I believe most things you write here in isolation are true (besides maybe stretching the muscle/brain analogy a bit, and missing the importance of scaling down synapses during sleep, rather than building new ones), but it's not supported by anything (references/citations), and doesn't directly address the question asked.

It's all just conjecture of course! I made an edit to clarify my answer. "It probably is possible to solve problems without active thinking, because thinking itself and creating new ideas is a rather passive process."

This sites does not like answers that are based on conjecture and require them to be backed up with evidence.

Thank you! Sorry I didn't realize I entered a professional setting. I thought it was a website that's open to the public.

Yeah, but we have a collective case of imposter syndrome. We don't believe anything we say if we don't back it up with external research.

Next question for this site: If I say something true but noone else has said it before, does that make it false?

@oliver_siegel I wouldn't describe this site as professional, but rather as more structured than you're thinking if you have in mind sites like Reddit where discussion is on-topic. StackExchange in general has a Q&A model where answers are for answers only, and not for just talking about a topic. Skeptics.SE has a specific model of scientific skepticism which is based on supplying referenced evidence in support or contradiction to notable claims made in the media and other sources. See also skeptics.meta.stackexchange.com/questions/5/…

I see... Also, it says the original question was asked 9 years ago. Seems like there had been some stagnation with this model that doesn't allow for creative problem solving. I thought I'd help a fellow internet surfer out by giving an answer to advance the development of useful knowledge. I do appreciate your feedback as to why the downvotes. I can live with it :)

@oliver_siegel There are 10,764 questions here; 1,377 don't currently have an answer that is either upvoted or accepted. So yes, not 100% of questions get answered in part because there's a higher barrier to answering than other sites. The intended benefit, which I think works well, though, is a much higher quality standard. Someone can post a question here when what they want is a well-researched answer, and won't get the types of answers they don't want.

Welcome to Skeptics! Please have a read of that link to better explain what we are about. Every community, even those open to the public, has customs and standards - ours are pretty explicit.

Even with references, the answer is still largely conjecture (feel free to provide a quote from the references that directly substantiates the claims to prove me wrong). To address your question: If you want to say something true and no-one else has demonstrated with empirical evidence that it is true before, then it isn't of interest to us here, because it isn't applying scientific skepticism, which is the purpose of the site.

Thank you @Oddthinking ! What would be the null hypothesis to sufficiently disprove the claim that problem solving during sleep is possible? How could we show that it's impossible?

@olli: First, I want to be clear that the burden of proving that this claim is falsifiable isn't on me. But I will play along.

It is worth clarifying what the claim is. I see the links in the original question talking about explicitly dreaming about a particular problem... is that the same as the claim in the question?

If I am learning a new skill, and I am plateauing in progress, I will often leave it sleep on it and try again another day. Why? NOT because I think I will have a dream (that I remember), but because:

* Sleep consolidates your memories, helping "muscle memory" and speeding up some skills (like repetitive sums, etc.) [citation needed]

* Because I am tired, and when I try again I will be rested.

* Because I might be stuck in a mental rut, and coming back to it afresh means I will have a different mindset. I will likely also review where I am up to, and in doing so may spot flaws in my reasoning or technique.

If the claim is "Putting a problem/training down when you are not progressing, and picking it up after sleeping will help you reach an answer faster (in effort, if not duration)" then we can come up with a null hypothesis easily, and try to test that.

e.g. Find a crossword that typically takes an hour to solve, and a sudoku that takes an hour to solve. Compare (a) attempting one on day 1, and the other on day 2 with (b) spending 30 minutes on both on day 1, and then finishing them on day 2.

If the claim is "if you sleep, you will naturally dream about the problem you have been working on", we can give people problems and then ask what they dreamt about.

[I have seen a paper like this - they put people in a skiing simulator game for an hour or so, and then let them sleep. They woke them up and asked what they had been dreaming about. People who were only allowed to sleep for a short time reported a lot of dreams about skiing. People who were allowed to sleep longer reported dreams about related areas - falling down mountains, and the like.

[Note: I have not provided a reference to this paper; discount your belief in it appopriately.]

If the claim is "By asking yourself to dream on a certain topic, you will dream on that topic?" we could produce a similar test.

Proving that you will NEVER dream on a particular topic is hard - not least because that means you don't even dream on the topic randomly! But we can test to a clinically significant level. If you only have "useful" dreams one time in 2000, there is no practical use in trying it.

I think this was the paper: academic.oup.com/sleep/article/33/1/59/2454584