Do you think I'm realistic/scientific?

Frank

Certified not a majority
Veteran
While I think I am, I sometimes get the impression that people like Chalnoth and DemoCoder think I'm not being scientific, as I'm ignoring perfectly valid scientific and statistical reasoning.

On the other hand, I think I do try to look at all sides of any argument (although I tend to pick a side to argument, and exaggerate), and I tend to disregard things that are mostly abstract or have little statistical significance. Which is being realistic to me. And I think I do have a very good grasp about most things technological.

Then again, i do experiment when my gut feeling is telling me different. And I do think I should listen to my feelings more. ;)

Anyway, I am interested to hear what you all think. And I don't mind if you take this opportunity to rant a bit about the things you don't like about my postings. I want to hear them.
 
Abstract thinking requires you think outside the box without any preconcieved biases or artificial restraints. Scientific reasoning can only get you so far. Great ideas are not always born from science or math. The best math or science in the world doesn't magically endow one with common sense. Forest for the trees.
 
Last edited by a moderator:
Well, personally, I don't think that "being scientific" is an inherent quality. It's something learned, and I've had a lot of training at it. I don't expect most other people to be as nearly highly-trained in physics, statistics, and so on as I am (hell, my job at present is statistical analysis of some cosmological processes).

I think the really important thing on these boards is rather knowing how to ask the right questions. After all, it's unlikely that you'll be able to convince people of much on these forums. I hang out here because I learn new stuff all the time. That's what makes it interesting for me.
 
Chalnoth said:
Well, personally, I don't think that "being scientific" is an inherent quality. It's something learned, and I've had a lot of training at it. I don't expect most other people to be as nearly highly-trained in physics, statistics, and so on as I am (hell, my job at present is statistical analysis of some cosmological processes).

I think the really important thing on these boards is rather knowing how to ask the right questions. After all, it's unlikely that you'll be able to convince people of much on these forums. I hang out here because I learn new stuff all the time. That's what makes it interesting for me.
People may not be convinced by these weak representations, by these words, of one's ideas. But words are not the only representation there exists, there are stronger, more vivid ones, that will surely be better able to convert the world into the ideal.
 
DiGuru said:
While I think I am, I sometimes get the impression that people like Chalnoth and DemoCoder think I'm not being scientific, as I'm ignoring perfectly valid scientific and statistical reasoning.

If you want to make a sound argument, you have to obey rules of logic. If you want to make an inductive argument about the world, you have to obey the rules of mathematics, even moreso than when you're reasoning in a closed deductive system. If you want to make an argument in the absense of information, we have Bayesian inferencing to tell us how to compute the credibility or likelihood of certain statements. It's as simple as that.

This is not to say one cannot use other modes of thinking. That one can't dream or "play" with ideas or use intuition. Otherwise, no one would ever invent anything new.

But one cannot prove or convince other people following Western philosophy of the correctness of your "gut", or "intuition" or "dream" unless one uses the right tools. Why? Because people's instincts, their guts, their dreams can be wrong, in fact, they are very often wrong.

We don't have mind reading capability as a species. So the only way for person A to convince person B that an idea is correct, is to write it down very precisely, step by step, how it was arrived at, so that person B may playback the chain of reasoning in their mind and verify it. That's all we have as a species. It was this invention that took us from being hunter-gatherer hairless chimps to modern civilization. The invention of writing, of rules of reasoning, or logic, and more importantly, a rigorous symbolic representation of numbers and logic. That last step is important, because philosophers of language and cognitive linguists know that even two human beings who speak the same written language don't neccessarily agree on the same definition of a word. Just ask two people on the street to give their definition of a few vocabulary words.

The thing is DiGuru, it's all well and fine and say "I feel in my gut there's life on other words", but it is quite another thing to start referencing evidence and trying to claim support for your gut. Once your cross that barrier, you have made a logical assertion. And now, the fight enters the realm of logic and mathematics. So if you want to avoid being accused of being unscientific, don't assert a proposition based on pre requisite evidence, or else, someone can use that evidence to show how the proposition isn't supported.

This gets into the intelligent design debate. Believers feel in their "gut" that abiogenesis is impossible without a designer. That's fine. It's a belief. But once someone says "life from nothingness violates the 2nd law of thermo" or "life's complexity couldn't have been by chance because the probabilities are too slow", then one has made logical assertions that will rise or fall purely based on a chain of reasoning flowing from logic and mathematics.

In my hounding of people in their loose usage of law numbers to justify whatever their gut told them, I merely wanted to point out that their conclusions have no justification at all that is supported by statistical reasoning, even though their language sounds like "statistics language" It's handwaving arguments based not based on any calculations whatsoever, and it's easy to show just by asking them to see their equations.

Without regard to reasoning in the absense of information, Bayesian inferencing, which is pretty much the best you can possibly do in statistics, shows you exactly how you should adjust the likelihood of a general statement being true, like "the sun will rise tommorow". Normal boolean logic doesn't work, because often you don't have enough information to disprove something.

Choose to ignore it if you wish. But if you really fancy yourself a scientific guy, why would you ignore a beautiful field like Bayesian inferencing? Wouldn't you be curious, and go look it up and see how it applies?
 
DemoCoder said:
If you want to make a sound argument, you have to obey rules of logic. If you want to make an inductive argument about the world, you have to obey the rules of mathematics, even moreso than when you're reasoning in a closed deductive system. If you want to make an argument in the absense of information, we have Bayesian inferencing to tell us how to compute the credibility or likelihood of certain statements. It's as simple as that.

This is not to say one cannot use other modes of thinking. That one can't dream or "play" with ideas or use intuition. Otherwise, no one would ever invent anything new.

But one cannot prove or convince other people following Western philosophy of the correctness of your "gut", or "intuition" or "dream" unless one uses the right tools. Why? Because people's instincts, their guts, their dreams can be wrong, in fact, they are very often wrong.

...

The thing is DiGuru, it's all well and fine and say "I feel in my gut there's life on other words", but it is quite another thing to start referencing evidence and trying to claim support for your gut. Once your cross that barrier, you have made a logical assertion. And now, the fight enters the realm of logic and mathematics. So if you want to avoid being accused of being unscientific, don't assert a proposition based on pre requisite evidence, or else, someone can use that evidence to show how the proposition isn't supported.

...

Choose to ignore it if you wish. But if you really fancy yourself a scientific guy, why would you ignore a beautiful field like Bayesian inferencing? Wouldn't you be curious, and go look it up and see how it applies?

Demo, I actually agree with you about most of all that, and I wasn't trying to prove there is life on other planets or wherever. And I even agreed with you that getting many negatives makes people adjust their expectations.

It's just, that even if you can make a scientifically sound statictic about something, even if you have only a single datapoint, that that isn't statistically significant. Even all the negatives aren't significant, as they only cover a vanishingly small part of the possibilities.

In the case of other life, I'm trying to walk the middle ground, as I don't think we can say anything with a good probability and significance either way. Although I do like to come up with possible scenarios that offer alternative viewpoints to the reasoning used by both sides.

To make it more concrete, let me build my own theory.

We know that the probability a star in this galaxy has a planet with intelligent life is at least 1 in 300 billion. It might be more, but that gives us a lower limit. But then again, we can also say, that the probability of an Earthlike planet to develop life is 50% (1 datapoint). Only, we don't know how many Earthlike planets exist in this galaxy, as we only know about one solar system. That gives us a probability that 50% of all stars have Earthlike planets, statistical speaking. And, again going from one datapoint, we have a chance of 50% that such a planet develops intelligent life. So, we can set an upper bound of 37.5 billion planets with intelligent life.

While I might say, that statistically it's very likely that 18.75 billion planets in this galaxy have intelligent life, I build all that from a single datapoint. Statistically insignificant. So I disregard it as a sound hypothesis.

Then again, I can think of plenty probable scenarios (with the same amount of datapoints) by extrapolating what we know to give a plausible reasoning for just about every conceivable scenario. And I do. Each way.


Edit: statistics from single datapoints only work on average, when you can use them on very many individual cases. Like insurance.
 
Last edited by a moderator:
The point is, what's the best you can do? If all you've got is a single datapoint, what's the best you can possibly do in terms of adjusting beliefs? That's the point. Of course, it's not a very strong adjustment, but statistics just tells you the limits on correlations you can deduce, including the direction of adjustment.
 
Shouldn't you guys first try to agree on what the definition of a "scientific method" is?

Once you have done that, you can see if that can apply to the "life in outer space" or if scientific methods cannot be used to make any predictions about it. Science cannot solve everything....

A scientific method would normally contain the following:

1. Collect objective data
2. Create a testable hypothesis based on the data
3. Make predictions based on the hypothesis.
4. Do tests to check if the predictions are correct.

It must be possible to "falsify" the hypothesis. Meaning that it must be possible to prove that the hypothesis is false.
The hypothesis must also be formulated in such a way that others can also test the hypothesis. (and possibly prove it is false)


Applying this to "life in outer space" is left as an exercise to the reader :)
 
Science can solve everything that is solvable. And in cases like this, falsifiability is merely a matter of wording the question correctly.
 
mjtdevries said:
Shouldn't you guys first try to agree on what the definition of a "scientific method" is?

Once you have done that, you can see if that can apply to the "life in outer space" or if scientific methods cannot be used to make any predictions about it. Science cannot solve everything....

A scientific method would normally contain the following:

1. Collect objective data
2. Create a testable hypothesis based on the data
3. Make predictions based on the hypothesis.
4. Do tests to check if the predictions are correct.

It must be possible to "falsify" the hypothesis. Meaning that it must be possible to prove that the hypothesis is false.
The hypothesis must also be formulated in such a way that others can also test the hypothesis. (and possibly prove it is false)


Applying this to "life in outer space" is left as an exercise to the reader :)
Yes, absolutely.

But, when you only have a single datapoint (or none, in the case of the other player in this field: religion), what more can you do but speculate? Especially because we have as yet not been able to conduct experiments that show any proof either way, let alone developed a theory that is inclusive and allows falsification.
 
Chalnoth said:
Science can solve everything that is solvable. And in cases like this, falsifiability is merely a matter of wording the question correctly.
Then again, the wording might make or break the theory. Even mathematics is no proof against that.
 
DiGuru said:
Yes, absolutely.

But, when you only have a single datapoint (or none, in the case of the other player in this field: religion), what more can you do but speculate? Especially because we have as yet not been able to conduct experiments that show any proof either way, let alone developed a theory that is inclusive and allows falsification.

Yes, but there is an assymetry: The burden of proof is on the one asserting the positive. We do not by default assume the existence of everything that can exist in principle.

Thus, it is not the burden of non-believers to prove that there is no God or no Great Spaghetti Monster, or Xenu. It is the burder of believers to prove it.

Disproof of non-existence is not a requirement.
 
DemoCoder said:
Yes, but there is an assymetry: The burden of proof is on the one asserting the positive. We do not by default assume the existence of everything that can exist in principle.

Thus, it is not the burden of non-believers to prove that there is no God or no Great Spaghetti Monster, or Xenu. It is the burder of believers to prove it.

Disproof of non-existence is not a requirement.
Yes, I agree. But that still doesn't allow definite statements either way. And even in the case of religion or God, I would have to say that I'm not sure. I don't know, no matter what my gut feeling tells me. Which can go either way, depending. Which is just what it makes so compelling.

So, what do I believe? I don't know. I believe in myself. ;)
 
DiGuru said:
Then again, the wording might make or break the theory. Even mathematics is no proof against that.
Well, in terms of the culture of science, sure. Some theories require a lot of work to get tested (or sometimes even developed), and if they're not well-presented, they'll never get tested.

One good example of a theory that has managed to sell itself really well is string theory: string theory is so vastly far away from complete that it may still be decades before there are people who feel they have an understanding of the whole of it. And it's also absolutely inconceivable that it could ever be tested with our current understanding. But despite this, it has garnered a pretty large following among theorists.

One example of a theory that wasn't sold very well was a bacterial origin for gastritis. Specifically, one doctor had claimed that he had found a bacteria that was responsible for all gastritis. Since mainstream medicine believed that no bacteria could survive within the stomach, and particularly since this doctor claimed that all gastritis was caused by this bacteria, he was shunned. Later he swallowed a vial containing the bacteria and subsequently came down with a rather acute case of gastritis, which finally got his idea some acceptance.
 
DiGuru said:
Which can go either way, depending. Which is just what it makes so compelling.
And which is exactly what makes such things so uninteresting to me ;) I'd rather go in search of knowable unknowns.
 
Chalnoth said:
One good example of a theory that has managed to sell itself really well is string theory: string theory is so vastly far away from complete that it may still be decades before there are people who feel they have an understanding of the whole of it. And it's also absolutely inconceivable that it could ever be tested with our current understanding. But despite this, it has garnered a pretty large following among theorists.
I don't believe in it. While I completely agree with the basic "look and feel" of it, I am at the same time totally sure they got only a glimpse, but couldn't make head nor tail from it. And improvised. Because it was extremely compelling, felt so very right, and followed them everywhere, even in their sleeps and dreams.

I should know. ;)

But whatever I read about it, some of it feels totally right (like religion, probably), and other parts just feel wrong.

I'm pretty sure that's how science feels for the people who stubbornly walked the walk, refined their ideas while doing so and came up with the really good things.

I'm just not one of them. I always wanted to be an inventor, and I totally grokked computers from the start. I succeeded in those, as I envisioned them.

Only, I want other things as well. But none of them include becoming a great scientist.
 
DiGuru said:
Yes, I agree. But that still doesn't allow definite statements either way. And even in the case of religion or God, I would have to say that I'm not sure. I don't know, no matter what my gut feeling tells me. Which can go either way, depending. Which is just what it makes so compelling.

So, what do I believe? I don't know. I believe in myself. ;)

Well, thats the whole point.... : Science is NOT about believing.

Sciense also isn't about making definite statements on everything possible.
For some things it is not (yet) possible to make definite statements. Scientists accept that fact.

People who cannot accept that, turn to religion for help.
 
String theory has a large following amongst serious theorists b/c it fulfills a number of theoretical prejudices to model building past say the standard model (which we more or less know is correct with 99+% certainty).

Yes st can very well be wrong, but theorists are different beasts than 'pure' scientists, we have to relax various strict rules somewhat temporarily to make progress (like say Occams razor) and even the immediate requirements of 'manifest' falsifiability and things like that.

If you look at quantum gravity, you will see its fairly rigid in many respects. When you ask the timeworn theorist question 'well what else can it be', you really are horrendously limited.

Further string theory is rather remarkable in the sense that it can be wrong as a theory of gravity, yet be 'true' in various mathematical senses (indeed it most likely is). AdS/CfT, the various stringy dualisms, string twistor theory, and so forth will likely always be with us, even if the generating formalism for gravity is eventually falsified.
 
Back
Top