I think ideally a PhD is an apprenticeship. Practice in my experience is unfortunately not like an apprenticeship, instead it's a low paid job that can be quite abusive. Plus many of the things you "learn" aren't good for research in my view, e.g., I think "publish or perish" causes many researchers to avoid projects with the largest impacts. This is treated as "smart" but I strongly disagree.
That's what makes PhD work a good example I think, because a low-paid job that can be quite abusive describes apprenticeship very well. It's Inherently reflective of the quality of the master. If the master is an abusive drunk, the apprentice is going to have a bad time. Also the apprentices historically had to do much grunt work like (in blacksmithing, to choose a profession) churning out nails. Produce or perish.
I think there are advantages to bringing back parts of apprenticeship, but it is not a system without flaws or weaknesses.
I personally don't mind the grunt work and don't consider that abuse. Grunt work is either outright useful or a learning experience.
If blacksmiths were academics they would produce a large number of faulty nails. (One might argue the vast majority would be faulty!) I doubt blacksmiths would find that acceptable, but many (if not most) academics don't seem to care too much about quality as long as the paper passes peer review.
A PhD can become abusive when the advisor wants to produce more work when someone has done more than enough for a PhD. At that point they have a full researcher getting paid only a fraction of what they could be.
PhD students are paid less than apprentices last I checked, too. According to this page apprentices are paid roughly 1/3 of what professionals are paid:
So far, at no point during my PhD was I ever paid 1/3 of what my advisor is, even if you factor in benefits. I haven't even received a raise (to my knowledge), but he has.
Again, ideally a PhD is like an apprenticeship, but my experience suggests that case is rare.
My experience mirrors yours. A Ph.D is an arrangement where your advisor can keep asking for more work, beyond all reason. Unlike a (say) blacksmith apprenticeship, one cannot just leave with "six years of apprenticeship in X"... one instead must quit one's Ph.D.
Yes, peer reviewed papers are nails in the analogy. (Non-peer-reviewed work might as well not exist to many academics in my experience, regardless of its quality.) There are good nails and bad nails, just like there are good papers and bad papers, though the criteria for what makes each good differ.
There probably is some innovation in nails still going on. I imagine it's mostly about manufacturing methods and materials.
The same most definitely happens with real apprenticeships. In countries like Switzerland, apprenticeships start after the 9th school year and take around 3 to 4 years. During the apprenticeship, the apprentices still have to visit school and write tests. This is a common point of conflict. Let's imagine you are a 17 year old who wants to become a cook and your boss makes you clean the kitchen until 2am and the next day you have a test on algebra at 7:15am. There are laws against this kind of stuff but enforcement is really spotty.
