Hacker News new | comments | show | ask | jobs | submit login
The Fallacy of Gray (lesswrong.com)
97 points by jonp 2379 days ago | hide | past | web | 27 comments | favorite



This article is great, I've always felt the same way but this essay framed it and worded it perfectly. Having my thoughts more organized on the subject will come in handy while discussing politics with certain friends.

"You mock the simplicity of the two-color view, yet you replace it with a one-color view"

I love it!


Good article. A thought on the "Science is/isnt based on faith/ a religion" : for people who don't have the capacity to falsify scientific theories, science can be similar to a mysticism. Especially with complex theories, experts are equivalent to augurs reading the sky for the common mortal. That is why it is good to have down-to-earth experiments illustrating scientific principles, like the double-slit experiment showing diffraction. The first time I saw it was a shock because it is a counter-intuitive phenomenon. Epistemology and history give a lot of meaning to science, that is why I think it should be an integral part of the way science is taught.


Likewise the folly of those who say, "Every scientific paradigm imposes some of its assumptions on how it interprets experiments," and then act like they'd proven science to occupy the same level with witchdoctoring. Every worldview imposes some of its structure on its observations, but the point is that there are worldviews which try to minimize that imposition, and worldviews which glory in it.

Er -- except no, you've missed the point entirely here, Eliezer.

The point isn't that you step inside the paradigm and judge one as being somehow "better" than the other. Your choice isn't the false dichotomy of everything either being all the same or there being shades of better or worse. The entire point of paradigms is that, inside of each of them, they conform to those values that create the paradigm. So everybody sitting inside a paradigm, no matter how well educated, is going to be praising that paradigm. This is why the word "science" is way too overloaded for discussions like this. Do you mean reproducible, falsifiable model creation through abdcution, deduction, and induction? Or are you referring to the much more common Bayesian best-guess method of science? Seems like you're talking about the Bayesians, but there's a big difference which you've glossed over.

And from the outside it doesn't look all the same, either. What we can observe about various paradigms is measuring the extent of their vocabulary and measuring the rate of paradigm change.

Epidemiology isn't witch-doctory because epidemiology has a vast system of symbols, models, and data around how people get sick. We can pick up these symbols, models, and data and do useful things with them. Witch-doctory has but a few. Epidemiology has changed many times in the last two hundred years. They pick up the same data, symbols, and models ad rework them into new paradigms, increasing the richness of the field. Witch-doctory has not.

We can't look outside the paradigm, but we can certainly make subjective observations about what's more or less useful to the outside observer. I can be a Zoroastrian and still pick up enough useful information from Physics to walk on the moon -- no matter what the status of the Grand Unified Theory is. At the end of the day, it's always going to be pragmatic distinction, not a epistemological one.


I'm not sure what you're trying to say here. How is Bayes' theorem not deductive?

Are you talking of the difference between theoretical and experimental science? Because those both operate under the same paradigm: that the universe is causal and consistent.


I just lost 8 hours straight on lesswrong. Thank you.


The Sophisticate: "The world isn't black and white. No one does pure good or pure bad. It's all gray. Therefore, no one is better than anyone else."

I have often stated and thought that the world to me is not black and white and that indeed no one does pure good or pure bad. But I can't say this view has ever lead me to the conclusion that therefore, no one is better than anyone else.

Turning this issue over in my head, I guess I always just assumed that things being on a grey scale - the operative word being "scale" - that this would surely mean that things existing at different points on that scale could not be considered as all being the same. Interesting to read that apparently this is not the view held by some.

In my personal experience, I find that those whose view is that the world is black and white tend to also be those who wish to impose this view on others. While those who see the world as grey (whether they fall into this trap of the "Fallacy of Gray" or not) tend to be less concerned with converting others to their viewpoint and if anything are more concerned with defending their viewpoint from attacks by the black/white people. I'm not entirely sure what conclusion to reach from this observation... Authoritarians tend to see black and white, while libertarians tend to see grey and sometimes the "Fallacy of Gray" possibly as a defensive overreaction?


I thought the first part was great. Loved "You mock the simplicity of the two-color view, yet you replace it with a one-color view".

Then it tanked. I couldn't help but feel that straw man arguments were used for the examples of those who claim science is a faith. Sigh. Ok, how about a few things that make some scientists and especially the media seem like they belong to a religious or evangelical movement rather than one of discovering the truth: * believing in something no matter what the evidence and coming up with bizarre explanations for things because the evidence contradicts the theory. * saying the science is "settled". Seems like a dogmatic view if I ever saw one. * reluctant to share data to someone who disagree * appeals to experts ("Like most scientists agree") to replace the actual need to show that data fits into a coherent theory

Global Warming scientists seems to fit into this camp more often than others.

But really, you look at different disciplines. In archaeology, some guy was just claiming he might have found the first spot that civilization started, and it was from 12000 years ago (if I remember right). How does he know? Why make claims for that? Why not just stick to "oldest known", rather than sensationalize it?

Why also does every show have to talk about either evolution or global warming? I watched a great Bobby McFarrin special on music, and all these scientists involved kept talking about evolution, and "since we know evolution is true then X" and the hilarious part was that all the scientists came to contradictory conclusions. :-) Why not just leave the whole "evolution is true" part out, and just follow the data where it leads?


Great post. For some people science is absolutely a religion. These folks don't know enough to try and falsify the theory they believe but they have no issue trashing people who aren't as convinced.


> people who aren't as convinced

...but who somehow feel that they do know enough to try and falsify a theory?

> for some people science is absolutely a religion

Quoth the OP: If science is a religion, it is the religion that heals the sick and reveals the secrets of the stars. It would make sense to say, "The priests of science can blatantly, publicly, verifiably walk on the Moon as a faith-based miracle, and your priests' faith can't do the same." Are you sure you wish to go there, oh faithist? Perhaps, on further reflection, you would prefer to retract this whole business of "Science is a religion too!"


>...but who somehow feel that they do know enough to try and falsify a theory?

You don't have to try to falsify a theory not to believe it. Agnostic is also a valid position (so long as you're not trying to influence others one way or another).

Also, your comment smacks of "you aren't qualified to do anything but agree". Personally, when I hear someone say that one would need years upon years of training before being qualified to even have an opinion I wonder what controls are in place to ensure this qualification process isn't tainting their thinking.

For example, long time programmers of some awful programming language will report that it "works just how I think", but in fact long repetition has programmed them to work and think this way.

>Quoth the OP

Ah, but you've misunderstood me. I'm not saying science is a religion itself, rather that some people put blind faith into it. I've seen plenty of people ridiculing others for not believing in some theory, but when engaged they don't really know anything more than "most people believe this".

I once saw a self-proclaimed Christian ridiculing a non-Christian about some moral issue. When I asked how many times he had read through the bible he admitted he had never read the whole thing. I don't see much difference between this and what I see from some laymen "champions of science" I've encountered.


If science gets credit for healing the sick, does it also take the blame for deaths caused by nuclear weapons? Blatant, public, and verifiable. Praise worthy? Not in my book.


> does it also take the blame for deaths caused by nuclear weapons?

In my view, yes, it does, and science needs to learn from that (and countless other mistakes). It's certainly not universally praise-worthy. But it does provide a way for ethical people to reliably improve the human condition.


Who said anything about praise? Yes, my deity can be angry, but my faith is still more true than yours, o Christian.


The same science that lets you heal the sick also tells you how to kill people.


Why is it hilarious that scientists disagree?


This analogy assumes you have a linear scale from black to white, and thus any two values can be objectively compared to each other. You have to be careful, however: Many discrepancies and disagreements cannot be likened to simple shades of gray. I'll take an example from the article:

The Sophisticate: "The world isn't black and white. No one does pure good or pure bad. It's all gray. Therefore, no one is better than anyone else."

Assuming they're talking about moral relativism, it can be argued this is a purely subjective matter. When it comes to morality, it's likely people will disagree on whether ActionX is a darker or lighter shade of gray than ActionY. Although his reasoning is faulty, the Sophisticate's conclusion is correct.

Instead of: "It's all gray. Therefore, no one is better than anyone else."

It should be: "No one can definitively claim one shade of gray is darker than another. Therefore, ..."


I don't quite agree. While in some cases, there is certainly subjectivity in morality, almost everyone would agree that certain things are definitively a darker gray than others. Almost no one would argue that murdering random people is morally superior to nursing an injured bird back to health. It is true that some people might, but I think that the vast majority is sometimes right in judging comparisons like that.


>> almost everyone would agree that certain things are definitively a darker gray than others

Yes, but we must ask ourselves: Does widespread agreement on an opinion turn that opinion into a fact? If 100% of humanity were to suddenly believe that red is better than blue, would that make it objectively true? I think the answer is no. The difference between fact and opinion is not reliant on the number of people who believe one way or the other.


"If 100% of humanity were to suddenly believe that red is better than blue, would that make it objectively true?"

No, but it does make it subjectively true for 100% of humanity.

In addition, as a sibling comment pointed out: it is possible to objectively decide which of two actions is better than another for any specific definition of better. For example, one definition of moral "betterness" is the quality of being conducive toward the survival of a given society.


I'm suspicious of this vague terminology. Instead of saying something is "better", why not just say "it's more conducive toward the survival of a given society"?


In this case moral relative depends on the definition of better. If you have a specific definition of better (EX: for one person, one group, society, the environment etc) you can actually come up with a relative ranking. However, it's the choice to use better in the most vague term that makes it a monochromatic argument.


He's exactly right. The answer is to stop looking for certainties and instead look for models - theories that produce results that are as right as you can get them. We don't ever expect to reach perfection, we just expect them to asymptotically approach it.


It's the fallacy or relativism, which is a special case of believing too much in syllogism.


I was interested at first, but it seemed more and more that the author was referring to a context that he thought was shared but was actually obscure. When using examples from novels I've read or exchanges between bloggers, I find some light introduction (e.g. who's the "they" in this "Player of Games" book?) or footnotes is the best way to avoid the "scary homeless person that yells at me when I get off the subway" vibe.


The context is shared on LessWrong, but it's a very large context, touching on multiple subfields of philosophy, mathematics, and biology (to name just the big ones). There's currently no better introduction than reading the major LW posts/essays.


> There's currently no better introduction than reading the major LW posts/essays.

At the present time, that's correct. Eliezer is working on at least one book (maybe more.. I'm not sure anymore) that should stand on its own better.


It's gray because you don't know where the line is.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: