Science analysis. A dangerous encounter may become unforgettable while an ordinary disappointment vanishes. That feels like a sensible bargain for a species that had to notice predators, injuries and hostile rivals. But an appealing survival story is not the same as proof of how human memory evolved—and some of the newest experiments complicate the slogan that our brains always remember bad things best.
In fact, a large 2026 study found something that sounds almost heretical in the age of doomscrolling: participants remembered positive images slightly more often than negative ones, even though the negative images elicited stronger arousal. Both categories beat neutral images. The researchers emphasized that emotional intensity, the kind of material shown and the mechanics of remembering all matter. Read the 2026 study.
The evolutionary argument: a missed meal versus a missed warning
Imagine two costly mistakes. You overlook a patch of edible plants and go hungry for a night. Or you overlook the movement of a predator in the grass. The second error may eliminate any chance to learn from the first. That asymmetry offers an intuitive explanation for why the mind might prioritize danger.
Evolutionary psychologists have long considered whether the cost of missing a threat could have favored greater attention to negative information. A landmark 2001 review by Roy Baumeister and colleagues, Bad Is Stronger Than Good, surveyed research on relationships, emotions, feedback, impressions and learning, finding many contexts in which unpleasant experiences carried disproportionate weight. A companion review by Paul Rozin and Edward Royzman mapped several forms of negativity bias, from attention to judgments.
That is strong evidence for a psychological tendency—not a fossil record of thought. We cannot directly measure what a particular ancient human remembered or show that a single predator-versus-food trade-off produced all modern biases. What we can test is how people now attend to, encode, retrieve and evaluate different experiences.
What neuroscientists actually find
Emotional experiences often receive more attention and are remembered better than neutral ones. Brain systems involved in this process include the amygdala and regions supporting attention and memory; they do not operate as one simple “fear switch.” The precise memory advantage depends on the task, how arousing a stimulus feels, and what details researchers test. A review by Elizabeth Kensinger explains how emotion may sharpen memory for an event’s focal details while leaving surrounding context weaker.
The newer 2026 neuroimaging research drew on a brain-imaging sample of 1,006 young adults and a larger behavioral analysis of 1,591 participants. People viewed positive, neutral and negative photographs and later received an unexpected recall test. On average, positive photographs were recalled most often, negative ones next and neutral ones least often. The positive-over-negative difference was modest, and the authors warned that visual and semantic features not fully controlled in the study could influence memorability. Still, the result challenges any universal claim that negative information must win.
It also reveals an important distinction: “bad grabs attention” and “bad is always remembered more accurately” are different propositions.
Vivid does not mean accurate
Some memories feel like recordings. They are not. In a 2003 Duke University study, 54 students recorded their memories shortly after the September 11 attacks, along with an ordinary recent experience. Follow-up tests found that the consistency of both memories declined over time. What distinguished the emotional memories was the participants’ continuing vividness and confidence—not special immunity to distortion.
That matters far beyond laboratory psychology. A memory can provide a powerful warning while still containing mistaken details. Strong conviction is not the same as an independently verified account, whether the subject is a personal confrontation, an eyewitness recollection or a widely shared public event.
The “warning signal” theory gets a test
One proposed mechanism goes further: perhaps people retain details from immediately before a threatening event so that similar cues can warn them next time. A sound, a smell, a glimpse of movement—potential signals of trouble.
In an open-access 2025 study, researchers asked 175 people to recall neutral, positive and negative or trauma-related autobiographical events, including the moments preceding them. Memories of the lead-up to emotional experiences were more vivid and detailed than those before neutral events. But the enhancement was not confined to trauma. The authors concluded that the results did not support a trauma-specific version of the warning-signal hypothesis; emotionally significant events of either kind may shape what comes to mind.
That does not prove the evolutionary idea false. It does make the simple explanation incomplete. Our minds may preserve signs of a wonderful surprise as well as signals of danger.
The overlooked counterweight: good memories fight back
Research on autobiographical memory makes clear that positive experiences can be vivid, accessible and influential. A 2022 review of positive and negative episodic memories describes how both kinds of emotional experiences can persist, come to mind readily and influence decisions. In some circumstances, pleasant personal memories are recalled more often than painful ones.
Age also changes the balance. A meta-analysis of 100 studies involving 7,129 participants found a reliable age-related positivity effect: compared with younger adults, older adults tended to favor positive over negative information in attention and memory, although the strength of that effect depended on research conditions. Not every study finds it; the pattern is a tendency, not a rule for every person.
Apep Bar has also explored why affectionate touch can form enduring emotional memories. The larger point is that human memory is not merely a warehouse for warnings. It helps us seek rewards, maintain relationships and decide which experiences deserve to be repeated.
So is the brain pessimistic—or selective?
“We evolved to remember the bad” is an attractive one-line explanation because risk really is asymmetric. But there is no single, all-purpose negativity setting. Researchers find biases in attention, evaluation, vividness and retrieval; these are related phenomena, not interchangeable measurements. Some experiments favor negative material. Others show that positive material can have the edge. Context, age, emotional intensity and the meaning of an event affect the outcome.
The safest conclusion is neither that our memories are defective nor that evolution deliberately made us miserable. Emotion changes what gets prioritized. Sometimes that may help us avoid a costly mistake; sometimes it can leave us overconfident about details or preoccupied with a past threat. And sometimes a joyful experience wins the competition.
The question worth remembering: if the same brain can preserve a danger signal and a cherished moment, what decides which one gets to tell the story of our lives?
Sources and further reading
- Baumeister et al. (2001), Bad Is Stronger Than Good.
- Rozin and Royzman (2001), Negativity Bias, Negativity Dominance, and Contagion.
- Williams, Ford and Kensinger (2022), The Power of Negative and Positive Episodic Memories.
- Rustand et al. (2025), Autobiographical Memory for Emotional Events: A Test of the ‘Warning Signal’ Hypothesis.
- 2026 study, Neural Correlates of Emotional Memory Enhancement: The Role of Valence and Arousal.
- Talarico and Rubin (2003), Confidence, Not Consistency, Characterizes Flashbulb Memories.
- Reed, Chan and Mikels (2014), meta-analysis of the age-related positivity effect.
Featured photograph: Documentary photograph of radiology staff conducting a brain MRI examination at Naval Medical Center San Diego. Photo: NavyMedicine / Flickr, Public Domain Mark 1.0. This real clinical photograph illustrates brain imaging generally; it is not from any study of emotional memory or negativity bias discussed in this article.
