Many atheists who recognise society's need for a moral philosophy propose that the soundest workable one is "do as you would be done by". (Nicolas Walter, Colin Ward.)

This philosophy is open to the criticism that what one person would be done by may be significantly different from what another person would, and yet it is hard to accept that a third person might rightly be treated differently by the two of them in accordance with how they differ from each other in how they themselves like to be treated.

Thus if Chris has a higher threshold for being imposed on by someone previously unknown than Mary does, the "do as you would be done by" philosophy would imply that if they both meet a previously unknown third person, Jamie, there is an amount of imposition on Jamie that would be acceptable from Chris but not from Mary.

It might be argued that this objection does not demolish the philosophy, since Chris may be able without violating the morality to impose on Jamie more than Mary is, but in turn he should accept being imposed on by Jamie more than Mary would. But this would seemingly be to disregard Jamie's own preferences.

For example, when he is eating a meal Chris might not mind if a stranger comes and picks a small amount of food off his plate, but Mary might strongly object if anyone does that to her. So is it acceptable for Chris to do it to Jamie but not for Mary to? Where does that leave Jamie and how he likes to be "done by"?

What answer has atheism got to this criticism?

  • 3
    I don't see what is specific to atheism about your question? The eating off of other people's plates example would be a conundrum in many religious systems as well. Also, your problem seems to apply to Kant's categorical imperative, I am sure there are some Neo-Kantians out there who have an answer to such a dilemma. Commented Mar 3, 2016 at 23:47
  • @AlexanderSKing - Which religious systems? And what dilemma? As for atheism, the question is about atheism.
    – user19558
    Commented Mar 4, 2016 at 1:26
  • 2
    Neither Christianity, nor Islam, nor Buddhism (the 3 I am familiar with) have any explicit rules on whether it is OK or not to eat from someone else's plate. All sorts of modern ethical dilemmas aren't covered by the prescriptions of any of the established religions. In all of these dilemmas, trying to go by "Do unto others as you would have them do unto you" would lead to the same criticism you mention, regardless of the faith(s) of the subjects involved. Commented Mar 4, 2016 at 1:31
  • The question isn't essentially about whether or not it is OK to eat from someone else's plate, or modern. Atheists who propose "do as you would be done by" as the soundest workable moral philosophy that meets what they believe to be society's need for one (the focus of the question) view it differently from theists.
    – user19558
    Commented Mar 4, 2016 at 2:41
  • I agree with you that constructing a moral code without a transcendent source of values is a challenge to atheists, but that larger issue doesn't fit the specifics of the technical question you are asking in this post, which is more about moral relativism than faith (or lack thereof). Commented Mar 4, 2016 at 5:12

4 Answers 4

The Golden Rule:   Do unto others as you would be done by.  
The Platinum Rule: Do unto others as they would be done by.

(Stated by Gerald M. Weinberg in one of his books, but I am sure it is much older than that.) Solves the entire problem. Oh, that wasn't the actual question. DOH! Well, I would want a satisfying answer, so I am posting it.


Kant offers an extension of the rule that gives it more stability. The point is to expand 'you' and 'do' into something slightly more abstract.

Kant suggests we identify 'you' as someone we might all be. We should consider ourselves satisfied with an action only if everyone, everywhere could be expected to play both the agent's role and the role of everyone affected. In fact, we should generalize the question so broadly that the particulars of the given situation are simplified into abstract rules.

If we cannot interchange viewpoints in this way, then the action is at least questionable, and consent should be sought. Even the seeking of consent is subject to this same perspective, so consent cannot be sought under duress or between people of substantial power differentials unless those mitigating factors are addressed in some way that is also generalizable.

We are to do in a particular instance as we would be willing to require everyone to do as a general rule.

Even if we as individuals easily allow someone else to take a bite from our plate, we should not do so because we might be quite hungry and that might be that last bite that would make the difference between making it to our next meal reasonably, and not doing so. We should pad out the situation in every possible way and extend as much autonomy to the individual as possible because their current state might be more similar to ours in some more extreme state.

We might be willing, in a given circumstance, to make the rule more specific to the given circumstance, but we should give and recieve permission to do so quite explicitly, and be careful that the request itself does not leverage some particular of the situation which we would, in some other role, see as unfair.

So we might ask about the bite, or we might have a standing rule with our spouse about such bites, but we should also consider anyone else present and how they might be affected.

Fortunately, one of the aspects of such an arrangement itself is that none of us would want to be held to too high a standard. There is a level of complexity where affordances are just beyond what we would put up with in a variety of other roles. This creates some space where we can be mildly immoral in absolute terms, but can expect to be forgiven unless our misbehavior is too consistent. (There are absolute and relative duties.)

Kant was himself not an atheist. But he did intend to cover the case of atheists among us, (as enforcing religious requirements upon the non-believing is itself not something that can be generalized in this way.)

  • +1. Thanks. That Kant was not atheist doesn't help us bring together his generalisn and devt of the golden rule with an atheist outlook, but cogitation on his effort certainly might. He goes beyond the idea of generalising from 1-on-1 relations. But might it be possible to go too far beyond, letting "general" segue into "universal" and abstract to get detached from concrete, such that on returning to earth one has to allow for "mild" immorality? I'm not persuaded by your para 5. Should the notion of "(us) as individuals" be transcended too? How does the notion of "role" work in your para 8?
    – user19558
    Commented Mar 5, 2016 at 11:35
  • Kant really thought you had to do to universal and then build in compromises. He had in mind an ethics that worked across intelligent species, and was not limited even to humans. But you needed an absolute basis to argue down from. The central aspect of morality for Kant was autonomy, so if you don't buy paragraph 5, this is not the solution you are looking for.
    – user9166
    Commented Mar 7, 2016 at 1:04
  • 1
    Roles are one of those compromises, humans not being telepathic, and time forcing decisiveness to be an issue. (Kant imagined this was not universally so) But as Rawls expansion of Kant makes even more clear, to be moral in this sense, a role must be assigned and shaped in a way that everyone could agree to, not knowing who got to play whose part in reality. He has us consider whether we would agree to the system before time assigned us our parts. If not, the systems in which we live should be continually reshaped toward something we could all have agreed to in that way.
    – user9166
    Commented Mar 7, 2016 at 1:08
  • Thanks for this. But I think the idea that time assigns us our parts, although I realise it's metaphorical, or at least metonymic, gets too far away from the social which after all is at the root of the discussion. Much more importantly, I think humans are telepathic.
    – user19558
    Commented Mar 7, 2016 at 1:39
  • Kant and Rawls have their own super-abstract metaphysics to wrap around their ethics. Sometimes the best way to do politics is as indirectly as possible, by changing the options we are given for the way to frame things. Kantian ethics has had a huge impact on the West, despite being intractable, insanely abstruse, and a little bit sci-fi. Meanwhile, our telepathy is not reliable: it won't lie for us, and only lies are clear enough to actually communicate. Otherwise we would dispense with the horror that is language.
    – user9166
    Commented Mar 7, 2016 at 1:54

Having given an answer that responds to the surface of the question, I would like to give a different one that I feel addresses its core.

Atheism does not need to meet this criticism at all. It should not accept the golden rule, or some moderate refinement of it, as a basis for ethics.

This principle is simply a warmed-over compromise between religions, abstracted out of what they tend to have in common. We like it because most modern atheists are from traditions where it is prominent, and because the idea is simple.

But atheism is, by its nature as a 'negative principle', pluralistic at its basis. There are many, many reasons to have a religion, and the motivations not to include multiple objections to each of those. There is not a single core of atheism, and atheists of different stripes, in general, should find ethical systems that represent their true motivations.

For instance, an atheism motivated like Dawkins', flowing from his understanding of our species' biology, suggests we learn from evolution and choose a tenable principle in accord with our nature as animals. This has a framing largely in line with Nietzsche's call to be 'beyond' good and evil and to make the 'art of the self' the primary ethical motivation.

Art does not need an audience, but it is always better when it expects and finds one. To make art of the self, then, involves other people, and so you cannot alienate or destroy them. You have a good use for their regard and their ideas, and for that reason, you need their respect.

The alternate form of this is 'follow the will to power', but it is too easily misunderstood. It is easy to see the power of destruction, but, in general, it is a short-sighted option that betrays a lack not just of wisdom, but of taste, including awareness of your environment and how you use it.

So, working dialectically, I might refine the Golden Rule by opposing it: "Do unto others as you, as an animal, naturally do. When that precipitates undesirable responses, adapt. But adapt as a social primate with a wide vision of the world: in a way that protects your cultural standing, contributes to maintaining your social structures and does not forestall the future development of your individual relationships."

The enneagram came up in response to the first answer, in those terms, that answer conveys position '2' -- serve others. Kant's position is ultimately '9' -- make peace. The most common contrast to Kantian ethics, Mills' utilitarianism is from position '5' -- analyze and optimize. Nietzsche's is closest to '8' -- fight smart.

(And despite its whole "Madam Blavatsky" Ouspensky/Gurdieev history, the structure is not as esoteric or niche as it seems. It is the same structure that a dominant school of psychoanalysts use to diagnose personality structure -- see McWilliams' "Psychoanalytic Diagnosis", which includes these same nine stable positions in the same order, but starting from a different point in the sequence, and therefore presenting the same intermediate options. She adds a tenth arguably 'unstable' position in the 'dissociative' core. There are easy mappings from psychoanalysis in general onto other personality theories.)

Given a structure like this, one might expect five (or six if McWilliams is right) more answers that agree better with different motivations to choose a moral path outside dogma.


Do they have a justified reason to propose that? I had an atheist oppose an answer to a similar problem, to whit "because it was polite". But that still was inadequate to give a truly atheistic basis of 'why' they could accept the cultural behavior theologically.

  • 1
    +1. Thanks for this. I think the reason is that they believe it to be the moral philosophy that most workably upholds respect for people in general. It's fundamentally social. In some cases, it's woolly too. Nicolas Walter proposed the principle with reluctance, as if sighing while conceding that we can have this as a moral principle if we really want one. Did you mean to type "propose" rather than "oppose"? Not sure what to make of your last sentence, because atheists should surely not be required to justify why they accept cultural behaviour theologically when they don't claim to.
    – user19558
    Commented Mar 5, 2016 at 11:12
  • If this were an answer, instead of a question, I would upvote it. It is the basis of the right response... Atheism should not have a single underlying ethical basis, or that basis will simply become a different kind of religion. Atheism should not take a defensive posture relative to ethics, falling back on religious leftovers, it should analyze its own motives and propose a real ethics of its own.
    – user9166
    Commented Mar 8, 2016 at 15:30

You must log in to answer this question.