The Failure of Falsifiability

Nothingness · February 7, 2008, 4:02am

Is the principle of falsifiability truly a good principle to base scientific theories on? A theory must be susceptible to scientific experimentation. If this theory cannot be empirically investigated in any way, then it is not falsifiable and therefore not considered viable. On the surface, it seems commonsensical when viewing it from a scientific standpoint. We cannot have invisible elves dancing around causing water to boil or speak of the spirit that moves all things. There is no way to test for these hypotheses. Yet when one probes deeper into the history of scientific theories or philosophies, one finds that falsifiability depends on the technology of the times.

Democritus proposed an atomic theory over two millennia ago. These atoms could not be observed or tested by experiment in any way. Why should one believe in these invisible entities that are supposed to be the constituents of the universe? It was unfalsifiable. Even Ludwig Boltzmann faced immense criticism for his belief in atoms in the late 19th century, but with new technologies the theory became falsifiable. If one were to hold to the principle of falsifiability, then one must have held that this was a non-sensical theory not worth giving any more precious time to. Of course, this theory has held up, and unto this day there is a big speculation on new particles waiting to be discovered in the massive accelerators.

A second instance of the failure of falsifiability has to do with the heliocentric theory in either the time of Aristarchus or later with Copernicus. The stars in the sky should move relatively as the year advances and one should be able to calculate the stellar parallax from one point in the year to another months later. This theory was seemingly falsifiable. It was said that stellar parallax would be observed, but the results were null. No stellar parallax was observed, and thus this theory was ruled out by satisfying the principle of falsifiability. It has since then turned out that the stars are very far away and that it is the case that heliocentricism was correct.

To conclude, falsifiability as a criterion can fail in at least two ways: (1) A theory can be deemed unfalsifiable and yet turn out falsifiable at a later time. (2) A theory can be deemed falsifiable and be ruled out by observation, but at a later time it can be observed to be the case. The commonality between these two cases had to do with the technologies of the times. The contingency of invention does not seem like a good arbiter of the nature of the universe; in fact, itâ€™s rather arbitrary to believe that our current technology will banish or uphold theories against the whims of time. Untimeliness clearly cannot factor in when considering scientific theories or theories in any other area of life.

ladyjane · February 7, 2008, 5:14am

yes
but that is childs play compared to Deans theorem
but get your mind around Deans theorem

gamahucherpress.yellowgum.com/bo … nality.pdf

Anthem · February 7, 2008, 5:56am

There’s nothing to get your mind around. colin’s mindless dribble is hardly passable as language, let alone a theory.

Nothingness - As a head’s up, ladyjane = colin leslie dean.

I see nothing wrong with the principle of falsifiability…prove what is within your means. Also, disprove what is within your means. I can’t prove or disprove the flying spaghetti monster or elves, and I’ll admit that. But from a scientific standpoint, they’re meaningless to talk about, that’s all.

While he’s probably right, you should have a problem with anyone who says the flying spaghetti monster certainly doesn’t exist, though.

ladyjane · February 7, 2008, 6:05am

just like irrational numbers
cant see any
cant point to any
so i guess lying spaghetti monster or elves are just as meaningless as numbers thus mathematics

Anthem · February 7, 2008, 6:12am

/reply failure

Keep to the topic.

Nothingness · February 8, 2008, 11:33pm

Objections and Replies
1)

Originally Posted by BrJLa The logic here is terrible. The theory was not “seemingly falsifiable”. It was (is) falsifiable. The theory was not “ruled out by satisfying the principle of falsifiability”. Being able to be tested satisfied the principle of falsifiability. Theories are not ruled out (or accepted) by satisfying this principle.

It was “seemingly” falsifiable because the theory was able to be tested and failed, but later on observations vindicated the theory. If one were abiding by the principle of falsifiability at that time, the heliocentric theory had no business re-entering into the discourse. The theory was falsified based on the technology of those times. In a future time, humans had the technology to actually show that it was not falsified. What I’m saying is that sometimes past technology can lead us to exile a theory that has no business being exiled because more accurate technology gives different results. I’m not sure if you can truely hold theories as falsified for all times.
2)

Originally Posted by Randor Being serious, there’s a big difference between things that just are not ever falsifiable, nor ever will be and things that are not falsifiable because of the current limits of technology.

The problem is that we don’t truely know when some things are falsifiable and some others not. We don’t know what future technologies can reveal to us about the world.

If the ancient Greeks held onto falsifiability, then atomic theory would not have been considered falsifiable. In the 19th century atomic theory would not have been considered falsifiable. Falsifiablists would not have considered the micro-organism theories of disease falsifiable. These all could not be observed or tested for with the technologies of those times. These ideas were as good as the idea of invisible demons causing disease.
3)

The somewhat discouraging conclusion to draw from observing these two arguments is that scientists can only be certain that their hypothesis or their assumptions are false, never that they are true. The best they can do in that respect is to accumulate more and more support by confirming additional, ideally surprising, test implications.

This is what I kind of have a problem with. I’m not so sure we really can be certain that some hypotheses are false.
I hold to a fuzzier boundary between what some consider as metaphysics and science. Some see science as nothing more than the continuation of western metaphysics. One can question the validity of Popper’s distinction between “scientific statements” and metaphysical “statements.”

From wiki Logical Positivism…

Quote:
First, though, Popper’s concern was not with distinguishing meaningful from meaningless statements, but distinguishing “scientific” from “metaphysical” statements. He did not hold that metaphysical statements must be meaningless; neither did he hold that a statement that in one century was “metaphysical” while unfalsifiable (like the ancient Greek philosophy about atoms), could not in another century become “falsifiable” and thus “scientific”. About psychoanalysis he thought something similar: in his day it offered no method for falsification, and thus was not falsifiable and not scientific. However, he did not exclude it being meaningful, nor did he say psychoanalysts were necessarily “wrong” (it only couldn’t be proven either way: that would have meant it was falsifiable), nor did he exclude that one day psychoanalysis could evolve into something falsifiable, and thus “scientific”. He was, in general, more concerned with scientific practice than with the logical issues that troubled the positivists. Second, although Popper’s philosophy of science enjoyed great popularity for some years, if his criterion is construed as an answer to the question the positivists were asking, it turns out to fail in exactly parallel ways. Negative existential claims (“there are no unicorns”) and positive universals (“all ravens are black”) can be falsified, but positive existential and negative universal claims cannot, although Popper thought himself these could be deemed as verifiable[12].

Originally Posted by Arlo And yah, when Popper and Carnap and others who invoked falsifiability or something akin to falsifiability, they didn’t judge whether something was falsifiable or not based on current technology (or even future technology); it was based on whether something was falsifiable in principle. So, a theory like “the dark side of the moon is made out of jelly” would be considered falsifiable even before we had been to the moon because one can imagine a way in which it could be tested, whereas “the great spirit moves all” would not be considered falsifiable because it doesn’t suggest anything which could be disproven. I think that’s what Randor was saying.It’s pretty presumptuous to title a 3 paragraph essay “The Failure of Falsifiability” without understanding such a basic part of the theory.

The time factor was something I thought brought up some problems to falsifiability.

If someone said in Ancient Greece “all is made of atoms” they would have had no way to test for it. It was unfalsifiable. It’s the same with the germ theory of disease before microscopes. It was unfalsifiable.

Knowing the limits of our technology gives us a conception of what can be considered as a falsifiable theory or not.

This either leads us to say “no, that particular theory is not falsifiable and must be trashed” or “that particular theory falsifiable,” and if this particular theory is not confirmed by the observations, then it gets deemed false. Little did we know at the time that the particular theory was not false, our technology just misled us due to its insufficiency.

SpaceDog · February 8, 2008, 11:36pm

The concept of Falsifiability was NOT around until the early 20th-Century.

Looking for causal chains, speculation, etc. has been around since the dawn of humanity - IMHO.

Ultimately, however one approaches science, we do know it needs to be: public and “testable”.

Nothingness · February 9, 2008, 1:36am

I’m not claiming that it was around before then. I just think some history sheds light on its inadequacies. The Ordinance of Time still Judges the world. Our conceptions of observational possibilities change with time. If falsificationalism were to become a totalizing practice, I think it would actually inhibit scientific progress.

Membrain · February 9, 2008, 8:28pm

My own two examples of where I’ve seen falsifiability fail, are the concepts of objectivity and infinity.

While I think that objectivity and infinity exist based upon a number of arguments, I have heard the counter-argument that because we, by definition, can’t measure either their existence or their non-existence, that the argument is itself meaningless.

I think therein lies some evidence to support your point.