I more or less agree with that, with the proviso that I think academia in general (not just the humanities) would benefit from easing up a bit on the insistence on "producing new knowledge". It's good to produce new knowledge, sure, but I think the way that's been pushed has led to a situation where people just publish a lot of papers without necessarily creating a lot of new knowledge. In part this is due to Goodhart's law and people optimizing for publications. In part though it's due to the two-tiered (tenure/non-tenure) academic job system.
Even in fields quite remote from humanities, we have, for instance, a bunch of people who need to be taught calculus and so on. And it would be fine for them to be taught calculus by someone who isn't "creating new knowledge" in mathematics. But you can get paid a lot more to create new knowledge while begrudgingly teaching calculus now and then than you can to just teach calculus with gusto.
Likewise in the humanities, I think your argument leaves open the possibility that there could be new knowledge produced there, but that we just shouldn't expect everyone who's teaching Intro to American Literature or whatever to be producing such knowledge.
In my view a good step would just be to significantly reduce the pay gap (and gaps in benefits, job security, etc.) between teaching jobs and research jobs. There are many people who love Moby Dick or basic calculus and could ably and happily teach it for years without feeling any need to write a novel or prove a novel theorem themselves. We'd all benefit if such people could get a steady job doing that.
Yes, simple lecturing jobs are fine, and they do exist, but as you said they are paid less. Because in truth this is the reality already, we just don't admit it.
The intention behind it is understandable though. Someone who has produced new knowledge tends to have a more flexible mind, they have felt that the walls of knowledge are soft and malleable and not some concrete slab. They work with the math even outside class, and have a real grasp on why things are defined in certain ways, having also defined new concepts and written new theorems and proofs and having faced dilemmas of how to construct it to be most elegant and compact and logical etc.
Now, of course today the research and the teaching are often on quite distant topics. Like teaching some basic computer science stuff like basic data structures and algorithms while you actually research computer graphics or speech recognition.
I'm not so sure that having produced new knowledge is so vital to teaching old knowledge, at least not at the scale that's required for a tenure-track job. It's good for a teacher to not be locked into a static view of old knowledge, but I don't think that requires anything like the breakneck amount of publishing that's expected in many fields today.
Even in fields quite remote from humanities, we have, for instance, a bunch of people who need to be taught calculus and so on. And it would be fine for them to be taught calculus by someone who isn't "creating new knowledge" in mathematics. But you can get paid a lot more to create new knowledge while begrudgingly teaching calculus now and then than you can to just teach calculus with gusto.
Likewise in the humanities, I think your argument leaves open the possibility that there could be new knowledge produced there, but that we just shouldn't expect everyone who's teaching Intro to American Literature or whatever to be producing such knowledge.
In my view a good step would just be to significantly reduce the pay gap (and gaps in benefits, job security, etc.) between teaching jobs and research jobs. There are many people who love Moby Dick or basic calculus and could ably and happily teach it for years without feeling any need to write a novel or prove a novel theorem themselves. We'd all benefit if such people could get a steady job doing that.