On the Symbiosis of Remembering, Forgetting and Learning

When I read the first paragraph's promise that "Conditions that produce forgetting often enable additional learning" and "learning and recalling some things is a contributor to the forgetting of other things" I had two initial thoughts: 1) Great, so maybe I don't have early onset dementia, but am actually just learning so much from life in the past 5 years that it just seems like I have early onset dementia and 2) I'm pretty sure this is not actually going to be what this paper is about, but I'm also pretty sure that this is still going to be my main takeaway from this paper.

The latter statement in #2 turned out to be true, so I think my thoughts/experiences provide evidence for the authors' points about memory and forgetting. But to be honest I'm not sure, because I could only focus on my initial takeaway. (And really, I'm less than half joking here).

Another positive takeaway I got from this paper: when we forget things, we can relearn them very efficiently. I suppose we ought to know this from experience (aka I'd assume if most of us took a pop quiz on some specific high school subject matter, we might not do great, but a quick review and we could relearn what we had forgotten). But it's nice to hear that there's some official scientific evidence for it. Because sometimes I feel I've forgotten everything I ever learned.

Also, to speak to Joseph's point: I was actually thinking while reading this that perhaps a specific therapy of retrieval-induced forgetting could be beneficial to, say, PTSD patients. But a quick Google scholar search suggests there is not research out there to support this conjecture. I suppose it would be exceedingly difficult to make that work in practice.

Also, to Fannie's side not about studying in different rooms: I do remember reading several years ago-- likely from some very reputable online source-- that it is helpful to test performance (or something in that vein) to vary the location where you study/work.

I was wondering, how does photographic or partial photographic memory works then? What part of the brain or the described model works differently to produce the effect of never forgetting?

Also, I think Franceska raised a good point about how much forgetting counts as forgetting? I don't remember a lot of algebraic formulas from advanced Mathematics, but it doesn't mean that I have forgotten the concepts. This relates to the discussion about forgetting promoting abstraction, but again -- the question is what level of abstraction are we talking about? I remember, the concepts and in this particular scenario, they become the abstraction that was promoted by the forgetting of formulas, but what determines the appropriate level of abstraction? These are some of the missing elements in the paper that make it a bit hard to digest.

Also, adding to the ongoing education discussion, someone pointed out the fallacies in Duolingo, which make sense within the framework created around this paper. However, it's still very popular and has helped a lot of people do long term language learning. What explains that phenomenon?

I think although the argument proposed in this paper "forgetting contributes to learning" is counterintuitive, the two explanations they gave were consistent with our understanding of learning. For example, if we learnt a math equation, and later we were asked to recall the math equation, if we have other environmental cues to recall the equation very easily, then we spent little effort in this retrieval practice, which didn't result in additional learning; however, if we don't have as many environmental cues, or we have completely forgotten how the equation looks like, we'll need to think of other ways, for example to retrieve information about how the equation was deducted, and this retrieval practice obviously resulted in more learning, and would be helpful for our future retrieval practice. So I think the two explanations they gave, including the encoding variability, and retrieval practice are consistent with each other in some ways.
One thing that has been discussed in previous posts is about "learning" vs. "memorizing", i agree that the tasks that they presented in this paper were more like memorization tasks, in which participants are required to memorize a word pair, etc. i think in simpler tasks, additional cues might help with memorization, but for more complicated tasks, such as solving a math or programming problem, the efforts they spent in reasoning about the process, or the efforts they spent in retrieving the information would result in learning.
This has been applied in some education practices in terms of the assessment we're using. We're not using only multiple choice questions for exams, there are more open-ended questions where students are required to articulate their reasoning, which is helpful to examine the cognitive processes of the students. In the cognitive tutor paper, we also see that the evaluation of cognitive tutors is not only using high-stake exams, but also using some performance-based tasks.
And I think it would also be interesting to see similar experiments but using more complicated tasks to see whether they could generate similar results.
And my takeaway from the paper isn't along the line of whether we should use block learning or interval learning, I don't think that's the core effects. I'll be more inclined to think along the line of how we can design the learning process to provide more cues that are conducive to learning, and how we can design the learning process to help students experience the retrieval practice more effectively before forgetting.