There are three things in tension here. 1. It is usually combinatorially impossible to answer all the questions when solving a problem - people just choose som

There are three things in tension here. 1. It is usually combinatorially impossible to answer all the questions when solving a problem - people just choose some random walk over the problem space and it doesn't always lead to ideal results. 2. It is socially unacceptable for "senior" people (who usually have better mental model of higher level concerns) not to project smug confidence in front of less "senior" people, so they pretend to know things they do not. 3. People have evolved to be lazy as fuck, so unless they absolutely must know something and can't do without it, they mostly won't even bother. Some won't bother even _if_ they can't do without. I don't have solutions to #1 and #3, and I doubt #3 can even be solved at all. But I do have a solution for #2 - it's the Socratic method (or as they call it at Microsoft "precision questioning"). It lets you save face by asking pointed, outcome-oriented questions of someone who likely knows more about the problem, without necessarily making your own lack of knowledge obvious. Besides that it also surfaces the assumptions, some of which routinely turn out to be wrong, and it's the easiest method I know for getting people to change their mind on things, because they feel like they've arrived at the answers using "their own" thought process. It is difficult to master, so it too runs into #3, however, but I think all engineers, senior and junior would benefit from learning this. Another upside of learning it is it makes it harder for unscrupulous others to manipulate you maliciously via Socratic questioning, since you know what they're trying to do.