Empiricism Is Not a Matter of Faith (2008) [pdf]
aclweb.org
aclweb.org
There is little incentive to publish your code. Refusing to give away your implementation does not in any way constrain your ability to publish, and giving away your code has only minimal benefits for your career. On the other hand, it's risky, since someone might find a bug in your implementation that changes your results. Additionally, a competing lab might show their algorithm is better than yours, or worse, improve your algorithm and publish a higher impact paper based on it, which might affect the profile of your publication. Finally, publishing your code means you have to package it in a way that is usable by others. Given these facts, it is not entirely surprising that most people would not publish their code.
Since science is pretty decentralized, it's hard to achieve the kind of large-scale change in behavior you'd need to make code sharing standard. The only people who could simply decide that people should publish their code and make it happen as a consequence are the funding agencies (e.g. NSF), who have only recently begun thinking about data sharing and have yet to make code sharing a priority.
One thing I'd like to see happen is a "viral" AGPL-like license for scientific code. Code would be freely redistributable, but if it is used in a commercial product or publication, the license should require that the modified work be freely redistributed. Given the choice between building on publicly available code or writing everything from scratch, I feel that most researchers would choose the former, even if it meant releasing their code as well.
The competition between labs is an interesting point. In a non- academic world we would see competition as good: better results, algorithm optimizations etc. Sadly I can see this as a deterrent to publishing code in the academic world though.
Publishing your code costs time in terms of getting it polished for public use, money by way of time. And it has very little benefit - you've lost exclusive rights to a publication platform, and Github doesn't show up in tenure portfolios.
a. People to cite what software they used way more heavily than they do (this includes me). b. The effort to be rewarded in and of itself. A paper, cited or not, is credit. There's a huge leap of faith that, after all the time prepping software for release that anyone will use it.
Faith by definition a "belief that is not based on proof".[2]
So you're arguing against the validity of formal logic or the scientific method or that a reached consensus cannot be trusted?
EDIT: I think you're overestimating the certainty that an empiricist would have about future events.
___
Do not confuse formal logic (which deals with sintax, not meaning, so it has no truth in itself) and the scientific method for which belief is utterly necessary.
Trust is not a bad thing. But trust requires faith. Faith is not a bad thing. "Consensus" is not an absolute source of unerring knowledge of reality.
Empiricism, often used by natural scientists, asserts that “knowledge is based on experience” and that “knowledge is tentative and probabilistic, subject to continued revision and falsification.”
Though assuming that causality is a thing seems to have worked out pretty well, you can also at least conceive a Looking Glass universe where the rules don't work the same and cause and effect are turned on their head.
Empiricist thought means that I do not have faith in, say, Ohm's law. I have simply observed it to be true sufficiently often that I have been able to formulate a set of rules based on which I can tell when it holds and when it doesn't, and that, if I have doubts at any time, I can test it. I carry on my everyday activity being convinced that it is true as a consequence of having observed it. In a sense, yes, I can figuratively say that "I have faith in classical electromagnetism", but not because I think classical electromagnetism is infailible, but because I have observed it making good predictions. I don't have faith that it will work, I have a) proof that it worked before and b) proof that the things now are just like they were before, when it worked. If I stumble upon a case when it doesn't, I will first check my understanding of classical electromagnetism and, after making sure it's correct, I'll try to see why the prediction was incorrect and review my options about the cause of the phenomenon I am observing.
Contrast this with my faith in Othin: I haven't observed him anywhere, I'm yet to see a giant and I can't really test whether he exists or not. I carry on my everyday activity being convinced that he exists, but only as a consequence of my faith. If this religious system makes a wrong prediction, I will typically review my understanding of it, but no my doubts. No giants around? Yeah, 'cause Thor HAMMERED THEM ALL!!! YEAH! No one has seen Othin in a thousand years? Well he likes to wander as a stranger, maybe they didn't know it was him.
Both mechanisms can fail, of course, but I think their possible failures stem from different origins and through vastly different mechanisms.
Can we observe an electron? Or quantum particles?
What about studying distance stars, or investigating past events?
And where do empiricist get to the point of deciding that observing the world is something they can rely on?
I don't genuinely hold any of these sceptical world views (nobody really does, do they) - these are mostly hard to answer/explain questions I keep getting thrown at me when I end up talking to serious creationists.
This implies that unless you can physically see it, then it's not observation. Trivial examples, such as the fact that you can tell that a door was open in a room because it's far colder than it should be and the door was the only way for cold air to get in, show that this is not generally as complicated.
This is pretty much what happened with classical electromagnetism, too. I didn't pick up the example with Ohm's law just because I like it: it actually predates the discovery of the atom's structure. It made valid predictions about how things happen, based on observing only effects, not the underlying structures. They actually got much of the underlying structure wrong, but the predictions based on nothing but observed phenomenons rather than wishful thinking about ether turned out to be correct, even though their mechanism was explained about a century later.
> Can we observe an electron? Or quantum particles?
Actually, yes for both, we pretty much photograph them to know they exist.
> And where do empiricist get to the point of deciding that observing the world is something they can rely on?
They don't; or, rather, if any of them claims they do, they're full of shit IMO. For all we know, the world could indeed be an intricate illusion, in which case we're only deciphering the strange mechanism of our illusion.
> I don't genuinely hold any of these sceptical world views (nobody really does, do they) - these are mostly hard to answer/explain questions I keep getting thrown at me when I end up talking to serious creationists.
Yeah, I know the feeling. I've been asking myself all these questions, too, and many more, for years, since I began to seriously study philosophy. The abyss stares back at me every day :-).
When someone does this, you've encountered someone with a broken brain. Someone incapable of critical thought and meaningful dialog.
Faith is not ok. Faith is a really bad way to come to conclusions. Faith is a measure of last resort. If people who believe in the supernatural had anything other than faith to cling to to validate their beliefs they would jump on it. Instead all they have is faith. All hundreds of thousands of belief systems out there with completely contradictory beliefs using the same flawed faith mechanism to live their life.
You talk to Christian apologists, as an example, for any amount of time you get all kinds of nonsensical rubbish that sounds like they're validating their beliefs on eye-witness testimony from people long dead, to validate a belief that heaps and mountains of empirical evidence you yourself can verify contradict. It's like you wouldn't believe a living, breathing witness in court if their testimony contradicted a DNA result, a single piece of evidence. Yet happy feeling supernatural mumble jumbled about being able to come back from the dead because some people thousands of years ago said they saw someone come back from the dead is ok? No, that's them relying on faith.
Faith is bad. You don't use it to fix real world problems. And no heuristics are not the same as faith, so don't go there.
Anyway, these people will use their keyboards and talk here about all kinds of things, but you know, press them and their mental gymnastics will fall like a house of cards, exposing their flawed mental states.
Empiricism doesn't require faith, and faith is a bad thing. Avoid faith whenever possible. Use a heuristic, and verify when you have the time, make sure it's sound.
You're backlashing against other people's bad logic, and in so doing letting it cloud your own thinking.
Your definition of "faith" for example seems to be "anything you believe even though you are not completely sure about it"? By that definition, some fundamentalist's belief in god is not faith, but when some physicist tells you that some house exists, say, then that is faith? Sounds like a pretty useless definition to me - plus I doubt that that is the actual definition you use, outside of providing a definition?
andrewflnr just told you what his definition of faith is, and that was manifestly not it. Try reading again.
I notice this easily because I favor this as the most useful of the myriad definitions of "faith" in real life. It isn't the only one, and it isn't the most popular one, but it's the most useful one. And that is the beginning and the end of the definition; it simply does not encompass "what you believe in even though you know it isn't true" or anything like that; it is an action-based definition that pays little attention to what people say, which, in this particular domain, is a major virtue. To see what someone has faith in by this definition (which, again, is not the only one), watch their actions.
As for your definition, I don't think it's particularly useful, because there already is a word for that, but well ...
@"what you believe in even though you know it isn't true": that strikes me as nonsenical. How could anyone consider something to be true ("believe" it) while knowing it isn't (thus presumably believing that it isn't true?)?
Practically and biblically, faith is defined by action. Anything else is probably just smoke.
Mhhhh.
And there you have it. pfortuny has declared"heuristics" in quotes and meaningless. Let's all black out that word with a felt tip marker in our dictionary. If you didn't have enough reasons to already not find pfortuny as credible, well now you have another.
But health and wisdom to you too.
And by the way, what does 'heuristic' have to do with 'truth'?
Alternatively I might be talking bollocks, I did first think of this while pretty drunk.
That does not mean they never will vary. They may vary at some point in the future. Although evidence strongly suggests it never will, that's not sufficient proof that it never will.
https://en.wikipedia.org/wiki/No_Free_Lunch_theorem https://en.wikipedia.org/wiki/No_free_lunch_in_search_and_op...
I'd put it as "All empiricists have faith that there is not much point in being an empiricist if there is a comedy god fiddling the numbers on purpose, purely to screw with the notion of empiricism, so therefore they cheerfully discount that notion as otherwise they wouldn't get much done."
If empiricism actually works and gives us (a good approximation of) the truth/reality, then it's probably a good idea to use it, as it allows us to influence our future (assuming free will or whatever, I don't want to get into that).
If empiricism doesn't work, then all bets with regards to influencing our future are off, anything you do is essentially random, and what happens as a result is pure chance. Behaving as if empiricism worked does not change that outcome, though.
That's why assuming that empricism works is the safe bet, no need to actually believe that it works - it's the one strategy that makes you better off if it works and doesn't change the expected outcome if it doesn't work.
sed 's/God/Science/g' PascalWager.template
Pascal's wager assumes that there either is no god or a god that has particular values, and then deduces from that that you should follow those particular values, because, if no god exists, it won't cost you anything to have done so, while you get a big payoff if that god actually does exist. There are two big flaws in that reasoning: First, it simply ignores the possibility that a god exists, but its values differ from what you assumed them to be, in which case you presumably would lose the game, and secondly, it's kinda obvious that following certain religious doctrine can have major costs.
My argument, on the other hand, does not assume any values (except perhaps that it's advantageous to be able to influence your future - find me anyone who seriously disagrees with that and also thinks that it matters whether you believe in some god or not (if you can't influence it anyhow, why should it matter?) ...). It simply shows that as a matter of fact, almost by definition, your only chance of gaining control over you future is empiricism. Possibly, it's impossible, but if it is possible at all, then empiricism is the only way that could possibly work - just assuming some randomly made-up laws of nature instead of extrapolating from past observations can not be a more reliable way to predict the future, in the worst case your extrapolations are also completely unreliable, in which case they are no worse than randomly made-up laws.
1. You correctly notice that Pascal is unable to imagine a reality where there is a God, but it is not the God he happens to worship. However you fail to notice that there are many many realities where neither empiricism is "almost by definition, your only chance of gaining control over your future" or indistinguishible from "randomly made-up laws". In particular I happen to believe that empiricism is a terrific tool to investigate a subset of reality, but it may be a poor tool to investigate some other subsets (religious/spiritual experience come to mind, but does not have to be the only ones. think of the cases where you investigate phenomena that are practically unreproducible within any humanly relevant timeframe).
2. You also see quite clearly that "following certain religious doctrine can have major costs", which Pascal just hand-waves away in his wager. But you fail to notice that by embracing empiricism as your one an only measure of truthfulness, you incur into a cost of opportunity by cutting off all other possible avenues of human knowledge.
Of this last point I offer two examples. First, in your response you mention that "if you can't influence it [God] anyhow, why should it [belief in such God] matter". This is not a matter of actual reality but a matter of values. You cannot argue if this is the good thing to do (ethics) or the most humanly fulfilling thing to do (esthetics?), but just whether this will result in a tangible benefit or not (economics).
In a more mundane way, there are lots of experiences of human life that are not easily measured with a scientific mindframe (I mentioned ethics and esthetics before). So we end up reducing "what is beauty" to whatever output of a long series of A/B tests happens to output, or "what is virtue" to whatever utilitarian explanation du jour.
As for your use of "investigation": Could you explain how you would investigate something that is not reproducible, or in general how you would investigate something in a way that leads to results that do not depends on empiricism?
I would think that as soon as you throw out empricism, you throw out investigation. If some event is completely random ("not reproducible"), that by definition means that you cannot know or deduce any rule for why/how it happened, which in turn means that you cannot even try to cause such an event or know where to look in order to observe it, so I don't see how you would "investigate" under such circumstances.
So, while I agree that there might be things you cannot investigate with empiricism, I would think that almost by definition, you cannot investigate those things at all, as investigation relies on empiricism.
2. No, I don't fail to notice that, very much the opposite. What I thought I stated very explicitly is that if empiricism does not work, assuming that it does work gives you the same expected outcome as assuming that it doesn't work. So, the actual outcome indeed likely would differ depending on whether you assume that empiricism works or doesn't work, so if you assume that it does work, you incur an opportunity cost of those possibilities that might lead to a better outcome - but you have no way to distinguish those from the choices that lead to worse outcome, and on average ("expected outcome"), assuming empiricism gives you as good an expected outcome as any other choice.
As for that "if you can't influence it" thing: I don't get your point, but I think I was a bit ambiguous, so let me rephrase my point: "find me anyone who seriously disagrees with that and also thinks that it matters whether you believe in some god or not and would argue about it (if you can't influence the future anyhow, and thus can't influence whether you believe or not, why should it matter?)".
As for your last example: I think you are contradicting yourself? I do understand it correctly that beauty also is one of those things that you think is not easily measured with a scientific mindframe? And yet, you explain one possible scientific approach to beauty? I would think it's almost trivial to measure all of those scientifically - in just the way you describe. It's all kinda about what people like - and the only reliable way to figure out what people like is empirical: You ask them and you observe their reactions to different circumstances. What would you suggest instead?
Really, I think you may want to think about what it actually is that you suggest to use instead of empiricism wherever you think empiricism does not apply. Just saying "something else that is not empiricism" is easy, but what would that look like? Have a close look at whatever you think the alternative is, whether there is any component of a known concept or a rule of how/why/when things happen or what some action will result in it. If there isn't, then how do you even formulate it? If there is, check whether there is any reason to prefer it over its opposite. If there isn't, then why don't you believe the opposite? If there is, that probably is where empiricism is hiding.