Sunday, June 30, 2013

Idea-ism: a rational basis for morality

In an earlier essay I proposed a moral principle, namely, that the interests of memes (or ideas) can form a rational basis for morality whose results coincide well with our intuitions.  In this essay, in response to some welcome prodding from occasional Ramblings contributor Don Geddis, I want to explore and expand on this idea.

Let me start by restating the problem I am trying to address: humans have a moral intuition produced by Darwinian evolution in response to the fact that life is a game in the game-theoretical sense.  Evolution optimizes for the survival of replicators.  Humans are hosts for two kinds of replicators: genes, which are encoded in DNA, and memes, which were encoded originally in human brains and then in various artifacts: clay tablets, papyrus scrolls, and now digital media.  Our genes and our memes have a mostly symbiotic relationship: our genes build brains capable of hosting memes, and our memes in turn invent things like agriculture and antibiotics, which in turn help to propagate our genes.  But occasionally the interests of genes and memes are in conflict, such as when our memes invent birth control or weapons of mass destruction.

One particularly interesting side-effect of the invention of brains capable of hosting memes is the development of self-awareness and consciousness.  Because we are conscious, we tend to assign a great deal of importance to consciousness, even going so far as to postulate that the fact that we are conscious is evidence of some sort of transcendent connection with the divine.  Even non-religious people put a lot of stock in consciousness.  For example, the comment that Don made that inspired this essay was:

[M]y genes want my body to die, but I choose not to adopt those same goals, on my individual level.  My intuition suggests that there should be similar conflicts with memes. Things that memes "want", which I may choose (on an individual level), not to adopt as my own personal goals.

Note the phraseology: "my individual level" and "my own personal goals" as distinct from the goals of both one's genes and memes.  This assumes that there is an entity that is distinct from one's genes and memes but is nonetheless an actor in the drama of life.  This hypothetical entity is often called the soul, but the modern fashion is to eschew dualism and call it the mind or the self.

Much ink has been spilled on the question of whether or not this entity exists, and if it does, whether it is actually a player on the universal stage.  But to talk about morality at all one must make this assumption.  The questions of moral behavior are, by definition, the questions of how to resolve conflicts between the desires of the self and the (possible) interests of other entities in the universe.  Moreover, in order to have even the prospect of resolving such a conflict one must assume that there is some sort of moral agency at work.  If we don't have free will then there is nothing to decide and hence nothing to discuss.  I am not taking a position over whether or not free will is actually metaphysically real.  All I'm saying is that in order to talk about morality at all you have to take free will as a working assumption, even if it turns out that the ultimate truth is that free will is an illusion.

The problem, then, is this: there is a part of our brain that produces the sensation of consciousness, rationality, and moral agency, and we wish to elucidate a principle by which the activities of this part of ourselves ought to be conducted.  Furthermore, we wish this principle to have the following properties: first and foremost, we want it to be useful, by which I simply mean that it should not be vacuous, not that it should score highly according to some quality metric.  Indeed, that would simply beg the question.  The whole point of this exercise is to describe a quality metric by which to measure moral decisions.  Second, the principle should be consistent with what is known about the nature of the universe.  And third, to the extent possible it should be consistent with our moral intuition.  Note that this property can never be fully met because there is a great deal of variation among people's moral intuitions (which is to be expected given that it is the product of Darwinian evolution).  Indeed, much of what I am proposing here hangs on the meaning of the word "our" in the phrase "our moral intuition."

All the moral systems that I know of fail to exhibit one or more of these properties.  For example, the ethical culture movement is based on the idea that one should live according to "ethical principles".  But this is vacuous, and hence not useful: it simply replaces the word "moral" with the word "ethical", which gets us exactly nowhere.  Moral relativism is, likewise, vacuous and hence unacceptable.  Religions based on divine revelation are not vacuous, but they fail on the second count: they are not consistent with what we know about how the universe really is.  Humanism is neither vacuous nor inconsistent with science, but it fails because it is not consistent with intuition.  It is species-ist, placing the interests of humans axiomatically above the interests of all non-human entities, which most people find unacceptable once they actually pause to reflect on what this actually entails.

It is worth digressing for a bit to talk about Sam Harris's moral premise as described in his book The Moral Landscape.  Sam's position is that, "morality is that which advances the interests of conscious beings," a premise that I actually do accept (but as a constraint, not a definition).  The problem is that Sam's elucidation shows that what he really means by "conscious beings" is "Western liberal intellectual members of academia who think like Sam Harris."  In particular, religious people somehow abdicate their claim to being conscious merely by virtue of being religious.  This is the fundamental problem with any humanistic premise: like it or not, religious people, even religious fundamentalists, are still human, still conscious beings.  There is no principled way to start with "the interests of conscious beings" as your premise and not accept the possibility that standing on the street corner screaming "God hates fags" might be moral.

So what I propose instead as a moral premise is: moral behavior is that which advances the interests of memes.

Before I go into a detailed analysis, let me point out that this is still a humanistic worldview.  The difference between humanism and meme-ism (actually, I think a better name might be idea-ism) is that idea-ism is humanistic as a consequence rather than a premise.  In the current state of the world, memes can only survive in human brains and artifacts.  Accordingly, humans have value, not axiomatically, but because they provide -- and produce -- habitat for memes.  Notice that this provides a principled way to resolve some very thorny dilemmas for the axiomatic humanist.  For example, as Bertrand Russell famously lamented:
I cannot see how to refute the arguments for the subjectivity of ethical values, but I find myself incapable of believing that all that is wrong with wanton cruelty is that I don’t like it.
On the idea-istic view, wanton cruelty is wrong because it destroys habitat for memes.  Likewise, it is immoral to (say) withhold education from women because education is a prime conduit for the replication of memes.  Withholding education is tantamount to forced sterilization with respect to the interests of memes.  (So is censorship.)

So let us evaluate this proposal with respect to my three criteria.  It's pretty evident that idea-ism is non-vacuous.  The mere fact that someone could potentially disagree with it is proof of that.  Likewise, it is consistent with science.  But is it consistent with intuition?

As mentioned earlier, no moral principle can ever be consistent with everyone's moral intuition.  That's actually a feature: any principle consistent with every human's moral intuition would be tantamount to moral relativism and hence vacuous.  And most people really do want to be able to distance themselves from the more distasteful behaviors exhibited by members of their own species.  So the consequences of idea-ism are likely to conflict with the moral intuitions of fundamentalists.  That is not in and of itself a failure.

Of course, idea-ism is first and foremost consistent with my own moral intuitions, and, I believe, consistent with the moral intuitions of western liberal intellectuals, a group of which I am (not coincidentally) a member.  But I think there's an argument to be made that idea-ism is not just a thinly disguised form of bigotry in favor of my own kind.  Ideas are special in a way that national identity, skin color, and even species-identity are not.  Ideas, and more fundamentally information, is woven into the fabric of life at the very deepest levels.  Life, at root, is the process of replicating information.  That this information has been predominantly encoded in DNA is a detail.  DNA just happened to be handy given the chemical inventory of this planet. but there's no reason to believe that life based on some other chemistry would be any less worthy.  But once you've made this leap and stopped being a DNA-ist, it is not a big leap from there to the realization that there's nothing particularly special about chemistry.  We are not that far away from having purely mechanical self-replicating systems.  Artificial intelligence is a bit further off, and artificial self-awareness probably further still, but there's no indication that these things are inherently impossible.  If and when it happens why not have a set of future-proof moral principles handy?

But you don't have to delve into the realms of science fiction to find plausible intuitive justification for idea-ism.  We are, in our present states, a symbiotic collection of both genes and memes.  At the moment, neither can survive without the other.  This basic fact of our present existence manifests itself in our yearning to connect with other meme-hosting entities.  This desire goes beyond our need to gather in physical groups in order to survive.  The exchange of memes is as fundamental to our nature as eating or sex.  We tell stories.  We engage in chitchat.  We write blogs.  As children we jump up and down and shout "look at me!" in the unconscious hope that something in our brain will escape and find a home in another.

And when we die, we want to be remembered.

All of these traits are fundamental to human nature.  I will go further: these traits are what define human nature.  We are not our bodies, we are our minds.  We can lose our arms and legs, our hearts, our lungs, even our ability to biologically reproduce and no one would question that we are still human.  But if we lose our minds we are dead, even our bodies are otherwise healthy.

Idea-ism is not an arbitrary premise.  It is a recognition of our true nature.  But more than that, it is a choice to take that nature and make it (or view it as) a purpose and hence a guide for making decisions.  Beyond being useful, rational, and intuitive, it's also noble.  It is an embrace of life in the broadest possible reading.

I hope you'll help me spread the Word.


Don Geddis said...

Interesting! Thanks for the expansion. Gives me a bit to think about :-).

On the narrow topic of "do we have free will?", I am generally satisfied with the answer from Eliezer Yudkowsky on Less Wrong. The answer is basically: any system that makes action plans -- us, an AI tic-tac-toe players, etc. -- needs to hypothesize that it "could" do action A, or action B. And then reason about what the opponent (or world) might do in response. Even if the AI game is completely deterministic, and completely predictable ... even then, as it is in the process of making the decision, it doesn't yet know what decision it will come to. That is free will. Not outside of physics (or even determinism) at all. Just a standard property of any decision planner.

Ron said...

Here's the problem with Eliezer's theory of free will with respect to the question of morality: Orcas eat seal pups. Before they eat them, they will often play with the pups, throwing them into the air like beach balls, e.g.:

You really should watch the video. The visceral sensation of actually seeing a seal pup tortured by an orca is very different from merely having it described to you.

Let us grant for the sake of argument that both orcas and seal pups are sentient creatures, and that being repeatedly flung into the air before being eaten causes the seal pup great distress. Furthermore, let us assume again for the sake of argument that this behavior doesn't serve any constructive purpose for the orca beyond entertainment. The seal has already been caught and dragged into deep water where it is completely helpless. The orca *could* just eat it, but it doesn't.

Is the orca's behavior immoral?

I would say that the answer depends not on whether the Orca's brain can be modeled merely as a decision-making machine. Clearly it can. Not only that, but an orca's brain is also clearly capable of constructing some kind of mental model of seal pups, otherwise it wouldn't be able to play with them.

However, it is an open question whether an orca can form a mental model of a seal as a fellow sentient creature. If an orca knows that a seal is sentient like it is, that it can feel pain like it can, and it just says to itself, fuck it, I'm at the top of the food chain so if I feel like playing with a seal then I'm going to damn well play with a seal, I would say that that's immoral (because it's knowingly destroying potential habitat for memes). On the other hand, if an orca's brain is not capable of forming a mental model of a seal as a fellow sentient creature, then I don't think it's fair to call the orca's behavior immoral. Morality requires (IMHO) the ability to recognize that other entities have subjective experiences.

You can construct similar scenarios with humans. If an adult hits someone without provocation, we tend to treat that behavior differently (from the point of view of morality) than when a two-year-old does the same thing, despite the fact that even at age two humans are clearly capable of formulating plans and making decisions, and even forming mental models of fellow humans. But clearly a two-year-old does not possess fully fledged moral agency (It's even arguable that a two-year-old doesn't possess free will.)

Don Geddis said...

Love the video. I had expected the orcas to toss the pups using their mouths and teeth, but instead they toss them with their tails. Very impressive!

I 100% agree with this: "Morality requires (IMHO) the ability to recognize that other entities have subjective experiences."

My own suggestion to you, is that "free will" is a philosophical distraction. Your two-year-old is a good example. You can easily have free will, while still not being capable of morality.

I suppose it's a necessary but not sufficient condition. If a hypothetical entity "didn't" have free will, then you might argue that it's unfair to morally criticize it ... because it's not capable of acting any differently? But then all that's important about free will (from a moral perspective), is that you can talk to the thing, and that it has the potential to change its behavior, based on your conversation.

Luke said...

What would you say is the connection between idea-ism, hypothesis formation, and creativity? Feel free to state that you think one has nothing to do with the others.

Ron said...

Idea-ism *is* a hypothesis. One of the consequences of the hypothesis is that creativity is valuable (because it produces a greater variety of memes). There might be more to say about this but I haven't given it a lot of thought.

Does that come anywhere close to answering your question?

Luke said...

I guess; from our discussions, it seemed like you had thought about idea-ism a lot more than it appears you actually have. I find that a bit odd, if you're trying on idea-ism as a new, better morality. What am I missing?

Ron said...

> I find that a bit odd

Why? This is a blog, not a peer-reviewed journal. There's a lot of half-baked stuff here.

Luke said...

> Why? This is a blog, not a peer-reviewed journal. There's a lot of half-baked stuff here.

The extent to which you link to it makes me think it's fairly important to you. Generally, I find that people develop ideas (hehe) that are important to them. Then again, you and I still hardly know each other.

Ron said...

It *is* important to me. But it's still an idea that is less than six months old, and I do have other fish to fry. Philosophy doesn't pay the bills.

Luke said...

I thought you had no more bills to pay. :-p Anyhow, know that I find the idea fascinating, that I do intentionally make lots of time for this stuff, and am around to discuss it whenever!

Ron said...

The only way you escape from having to pay bills in this world is to die, or move into the wilderness. It's true I don't have to work for the man to pay the bills, but I still have to pay the bills, change the odd light bulb. Oh, and there's this startup company that I'm supposed to be running.

> am around to discuss it whenever

Awesome! That's exactly the reason I put half-baked ideas up here for the world to see, to solicit feedback and criticism so the ideas can be improved. That's how science works! :-)

Luke said...

When we sit down next, I look forward to getting an idea of your priorities in life. I also look forward to hearing about the startup you're doing. :-)

Publius said...

>So what I propose instead as a moral premise is: moral behavior is that which advances the interests of memes.

This seems little different than Consequentialism.

If we could prove that memes would flourish and expand if we just enslaved 2% of the population -- would that be ethical?

Under your system, the moral worth of an action is premised on the predictive capabilities of the meme. What if the meme gets it wrong? The future is hard to predict.

Your ethical theory is alienating because it requires memes to put too much distance between themselves and their own projects and commitments. Your idea-ism requires moral agents to take a strictly impersonal view of all actions, since it is only the consequences, and not who produces them, that are said to matter. This demands too much of moral agents—since idea-ism demands that it be willing to sacrifice any and all personal projects and commitments in any given circumstance in order to pursue the most beneficent course of action possible. This fails to make sense of intuitions that it can matter whether or not a meme is personally the author of a particular consequence. For example, that participating in a crime can matter, even if the crime would have been committed anyway, or would even have been worse, without the meme's participation.

Ron said...

> This seems little different than Consequentialism.

The difference is that I am taking a very specific position on how to assess the consequences. Consequentialism by itself doesn't do that.

> If we could prove that memes would flourish and expand if we just enslaved 2% of the population -- would that be ethical?

That's an excellent question. My answer is that I predict you cannot prove this. If you could, I would consider idea-ism to be "falsified". (I put falsified in scare quotes because it's a theory of morality and not of objective fact, so it can't be objectively falsified. But it would mean that it does not adequately capture my moral intuitions.)

> idea-ism demands that it be willing to sacrifice any and all personal projects and commitments in any given circumstance in order to pursue the most beneficent course of action possible

No, that's not true. This is one of the reasons that choosing memes as the basis of assessing consequentialist outcomes is such a good idea. Pursuing personal projects is empirically one of the richest sources of new ideas. So putting value on individual agents is a direct consequence of idea-ism. Individuals have value under idea-ism, but as a conclusion, not as an axiom.

> participating in a crime can matter, even if the crime would have been committed anyway

Counterfactual reasoning applied to morals is fraught with all manner of logical peril. How do you know that the crime would have been committed anyway?