13

Fundamental to science is the concept of hypotheses being falsifiable. A falsifiable hypothesis, naturally, is one which could be proven wrong by empirical experimentation or observation.

Karl Popper advocated for "critical rationalism," and built much of his argument around the idea of falsifiable statements.

However, can we make meaningful falsifiable statements in the form of "the scientific method is right" or "the scientific method is good?" In other words, are positions based around falsifiability, themselves, falsifiable? Could we one day do an experiment to show that the scientific method does not lead us towards truth?

It strikes me as though the scientific method advocates use of falsifiable hypotheses except in the case of advocating the scientific method itself, but I cannot tell if that is because of how I interpret how one is expected to apply the scientific method, or if it is indeed intended to be treated as the exception that proves the rule.

Cort Ammon
  • 17,336
  • 23
  • 59
  • is he saying that the scientific method is right? or that every scientitic theory is falsifiable? the question is too cute as it stands, i think –  May 26 '17 at 15:07
  • 1
    @user3293056 He was advocating for the position that living one's life according to a philosophy which relied on the scientific method and falsifiable hypotheses, but its unclear to me whether such positions can withstand the rigors of their own arguments. Could I ever use empirical testing to test whether the scientific method, itself, should be followed? – Cort Ammon May 26 '17 at 15:13
  • wow, ok. sorry for ignorance, i find that startling. –  May 26 '17 at 15:16
  • 2
    We can hardly assert the "the scientific method" is a scientific theory... it does not "describe" but prescribe. – Mauro ALLEGRANZA May 26 '17 at 16:32
  • Once I had doubted about a third group of people in this SE. See: https://philosophy.stackexchange.com/questions/37811/is-the-usage-scientific-actually-scientific – SonOfThought May 26 '17 at 16:39
  • Is this question somehow related to the "Halting Problem"? – dreftymac May 28 '17 at 17:04
  • 1
    @dreftymac Not intentionally, though there are halting problem like issues that do arise in the scientific method if you actually try to repeatedly test something until it is falsified and you want to know whether you will ever stop. – Cort Ammon May 28 '17 at 17:20
  • Not sure if Quine would use the word "falsifiable", but [naturalized epistemology](http://www.iep.utm.edu/nat-epis) subjects methodology to the same process of revision-by-experience as the science itself is subject to. Some parts of what was once considered part of scientific methodology (determinism, cognitive physicalism, etc.) were indeed drastically revised. More broadly, there is no such thing as "the" scientific method, just a loose collection of vague principles made specific in specific contexts, and always up for revision. – Conifold May 28 '17 at 21:31
  • Popper does not get to define the scientific method. People from Galileo to Darwin have had little concern for falsifiability, and used other criteria to generate great science. This is simply a convenient oversimplification. You might want to see Kuhn, Lakatos, Toulmin, Feyerabend for arguments that science does not follow these rules, and that the rules it follows cannot reasonably be captured by an objective criterion. –  Jun 01 '17 at 19:59
  • [Obligatory xkcd](https://xkcd.com/54/). – Nat Jun 02 '17 at 03:07
  • @jobermark note that Popper wouldn't say that scientists use falsifiability "criteria to generate great science," either. For that he prescribes bold conjectures using bold imagination! – ChristopherE Jul 11 '17 at 23:42

5 Answers5

8

No, and we should not want it to be falsifiable, nor expect it to be on its own terms.

Philosophers of science generally — though not universally — dispute the idea that there is a scientific method, as opposed to many scientific methods. But leaving that point aside, if there were a single scientific method ...

How could a scientific method be falsifiable, and why should it be?

To be falsifiable it would need to make a lot of predictions (or retrodictions) about what we should observe. I don't think any scientific-method candidate itself makes predictions. Let's take Bayesianism as an example. It offers a method for updating beliefs in light of new evidence. What does it say we should observe? Nothing. It is silent about that.

Indeed why would we want a method to make predictions? It is a category mistake to want this from a method. Recipes for cooking pie or soup don't generally make predictions (or if they do, it's not essential to what they are). They offer advice about what to do.

It may be right that the scientific method "advocates the use of falsifiable hypotheses," (though theoretical science involves hypotheses not open to falsification). But if so, it advocates using them as scientific hypotheses, not as exhausting the kinds of useful or meaningful sentences more generally.

In Popper's case in particular, he did not describe falsifiability as a criterion of meaning — not even in the narrow sense of "cognitive significance" used by some positivists. So, no scientific method, including any form of falsificationism fails to be meaningful, or could fail to be meaningful, by virtue of unfalsifiability.

So, in short, no candidate for being a scientific method is likely to be falsifiable, and I can't think of any reason that should worry anyone.

ChristopherE
  • 5,478
  • 1
  • 18
  • 31
  • Bacon made a "prediction" that "induction by simple enumeration" would work for experimental science. That was arguably "falsified" by the practice of science. The same happened to "look for a cause when different outcomes occur in the same circumstances". Any methodological practice is in a sense a falsifiable hypothesis, or meta-hypothesis if you prefer, and many have been falsified. Perhaps, the word is used loosely, but it is not like any theoretical principles can ever be falsified in any straightforward sense either. – Conifold May 29 '17 at 03:48
  • 1
    Sure, but what you've offered is a rejection of a hypothesis made *about* a method. And as you seem to be aware, you've slipped from falsification to a more general rejection. If we use this broader idea, I think the question loses the air of paradox it has when phrased in the original way. That is, it becomes "can a method be rejected" And why would anyone say no to that? – ChristopherE May 29 '17 at 12:01
  • "Rejected" is not quite the word, one can reject something for any number of reasons, because they dislike it, for instance. The rejection has to be rationally based on experience, which is what falsification connotes, and there are many accounts of "the scientific method"(Popper's included) that do neglect to include the testing methodology they recommend into the process of testing. Even from purely technical standpoint testing and adopting/rejecting methodology rationally is non-trivial. – Conifold May 30 '17 at 15:54
  • 1
    However, that falsification "connotes" the same thing as something else does not make it the same as that other thing. Swimming and drowning both connote getting wet. If that something else doesn't involve showing that something isn't true, it isn't falsification, which is what the question is about. – ChristopherE May 30 '17 at 16:01
  • The difference is purely verbal. When we say that theory isn't "true", we simply mean that it does not "work" when tested. But that equally applies to methodology, even if we use "good" instead of "true". – Conifold May 30 '17 at 22:51
  • 1
    @Conifold only if you're a pragmatist about truth, and I am not and Popper is not. – ChristopherE May 30 '17 at 22:52
  • One may believe that theories reflect reality or not, but when it comes to testing what matters is what one does, so philosophical position on truth is moot. The substantive issue is whether one has a plausible account for revising methodology, not whether to call it falsification. If not, it should be a worry. – Conifold May 30 '17 at 23:01
  • 1
    Sure, if that's the substantive issue for you, you should pose it in a new/different question (that's not about falsification)! – ChristopherE May 31 '17 at 01:15
2

no. to think that a method could be "falsifiable" is a category error. propositions are true or false; methods are neither. "the scientific method" (never mind the fact that most scientists and phosophers do not believe there is such a thing) is not a proposition, so it cannot be either true or false, and thus cannot be falsifiable.

  • But what of the idea "the scientific method is the only [accepted] path to truth". That statement should be falsifiable. – Marxos Jun 01 '17 at 19:12
  • @TheDoctor Sure, that statement's falsifiable if it's a social observation, i.e. a statement about what a group of people accept as a path to truth. Then you could validate/falsify it by doing a survey. – Nat Jun 02 '17 at 02:50
  • 2
    @TheDoctor: statements about methods are of course falsifiable. that does not make the methods falsifiable. –  Jun 02 '17 at 21:23
0

You may find Jarvie's analysis of the normative component of Popper's 'falsificationism', cited here

enter image description here

On Twenty-Five Years of Social Epistemology: A Way Forward

Seems to me like the normative component is falsifiable, as we can discover that his suggestions make for bad science: surely he allows scientists room enough to say that a scientific theory is just bad.

Not sure about the descriptive component.

But anyway, his 'falsificationism' is not consistently wedded to deductive justification, so it probably doesn't matter. i.e. the weary falsificationist could merely say that it's not true that real science is such and such, but the claim is justified etc..

0

The scientific method itself cannot be right or wrong. It's a checklist a scientist must complete before publishing their findings. What is falsifiable is the assumption1 behind it, proposing that the scientific method would produce useful results.

For example, science's sucess depends on the laws of nature stayng immutable.

1 That assumption itself is irrational. As such, it is neither a scientific theory, nor it is referenced by the scientific method. Instead, this and other assumptions about reality add up to Søren Kierkegaard's "leap of faith." Specifically, it's a leap out of "cogito ergo sum" trap and towards the One and only explainable reality that we all share and are a part of.

Yuri Zavorotny
  • 590
  • 2
  • 10
-1

It is fair to say that the principle of falsifiability is not in itself falsifiable. This was the undoing of logical positivism.

However in the context of natural science, it is more accurate to say that science is nowadays defined as the application of falsifiability to the study of nature. If you question falsifiability, you are kissing science goodbye.

Other modes of study are possible, but they are unscientific. That is not to say that they are necessarily irrational or never yield useful insights to the investigator, just that those insights need not be falsifiable and anybody who suggests they are is indulging in pseudoscience. Most interpretations of quantum mechanics and some even more abstruse theories such as string and M-theories fall victim to such criticisms.

Guy Inchbald
  • 2,532
  • 3
  • 15
  • This is not how science is *defined* by any stretch of the imagination. Science is defined by the application of methodological 'best practices' within a given topic area. Falsification isn't a scientific principle as much as a rule-of-thumb used to avoid irritating discussions. – Ted Wrigley Jul 27 '20 at 17:48
  • @TedWrigley "How is science defined?" is a question all its own. I contend that your "best practices" necessarily involve falsifiability. – Guy Inchbald Jul 28 '20 at 07:57