Does science need to be proven?

[todofwar]

So, I stumbled upon this little essay https://www.edge.org/response-detail/25322 in which the author argues you don't need the ability to falsify a theory for it to be valid scientifically. I would argue that if you have no way to test a theory it doesn't belong in science, but I also can't think of a way to bring complex mathematics back into the fold of science with this harsh criterion so I am a bit torn. Any thoughts? 

[insert_name]

26 minutes ago, todofwar said:

So, I stumbled upon this little essay https://www.edge.org/response-detail/25322 in which the author argues you don't need the ability to falsify a theory for it to be valid scientifically. I would argue that if you have no way to test a theory it doesn't belong in science, but I also can't think of a way to bring complex mathematics back into the fold of science with this harsh criterion so I am a bit torn. Any thoughts? 

simple, the math is the proof.

[todofwar]

Just now, insert_name said:

simple, the math is the proof.

But math alone can't be proof. There can be many different sets of equations to describe something, in the absence of an observable you're left with no way to choose between competing theories.

[lobe]

I think theories should be empirically falsifiable, it is how we can set criteria to determine what is real. However I think this essay is stating because of the disparity between our understanding of the universe and our ability to make experiments to test them falsification should take a back seat until we reach that point. I can agree with the point on how we are not able to currently experiment, but we shouldn't just blindly accept theories without evidence or attempts at least to falsify that theory. I think the best example would be particle physics, where particles are more often then not predicted before they are discovered.

As for complex math, it generally is based on principles we can see in real life, such as addition and subtraction.

[cubinator]

sci·ence

ˈsīəns/

noun

1. the intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment.

    

    

   Forgive the weird font (mobile editor strikes again), the portion below is mine:

   If you can observe it, you can perform science on it. Right now there are lots of things that we are theorizing about which we have no way to observe currently (i.e. string theory, dark energy, multiverse, etc.). So does that mean it's not science because we can't observe them?

    

    

   Technically, yes. If you can't observe something then you can't know for sure what it's like. So we call those fields regarding things we don't yet have the technology to observe "theoretical" fields. One day we might make a dark energy detector, and then we'll be able to observe, and gather actual information about what dark energy is really like. But just because we can't observe it yet doesn't mean that developing ideas about dark energy isn't important; Someone needs to figure out how to build a dark energy detector, and that will only happen if they have an idea about how dark energy could be detected that just happens to be correct. I've got nothing against theoretical fields and I think they're important to the advancement of knowledge, but unfortunately there's a lot of stuff we can't actually test yet.

[lukethecoder64]

If a guy named Albert walked into your office and said he had figured out how everything moves (BUT it changed our entire definition of time) and he only had math to back it up, would you believe him?

No.

But if he told you he could prove it by preforming these observations and they turned out exactly how he said they would, would you believe him then?

Yes.

Edited May 3, 2016 by lukethecoder64

[lobe]

2 minutes ago, lukethecoder64 said:

If a guy named Albert walked into your office and said he had figured out how everything moves (BUT it changed our entire definition of time) and he only had math to back it up, would you believe him?

What predictions does his theory make? Is there a simple experiment the theory should make possible?

[todofwar]

6 minutes ago, cubinator said:

sci·ence

ˈsīəns/

noun

1. the intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment.

    

    

   Forgive the weird font (mobile editor strikes again), the portion below is mine:

   If you can observe it, you can perform science on it. Right now there are lots of things that we are theorizing about which we have no way to observe currently (i.e. string theory, dark energy, multiverse, etc.). So does that mean it's not science because we can't observe them?

    

    

   Technically, yes. If you can't observe something then you can't know for sure what it's like. So we call those fields regarding things we don't yet have the technology to observe "theoretical" fields. One day we might make a dark energy detector, and then we'll be able to observe, and gather actual information about what dark energy is really like. But just because we can't observe it yet doesn't mean that developing ideas about dark energy isn't important; Someone needs to figure out how to build a dark energy detector, and that will only happen if they have an idea about how dark energy could be detected that just happens to be correct. I've got nothing against theoretical fields and I think they're important to the advancement of knowledge, but unfortunately there's a lot of stuff we can't actually test yet.

That is the current state of things, but what this essay says to me is he wants us to abandon the need for the dark matter detector in the first place. There is a reason we waited on the Nobel prize for Higgs until after they found his famous Boson. Not to use a terribly misunderstood phrase, but until we have proof things remain just theories. 

[lukethecoder64]

2 minutes ago, lobe said:

What predictions does his theory make? Is there a simple experiment the theory should make possible?

https://en.wikipedia.org/wiki/Tests_of_general_relativity

[todofwar]

2 minutes ago, lobe said:

What predictions does his theory make? Is there a simple experiment the theory should make possible?

That's the point, we need experiments or observables. What the essay is saying is that we should abandon the need for such experiments. 

[Findthepin1]

I say it is all math, and if you end up with multiple competing theories, choose the one that has the least discrepancies from the stuff that has been proven. You can prove it eventually if you want, but science is best guess. And the theory that has the least discrepancies or the most evidence is the best guess in this case. 

Edited May 3, 2016 by Findthepin1

[Bill Phil]

Science is a tool. A process. A process by which the facts are made bare. But, as with any tool or process, it's not necessarily accurate. Refining the tool and constantly using it to find the facts is what it's all about. But theories are only a part of the process of finding the facts. They're incomplete to an extent, if not false. But they are tested. 

Some scientist said that science is trying to disprove something while pseudoscience is trying to prove it. It's usually easier to disprove, anyways.

6 minutes ago, Findthepin1 said:

I say it is all math, and if you end up with multiple competing theories, choose the one that has the least discrepancies. 

Use Occam's Razor.

[Robotengineer]

Science can't be proven or disproven, it's not a statement, a theory or a postulation. Theories should be falsifiable, or at least have some sort of application to the real world.

6 minutes ago, Bill Phil said:

Use Occam's Razor.

Someone needs to start a razor company called Occam's, and have and ad saying that '4/5 scientists shave with Occam's!' I can't believe someone hasn't done it already.

[PB666]

1 hour ago, todofwar said:

So, I stumbled upon this little essay https://www.edge.org/response-detail/25322 in which the author argues you don't need the ability to falsify a theory for it to be valid scientifically. I would argue that if you have no way to test a theory it doesn't belong in science, but I also can't think of a way to bring complex mathematics back into the fold of science with this harsh criterion so I am a bit torn. Any thoughts? 

Written by a string theorist? Which camp, the 7 dimension, 11 dimensio, or 13 dimension string theorist. 

Theories look for patterns of consistencies. BTW, if you want to follow unprovable theories, theology can fix you up, one heretical branch of theology uses occams razor and the scientific methods as its basis. Going back to the unprovables is a kind of reverse evolution. 

The universe is a quantum universe, which suits me fine because im a statistian. Once you dive into the quantum view you immediately have the possibility of many outcomes, but at least on the relatvavistic scale they coalesce in the observable through mass action. The outcome of quantum scale physics are a set of parallel probabilities. The most probable outcomes end up being realities except when the scale is reduced. An example of small scale events are radiation induced mutations. In this case you cannot predict the decay of the particle, its direction of irradiation or even that it will ionize some nucleotide, its a spontaneous corruption of a common process, which we accept as this thing called evolution. 

But quantum interactions accumulate on the mass  scale and have predictable outcomes as binomial probaility distribution increases, the relative width of the confidence range decreases until at population its a single unique value. If the deBroglie wavelength is smaller than the smallest dimension, then you should be able to test a hypothesis about the object.

If you are dealing with mechanics on the planks scale, or with particles you need alot of observations, cause we cant observe quantum time or length. So far as yet we are unable to test the existance of gravitons, its science. Gravitational constant may not be constant. 

in circumstances were the universe does not give you a consistent answer, proof or disproof, use statistics. If you cannot prove where something is,  disprove all the places something cant be. Examine the universe from different perpectives until the only place where credible can be found is the place not eliminated. 

I would also argue that the process of analysis is subject to statistics, there is a recent paper on dinosaurs where the  collective of discoveries is evaluated to propose diversity and estimating gaps, the binomial probaility distribution/poisson distributions are very good at this. I have frequently used these to postulate gaps in human evolution, which oddly get support from genetics a decade later. This problem is that fossiles appear  like  quanta,  but they do not tell the state of the system but the state of an individual in an unknown number of systems, so in looking for systems it is more likely to satisfactorily estimate the number of systems without knowing the state of individuals in a system at any given time. This creates place holder hypothesis which float over typhonomics, as new data appear they can be binned into the placeholders and need not interfere with the state of other existing systems unless warranted. 

There is a bias in many sciences that becomes profoundly evident in paleontology, something does not exist outside the known bounds until proven, unfortunately this creates a rather large ascertainment and cladistic bias. When you try to mesh physical/paleoanthropology and  ancient DNA genetics, which is not an interpretation of how things evolved, but the blueprint of evolved or evolving state (and it begs the environment, phenotype, genotype head butting contest) one runs into mutual disbelief. An example of this is desinovan, its DNA is shared in a mode of southeast asians, and peaks in papua new guineans now claimed to be a single admixture event, but physical anthropology has no evidence of desinovans in the region. And the physical evidence from the region is of some other dwarfed species. Its sometimes not smart to be too constrained by known evidence when it is certain that the known evidence is fractured, its not a license to contrive evidence, but its like creating imaginary numbers to solve problems. The problem is that PA and ancient DNA analysis has filled in some of the gaps near the placholders i created almost two decades ago, but still haven't filled in the most important gaps, so as random data pours in you are still surprised, even if you create placeholders to catch the surprises. Science is sometimes a tempest. 

With quantum mechanics you do not need to prove everything with perfect observation, but as stated above, we can get carried away, ignore occam, and create a theory in which the universe has 13 dimensions in it. But we also have to be aware of the contrary, the limits of observation are massively larger than the scale tick-tock and step of the universe, theres alot that can happen in that quantum equivilent of space time that we cannot observe. It goes without saying that the smaller the scale of observation, the more indirct those observations become until at some point on has to question observation itself. 

[YNM]

Yes, science needs to be proven with data. Otherwise it's just philosophy. There's a reason why we have scientific data and not philosophical data.

And that is why I admire Hubble, not Hawking. Observe ! Experiment !

[Shpaget]

Some would say that, in science, you can't prove anything. Just because the experiment you are running ends up agreeing with the hypothesis, it doesn't necessarily prove causality. You can only be a little bit more confident in the relation.

At some point the confidence level is so high that the causality is taken as a fact, but a new scientific finding can alter the model. It happened before, it's likely it will happen again.

Before we came up with relativity, Newtonian physics were all the physics we needed and were considered a done deal. Then came along Mr. Albert and basically told us that we were wrong.

[Shpaget]

28 minutes ago, YNM said:

And that is why I admire Hubble, not Hawking. Observe ! Experiment !

Oh, forgot to address this...

I believe both approaches are valuable. In the case of Hawking, you must admit that he has quite a handicap and is limited in what experiments he can perform, so he specializes in what he can do - think and do math.

There are some people that don't like to calculate stuff and just build something to see if it works. I work with one such individual. Sometimes it's hard to talk to him since he doesn't like to do math for the projects he works on and occasionally gets stuck on some detail. Just a few days ago he complained to me that some induction coils he worked on were not performing as he hoped they would and drew me a sketch. Even though I know only basics of electronics I immediately saw the problem, because I know that induction is dependent on the angle of coil in the magnetic field. At that point he had already spend several days troubleshooting the odd behavior. He eventually figured out that if he rotated the coil he got better performance, but was perplexed by it.

[YNM]

Hmm... Yes, it's an equally good thing to do a few thoughts on what's possible and what not. Probably if general relativity didn't published people won't search for deflection of apparent star position during solar eclipse. But in the same way, Hubble was at first opposed when finding the truth of galactic distances. Observation and Experiment do need some Hypothesis beforehand, but if you have a mathematically proven theory without any real-life proof it's nothing more than a hypothesis. Case in point are string theory.

I respect Hawking on the fact that he contracted ALS, and is a great thinker. But not so for being someone who like to prove.

And on the point of data matching your hypothesis doesn't always means your hypothesis is correct, I do support that. Even an observationally proven theory still needs to be proven further. Who knows if something more complex lies behind the already complex theory ?

[sevenperforce]

It would be easy to say that science can still be considered science if it is testable in theory, even if it can't immediately be tested. That would cover most cases. Might not get at the heart of the issue, though. 

Perhaps a more holistic approach would be to say that if you are really doing science, you'll be looking for ways to test it already, merely by working through the process. Science is a process of learning and refinement and investigation with some pretty telltale markers. So I suppose that even if a particular line of investigation defies attempts at experimental design, it can still conceivably constitute science based on the process being implemented. Perhaps you already have all the data that could possibly ever be collected and you can only attempt to describe what you already have, with no opportunity to ever collect a new dataset. You can still conduct science with that. It's probably just going to be a bit more challenging. 

The important thing, I think, is to not lose sight of the point where science stops being science and becomes pseudoscience. For example, there was a time when the theory of universal common decent by evolution was still in its infancy and alternative theories could have still been viable. But that time is long passed. To persist in advancing something like creationism is not science, no matter how thick a veneer of sciencey investigative activities you layer onto it. The same is true for the vaccine-autism connection, for homeopathy, and for a thousand other foolish notions which could theoretically have begun with scientific intent but which now persist doggedly in their illegitimacy.

[PB666]

11 minutes ago, sevenperforce said:

It would be easy to say that science can still be considered science if it is testable in theory, even if it can't immediately be tested. That would cover most cases. Might not get at the heart of the issue, though. 

Perhaps a more holistic approach would be to say that if you are really doing science, you'll be looking for ways to test it already, merely by working through the process. Science is a process of learning and refinement and investigation with some pretty telltale markers. So I suppose that even if a particular line of investigation defies attempts at experimental design, it can still conceivably constitute science based on the process being implemented. Perhaps you already have all the data that could possibly ever be collected and you can only attempt to describe what you already have, with no opportunity to ever collect a new dataset. You can still conduct science with that. It's probably just going to be a bit more challenging. 

The important thing, I think, is to not lose sight of the point where science stops being science and becomes pseudoscience. For example, there was a time when the theory of universal common decent by evolution was still in its infancy and alternative theories could have still been viable. But that time is long passed. To persist in advancing something like creationism is not science, no matter how thick a veneer of sciencey investigative activities you layer onto it. The same is true for the vaccine-autism connection, for homeopathy, and for a thousand other foolish notions which could theoretically have begun with scientific intent but which now persist doggedly in their illegitimacy.

Oh gee, psuedoscience, have you ever read a clinical case report. There is alot of just technical blah, blah out there. The big problem right now in medical science is getting someone to gather all the clinical data and create associations rather than case reports, we got magic drugs out there being used for off- label applications with no scientific justification, and they're working, just noone is proving that they work. 

Science is that which expands on the boundary of science, that includes things like the prediction of hawking  radiation and other preproven hypothesis, right or wrong. Science is a process, you can report on the beginnining, the middle or in a review, the end. Once you've reached an agreed upon conclusio, its knowledge, not science. Higgs plus five others equal hypothesis, LHC and other colliders = methods, billions of collisions and statistics = data, meta analysis =!conclusion. 

The problem of hypothesis creation is its a really good idea to add some test or have some test in mind when creating the hypothesis, otherwise you will spawn things like 5, 7, ,11, and 13 dimensional string theories which undermine the credibility of the original hypothesis. Dark matter and dark energy are two examples, the testing right now is not proving what they are just what they are not. 

[todofwar]

Higgs is a good illustration of how it should be. We have a feature of the universe, mass, but no good basis for it from first principles. Higgs proposes a solution, and scientists try to figure out how to prove it. But this all started from an observed phenomenon, mass. Another example is the graviton. We have no idea how to look for it but it is related to gravity which we do know exists. And it fits in best with the particles and forces we have observed. If something can explain gravity without a graviton, we will need to find something that breaks the symmetry of the two theories and test it. But what the original essay is getting at is theoretical physics can make predictions about phenomena we haven't observed, probably can't observe, and is giving theories that in theory would require a Dyson sphere around a supernova to get enough energy to test. Once you get that far removed from the real world I find it hard to keep calling it science. I could say the 13th dimension is the cow level from Diablo and no one can really prove me wrong. 

[YNM]

7 hours ago, Shpaget said:

Before we came up with relativity, Newtonian physics were all the physics we needed and were considered a done deal. Then came along Mr. Albert and basically told us that we were wrong.

The amazing thing is, Newtonian Gravity is well-documented within General Relativity. That's a real science progression. Much like how heliocentric view contains geocentric view, and the later non-centered view can still contains everything before it, to a fair extent.

A new theory never annihilate the old one - it just refine them.

(that's why the russian spam guy days ago quickly got perma-banned.)

Edited May 3, 2016 by YNM

[PB666]

4 minutes ago, todofwar said:

Higgs is a good illustration of how it should be. We have a feature of the universe, mass, but no good basis for it from first principles. Higgs proposes a solution, and scientists try to figure out how to prove it. But this all started from an observed phenomenon, mass. Another example is the graviton. We have no idea how to look for it but it is related to gravity which we do know exists. And it fits in best with the particles and forces we have observed. If something can explain gravity without a graviton, we will need to find something that breaks the symmetry of the two theories and test it. But what the original essay is getting at is theoretical physics can make predictions about phenomena we haven't observed, probably can't observe, and is giving theories that in theory would require a Dyson sphere around a supernova to get enough energy to test. Once you get that far removed from the real world I find it hard to keep calling it science. I could say the 13th dimension is the cow level from Diablo and no one can really prove me wrong. 

We know how to look for gravitons, merging black holes, but as quantum can we detect them, they impart so little energy its hard to see a graviton in the same way one might observe a photon. If you can imagine the earth has 10^50 elementary sized particles  interacting with space-time at the surface to produce 10 N of force or equiv of 50watts/kg of power, thats something 10E-48 J per particle per second for a kilogram. I think that  the detector is trying to do  is to sense a , 'water molecule in a brief flood'when two balck holes merge, the question is whether we have sensitivity to detect the unitary particle. This is a ral delimna because gravity is a scalar field either it is not quantitizable like other particles, or thevparticles may not be discrete entities, we see them, but different size everytime. 

When you say look, in this case you really mean resolving the signal from the noise. If you are a light speed  ship traveling by two merging black holes and the fine scale detector only weighs a gram, the signal is going to be rather intense, but it may not be quantizable. When you are on earth, the noise is very intense, you have to observe individual jumps away from the noise as a wave of particles pass. Not easy stuff. 

Note on the wiki, there is no energy assigned to the graviton. 

[todofwar]

13 minutes ago, PB666 said:

We know how to look for gravitons, merging black holes, but as quantum can we detect them, they impart so little energy its hard to see a graviton in the same way one might observe a photon. If you can imagine the earth has 10^50 elementary sized particles  interacting with space-time at the surface to produce 10 N of force or equiv of 50watts/kg of power, thats something 10E-48 J per particle per second for a kilogram. I think that  the detector is trying to do  is to sense a , 'water molecule in a brief flood'when two balck holes merge, the question is whether we have sensitivity to detect the unitary particle. This is a ral delimna because gravity is a scalar field either it is not quantitizable like other particles, or thevparticles may not be discrete entities, we see them, but different size everytime. 

When you say look, in this case you really mean resolving the signal from the noise. If you are a light speed  ship traveling by two merging black holes and the fine scale detector only weighs a gram, the signal is going to be rather intense, but it may not be quantizable. When you are on earth, the noise is very intense, you have to observe individual jumps away from the noise as a wave of particles pass. Not easy stuff. 

Note on the wiki, there is no energy assigned to the graviton. 

But there is a possible observable, it's a matter of engineering to find out how to detect it. Maybe with a probe we launch outside the solar system to get away from the noise of the Sun, which is hard, but there is an experiment to be designed. 

[PB666]

2 minutes ago, todofwar said:

But there is a possible observable, it's a matter of engineering to find out how to detect it. Maybe with a probe we launch outside the solar system to get away from the noise of the Sun, which is hard, but there is an experiment to be designed. 

They moved thier exoeriment to L2. I think they are hoping to see a rare event, like a supernova, the problem with detecting gravitons, not the virtual kind, but the kind that pulse from major chnges in mass, the closer you are to them, the less likely you are to survive them.