What Is Science?

• Science cannot prove anything true. That's what mathematics does. Formal logic (propositional calculus, predicate calculus, and its relatives) can be viewed as mathematics stripped of content. Mathematics, and its relatives, are incredibly sensitive (by empirical standards) to initial assumptions and conditions.
• Statistics is the branch of mathematics closest to science. (In some universities, it even rates its own department). Statistics answers the question, "How plausible is it that these two data sets really are about the same thing?" While the creation of the theory is mathematics (and thus proves things true), the application of statistics is about whether two things are acting differently. This is why proposed pharmaceutical treatments are compared against placebos.

  Whether a placebo actually has clinical effects against specific pathologies/symptoms is controversial. From the viewpoint of scientifically evaluating a pharmaceutical treatment, it doesn't matter. If there is no measurable difference in effect (on a specific condition) between a isotonic saline solution and a solution of medicine X, why should you (or your health insurance plan) pay for that solution of medicine X? If medicine X is measurably worse (on a specific condition) than that isotonic saline solution, shouldn't it be malpractice to use it (on that specific condition)?

• Science is not about how to think about the world around us. That is the job of religion (or lack thereof), and philosophy. [This is a Western distinction...Eastern cultures often consider these as one subject. Use both approaches as each proves convenient.] Some choices of religion and/or philosophy define science as impossible. While the linguistic distinction that enabled the discussion of science as a separate discipline from magic is first recorded in 4th century BC Greece, the first era (and region) when a religion/philosophy that considered science possible seems to have been Renaissance Europe somewhen between 1500 and 1550. [The last comment is based on a change in recorded worldview: a consistent attempt to cooperate with nature, rather than defy nature, when attempting to gain control over natural phenomenae. Yes, I'm glossing over a lot of history here.]

  Obviously, one's choice of religion and/or philosophy will seriously affect what questions one would subject to science. Red China, for instance, has a much easier time scientifically investigating applications of acupressure and acupuncture than the United States.

  Disasters can happen when something designed using science is abused by operators who are oblivious to the consequences of flouting the design basis. An example is the Chernoybl reactor meltdown/explosion (1981, Ukraine -- the Soviet Union was the government in effect then). This happened because the plant supervisor deliberately, with reasonable understanding of what could happen, ran the reactor without coolant in order to keep his job.

  A more recent example (Japan, 1999 or 2000), again from the nuclear power industry, happened when a nuclear power plant holding company doctored its safety manuals to authorize the direct mixing (by hand) of uranium oxide solutions with insufficient safety factors in quantity. That is, the company directly authorized violating Japanese safety regulations as standard procedure. Three misinstructed workers (following standard procedure), ignorant of what could happen, mixed together a sufficient quantity to cause a momentary uncontrolled fission reaction. This wasn't enough to heat the water -- but the bright blue flash came with a radiation dose sufficient to immediately knock them out. (The least-exposed worker took about 21 sieverts [very roughly 35,000 rem]. 5,000 rem is sufficient to cause cerebrovascular collapse within 24 hours.) The most advanced hospitals in the area were unable to keep the other two workers alive.

• All scientific statements have a way to show that they are wrong, by definition. In order to do this, it usually is necessary to severely control the circumstances. As a result, science is concerned with how things act now. Scientific facts (i.e., all attempts to show them false haven't worked) may be extrapolated (mathematically or logically) to try to understand the past or predict the future.
• A necessary metaphysical assumption, when extrapolating scientific facts to understand the past (or predict the future), is that the "rules don't change". That is, we smoothly extrapolate what we believe works now as far as we can in space and time. For instance, the steady-state cosmology explained the apparent equal prevalence of galaxies in all directions (after accounting for the blocking effects of our own galaxy, the Milky Way). However, it could not deal with Olber's Paradox (Heinrich Olbers, 1758-1840): why is the sky dark? After all, if space and time are truly unbounded, should not all lines of sight eventually intersect a star (or something sufficiently heated by a star)? Then testing of radio telescopes found a redshifted, practically isotropic thermal radiation (in the radio range) across all of the sky...this provoked the Big Bang cosmology, which had to be modified (by inflation) to explain why the radio background radiation is in thermal equilbrium. Note: the Big Bang is also a smooth extrapolation of the current known physics.
• This, incidentally, explains how the life cycle of stars can be discussed: while we do not have the experience base (including both historical and collective personal) to watch the whole process, professional astronomers (with supercomputer assistance for some aspects) can combine the observations of millions of stars, with supercomputer emulations of these stars, to make informed estimates of things such as lifespan from other things like observed spectrum, brightness, and parallax (or lack thereof). I have already used this assumption once: when explaining the use of the Hertzsprung-Russell diagram to estimate relative distances of star clusters. We also assume, when doing these extrapolations, that we do not find unexpected discontinuities in spacetime either backwards (religion: statements about creation) or forwards (religion: statements about eschatology).
• Black holes are expected discontinuities in space-time, as is the Planck time for the Big Bang (when the projected density of the universe approximates a collection of point masses whose quantum wavelength equals its Schwarzschild radius]: the mathematical model tells us that it is incorrect! For black holes, the mathematics breaks down well inside the event horizon.
• To visualize what an unexpected discontinuity is, drop a ball from a (small) height. There is no reason to expect the ball to hit the ground unless you look at where the ball is going (the ground is in its path). From the viewpoint of science, as long as there is no physical evidence of a discontinuity, it is meaningful to assume there is no discontinuity when doing these extrapolations. The near-isotropy of the background microwave radiation, with other considerations, suggests that we (or at least those who take the time to learn all of the physics required) can do so at least as far back as when matter ceased to be a plasma. This is as far back as we can see physically: a time when the average temperature of the universe was roughly 3,000K. All portions of the universe further away than this point are not visible to us with any sort of electromagnetic telescope (optical, radio, x-ray, gamma-ray, etc.). [This is how the Big Bang theory disposes of Olbers' Paradox.]
• Theoretically, one could look for evidence of an unexpected discontinuity in spacetime. Obviously, such an exercise is much more feasible looking into the past, than trying to work forwards. The microwave background radiation, and other considerations, suggest that there is probably no cosmological-scale evidence of an unexpected discontinuity in space-time.

Return to the Signs In The Heavens page, or the Math page, or the main page.