An earlier question Can explanatory historical accounts (as opposed to descriptive ones) ever be falsified? seems to contain the opinion that descriptive accounts can be falsified. In connection with the question Did Jesus live? (Skeptics.SE) this particular descriptive issue is discussed. A recent book with facts described in Einhorn L. A Shift in Time, discusses whether the Gospels places Jesus’ missionary period 15-20 years too early. I consider that a theory and look for a proper terminology in terms of falsifiability.

I offer a similar theory where the outcome lies in the future: determination of compliance for high level radioactive waste with national standard. In the US it would be the NRC’s standard, based on EPA’s Public Health and Environmental radiation Protection Standards for Yucca Mountain, Nevada (40 CFR Part 197). The details are irrelevant for our purposes. For the sake of argument we assume that the issue is whether in the distant future a person might suffer consequences below or in excess of a certain dose. The theory would be the applicant’s safety assessment and claim of compliance with the standard.

I see here the same characteristics as the first “theory”. Would you accept that

1 In both cases theories are used;

2 the theories are testable in falsifiability terms, in principle;

3 the problem arises in the implementation of the test?

  • To answer whether the problem arises in the implementation, the text should describe what that means. How will the test be implemented? Nov 22, 2017 at 5:16
  • We might say it can be tested in principle, in that reality will be (or have been) one or the other. However we can’t test it if it is a thousand years in the future. The past is more tricky – sometimes it can be (e.g. C-14-) tested. However many events are less known. If I say Julius Caesar had chicken for dinner at a certain date (within his life span) that would be true or false but we might not ever find out unless he was at well-known banquet that day. Nov 22, 2017 at 12:16

1 Answer 1


Just because a theory is not directly testable does not mean that it's not falsifiable.

For example, take the Big Bang. We (hopefully) cannot directly test the Big Bang. What we can do, however, is look at various models of how the Big Bang developed and look for interesting consequences that are testable. With enough predictions that enforce some theories and discount others, we could eventually get to a single theory that meets all the predictions. We could then be reasonably sure (though not certain) that we know what occurred.

To your example. We cannot test what will occur in 1000 years. However, we may be able to use our theory to make other, shorter term, predictions that we can test. If the predictions match the test then it will enhance our confidence in the theory. If they do not, then it will either diminish, or destroy, our confidence.

And this, in practice, is one of the key benefits of falsifiability. Disproving a consequence of a theory can be as powerful as disproving the theory directly. This means that the theory, itself, does not need to refer to a repeatable, testable process. Only that it leads to consequences that are repeatable and testable.

  • I note you take a practical view looking at model results. That is perhaps all Popper intended. The historical issues are perhaps a bit more tricky but I assume you take the same view for those cases. Nov 22, 2017 at 13:44
  • @MikaelJensen Indeed. I don't believe it's inherently more difficult to develop a testable, historical theory. But they're trickier due to the increased likelihood of the contamination or destruction of evidence. When it comes to falsifiability, then a results oriented approach tends to make the most sense. Other factors, like theoretical elegance, can have value but rarely help much in disproving something.
    – Alex
    Nov 22, 2017 at 14:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .