Rapoport's Rules Meet the Outsider Test

Rapoport’s Rules for Debate

According to the English Wikipedia, Daniel Dennett (March 28, 1942 – April 19, 2024) “was an American philosopher and cognitive scientist. His research centered on the philosophy of mind, the philosophy of science, and the philosophy of biology, particularly as those fields relate to evolutionary biology and cognitive science.” Dennett was and remains well-known in atheist/freethinking/skeptical circles as one of the so-called “Four Horsemen” of New Atheism, alongside Richard Dawkins, Christopher Hitchens, and Sam Harris.

In this post I draw from Chapter 3 of Dennett’s book Intuition Pumps and Other Tools for Thinking (2013). The particular intuition pump, or tool in that chapter is what Dennett called “Rapoport’s Rules for Debate”. The Rules are Dennett’s suggestion for how to disagree with someone productively. In this article I’ll explore the practicality of the rules, and how one might apply them to John W. Loftus’ Outsider Test for Faith.

Dennett’s version of Rapoport’s Rules attracted considerable commentary, as this DDG Web search shows. Quoting from Dennett’s original version:

Chapter 3. Rapoport’s Rules

Just how charitable are you supposed to be when criticizing the views of an opponent? If there are obvious contradictions in the opponent’s case, then of course you should point them out, forcefully. If there are somewhat hidden contradictions, you should carefully expose them to view—and then dump on them. But the search for hidden contradictions often crosses the line into nitpicking, sea-lawyering, and—as we have seen—outright parody. The thrill of the chase and the conviction that your opponent has to be harboring a confusion somewhere encourages uncharitable interpretation, which gives you an easy target to attack. But such easy targets are typically irrelevant to the real issues at stake and simply waste everybody’s time and patience, even if they give amusement to your supporters. The best antidote I know for this tendency to caricature one’s opponent is a list of rules promulgated many years ago by the social psychologist and game theorist Anatol Rapoport (creator of the winning Tit-for-Tat strategy in Robert Axelrod’s legendary prisoner’s dilemma tournament).

How to compose a successful critical commentary:

1. You should attempt to re-express your target’s position so clearly, vividly, and fairly that your target says, “Thanks, I wish I’d thought of putting it that way.”
2. You should list any points of agreement (especially if they are not matters of general or widespread agreement).
3. You should mention anything you have learned from your target.
4. Only then are you permitted to say so much as a word of rebuttal or criticism.

(Some useful links: Dennett uses sea-lawyering in the sense of rules-lawyering. Robert Axelrod is a political scientist famed for his work on the evolution of cooperation (which incidentally provides a naturalistic rebuttal to the Christian apologist’s claim that “without God you can’t have morality”) and game theory. The Prisoner’s dilemma is a classic thought experiment from game theory which roughly models the endless moral dilemmas we all face daily - should we attempt to cooperate with others for mutual gain, or should we “defect” by grabbing what we can and saying to heck with other people? The “legendary tournament” refers to a tournament Axelrod organized and then reported on in his 1984 book The Evolution of Cooperation. The tournament established Tit for tat as the winning strategy in an iterated prisoner’s dilemma game. Anatol Rapoport was its creator. That’s a lot of interesting stuff, and important to have in the skeptic’s toolkit, in particular for understanding how morality evolved without any detectable help from any god, but it’s incidental to Rapoport’s Rules themselves. This brief hyperlinked summary cries out for a blog post of its own, so stay tuned for that. I will try to write it before Jesus comes back. Do you like those odds?)

The Challenge of Following the Rules

The remainder of the chapter describes Dennett’s own mixed views and results of trying to apply the rules. Excerpt:

Following Rapoport’s Rules is always, for me at least, something of a struggle. Some targets, quite frankly, don’t deserve such respectful attention, and—I admit—it can be sheer joy to skewer and roast them. But when it is called for, and it works, the results are gratifying.

To further underscore the limits of the Rules, a quote attributed to Dennett, but evidently not reliably sourced from his works, is:

“There’s simply no polite way to tell people they’ve dedicated their lives to an illusion.” (From the New York Times, “Philosophy That Stirs the Waters” (29 April 2013).)

Regardless of the quote’s provenance, it strikes me as self-evidently true. Not even the strictest adherence to Rapoport’s Rules can completely cushion the blow when you tell a person of faith how they’re wrong. People who are raised in a religion and then come to doubt it often describe the experience in crisis terms.

So if you looked at the Rules and thought “Are you kidding me?”, Dennett feels you. Imagine, for example, trying to follow the Rules with a Flat Earth proponent. Rule 3 looks difficult, as you are unlikely to learn anything from a Flat Earther, except by counterexample (i.e. you might learn how not to think). And how much effort can you justify for Rule 1, where you try to clearly re-express a completely nonsensical position? This wouldn’t be far removed from applying the rules to the delusions of a paranoid schizophrenic.

Why Skewering the Opponent Feels Good

Dennett mentions the “sheer joy” of skewering an opponent, but he doesn’t delve into the evolutionary psychology behind this joy. Suffice it to say that our long tribal past shaped our brain circuits to function in hierarchies based on dominance and submission. Part of that seems to be our strong instinct to reward others who serve the interests of our tribe, and punish those who threaten it. Our ancestors relied on social cohesion in fairly small groups (a Dunbar’s number of individuals or fewer) for their survival. Differences of opinion threaten cohesion, as any observer of politics knows. Defending cohesion by policing “incorrect” beliefs (i.e., beliefs that violated tribal norms) was therefore fitness-enhancing for our ancestors. The human brain has a reward system that generates feel-good chemicals when we do something “beneficial” like that. “Owning” one’s opponent in such a way that the opponent acknowledges being “owned” seems to fall into that.

The “struggle” that Dennett mentions is obvious enough: it’s hard to learn as much about an opponent’s position as the opponent knows. Or, as is often more likely, as much as the opponent’s trusted “experts” know. For example, “dittohead” was the nickname for a follower of the late Rush Limbaugh, due to the follower’s habit of saying “Ditto Rush, ditto!” Since Limbaugh amassed a fortune of some $600 million off his schtick, and few of his dittoheads did, it’s probably safe to conclude that Limbaugh knew more about Limbaugh-ism than most of his dittoheads knew. So to do justice to Rule 1, you might have to study the source of the beliefs that your opponent imperfectly parrots. That’s going to involve work and time, if those beliefs are detailed and/or unfamiliar. And when the beliefs are distasteful to you, the Rules demand Stoic resolve.

How to Follow the Rules, Or How Not to Re-Invent the Wheel

When you already know a belief is highly likely to be wrong (for example, when the belief has failed to win over the qualified experts), I advise against reading too much of the original accounts of it. It’s more efficient to read the scientific and scholarly rebuttals of it, unless (and this is important) you happen to be one of those scientific or scholarly experts doing the rebutting. In that latter case, you (presumably) have the expertise necessary to evaluate the cranky claims without being suckered by them. If the real experts have done their job, the rest of us can simply read the experts’ application of Rule 1 to the claims. Fortunately, for almost any anti-scientific claim that matters, there are already books, FAQs, and Web sites that debunk it.

For example, for climate science denial, see the Skeptical Science Taxonomy of Climate Myths. For creationism, see the Talk Origins FAQ. In addition there are many books written by scientists that rebut the talking points of science deniers. You rarely need to spend your own time re-inventing those figurative wheels, unless you are the first skeptical person to encounter some new original denier argument, or you’ve thought of a superior reply to some old talking point.

For religion, you’ve got John W. Loftus’ books (see the list in the sidebar) and the rest of Dr. David Madison’s Cure-for-Christianity Library© (the curated cure, as it were). While learning to think critically is important, I think learning to read - as a lifelong habit - is even more important. (The most efficient way to learn critical thinking is to read books about it. Thinking therefore cannot substitute for reading.) Very few people can think so well that they can independently duplicate the many lifetimes of thought encapsulated in books. And the more you read, the better you will probably think, since you’ll be working from a larger base of knowledge and examples. Dennett’s whole book purports to teach us better ways to think, and I think he did a pretty good job - a lot better than I could have done, had I remained functionally illiterate.

And if you read enough, who knows - maybe John will invite you to guest-blog!

Get an AI Boost

Dennett’s book came out in 2013, and a lot has happened since then. Notably, on November 30, 2022, ChatGPT burst onto the scene. A number of other public-facing AI chatbots quickly followed, such as Google Gemini, Microsoft Copilot, and others. So now we have a new tool to do the heavy lifting for us.

While these chatbots have made some widely-publicized missteps, and always require thorough checking of their results, they are good enough to save you some work. For example, Gemini claims that it can help you follow Rapoport’s Rules when you disagree with someone. With the current text-chat interface to an AI chatbot, you’d have to do some extra editing work. You’d have to feed your interlocutor’s talking points to the chatbot as prompts, and ask for advice on how you might factually and constructively respond to your opponent. No chatbot, as far as I know, can directly “listen in” on a conversation between two people and offer constructive advice. However, I think it’s just a matter of time before we have that. If you have the programming chops, you might be able to mash something up.

Someday we might have AI personal assistants, to which we may delegate (some, many, all?) our difficult conversations. That is, instead of you having to directly confront your wingnut friend, family member, or co-worker, you’ll be able to tell your unflappable AI to talk to their AI and work something out. This is by analogy to “I’ll have my people call your people,” something a billionaire might say to a fellow billionaire.

The Rules Meet the Outsider Test

John W. Loftus’ Outsider Test for Faith deserves a book to do it justice. John wrote the book, so read it if you haven’t already. You can also read many blog posts about the Test, including this post with links to many other posts.

For this blog post, I’ll go with a short summary of the Test from the book blurb:

Author John W. Loftus, a former minister turned atheist, argues we would all be better off if we viewed any religion–including our own–from the informed skepticism of an outsider, a nonbeliever. For this reason he has devised “the outsider test for faith.” He describes it as a variation on the Golden Rule: “Do unto your own faith what you do to other faiths.” Essentially, this means applying the same skepticism to our own beliefs as we do to the beliefs of other faiths. Loftus notes that research from psychology, anthropology, sociology, and neuroscience goes a long way toward explaining why the human race has produced so many belief systems, why religion is culturally dependent, and how religion evolved in the first place. It’s important that people understand these findings to escape the dangerous delusion that any one religion represents the only truth.

Alternatively, from John’s Facebook page:

It’s time to take the Outsider Test for Faith! It challenges adults to doubt their own culturally indoctrinated childhood faith for perhaps the first time, as if they had never heard of that faith before. It calls on them to require of their own religious faith what they already require of the religious faiths that they reject. It forces them to rigorously demand logical consistency with their doctrines, along with sufficient evidence for their faith, just as they already demand of the religions that they reject.

In the remainder of this post, I will try to show that for many persons of faith, the Outsider Test for Faith is a final step in a potentially long process, similar to a long-time skier finally skiing down a double-black diamond slope. (John says that his own journey out of faith took twelve years, but then he didn’t have the advantage of all the books he would later write!) If you put a novice on skis and shove them down that slope, it probably won’t end well. Before you can ski the hardest slopes, you need to start on the bunny slope and work your way up. Also, I’m not sure that persons with a particular religious faith consciously demand much of anything from the religious faiths that they mostly ignore. I suspect most of them have never seriously thought about why they reject all other faiths. It’s probably more that they just ignore rival faiths rather than logically refute them. So the “bunny slope” for them might be to start with thinking seriously about how they “know” all other religions are “wrong”. Where did they get that idea? How did they evaluate it? How might they defend it, if they needed to? After all, if you are a Christian, living in a majority-Christian nation, you probably don’t have many people demanding to know why you are not a Muslim, not a Hindu, and not a Cargo Cultist. You rarely if ever have any need to defend your non-belief in all other religions.

Analogy with Alcoholism

I couldn’t agree more that we would all be better off if we would view all of our beliefs from the outside. Similarly, I agree that alcoholics would be better off if they would stop drinking. But the alcoholic’s experience of alcoholism - in particular, of the frequent overpowering urge to get soused - is very different from the outsider’s perspective of alcoholism as a plainly destructive behavior that no reasonable person would engage in.

I suspect that religious belief can (and often does) have a hammerlock on the mind of the religious believer which is similar in intensity (if not the same in all particulars) as booze has on the mind of the alcoholic. And so we might not get very far just by telling believers to take the Outsider Test, any more than by telling alcoholics to quit drinking.

I’m also reminded of a college math professor who bewildered the class with some complicated derivation, noticed signs of the bewilderment, and admonished the class with “Assume this is not confusing!” It was funny at the time, both because it was nonsensical (how does one make confusion vanish by assuming it is not there?), and because math professors use the imperative “Assume…” a lot. So I’m just cautioning not to be like that professor with the Outsider Test.

Beware the Psychologist’s Fallacy

This is where we run into the psychologist’s fallacy. I refer specifically to the English Wikipedia’s alternative statement of the fallacy:

In this alternative form, the fallacy is described as a specific form of the “similar to me” stereotype: what is unknown about another person is assumed, for simplicity, using things the observer knows about themself. Such a bias leads the observer to presuppose knowledge or skills, or lack of such, possessed by another person. For example, “I (or everyone I know or most people I know) don’t know very much about chemistry. Therefore I can assume that this other person knows very little about chemistry.” This assumption may be true in any number of specific cases, making inductive reasoning based on this assumption cogent, but is not applicable in the general case (there are many people who are very knowledgeable in the field of chemistry), and therefore deductive reasoning based on this assumption may be invalid.

The application to the Outsider Test should be obvious: it’s simple for a person to take the outsider’s perspective to anything they are in fact an outsider to. But it can be very difficult for the insider to take the outsider’s perspective to that same thing. The psychologist’s fallacy kicks in when we try to model the insider’s mind from our perspective as an outsider! What is simple for us may be as hard for them as it is hard for the alcoholic to stop drinking.

And, perhaps trickily, the psychologist’s fallacy is itself a kind of failure to take the Outsider Test with respect to one’s own mental modeling of another person. If something is easy for me to do, I might struggle to put myself in the position of someone for whom it is difficult or impossible. (And if you find any of that confusing, assume it is not!)

Analogy with Learning a Second Language

I suspect there may be underlying neural reasons for the common inability to take the Outsider’s stance. There might be an analogy with learning a foreign language. Young children seem to have a “language acquisition device” in their brains - brain circuits that specialize in learning a language simply by hearing it spoken. (See also the Critical period hypothesis.) Linguists have studied this for a long time, and computational linguists would dearly love to give computers the same skill. It would lead to the Universal translator common in science fiction works.

The young child’s language acquisition device, if there is one, might disappear by adulthood, perhaps due to synaptic pruning. It seems very likely that this is true for many people when it comes to mastering spoken language accents. Adult language learners can usually master the vocabulary, orthography, and grammar of a foreign languge with enough work. But learning to speak the language out loud like a native is often considerably harder. A small percentage of adults seem to retain much more of this ability, and some of them become famous as comedians and impressionists. See for example Robin Williams and Frank Caliendo.

The comedian Bill Burr and others have noted the difference between beliefs you first learned at age 4, versus beliefs you first encounter as an adult. If you first encounter Mormonism, Scientology, or Cargo Cults as an adult, they probably seem insane to you. (Unless your name is Tom Cruise, apparently.) Maybe - and I’m just speculating here - the human child has a “belief acquisition device” that works like the purported language acquisition device. If so, you might never learn a belief in the same way that you learned your childhood belief, much as you might never learn a second (or third) language in the same way you learned your mother tongue. However, plenty of people retain sufficient cognitive flexibility to learn languages, and to question their childhood brainwashing later in life. But - importantly - people may vary with respect to this flexibility. Some people may be like Robin Williams with accents, but with respect to beliefs - able to “try out” many beliefs later in life - while many more people struggle to get outside their early brainwashing. For some people, getting outside might be impossible or nearly so, like trying to get Melania Trump to speak American English like an American.

Getting Outside with Rule 2

Recall the second of Rapoport’s Rules:

2. You should list any points of agreement (especially if they are not matters of general or widespread agreement).

Rule 2 might offer a path to making the Outsider Test progressive.

Many domains of skill acquisition rely on progressive training. If you’ve spent much time in gyms, learned to play a musical instrument, or you went to school, you know the drill. You don’t go straight to the limit of human performance in whatever it is. Instead, you start with something really simple, like lifting a very light weight, or learning to play a musical scale. Then as your strength or skill improve, you move progressively to more difficult things.

The Outsider Test for Faith asks the person of faith to do something that might be very hard for them - taking an outsider’s perspective to something they are actually an insider to.

Rule 2 says to find points of agreement with your interlocutor. For a person of faith, that’s simple: literally all other religions besides the one they believe. As an atheist, you believe in zero religions, out of the many thousands on offer. The person of faith almost completely agrees with you - they think all those thousands of religions are wrong too, save for one. That is, the person of faith agrees with most of your conclusion, but probably for very different reasons.

Therefore a productive place to start with a person of faith might be to highlight what you agree on - that all other religions are bunk - and then explore their reasons for rejecting all other faiths.

There’s a good chance your person of faith has never seriously thought about this before. The bible tells Christians in 1 Peter 3:15:

“Always be prepared to give an answer to everyone who asks you to give the reason for the hope that you have.”

But notice what the verse does not say: “…and also give the reason(s) why you reject the countless competing hopes (i.e., religions)”. It seems that the standard religious response to the dumbfounding fact of religious diversity is simply to ignore it as best you can. The bible does rip on other religions a bit, such as idolatry. But the bible doesn’t come close to giving a logical or scientific refutation of them. Instead Christians historically engaged in demonization. This was one of the main tactics, along with warfare, that Christians used to destroy paganism. Instead of arguing against the existence of pagan gods, Christians declared the pagan gods to be lying demons. This process continues where exclusionary monotheistic Muslims encounter polytheists such as Hindus.

When discussing religion with persons of faith, try to be aware of their tactic of framing the argument in terms of positive arguments for their particular faith, rather than in terms of negative arguments against all competing faiths. This was on display in the four-way debate video that John W. Loftus posted about the Virgin Birth. John’s Orthodox Christian interlocutors demanded that John clearly define what he would consider to be sufficient evidence for their religious claims. But they did not mention that they must think that no competing religion has met the same standard of evidence for them. So they must know what “evidence” is, well enough to conclude that no other religion has it. Perhaps they have just never thought this through before.

Therefore, perhaps when trying to lead a person of faith to the Outsider Test for Faith, spend as much time as necessary on Rule 2, your points of agreement. For example, pick a competing religion (Scientology, Cargo Cults, some different Christian schism) and explore your interlocutor’s reasons for rejecting that religion. Those Orthodox guys might still be beefing with Roman Catholics over that bizarre filioque thing. Back in the Middle Ages, they used to slaughter each other on battlefields over it. As a double-outsider (raised in flavors of Protestantism before rejecting the whole clown show) I struggle to make any sense of that dispute. At least when devotees of Star Trek or Star Wars get into bitter disputes over the combat capabilities of different starship classes, or some other inside baseball minutiae, at the end of the day they understand it’s all made up. (At least I hope they do. Some might be taking Jedi-ism or Klingon religion a bit too seriously.) But you might get somewhere with folks on either side of the Great Schism of 1054 by asking them to explain exactly why they aren’t on the other side of it. I wonder how many Christians take care to “always be prepared to give an answer” for that!

I suspect that after spending time on Rule 2 with your religious interlocutor, you’ll conclude that while you both agree that all competing religions are bunk, you each have different reasons. As an atheist who is also a metaphysical naturalist, I note that all supernatural claims have the same epistemic status (and comedy value) as perpetual motion machines, water-fueled cars, Bigfoot, Elvis sightings, and phlogiston. Perhaps the only reason I don’t laugh out loud at my native religion is that I first heard it at age 4. So while I don’t believe it now, there is still a lingering and unjustified residue of respect for it lurking in the depths of my feeble brain. A joke isn’t funny when you’ve heard it enough times, I suppose.

However, I suspect that you might sneak your arguments against religion in there by framing them as arguments against religions that your interlocutor rejects. It might not sound as threatening if I point out how science all but refutes the miracle claims of a religion that the person of faith doesn’t believe. And maybe, if the person of faith learns how to think scientifically about “wrong” religions, some long-unused part of their brain might realize they can think the same way about their own religion.

Summing Up

1. The Outsider Test for Faith is a powerful tool for examining one’s received faith. Such examination should quite often refute faith, or at the very least show that no particular faith has any real claim to being special among all competing faiths.

2. But the fact that religions continue to fool billions of marks suggests that few persons of faith take the Outsider Test.

3. For many persons of faith, taking the Outsider’s perspective might be difficult. It may be as difficult for them as learning to speak a foreign language like a native, or for an alcoholic to quit drinking.

4. Religious people seem to greatly prefer offering positive arguments for their own religion, instead of negative arguments against all competing religions. But to reject all other religions, they must have sufficient reasons to do so!

5. Rapoport’s Rules for Debate suggest a way forward: a progression toward the Outsider stance to one’s own religion, starting with what any religious person can already do: take the Outsider’s stance toward another religion, for which they actually are an outsider.

6. You can then compare your reasons (the atheist’s reasons) for rejecting that religion, with the religious person’s reasons for rejecting all religions besides their own. The key difference is that the atheist’s reasons rule out all religions, while the religious person’s reasons somehow rule out all religions except for their own. You might find that few religious people have actually thought this through.

7. You can probably ask AI for help at all points in this process.