- by Greg Glassman
Modern science is source and repository of man’s objective knowledge. [MS001] Scientific knowledge is siloed in models. [MS002] A model maps a fact to a future unrealized fact as a prediction. [MS003] A fact is a measurement. [MS004] A measurement is an observation tied to a scale with an expressed error. [MS005] An observation is a registration of the real world on our senses or sensing equipment. [MS006]
A model’s validation derives entirely from its predictive power. [MS007] There are four grades of model ranked by predictive strength and they are conjecture, hypothesis, theory, and law. [MS008]
A conjecture is an incomplete model or analogy to another domain. [MS009] A hypothesis is a model based on all data in its specified domain, with no counter examples, and incorporating a novel prediction yet to be validated by facts. [MS010] A theory is a hypothesis with at least one nontrivial datum. [MS011] A law is a theory that has received validation in all possible ramifications, and to known levels of accuracy. [MS012]
Predictive power is evidence of and reason for sciences’ objectivity, the sole source of science’s reliability and the demarcation between science and pseudoscience. [MS013] Predictive power as determinant of a scientific model’s validity provides the basis for any rational trust of science. [MS014]
Models are predictions mapping a fact to an unrealized fact where the current fact constitute the premises and the unrealized fact the conclusion of an inductive argument. [MS015] Induction derives conclusions from premises with probability and not certainty. [MS016] All scientific knowledge is therefore the fruit of induction validated by predictive power which is a measure of probability. [MS017]
It is important to note that validation comes independent of method. [MS018] Models derived from inspiration or perspiration both rank by predictive power alone. [MS019] It also warrants mentioning that modern science is an extension of logic and therefore must be consistent with the rules of logic, language, and mathematics. [MS020]
References
2 Comments
Leave A Comment
You must be logged in to post a comment.
This could be taken one step further in that the predictive power of a model can only be tested if the experiment is replicated under the same (or similar) conditions and the facts are replicated within the same (or similar) margin of error. This puts extra weight on the method of measurement, i.e. how the facts were obtained.
Greg: I find this quite fascinating. Some of your statements I have discussed in the first chapter of my not-very-widely-read book, “Science Under Attack: The Age of Unreason” (2017) which is available on Amazon. My blog of the same name covers a variety of specific attacks on science, primarily on the topic of climate change.