- Justifications are always with respect to some underlying set of customs for what constitutes an adequate explanation. These customs aren’t a priori “correct” and differ between cultures and individuals. They have various meta properties, justified by some means (but note Postulate 1—there is no escape).
- A justification custom may be shown to be inadequate for a task. For instance if I design a bridge, produce a guarantee of some properties, and on building a bridge closely following the design it’s shown that those properties don’t hold, then I may have to reconsider the customs I follow. (Cf. the Millennium Bridge comedy.) If I design a logical calculus then I may hope that from consistent premises it can only prove one of p or not-p: not both. So a proof of both from consistent premises would indicate that something is wrong. (Cf. Frege’s little error.)
- Failure to justify oneself adequately, or failure to comply with customs in other ways, results in punishment or attempts at correction (e.g. imprisonment, detention under section). (See e.g. Martin Kusch’s work.)
- People often invent explanations of behaviour which are inconsistent with their behaviour or the actual causes of their behaviour. (See the large body of work produced in social psychology.) The “actual causes” mentioned here must be described in a framework for explanation (see Postulate 1).
- Choosing a particular style of explanation does not rule out the possibility that one believes other styles of explanation are necessary and useful. Take the framework of “cognitive processes”. This emerged from the viewpoint that it’s helpful for explanatory purposes to abstract away from the exact medium in which data enters a human, so ignoring photons, neurotransmitters, ion flows. The other viewpoints relevant to human psychology include fMRI, animal studies, genetics, sociological theories. Many interesting studies combine multiple viewpoints.
- Someone’s past experience and education may cause them to decide that, for them, certain classes of explanation are inadequate.
- All models are incomplete, otherwise they’re not models. The map is not the territory (Korzybski). From a Koan: “If you call this a short staff, you oppose its reality. If you do not call this a short staff, you ignore the fact. Now what do you want to call this?”
- Theories aren’t used only to make predictions, they also give a framework in which to discuss phenomena with a common language. In particular this makes it possible to simplify and unify empirical evidence.
- That something we enjoy, e.g. music, art, love, is given a scientific explanation does not necessarily mean that we can no longer enjoy experiencing that thing. A musician who understands the physics of sound propagation and the neuroscience of sound perception need not have a lesser enjoyment of music than someone who does not. However it could well be the case that sometimes scientific explanation destroys the efficacy of the phenomena under investigation. An example that comes to mind is the placebo effect, but I’m unaware of any evidence which suggests its effectiveness is reduced in an individual treatment when that individual is aware, at the more general level, of the existence of placebos. Another example could be the case of religious experience: it’s interesting to ponder what would happen if people discovered they could get the same sense of hope without reliance on the existence of a God Out There.
- Evolution by natural selection does not explain why we are here in some grand metaphysical sense. It explains how organisms which reproduce, communicate (by genetic material as well as, e.g., via talking*, education*), and vary, can adapt within* and between generations to their environment and continue to reproduce.
* You may disagree with these inclusions. Perhaps I’m conflating memes and genes.