A verdict answers only the question you asked
Suppose you visit a café to try its coffee. Later, a friend asks whether it would suit an online job interview. “Good coffee” is little help. Remembering the blender behind the counter and the person taking a speakerphone call gives you something you can use.
This is an illustrative example, not an experiment. It captures a problem studied by Jonathan Nicholas and Marcelo Mattar: when an experience happens, you may not yet know which of its details a later decision will require. Their public author manuscript examines how memories of individual events can support choices when the relevant question changes.
A summary is efficient because some work has already been done. An overall judgment compresses what happened into an answer. An episodic memory can preserve connections among details of a particular occasion, allowing some of that work to happen later. Neither approach guarantees a correct decision. They place different demands on what you keep and when you use it.
Two ways to solve the same points problem
Nicholas and Mattar used five experiments with participants recruited from a university subject pool. In the first four, people saw six items per round, each paired with a positive or negative reward. Items had several features. After a distractor task, participants accepted or rejected offers defined by a shared feature. An offer’s value was the sum of the rewards attached to matching items.
Here is a simplified worked example of that rule. A blue butterfly earns two points; a blue cup loses one. If those are the only blue items, the blue offer is worth 2 − 1 = 1 point, so accepting it is beneficial. This example explains the arithmetic, rather than reproducing a particular participant’s trial.
You could solve it by keeping a running blue total while the items appear. When the offer arrives, you already have an answer. Alternatively, you could recall the butterfly and cup, retrieve their rewards, and add them then.
Those routes produce the same answer when everything goes well. They differ when you are suddenly asked about animals. The blue total cannot tell you how much the butterfly contributes to an animal offer. Remembering the item together with its properties and reward potentially can. You would still need to remember any other animals before calculating the complete offer.
The authors compared decision times, choices and subsequent memory reports to distinguish these strategies. Patterns consistent with retrieving individual episodes supported their account. These are behavioral inferences about a strategy, rather than direct observations of each thought during a choice.
Knowing the question changes the useful strategy
In later experiments, the researchers changed when participants learned which features would matter. Information provided before the items appeared made it possible to prepare the relevant totals. Information provided afterward left participants needing a way to evaluate features they had not specifically prepared for.
In the fourth experiment, preparing a single expected offer was especially efficient. On the final round, however, participants also received unexpected offers. For participants cued before seeing the items, performance on unexpected offers was worse than on expected ones, but comparable to earlier performance when they had not known the relevant feature in advance.
That pattern supports the idea that preparing an answer need not eliminate access to individual memories. Some information remained usable when the original preparation no longer fit. It does not mean the unexpected questions were answered perfectly, or that every irrelevant detail survived.
The trade-off is easy to recognize outside the task. If you repeatedly choose lunch on price, a remembered price ranking can save time. If a visitor asks which restaurant has a step-free entrance, that ranking answers the wrong question. A specific remembered visit may help, provided you actually noticed the entrance. The restaurant example extends the logic; the experiments did not test restaurant recommendations.
How details can belong to one event
Remembering an event involves more than independently recognizing its ingredients. The useful connection is that this sound, this object and this location occurred together.
In a separate brain-imaging study by Cooper and Ritchey, 27 people learned objects associated with a color, a scene location and a sound. Early activity in frontal and perceptual regions related to later memory for features. Toward the end of an event, memory-related activity in the left hippocampus was particularly sensitive to binding spatial information with other features.
This supports a role for the hippocampus in integrating event details. It does not establish a literal storage folder for each experience. Nor does it show that the café example follows exactly the same neural sequence: learning controlled combinations during a scan differs from visiting a noisy café.
“Not needed yet” differs from “never noticed”
A detail can be irrelevant to your current goal while still receiving attention. Hearing a blender while judging the coffee is different from failing to notice an electrical outlet behind your chair.
Uncapher and Rugg’s selective-attention experiment asked volunteers to attend to color or location and later gave them a surprise memory test. The findings indicated that attention biased which contextual information was encoded. A future need cannot make memory a dependable source for something you never registered.
This also separates the topic from curiosity and remembering an anticipated answer. Here, the interesting case is information that becomes useful before you had a reason to prepare an answer about it.
Use the finding without promising perfect recall
The decision experiments concern controlled tasks within single sessions, with young university samples. In the first four experiments, repeated memory tests gave participants an explicit reason to remember. The fifth used naturalistic photographs and an unexpected memory test, broadening the task while still falling short of everyday life over months or years.
The authors also acknowledge that real memories are compressed and reconstructed. Their account is evidence for flexibility under the tested conditions, not proof that all experiences remain intact somewhere, waiting for the right question.
For an important practical decision, keep observations separate from your verdict. A short note saying “blender audible at the back table during my lunchtime visit” preserves context that “nice café” discards. It also leaves room for the café to be quieter at another time. This is a practical application of the distinction, not a memory-training method tested by these experiments. If your friend’s interview depends on it, checking the current conditions is more useful than turning a remembered impression into a guarantee.
