote:As others have noted, this gets so fuzzy as to be useless. Yes, humans do have a genetic predisposition to learning language (the ability to learn language does not appear to be unique to humans).

Quick note on this: Given that learning language is not unique to humans, and that animals from various species (and not all mammals) have excelled in areas of language, then how can we argue that it is a genetic predisposition?

lucaspa wrote:But does that mean that ALL learning is to some extent under genetic control? It gets real fuzzy when we get to things like learning to use a microscope or playing raquetball or learning the equations for quantum mechanics. Fuzzy to the point that saying learning these is under genetic control is basically worthless. No, learning raquetball is not under genetic control.

The concept of "exaptation" is useful here. In exaptation a trait evolves for one thing, but the trait can do other things. A classic example used by Gould was that we evolved balance as part of bipedality. That sense of balance is useful for skateboarding. But we did not evolve to skateboard.

Exactly, well said

lucaspa wrote:

I think the OP was about whether some parts of morality are relatively invariant,

There's 2 parts to this:
1. How much of morality is due to evolution. Yes, it does appear that some of what we consider "moral" behavior is genetically programmed. One way to determine this is whether what is considered "moral" is invariant among widely different cultures. The trolley example seems to be one.

Well no, hold up a second there. The first step to figuring out whether something is genetically programmed is to check to see if it's culturally universal or not. This is the first attempt to falsify the hypothesis, because for it to be genetically programmed it necessarily must be culturally universal, however, being culturally universal is not sufficient to determine that it is genetically programmed.

This is because many of our culturally universal behaviors are a product of learning, as species-specific constraints lead to the same behaviors forming in individuals from the same species that are spread out across seemingly different environments. For example, all humans eat soup from a bowl. This isn't because we have a "soup-eating-behavior" that has evolved in our history, but simply because we are all subject to the forces of gravity. This is a simple example, but it also applies to other things (like language).

If we can demonstrate that a behavior is culturally universal, then we have simply demonstrated that it is not impossible for it to be a genetically programmed behavior. The next step is to test whether it is learnt or evolved.

lucaspa wrote:It is unlikely that this type of thinking is "learned" because the example is so hypothetical. It's not something parents are going to sit down and say. "Now remember, if you are ever faced with the situtation of throwing a trolley switch so it kills one person but saves 5, throw the switch. But if you have to throw someone in front of the trolley, then that is bad." Instead, it seems to be genetically based.

Wait, wait, wait... Are you suggesting that every learnt behavior must be explicitly taught, and a uniform response must be preceded by a predetermined stimulus?!

No, most of our behavior (especially our complex behavior like morality) comes about through rules and generalisations that produce novel behaviors and thought processes. Your comment is like saying that Einstein couldn't have learnt physics because nobody ever sat him down as a kid and said, "E = mc2". Learning works through the same selectionist process as evolution, so it's useful to think of novel behaviors (and emergent conclusions) being a result of a form of "learning exaptation". For example, this is how learning accounts for children coming up with novel sentences. Obviously stimulus-response psychology is ridiculous, but nobody has accepted that since Skinner demolished it nearly a century ago.

lucaspa wrote:How would that particular idea of what is moral evolve by natural selection? So far I haven't heard a plausible scenario.

There are a number of ways in which a general moral predisposition could come about, but it would likely be something like "imprinting", which would have a genetic component with a significant learning component. Imprinting itself would be far too rigid to explain the actual progression of moral behaviors in children and people though. The only way I could imagine a basis of morality being evolved is through a very general "altruism" behavior, but there are significant problems with this and doesn't seem to account for a large amount of behaviors. Altruism itself is better explained through learning, which in turn accounts for morality.

lucaspa wrote:The ability to detect cheating also appears to be genetically programmed into us. Along with that is the morality that cheating is immoral. For a species who lives in small social groups where individual survival is dependent upon the behavior of others in the group, it is easy to see how this morality evolved.

Not at all, we know for a fact that there is no innate cheater-detection module in us. The faulty work of Cosmides was overturned quite a few years ago now. The behavior is better explained by a more general logical process (which in itself may or may not be evolved).

lucaspa wrote:

zoon wrote: Conversely, all learning has to be to some extent under genetic control (and evolved through natural selection), since the complex structures which make learning possible are built up by the proteins which are coded for by genes. For example, although nobody can speak a language without learning it, it’s clear that the ability to learn language is innate in humans and not in other species.

As others have noted, this gets so fuzzy as to be useless. Yes, humans do have a genetic predisposition to learning language (the ability to learn language does not appear to be unique to humans). But does that mean that ALL learning is to some extent under genetic control? It gets real fuzzy when we get to things like learning to use a microscope or playing raquetball or learning the equations for quantum mechanics. Fuzzy to the point that saying learning these is under genetic control is basically worthless. No, learning raquetball is not under genetic control.

The concept of "exaptation" is useful here. In exaptation a trait evolves for one thing, but the trait can do other things. A classic example used by Gould was that we evolved balance as part of bipedality. That sense of balance is useful for skateboarding. But we did not evolve to skateboard.

That's a good analogy. I like.

I think the OP was about whether some parts of morality are relatively invariant,

There's 2 parts to this:
1. How much of morality is due to evolution. Yes, it does appear that some of what we consider "moral" behavior is genetically programmed. One way to determine this is whether what is considered "moral" is invariant among widely different cultures. The trolley example seems to be one. It is unlikely that this type of thinking is "learned" because the example is so hypothetical. It's not something parents are going to sit down and say. "Now remember, if you are ever faced with the situtation of throwing a trolley switch so it kills one person but saves 5, throw the switch. But if you have to throw someone in front of the trolley, then that is bad."

Come on. You're being facetious. No one would suggest that to learn something one must actually learn the precise scenario.

Instead, it seems to be genetically based. How would that particular idea of what is moral evolve by natural selection? So far I haven't heard a plausible scenario.

I speculated earlier essentially what you have talked about above with "exaption". Our moral sense was a logical flow on from our likely evolved behavioural trait of communalism. That is, we evolved some sense that cooperating with our fellow conspecifics was beneficial to the individual, therefore communalism was the result. For communalism to work, there would need to be a set of rules that would ensure harmony and productivity in the group. I don't think it's that hard to imagine that this sort of thing could have been learnt by the older members throughout their lives, and passed on to the younger members as social learning.

I guess it could have been a result of direct genetics, but as Samsa said, the onus is on the Evolutionary Psychologists to prove this with more than just-so stories.

Interestingly, though, I chanced across a documentary the other night about wolves/dogs and the reasons for their domestication success. I didn't actually see all of it, but the thesis was that the "instinct"/suitability for domestication was more or less wholly genetic. An old commie Russian experiment on selective breeding in Foxes seemed to prove this point nicely. I don't know how scientifically rigorous this thesis was, but on the surface it seemed sound. It was interesting because it purportedly showed clearly how a behavioural trait (i.e. affinity for humans and other species) was controlled so strongly via genetics.

Mr.Samsa wrote:Well I think, at the very least, an explanation based on evolution would need to demonstrate how it is superior to general cognitive processes and attempt to demonstrate that the results could not be explained by learning.

I don’t quite follow this: Presumably general cognitive processes arose precisely because, in some circumstances, they conferred an advantage over purely innate predispositions (which innate predispositions did not consequently vanished).Therefore there must be quite numerous innate predispositions that are not superior to general cognitive processes, at least not always.

As to “...and attempt to demonstrate that the results could not be explained by learning.”: Are you saying that if a behavior could be explained by learning it must follow that an evolutionary explanation must be discarded? Wouldn’t that be like saying that if sneezing and a running nose could be explained by a common cold, an allergy must be discarded?

Mr.Samsa wrote:Quick note on this: Given that learning language is not unique to humans, and that animals from various species (and not all mammals) have excelled in areas of language, then how can we argue that it is a genetic predisposition?

Why would a given behavior have to be unique to one species to be considered a candidate for an evolutionary explanation?

Mr.Samsa wrote:For classically innate behaviors it can be a bit easier, for example, innate behaviors have a number of necessary characteristics which can easily be distinguished from other behaviors: they are rigid, universal across all individuals, automatic, elicited by a specific stimulus, etc.

True, but I presume you’d agree that innate predispositions need not be thus constrained? In other words, the characteristics you list would indicate with near certainty an evolved behavior, but it doesn’t follow that all evolved behaviors must exhibit them; it only follows that it can be very difficult to establish that a less rigid behavior is primarily the result of an evolutionary adaptation.

Mr.Samsa wrote:Altruism itself is better explained through learning, which in turn accounts for morality.

With your soup example you illustrated very well why universality does not necessarily mean innateness. However, I fail to see how the intense preference that humans, along with so many other species, exhibit for their offspring and other relatives is in any way comparable. From the perspective of our conscious and deliberate reasoning, to which the imperatives of our selfish genes are hidden, I see no logic in having so marked a partiality for one’s own children, nothing inherently advantageous to the parents, no general cognitive process through which one would inexorably conclude that that’s the most sensible behavior. I’d be interested then in your elaborating on how altruism, at least with respect to kin, is better explained through learning.

CdeLosada wrote:

Mr.Samsa wrote:Well I think, at the very least, an explanation based on evolution would need to demonstrate how it is superior to general cognitive processes and attempt to demonstrate that the results could not be explained by learning.

I don’t quite follow this: Presumably general cognitive processes arose precisely because, in some circumstances, they conferred an advantage over purely innate predispositions (which innate predispositions did not consequently vanished).Therefore there must be quite numerous innate predispositions that are not superior to general cognitive processes, at least not always.

Not necessarily, no. Most general cognitive processes are derived from a very basic learning process that must have existed in all living things (i.e. simply put, to approach the good and stay away from the bad). This like remembering, attention, stimulus control, choice, self-control, altruism, etc, are all derived from that very basic principle. Since these cognitive processes are by-products of the innate pain-pleasure distinction, there is no selection pressure on them and they can't confer an "advantage" at all. It seems more likely that general cognitive processes that were successful produced organisms that developed innate predispositions which strengthened some of these (e.g. more sophisticated brains to handle more memories). However, the selection pressure would not be on the process of remembering, but on the brain structures controlling memory.

I may have misunderstood what you were asking though. If my answer doesn't make sense, then I might need you to clarify what you mean.

CdeLosada wrote:As to “...and attempt to demonstrate that the results could not be explained by learning.”: Are you saying that if a behavior could be explained by learning it must follow that an evolutionary explanation must be discarded? Wouldn’t that be like saying that if sneezing and a running nose could be explained by a common cold, an allergy must be discarded?

Not quite - it would be like having two mutually exclusive options that could explain a phenomenon, and one explaining it better than the other. The point is that science works by accepting whichever theory is the most useful (i.e. the one that generates the most predictions, can be falsified, explains the most data, etc), and so if learning is more useful than an evolutionary explanation, then we accept the learning theory.

If it proposed that a behavior has both an innate and learnt component, then it needs to be demonstrated that this is the case. Note that this is different from saying something "cannot be explained by learning", in that there is no mutually exclusive option. For example, if "cheater-detection" is said to be an innate ability, then finding that it is better explained by learning means we have to reject the evolutionary explanation; however, it might be possible that the learning mechanism underpinning cheater-detection has an innate component (like a logical rule which makes learning cheater-detection easier).

CdeLosada wrote:

Mr.Samsa wrote:Quick note on this: Given that learning language is not unique to humans, and that animals from various species (and not all mammals) have excelled in areas of language, then how can we argue that it is a genetic predisposition?

Why would a given behavior have to be unique to one species to be considered a candidate for an evolutionary explanation?

It wouldn't be a "death blow", but it would dramatically weaken the support for the argument. If organisms from vastly different classes (mammals, birds, etc) then we either have to consider that language is learnt, or that the underlying innate mechanism evolved right back to at least the chordata phylum, and arguably further into the original kingdom split of animals themselves. This would make the innate predisposition for language pretty general, and very few language nativists would accept such a claim.

CdeLosada wrote:

Mr.Samsa wrote:For classically innate behaviors it can be a bit easier, for example, innate behaviors have a number of necessary characteristics which can easily be distinguished from other behaviors: they are rigid, universal across all individuals, automatic, elicited by a specific stimulus, etc.

True, but I presume you’d agree that innate predispositions need not be thus constrained? In other words, the characteristics you list would indicate with near certainty an evolved behavior, but it doesn’t follow that all evolved behaviors must exhibit them; it only follows that it can be very difficult to establish that a less rigid behavior is primarily the result of an evolutionary adaptation.

Indeed, innate predispositions are harder to identify than the 'classically' innate behaviors, but if something is proposed to be an innate mechanism (e.g. cheater-detection) then it's easy to test. It's just that innate predispositions require vastly different tests because we're no longer testing whether the trait exists or not in the species, but rather whether there is a significant difference in some aspect of the ability (e.g. the time it takes to learn it). For example, it's argued that we have an innate predisposition towards learning a specific relationship between certain stimuli; that is, the taste-aversion relationship (i.e. when we eat a poisonous plant, we learn to avoid it in the future because it makes us sick). For this "Garcia effect" to be an innate predisposition, we have to demonstrate that it is fundamentally different to any other learning process (because if we learn taste-aversion in the same way Pavlov's dogs learn to associate a bell with food, or whatever, then there is no reason to presuppose an innate component beyond the ability to learn). And this is what we find. Unlike other associative learning relations, we find that taste-sickness pairings only require one instance for the relationship to be learnt whilst others take many trials, the cause-effect of the stimuli can be separated by up to 24 hours whereas other associations require a delay no longer than seconds, and sickness relationships are harder to form using other stimuli (e.g. it's hard to train a rat to avoid food using a light cue to signal sickness).

CdeLosada wrote:

Mr.Samsa wrote:Altruism itself is better explained through learning, which in turn accounts for morality.

With your soup example you illustrated very well why universality does not necessarily mean innateness. However, I fail to see how the intense preference that humans, along with so many other species, exhibit for their offspring and other relatives is in any way comparable. From the perspective of our conscious and deliberate reasoning, to which the imperatives of our selfish genes are hidden, I see no logic in having so marked a partiality for one’s own children, nothing inherently advantageous to the parents, no general cognitive process through which one would inexorably conclude that that’s the most sensible behavior. I’d be interested then in your elaborating on how altruism, at least with respect to kin, is better explained through learning.

The problem here is that there is very little research supporting the idea that altruism is an innate mechanism; there are just-so stories, and observations in other species, and plausible mechanisms to explain it, but no real experimental evidence to demonstrate it. So learning theories essentially win by default, as long as they can provide some mechanism (remembering that science is not about what is the "true" explanation, but simply which is the most useful. So even if altruism was an innate behavior, science would reject it if another "false" theory was more useful).

There are a number of suggestions for how altruism can come about, but I think this paper is a good start: Prisoner's dilemma and the pigeon: Control by immediate consequences. The author demonstrates how a very general principle (good=approach, bad=avoid) can explain self-control in an individual, which in turn accounts for altruism. Essentially, all an individual needs to be altruistic is to be self-interested, which organisms (by necessity) already are. From reading up more on the Prisoner's dilemma game, it's not hard to understand why altruistic acts would be more common among kin and less common among strangers, because people you have "played" with before in an iterated game will produce more predictable results and there are contingencies in play to force them to cooperate, whereas in single-shot games with strangers, it is more advantageous for them to defect and thus more likely for them to screw you over.

Thank you for your thorough answers Mr.Samsa. While some further observations and questions certainly come to mind, I really don’t have the time, and at any rate I now have a better understanding of your take in this matter, which was my main intent. I visit the forum rarely, and then only briefly, but when I do I usually find myself stopping to read some of your posts, which are always interesting.

Cheers.

CdeLosada wrote:Thank you for your thorough answers Mr.Samsa. While some further observations and questions certainly come to mind, I really don’t have the time, and at any rate I now have a better understanding of your take in this matter, which was my main intent. I visit the forum rarely, and then only briefly, but when I do I usually find myself stopping to read some of your posts, which are always interesting.

Cheers.

Glad I could help to clarify some of my thoughts for you

Feel free to post questions or comments and come back to them at a later date, since all these posts remain on the forum there is no need for immediate responses.

Sorry for bumping an old thread, but I came across an interesting paper on the trolley problem which seems to argue the same points that I did above (but in more detail and clearer than my attempt):

Moral Principles or Consumer Preferences? Alternative Framings of the Trolley Problem by Rai and Holyoak

Abstract:
We created paired moral dilemmas with minimal contrasts in wording, a research strategy that has
been advocated as a way to empirically establish principles operative in a domain-specific moral psychology.
However, the candidate ‘‘principles’’ we tested were not derived from work in moral philosophy,
but rather from work in the areas of consumer choice and risk perception. Participants were
paradoxically less likely to choose an action that sacrifices one life to save others when they were
asked to provide more reasons for doing so (Experiment 1), and their willingness to sacrifice lives
depended not only on how many lives would be saved, but on the number of lives at risk (Experiment
2). The latter effect was also found in a within-subjects design (Experiment 3). These findings suggest
caution in the use of artificial dilemmas as a key testbed for revealing principled bases for moral
judgment.