“Seeing into the future” can describe several very different things: imagining what may happen, forecasting from present evidence, having a strong hunch, or claiming knowledge of an event that ordinary information could not reveal. These should not be treated as interchangeable.
In paranormal research, precognition refers to the last of these claims. A vivid image or convincing feeling does not by itself demonstrate it. This article revisits psychologist Daryl Bem’s much-discussed experiments, explains why later testing matters, and offers a practical way to reflect on future events without treating uncertainty as a personal failure.
Prediction, intuition, and precognition are different
If you expect a late bus because traffic is heavy, you are making an inference. If you picture yourself enjoying a new job, you are imagining a possibility. If you feel that a friend may call after several recent conversations, there may be a familiar pattern behind the feeling. None requires information to travel backward in time.
A precognitive claim asks for something stronger: knowledge about a future event that was not available through ordinary clues. That distinction changes what evidence is needed. Being right is not enough by itself; we also need to ask what the person knew, how specific the claim was, and how often comparable guesses were wrong.
What Bem’s 2011 paper actually reported
Daryl J. Bem’s author-hosted version of “Feeling the Future” reports nine experiments involving more than 1,000 participants. The experiments reversed the usual order of psychological tasks to investigate whether later events appeared related to earlier responses. Bem reported statistically significant results in eight of the nine experiments and interpreted the pattern as evidence for anomalous influence.
This was a published research claim, not a demonstration that anyone could reliably foresee personal events. The paper’s tasks included approach or avoidance, priming, habituation, and recall. Its author-hosted copy also notes that it may differ from the journal’s final version. A careful account should retain that context rather than describing the findings as an unquestionable discovery.
The memory experiment: practice came after recall
The unusual question was not whether normal practice improves later memory. In the recall task, people first studied words and completed a memory test. Only afterward did the computer select some words for additional practice. The proposed effect was that words selected for later practice had already been recalled more successfully.
Ritchie, Wiseman, and French’s 2012 paper describes three preregistered independent attempts to repeat Bem’s ninth experiment. Across 150 participants, those attempts did not produce significant effects. This directly challenges the idea that the reported recall result was a dependable finding. It does not mean that ordinary learning or memory is mysterious evidence of precognition.
Later findings need to be part of the story
The literature is not represented accurately by quoting only an original success or only one failure. Bem and colleagues later reported a positive meta-analysis of 90 experiments. A meta-analysis combines results across studies; its conclusion still depends on the studies, reporting, and analytical choices involved. It is not a guarantee that a particular effect will appear again under a specified protocol.
A 2025 project by Walleczek and colleagues tested the effect claimed in Bem’s Experiment 1, not the word-recall experiment. Across three studies it involved 26,483 participants and 420,472 critical trials. The first study did not reproduce the original precognitive effect. An exploratory opposite-direction finding was confirmed in the second study, but the third did not reproduce that finding. The authors considered a method-related anomaly more plausible under conventional standards than a stable psi explanation.
These results are reasons not to present Bem’s original report as settled proof. They address particular experimental tasks, however, and should not be stretched into a claim that one project tested every spiritual belief or every unusual experience.
Why a striking result is not the end of the question
A statistical result must be interpreted in the context of a test and its assumptions. It is not a direct reading of the probability that psychic powers exist. Very large “odds against chance” headlines can hide important questions about how outcomes were selected, how many analyses were tried, or whether later studies found the same thing.
For a reader, a useful research checklist is:
- Timing: was the prediction fixed before the outcome was known?
- Specificity: was there a clear rule for a match, rather than a broad resemblance?
- Ordinary clues: could participants or experimenters obtain relevant information another way?
- Complete reporting: were unsuccessful attempts included?
- Replication: did an independent team find the effect again under a clear protocol?
- Scope: does the conclusion stay within the task that was actually tested?
You do not have to solve the entire research debate to use these questions. They help separate an interesting claim from a reliable basis for action. Our guide Are Psychics Real? also distinguishes personal meaning, service quality, and evidence.
A prediction journal that does not rewrite the past
If you enjoy reflecting on impressions, keep the exercise low stakes and time limited. Do not make alarming predictions about another person’s health or safety into a game. Choose something ordinary that can be checked without interfering with anyone else.
- Write the exact claim. “I will receive a message” is broad. Identify the sender, channel, event, and deadline if those are genuinely part of the original impression.
- List existing information. Did that person promise to reply? Is the event already scheduled? Have you been checking for related clues?
- Set a matching rule. Decide beforehand whether an approximate date, different person, or similar object would count. Avoid making the rule flexible after the fact.
- Leave the entry unchanged. Add a dated result underneath it. Do not silently correct the prediction.
- Keep the whole record. Include misses and unresolved entries. An uncheckable outcome is not automatically a success.
A private journal is a reflection tool, not a validated research protocol. It can make your own record more honest, but it does not rule out all ordinary explanations or establish an ability.
A fictional example: the interview message
Fictional illustration, not a documented prediction: Noor has interviewed for a course and has been told that decisions should arrive during the week. Noor writes, “I feel I will get a positive email on Tuesday.” No email arrives Tuesday. An acceptance arrives Friday.
The result contains a positive outcome but not the specified timing. Noor records that distinction instead of changing Tuesday to “soon.” The existing expectation of a decision also belongs in the record. A welcome email can be meaningful without proving that the future was seen in advance.
Noor’s practical preparation was useful either way: checking the response deadline, keeping documents ready, and considering another course. Those actions did not depend on the impression being correct.
Replace a demand for certainty with a small plan
When the future feels unclear, ask what you can prepare for rather than insisting on one certain outcome. Try a short worksheet with three possible results: the hoped-for outcome, an ordinary delay, and an outcome you would find disappointing.
- Known facts: what is confirmed today?
- Unknowns: what has not yet been decided or communicated?
- Useful action: what small step helps under more than one outcome?
- Decision point: when will you review new information?
- Boundary: what will you avoid doing without better evidence?
For example, you can prepare a question for a meeting without knowing its result. You can decide when to follow up on an application without repeatedly buying predictions. You can ask someone about their intentions without assuming that an inner image reveals their private thoughts.
A measurable goal might be “send one appropriate follow-up on Thursday, then review the actual response.” That is something you can evaluate. “Become completely certain about next month” is not a fair requirement to place on yourself.
Keep important decisions anchored in present evidence
A strong impression is not a reason to abandon medical care, risk money you cannot afford to lose, accuse someone, or ignore a clear boundary. Gather relevant facts and obtain qualified advice where the stakes call for it. If a reading becomes coercive or pressures you to keep paying for certainty, you can stop.
If future-focused checking becomes frightening or disrupts sleep and daily life, step back from the exercise and seek appropriate support. For a practical discussion of uncertainty, read Should the Future Scare You? Curiosity about the future does not require surrendering your choices in the present.
Frequently Asked Questions
What does precognition mean?
Precognition is the claimed awareness of a future event that could not be anticipated through ordinary information or inference. It is different from planning, forecasting, remembering a pattern, or imagining a possibility.
Did Bem’s experiments prove that people can see the future?
Bem reported results he interpreted as evidence for anomalous future influence, but that is not the same as settled proof. Independent replication results and methodological questions must be considered alongside the original report.
What did later replication studies find?
Three 2012 attempts did not reproduce the reported recall effect. A large 2025 project on a different Bem task did not establish a stable precognitive effect. These findings concern particular experiments, not every possible spiritual claim.
Does a dream that later resembles an event count as proof?
A resemblance alone is not proof of prediction. Compare a record made before the event with the outcome, including mismatches, prior knowledge, vague details, and other predictions that did not match.
How can I keep a useful prediction journal?
Record exact wording, date, deadline, prior information, and a matching rule before the outcome. Keep every entry, score misses honestly, and do not turn a personal journal into a claim of scientific validation.
Can I use an impression to make an important decision?
Treat it as a prompt to gather information, not as a substitute for evidence or qualified advice. Avoid irreversible decisions, accusations, major spending, or delaying care because a prediction feels convincing.
What if thoughts about the future become frightening?
Pause prediction checking and focus on current facts and manageable actions. If distress disrupts sleep or daily life, seek appropriate professional support rather than trying to obtain certainty through repeated readings.