Pretty much the title
In: 210
There are always gaps in our understanding of mechanisms, and there always will be – such is the nature of scientific discovery. There is always another layer underneath a mechanistic explanation. If I ask you how a car works, you may tell me about an internal combustion engine. Then I ask how that works, and you will explain engine design along with thermodynamics and maybe Newtonian mechanics. Then I’ll ask how the fuel burns, and you’ll explain combustion chemistry and oxidation. Then I’ll ask how that works, and you’ll explain physics principles like potential energy, heat, entropy, etc. Then I’ll ask how that works, and maybe you’ll get into quantum mechanics or something. Eventually, we are going to reach a point where I ask “how does that work?” And you’re forced to answer, “we haven’t figured that out yet.” But here’s the thing: when we do figure it out, there will be an answer to my question, but also another question of “well, then how does THAT work?” And so on.
Pretty sure I can still establish that stepping on the gas makes the car go 😉
Yeah sure. For instance we know that tylenol treats inflammation and fever, but we have absolutely no idea how. There are lots of cause-and-effect relationships where we don’t understand the full mechanism behind it. The main issue is being careful to actually prove that it’s “cause and effect” and not “correlation”. Because if it’s just a correlation, there could be other important factors we’re missing that are actually causing the effect.
Certainly. Pharmacology offers numerous examples of this. A drug might cause an effect, and what this effect is can be measured and quantified extremely well, but the exact mechanism can be not be understood.
This is most of medicine. They still have no clue how anesthetics work, They don’t even know exactly how aspirin works.
Of course. You don’t need to understand how storms are formed to notice that one is coming.
It depends on what you classify as “science”. For the most exacting definitions, there is no causation, just “necessary and sufficient predicates”: if such and such occurs, then some certain thing occurs. “Causation” is, in this view, nothing more than a statistically significant correlation, and scientific theories (which are the formulas used to make quantitative predictions, not the explanations of “why” those formulas work) are *effective theories”, which means that whether the supposed entities that are used to describe “how” something happens are real is irrelevant, and potentially fictional; all that matters (pun intended) is whether the formulas work.
Koch’s postulates are 4 parameters used to determine if a particular bacterium causes a disease, but doesn’t address any of the mechanics of infection
https://en.m.wikipedia.org/wiki/Koch%27s_postulates
If you can control all confounding variables, you can establish causation by showing correlation. The “controlling all confounding variables” part is very difficult. There was a lot of chemistry and medicine in the 19th century that, mostly, worked without actually knowing what was going on.
There’s a class of drugs called SGLT-2-Inhibitors. They were originally devised to help with diabetes, as they block re-absorption of sugar in the kidney, thus leading to more sugar being excreted, lowering blood sugar levels.
Surprisingly, it was found out that these drugs also help with heart failure. It’s clear they do, based on numerous clinical trials for various competing drugs – but while there’s a ton of plausible mechanisms, there’s no smoking gun for either of them. Indeed, the effect may actually be mediated in parallel over several causative mechanisms.
When you do a well-designed clinical trial, you can be relatively sure that there is an effect (and you can quantify how sure you are). But it doesn’t tell you anything about the actual mechanism – that requires other kind of research.
Eating burnt meat statistically cases more cancers. It is unknown why. (take with a grain of salt, i have not fact checked myself on this one)
Latest Answers