What is Falsification, and Who is This Popper Fellow, Anyways? -Part 2- Objections to Popper

-a younger Popper- 

Part 1: The Problem of Induction

Has Popper left us where we started?*

If the raison d’etre of Popper’s solution was to argue away the problem of induction by claiming deduction was all that science required, he hasn’t done a very good job of convincing me. There are a number of problems with Popper’s view:

  1. Dogmatism: Popper assumes there must always exist a true counter example to every hypothesis; we have no a priori reason to assume this is true.
  2. Vagueness: what are “bold conjectures”? What is “rigorous testing”? How do we come up with such bold conjectures in the first place, and devise such tests (according to Popper, this is a problem for psychology, not philosophy)?
  3. Popper’s flavor of induction, elaborately disguised as corroboration, fails to provide a rational means of choosing a corroborated theory over an infinite number of other possible, unrefuted hypotheses. For example, if one plots the number of posts I’ve written for Terry per month over the past 5 months, you get something that looks like:

Monthly posting figure 2

The linear equation that describes this relationship could then use to predict the number of posts I might write in my sixth month. Now, intuitively we might see this and think, “Yea, that looks linear enough. Its probably a linear relationship.” Unfortunately, we have no rational bass for making this assertion. For example, the following two lines are just as likely to represent the true relationship between my months having blogged, and the number of my posts to appear each month:

Monthly posting figure

Popper has not given us any reason to accept the linear relationship over an infinite number of other relationships.

Balderdash! Malice! Dave, what about testing the hypotheses? You can’t tell me, with all sincerity in heart and mind, that testing does nothing to corroborate a theory, can you? Surely, filling in the spaces between those points would give the hypothesis more clout, right?”

Yes, I can. And no, it wouldn’t. Under Popper’s criteria, namely that induction cannot be rationally used to prescribe predictive content to a hypothesis (i.e. that a hypothesis has predictive power).  We’re back where we started: the problem of induction (i.e. how can we assume regularity?).

Now, you might be tempted to cite Occam’s razor, the notion that the simplest explanation (that is, the explanation requiring the least number of assumptions, an idea at the heart of modern reductionist science) is most likely to be true. However, simply proposing a concept to be true without any reason why (except to make life easier) is not the aim of philosophy, but rather that of pragmatism.

“No no no Dave, you’ve got it all wrong.  Popper wants us to choose a corroborated hypothesis over a non-corroborated one because its the best choice available.  I would rather rely on a corroborated theory over a non-corroborated one. Furthermore, a corroborated hypothesis likely gives us something that is closer to the truth than uncorroborated hypotheses – at least its been proven in the past!”

A student of Popper, J. Watkins, made the former argument, while Popper made the latter.  Neither is convincing, and both really do require inductive inferences to choose one particular corroborated hypothesis over another (even though both would argue they don’t, with little avail).

“Alright smart ass, let me ask you this.  If you were dieing, and you were given two options, either to take an untested experimental drug or a tried and tested, safe drug known to work very well, which would you take? It’s obvious, the tried and tested one!  How can you sit there and tell me corroborated doesn’t mean anything?”

Great question.  I feel this is where the practice of science, and the philosophy of science separate.  Although logically circular, the largely inductively based modern scientific method has been successful in producing highly predictive hypotheses, even though the preceding arguments suggest it is irrational for science to have proceeded as it has.  As such, since the primary driving agent of scientific progress is arguably the individual scientist, it appears the psychology of the experimenter plays an important, although not intrinsically rational, role in the success of a given hypothesis.

*Note: I am neither a philosopher, nor a philosophy graduate student. I simply find this stuff fascinating. If you take offense to anything in the above article and feel obliged to correct me in the comments field, please do so with the most honorable intentions in mind. Crushing my spirit will do neither of us good (unless you sustain on spirit juice, perhaps mixed with some simple syrup, crushed ice, and mango puree. In that case, I can’t really say no, can I?).

Related Topics


Dave Semeniuk spends hours locked up in his office, thinking about the role the oceans play in controlling global climate, and unique ways of studying it. He'd also like to shamelessly plug his art practice: davidsemeniuk.com