You can think of human knowledge acquisition as the construction of mental models / patterns.
A model works by being sufficiently true / realistic, to help you to predict real-world effects with enough accuracy, to inform your decision-making.
Models whose accuracy / veracity can be tested in real life, are considered “scientific” (taken loosely).
Those that cannot be tested, are simply called “unscientific”. “Metaphysics” is also a term used to describe such models. A high proportion of religious dogma also counts under this category, especially those that talk about “facts” about the past, present, future, etc., that are beyond human observation / testing. (E.g. “believe X untestable fact to get Y untestable effect in the afterlife”, “God answers prayers”, etc.)
Then there is another class of thoughts / ideas I call empty word games. Strings of words that people mistakenly believe mean something when it’s complete gibberish.