Monday, May 2, 2011

About this Blog

Instead of giving this blog the title: Evidence Trumps Common Sense, the title of this blog might more properly have been:

"Prospectively obtained evidence, obtained in the context of an experiment to test a hypothesis, is more reliable than common sense as a basis to judge the validity of that hypothesis."

But that title would have been a bit too long and perhaps still not clear.

So, let me explain.  Human beings make decisions continually.  What to prepare for breakfast, how to walk across the room to avoid obstacles, when to step on the brake to stop at a red light, which Presidential candidate to vote for, etc.  Our brains evolved to equip us to make these decisions.  Our brain circuitry incorporates a basic structure dictated by our genes but also incorporates a massive amount of information generated from our entire life experience of interacting with the external world.  Using these resources one’s brain constructs an internal model of the external world.  Each of us utilizes this internal model as a guide in making decisions and anticipating what the consequences of our individual actions might be.  For example, one might anticipate that if one went up to a big strong fellow, whom one didn’t know, and threw a pie in his face – the consequences might not be pleasant. As a consequence, one would likely make a decision not to take that action, even if one had a perverse inclination to do so.  One similarly might decide not to swim in an alligator infested swamp.

Common sense is the judgment one’s brain makes about a certain situation based on applying the brain’s internal model of the outside world to that particular situation.  Our common sense is often expressed as a set of beliefs, as judgment, as experience, or as philosophy.  We couldn’t function on a day to day basis without common sense.  However, common sense is not infallible.  One’s range of experience may not encompass every new situation, and even if one has appropriate experiential data, one may not have interpreted that data correctly.

Here are a few examples where common sense has failed:
  • As mentioned in an earlier post, our common sense dictates to us that time is immutable  - a clock placed on a ship or an airplane should run at the same pace as an identical stationary clock.  Yet Einstein’s theory of relativity violates common sense in demonstrating that clocks advance at different rates when they are moving with respect to each other.   The difference in rate is only appreciable when the relative motion of the clocks occurs at a speed close to the speed of light.  As a result, an astronaut sent out on a journey on a spacecraft that can accelerate up to a speed close to the speed of light might return home one year older than when she left.  However, the earth she returns to could be 100 years older and, under such circumstances, her entire generation would have died out.
  • We observe that all functioning complicated machines have been carefully designed by intelligent engineers.  Thus our common sense dictates that living things, which are fantastically complicated machines, must have been similarly designed.  However, Darwin’s theory of evolution demonstrates that, in fact, living things were not designed, but evolved through a process of random variation and natural selection occurring over billions of years.
  • We observe that alcohol abuse is the cause of much illness, death and social dysfunction.  Our collective common sense thus dictated at one point in US history that alcohol should be prohibited.  However, prohibition in practice fostered widespread crime and corruption and failed to eliminate alcohol consumption and, as a result, was ultimately repealed.
Does this mean that common sense is usually wrong?  No, in fact in most instances it is probably correct.  However, our experience is too circumscribed, our powers of analysis too limited, and the world is too complex for us to rely on common sense in all circumstances.  We should realize that common sense is just one basis for coming up with a hypothesis for what the outcome of a specific action might be.  Whenever feasible this hypothesis should be tested with new data generated in an experimental test of that hypothesis. This is particularly so if the impact of the contemplated action is likely to be large.

In this regard, not all evidence is equal.  We look back in time at existing evidence to generate a hypothesis (in many instances existing evidence gets incorporated into our common sense).  A pattern we discern in the existing data goes into our hypothesis.  However, retrospectively detected patterns are not proof – random variability generates things that look like patterns but in fact are just the result of randomness. A hypothesis can only be properly tested with, new prospectively obtained, data generated specifically to test that hypothesis.

It is not feasible to conduct an experiment to test every hypothesis.  However, particularly when making really important decisions or setting major policies we should attempt to do so.  Examples include government decisions about health care, public health, national security, energy policy, financial regulation, drug policy, crime prevention, etc.   Policy decisions should be driven by evidence, hypothesis testing, science and not just politics and political doctrine.  At the very least we should realize that because our common sense dictates a certain belief, that does not make that belief true.  We should be open to examining and testing all of our beliefs.

See also More About this Blog.


  1. How are common sense and induction related?
    What do you think about C.S. Peirce's epistemological views?

  2. This comment is in response to Anonymous who asked about induction and C.S. Peirce. With regard to induction I believe that there are two principal uses of the term. Mathematical proof by induction is a rigorous procedure (and is actually deductive in nature). In philosophy, inductive reasoning involves coming to a conclusion which may be likely to be true based on various observations, but there is also a possibility that the conclusion is not true. The process of hypothesis formation, perhaps ideally, may involve both critical analysis and inductive reasoning. One critically analyzes existing evidence and by means of inductive reasoning comes up with a hypothesis that appears compactly to account for that evidence. The "conclusion" here is the hypothesis which should next be tested prospectively. If one is going to go to the trouble of testing a hypothesis, one would like to formulate a hypothesis which one thinks is most likely to be correct.

    "Common sense" is generally arrived at by an intuitive, and often subconscious, process. "Common sense" is what one's brain comes up with by integrating one's life experience. Many hypotheses are based on "common sense" alone. It is better, at least, to include critical analysis and inductive reasoning when one is formulating a hypothesis.

    I don't know enough about C.S. Peirce's work to comment on it.