Sleep & insight · Research examined

At the edge of sleep, an idea can change shape

By Andrey Tadevosyan · Published 5 September 2026

You are almost asleep when an image arrives: a doorway, a phrase, an unexpected solution to something that felt stuck. It has the force of a discovery. By morning, it may be hard to explain why.

For someone interested in manifestation, this raises an appealing question. Is the drowsy boundary a place to rehearse a desired future, discover an idea, or learn something about events outside yourself? Those are different questions. Research has begun to investigate the creative one.

The useful surprise is that the evidence is more interesting than a simple instruction to enter the “right” state.

What researchers have actually tested

In 2021, Célia Lacaux and colleagues studied a number task with a hidden shortcut. Participants worked on the task, took a monitored break, and then returned to it. People who entered the first stage of sleep, N1, without progressing to N2 were more likely to discover the shortcut afterward than those who stayed awake or reached N2. Sleep stages were identified from recordings; people were grouped by the sleep they actually reached. This was a finding about a particular kind of insight, rather than a randomized assignment to three sleep stages. Lacaux et al., Science Advances.

In 2023, Adam Haar Horowitz and colleagues used a different approach. A device helped deliver prompts around sleep onset, including a suggestion to think about a tree. Participants then completed three creative tasks related to that same theme. The sleep-with-incubation group performed better on the combined creativity measure than the other conditions. The assessments followed immediately; the study does not establish a lasting improvement in general creativity. Sleep timing was inferred by the device rather than verified with EEG. The intervention involved repeated prompts and awakenings; an ordinary relaxation recording does not reproduce it. Haar Horowitz et al., Scientific Reports.

A preregistered 2025 study complicates the story. Anika Löwe, Marit Petzka and colleagues used a perceptual task with a hidden rule. Their final analyses included 68 participants. Insight was more common among those who reached N2; N1 did not show the predicted advantage. Because the task differed, this is not a direct repeat of the earlier number experiment. It does challenge the broad claim that the earliest stage is uniquely best for insight. Löwe et al., PLOS Biology.

Read together, these studies offer a reason to investigate sleep and creativity. They do not establish one universal state, audio track or bedtime routine that reliably unlocks it.

An idea and a prediction have different jobs

Imagine that you wake with a new opening for a story. You can write it and decide whether it works. The idea earns its place through what you make with it.

Now imagine waking with the conviction that a particular person will contact you tomorrow. That is a claim about an external event. To examine it, record the person, the deadline and what would count as contact before checking the result. A message from someone else, three weeks later, would be a different event.

Both experiences can be memorable. Their vividness does not settle their accuracy. None of the studies above tested whether imagining a future makes unrelated events occur or supplies reliable knowledge of them.

This is also the distinction to carry into Neville Goddard’s teachings. A spiritual account of assuming a fulfilled desire is worth examining in its own terms. A laboratory result about solving a task cannot validate that entire account by association.

A small practice: keep the fragment, then use it

This is an original journaling exercise, not a replication of the studies or a method for detecting a sleep stage. Use a naturally remembered fragment; there is no need to interrupt or shorten your sleep.

1. Record what arrived. Write the image or words before explaining them. “A bridge made of paper” is a description. “I received a warning” is an interpretation.
2. Name the question it might help. If you cannot identify one, you can simply keep the fragment. Every experience does not need a task.
3. Make one waking version. Sketch the bridge, try the sentence, rearrange the outline. Give the idea something concrete to succeed or fail at.
4. Review what happened. Was anything useful produced? Did the idea change once you worked on it? Record that too.

A fictional example: a writer dreams of a room with no doors. In the morning she treats it as a possible opening for a story. She writes a paragraph, discovers that the room works better as a train carriage, and keeps the revision. The dream did not need to contain a finished answer to be useful.

If your fragment includes an external prediction, put that in a separate record using the prediction guide. If it suggests something you want to do, take the action into GROUNDED’s practice studio.

The open question is worth keeping: when does an altered way of thinking help us see something we missed? A good notebook leaves room for the experience, the unsuccessful ideas and the unexpected result.