Until recently, I was assured, for years, that there is only one way for a theory to be "scientific" (my definition):

An hypothesis to solve a defined practical problem, which must be falsifiable;
A scientific question should be asked (tested) to reinforce the falsifiable theory as plausible or implausible.


Recently I have learned that this "Popperian" methodology isn't plausible in what I can define as generally totally abstract sciences such as Formal Logic, Mathematics and Computer science where the "falsifiable or unfalsifiable" paradigm doesn't catch because everything is "abstract" anyway.

Are there two types of scientific theories (one materialistic and one mathematical)?

  • 3
    Correct: mathematics and logic do not fit under Popperian's category of falsifiable. Feb 21, 2020 at 14:28
  • See The “science” of pseudoscience: "The English word “science” is primarily used about the natural sciences and other fields of research that are considered to be similar to them." Feb 21, 2020 at 14:31
  • 1
  • See K.R.Popper, Conjectures and Refutations, page 67 : "The Greeks' discovery of the critical method gave rise at first to the mistaken hope that it would lead to the solution of all the great old problems; that it would establish certainty; that it would help to prove our theories, to justify them. But this hope was a residue of the dogmatic way of thinking; in fact nothing can be justified or proved (outside of mathematics and logic)." Feb 21, 2020 at 15:17

1 Answer 1


I will confess up front that I am honestly confused why people still cling to Popper's work.
Even Popper himself eventually came to admit that his philosophy of science wasn't much more than an abstract aspiration, having little relation to the actual practice of science. Popper had some useful insights, sure, and an important place in the history of the philosophy of science, but seriously... Time to let that old dog rest.

Whinging aside, if we ask how many 'kinds' of science there are, we're going to get different answers depending on who we put the question to:

  • Popper would say there's one thing called 'Science' that is done within certain limits, and nothing else applies.

  • Kuhn would argue that every field creates a version of 'science' around its own central paradigm, and that even these paradigms can occasionally shift so that fields might have different understandings of 'science' at different points in their history.

  • Feyerabend (whose name I always misspell, dang it) would argue that there is no 'science' in the general sense; there is only the singular behavior of individual scientists in their practice, with a wide multitude of 'sciences' that results from all their efforts.

Popperian falsification as non rigorous

The common element of most philosophies of science is the observation that science is an ongoing practice: an activity/verb, not a thing/noun.
In this sense, when we try to think about whether something is 'scientific,' we don't mean whether it is/isn't within the class of objects labeled 'science.' Instead, we are asking whether the topic is approached in a manner that conforms to certain analytical/methodological constraints. This was part of the problem with Popper's falsificationism: the act of 'falsification' is not itself methodologically rigorous.
Just to take your example, let's say I wanted to 'scientifically' test math. Simple test: I decide to see whether 1+1=2 in some concrete context.

  • I take one apple and I put it on a table
  • I repeat that process;
  • I count the apples on the table.
  • If there are more or less than two apples on the table, then I've falsified addition!

Of course, there are two apples on the table (unless I got hungry), so addition survives the first effort at falsification. So I try again:

  • I take one water and pour it into a bowl
  • I repeat the process
  • I look in the bowl and count the waters I have
  • If there are more or less than two waters, then I've falsified addition!

Oh, err... Everyone will object that I've made a mistake by using 'waters' (which are not properly 'countable') and so addition has not actually been falsified. Popper would be forced to say, here, that this kind of manipulation — where mathematicians redefine the subject matter with new concepts (like 'countability') in order to preserve the core theory — means that mathematics is unfalsifiable, and thus not scientific. But as a process this is a perfectly normal scientific activity, where researchers redefine and recategorize observable data in order to make theories more exacting and coherent. So don't place quite so much credit on Popperean accounts of 'science.'

  • 1
    Whether it's Popper's precise formulation or not, there must be an aspect of testable predictions for a theory to be scientific. It is not necessarily Popper-like "falsifiable" but it must be distinguishable from other theories, at least in principle. Also, your examples don't test addition, they test the mapping of the operations onto addition.
    – puppetsock
    Feb 21, 2020 at 19:16
  • 2
    @TedWrigley What do you mean Einstein had no conception how to measure predictions? His original paper on GR suggested the bending of sunlight during an eclipse as a test. en.wikipedia.org/wiki/… And his original SR paper explained several existing measurement problems.
    – puppetsock
    Feb 21, 2020 at 21:26
  • 1
    @TedWrigley Wow dude! You've re-invented the "no true Scottsman" fallacy. Well done.
    – puppetsock
    Feb 21, 2020 at 21:35
  • 1
    @user4894: "whinge | (h)winj | British : verb [no object] to complain persistently and in a peevish or irritating way." Feb 22, 2020 at 21:01
  • 1
    Fortunately, science will still get done, even if the philosophy is not complete and universally agreed upon.
    – sdenham
    Feb 23, 2020 at 4:44

You must log in to answer this question.