What is science? The guiding principle of scientific inquiry was perhaps best summed up by a non-scientist, the novelist Arthur Conan Doyle, who placed into the mouth of his character Mycroft Holmes these words:
What is Science?
"To solve a mystery the first thing it to eliminate the impossible options. Whatever is left, however improbable, is the truth."
The modern practice of science is based on a method that employs that principle, with a little twist.
Key components of the scientific method:
Key concepts and vocabulary. Understand these or you're toast:
- Observations: Any impression of the physical universe that comes to us directly through our senses or our indirectly through instruments. E.g. "I looked out my east window this morning and saw the Sun rise."
- Repeatability: The ability of two different individuals operating under the same circumstances independently to make the same observation. E.g. "My neighbor down the block also saw the Sun rise in the east."
Some types of observations are inherently non-repeatable. E.g. "My wife is the most beautiful woman in the world." Sweet, but unlikely that all observers would have the same impression.
- Patterns: Tend to become apparant when we make and compare many observations. E.g. "The Sun appears to rise in the East every morning." Humans are compulsive pattern recognizers, both seeing real ones and confabulating ones that aren't really there. (The basic principle of Rorschach tests.) The scientist's task is to distinguish the two.
- Hypothesis: A formal statement about a pattern of observations whose truth or falsehood is initially unknown. E.g. "The Sun always rises in the East." Remember, a hypothesis is a statement about a pattern of events, not a single observation.
- Hypothesis falsification: the scientific method does not absolutely prove anything. It can, however, disprove things. We do this by looking for observations that are inconsistent with our hypothesis. When a such an observation is made and verified, we consider the hypothesis to be disproven or "falsified." E.g. If the Sun were to rise in the North one day, the example hypothesis given above would be falsified. Obviously, being falsifiable is different from being false.
- Falsification is definite - "proof" is unattainable: When an inconsistent observation is made, the hypothesis is definitely falsified - i.e. it's dead dead dead. No amount of consistent observations will bring it back to life. But even if we see ten, a hundred, or a gajillion observations that are consistent with our hypothesis, we can never be completely sure that the next observation won't falsify it. Thus, one never definitively "proves" a scientific hypothesis.
- Non-falsifiable hypotheses: Hypotheses that are inherently impossible to falsify, either because of technical limitations or because of subjectivity. E.g. "Chocolate is always better than vanilla." [subjective]. Or "There are living beings in the Andromeda galaxy." [beyond technical grasp] Another term for non-falsifiable hyoptheses is "speculation." I'm not saying that non-falsifiable hypotheses are bad or not worth pursuing. Lots of great art, literature, philosophy, and religious thought is deeply (and rightly) concerned with them. They just aren't a basis for scientific inquirey. They are beyond the grasp of science.
- A problem: Suppose you have a falsifiable hypothesis that consistently resists all attempts at falsification. At a certain point, it becomes sort of perverse to not provisionally accept it as "truth." Often, we see groups of such hypotheses dealing witht he same general issue.
For example, humans have been watching the sun rise for over 100,000 years, and is always rises in the East. Not only that, the moon, stars, and planets also rise in the East. Such an amazing congruence of provisional "truths" about celestial bodies gets us thinking about the underlying mechanisms governing them. From this we get theory.
- Theory: Set of universal rules that explain wide ranges of falsification-resistant hypotheses in terms of underlying processes. E.g. We now have theory that explains our earlier hypothesis about the Sun rising in the east: "Inertia causes the Earth rotate in the same direction on its axis once every day, causing observers on its surface to observe celestial bodies appear to rise in the East every day." These rules allow us to understand not only the sunrises we witness directly, but the rising of all celestial bodies everywhere throughout time.
WATCH OUT! In common speech, people often use the terms "theory" and "hypothesis" as synonyms. That's not the proper scientific way! When we refer to the "theory of the expanding universe," we are not suggesting that there's anything hypothetical about it. On the contrary, we are saying that the patterns of observations that lead us to conclude that the universe is expanding are so well attested that they merit the elucidation of the underlying processes governing them. In the exact same sense, when a music professor teaches "music theory," they aren't in any way suggesting that maybe music really doesn't exist.
- Overturning theories: Still, because hypotheses can be falsified, the theories to which they give rise can be overturned. The difference is one of scale. Hypotheses are falsified every day - trivial events. The falsification of a scientific theory is a big deal that gets into newspapers and history books.
Two ways to overturn a theory:
- Observation: New observations are made which are inconsistent with the theory. E.g. Einstein's theory of general relativity superceded Newton's theory of universal gravitation because it better explained several astronomical observations.
- Parsimony: A simpler theory is preferred by the principle of parsimony (= simplicity). When two or more posssible answers exist to a question, the simpler one is generally preferred. E.G. Copernicus' heliocentric astronomy superceded Ptolomey's geocentric astronomy because it was simpler. Actual observational rejection of Ptolemey's astronomy didn't come until a century later.
- Non-falsifiable hypothesis