Original studyPsych, 2026Read the paper →
A car horn, a starting pistol, or a ping from a phone can make us react faster to whatever comes next. Scientists have long assumed that this brief burst of readiness, known as phasic alertness, is triggered automatically by the warning itself. A new German study suggests the brain is more selective than that, and that a warning only speeds us up when we expect something to follow.
Researchers at Bielefeld University and HMU Health and Medical University carried out two experiments in which volunteers made rapid choices about targets on a screen. The study, published in the journal Psych, played a short tone before some targets and no tone before others, and measured how quickly people responded.
The key twist was the reliability of the warning. In one set of sessions, the tone was followed by a target on 90% of trials. In other sessions, it was followed by a target on only 10% of trials, so the sound carried almost no promise that anything was coming. Phasic alertness is considered a core part of the brain’s attention system, and understanding what drives it matters both for clinical research on attention problems and for everyday safety.
When the tone was reliable, it produced the classic alerting effect. In the first experiment, people responded about 47 milliseconds faster after a tone than without one. When the tone rarely predicted a target, that benefit disappeared entirely, falling to around two milliseconds.
The second experiment was designed to rule out a simple explanation, namely that people in one condition had to respond more often overall. Even with the number of responses balanced, the reliable tone sped reactions by about 92 milliseconds on average, while the unreliable tone gave a much smaller benefit that did not differ significantly from zero.
The pattern was remarkably consistent. Every participant in both experiments showed larger alerting effects when the tone predicted a target, and accuracy remained high throughout, at more than 98% correct responses in the second experiment.
The experiments used a small number of people tested over a very large number of trials, nearly 36,000 in the first experiment and almost 17,000 in the second. This design is common in perception research and gives precise estimates for each person, although larger and more varied samples would help show how widely the findings apply.
The authors conclude that phasic alertness is not purely stimulus driven. Even if a warning sound triggers alertness automatically, that readiness only helps behaviour when it fits with what the brain already expects. The findings could inform the design of warning systems in cars, aircraft cockpits, and hospitals, where alarms that rarely signal anything important may lose their power to speed up the response that matters.
