Saturday, July 16, 2011

Scientific Method in Practice (Pt. 6)

In this series of posts, I'm re-reading Hugh G. Gauch, Jr.'s philosophy of science textbook Scientific Method in Practice (Google Books).

[Series Index]


Scientific presuppositions are about leaving certain philosophical debates to philosophers by assuming common sense answers instead of expecting scientists to either:

1) Solve these debates for philosophers, or...
2) Conduct scientific inquiry in a way that doesn't take sides.

That second option might seem fair and reasonable, but we're talking debates like:
  • Does anything exist outside of our minds?
  • If event A happens a million times in a million opportunities, is event A more likely than not to happen at the next opportunity?
  • If multiple explanations can fit the facts, is it ever reasonable to prefer one over the other?
The common sense answer to all three is: yes. (And I don't just mean 'common sense' as naive prejudice; practically everyone lives their life under these assumptions.) Let's look at how the supposed virtue of staying open on these questions leads to a very strange approach to life in general, and how presupposing their common sense answers helps to address the issues raised in Part 5 of this series.

If multiple explanations can fit the facts, is it ever reasonable to prefer one over the other?

Your cellphone has gone missing. One explanation might be that you put it down somewhere and forgot about it. Another explanation might be that a team of foreign spies stole it while you were in the shower. Yet another explanation is that a time traveling alien collected it for a museum.

All three explanations fit the facts. As far as purely deductive logic goes, there is no reason to prefer one over the other. Yet we aren't paralyzed with indecision every time someone points out an outlandish, alternate explanation that could possibly be true. We prefer simple and ordinary explanations.

Remember the problem of underdetermined theory from last time? Same thing. It may be possible to think of outlandish alternatives for any set of observations, but it's inconsistent to let this paralyze scientific inquiry if we accept the notion of explanation-preference in general.

If event A happens a million times in a million opportunities, is event A more likely than not to happen at the next opportunity?

To reuse an example, suppose you have a small bag of marbles from which you keep drawing one marble, checking its color, and putting it back into the bag. After a million times, every drawn marble has turned out to be red. Is the million-and-first marble more likely to be red than non-red? Nearly everyone would answer 'yes!' because we think repeated trials count for something. Maybe not certainty, but surely they count for probability, right?

Or consider traffic lights. We trust that traffic light controllers are reliable enough that a green light on our own street almost certainly means a red light on the cross-street.

Yet Karl Popper's falsification only view of scientific learning denies any probability-increasing role for repeated results.

Does anything exist outside of our minds?

I don't need to explain how strongly we assume things exist outside of our own minds, which is why I find the popularity of Thomas Kuhn's incommensurable paradigms portrayal of science so puzzling (pun intended)!

If you'll recall, Kuhn portrayed scientific progress during normal periods as being judged against a set of puzzles scientists happen to be interested in at the time. After a paradigm shift, the puzzles and the way of judging can be so radically different there is no way to say science has progressed overall.

But if there is a mind-independent reality, we do have something to serve as a 'common measure' for scientific beliefs across history. Progress is a matter of how true these beliefs are.

A Fundamental Assumption
The physical world is orderly and comprehensible.1
This may be the only basic presupposition (or pair of presuppositions) science needs. It implies a world 'out there' which we are capable of understanding. This wouldn't be possible if our senses were utterly unreliable, or if past experience didn't count for anything, or if every logical possibility were on equal footing.

Taking Aim at Science

If a philosopher questions science's ability to get at the truth, the first thing to do is ask whether the presuppositions above are being challenged. If so, then science is not the primary thing under attack.
Science begins with the presumption that the physical world is comprehensible to us [....] Therefore, any argument that assays to take down common sense and thereby to establish radical skepticism is simply outside science's purview. Science has many theoretical and practical tasks, but refuting skepticism is not one of them. A legitimate attack on science's rationality must target science alone, not science and common sense both.2
So while science does require presuppositions, they aren't controversial ones. In fact, failing to make these basic assumptions would be far more questionable. Gauch sums this up neatly with his 'reality check':
Moving cars are hazardous to pedestrians.3
Do you accept this statement as objectively true and knowable? If so, you already accept the kind of mundane assumptions science needs to get started.

1. Gauch, H. G., Jr. (2006). Scientific method in practice. Cambridge: Cambridge University Press. p. 138
2. ibid. p. 106
3. ibid. p. 134


  1. Please unpack the "Fundamental Assumption". It appears to me from your post that it is really a host of assumptions - is this Gauch's position? I would enumerate the ones you raised as:

    1. Realism: There is a Reality that is independent of our beliefs about it.

    2. Rationality: The proper way to evaluate Truth about Reality is by Reason and logic.

    3. Reliabilism: Our senses are Reliable, in the sense that they reliably produce similar experiences/sensations for similar data.

    4. Internalism: Our memories of our experiences and the stability and rationality of our mental processing are generally reliable.

    5. Orderliness/Externalism: There are regularities in the world.

    and putting it all together,

    6. Comprehensibility (The Fundamental Assumption): The application of Reason by our Internal reasoning to the Reliable data our senses produce from the Orderly Reality can increase our knowledge of truth about [i.e. correspondence of our beliefs with] Reality.

    The above can be put more accurately. Note in particular that assumptions 2 and 3, at least, make a "Local Time & Causation" assumption - we are assuming that we're discussing the the Reason of an agent from within a causal world: the application of Reason to a stream of causal data in time. It is also not clear to me whether we can really make do without an explicit "Realism" assumption - the assumption that our "Folk Physics" is close-enough to the Truth so that e.g. notepads really do preserve what's written on them.

    The gist of my comment is that there isn't really just one Fundamental Assumption. There are a whole host of such assumptions, and I'm disappointed Gauach didn't elucidate them more.

    Also, I don't think science really assumes that the world is Orderly; it just assumed there are *some* regularities, and tries to establish them and build on that.

    I likewise don't think science assumes that the world is Comprehensible; it just assumes that to the extent that it is, the Scientific Method is the way to obtain that comprehension.


  2. Yair,

    The 'fundamental assumption' is called 'Science's Comprehensive Presupposition' by Gauch and he writes that "science's presuppositions do not individuate uniquely" (p. 138). So there would be a variety of ways to unpack the implications here.

    I'm more interested in knowing whether science requires any presuppositions that are independent of the comprehensive one, than arguing about how best to break it down.

    In broad strokes, the idea is that science can ignore debates about radical skepticism from philosophy and build knowledge about the world from a very mundane starting point.

  3. The idea that "science's presuppositions do not individuate uniquely" is... interesting. I can't see how it is true, fundamentally. I can see better or worse formulations and minor variants, but basically it appears to me that their essence is pretty much the same in either case. Maybe I'm wrong, though. I certainly haven't seen any proper discussion of this issue.

    As to whether any presuppositions are needed beyond the "Comprehensive" one, well, I can't see any per se, but one deep presupposition that is raised explicitly in physics is that the past happened in a "reasonable" way. This is not the same thing as assuming that it was Ordered. What I'm thinking of here is the idea of Boltzmann brains, in the context of cosmology and the second law of thermodynamics.

    You also need to reject more than just "radical" skepticism. You need, for example, to willingly extend the demand for the maintainance of the Order science reveals to far more than the "very mundane starting point" or common sense. It isn't trivial to assume that life didn't begin with a poof of magic by God a few thousand years ago, or that Israel suffered draught because of climatology instead of God's wrath. These extensions go far beyond mere "common sense" or the mundane starting point.


  4. On second thought, I find that science doesn't quite presuppose Order and Comprehension. Rather, it hopes for them. It supposes variants of them tentatively, as hypotheses, rather than as presuppositions.

    For a fuller explanation, I wrote a post.