Think of the human activity most obsessed with the clock: commercial aviation. Departure times, landing slots, a forty-minute connection. If there is anywhere the clock is king, it is there.
Now look at what a pilot actually has to remember to do. Most of the checklists are not tied to the clock at all. There is no "at 14:32, lower the landing gear". There is "before start", "before take-off", "approach". The trigger is the phase of flight, not the time.
It is an odd detail. The industry that lives by the timetable organises its memory around events. Meanwhile the rest of us, who fly nothing, type "buy bread, 6pm" and are surprised when 6pm finds us stuck in traffic, miles from any bakery.
The difference, in one sentence
It looks like a formatting choice, like blue ink or black. It is not. These are two different memory mechanisms, and one of them is a lot more expensive.
What each one really is
Psychologists call the ability to remember to do something later prospective memory, and they usually split it in two. The split became standard through the work of Gilles Einstein and Mark McDaniel from 1990 onwards: there are time-based tasks (do something at 3pm, or in ten minutes) and event-based tasks (do something when a cue shows up).
Their first study already held a surprise. They tested younger and older adults on an event-based task: press a key whenever a particular word appeared. The older adults did worse on ordinary memory tests, such as recalling word lists, but not on this one. The cue in the environment seemed to make up for what memory was no longer doing on its own.
Five years later the same team put the two kinds side by side. On the event-based task, again, no age difference. On the time-based task, the difference appeared. The authors' explanation is the heart of this piece: time-based tasks depend on self-initiated retrieval. Nobody taps you on the shoulder. You have to interrupt whatever you are doing, on your own, to check the clock. A cue taps you; a clock waits to be consulted.
The badge experiment
A lab is one thing, an office another. So Abigail Sellen and colleagues took the question into a working research centre. Twelve people wore electronic badges that logged where they were, and over several days had two missions: press the badge button at an agreed time, or press it on arriving at an agreed place.
The place task was carried out 51.9% of the time. The time task, 33%.
The counterintuitive part comes next: people thought more about the time task. It took up more room in their heads and delivered less. The place hits, meanwhile, came almost at the door: most happened within seconds of walking into the agreed space. And spontaneous thoughts about the tasks came up more in places of passage, such as stairwells, than where people sat down to work.
Where cues fail
If the story stopped there, the conclusion would be easy: swap every time for a cue. But cues have two flaws that clocks do not.
The first is that the event may never happen. "When I pass the pharmacy" is great if you pass the pharmacy. If that week you do not, the intention dies without a sound. The clock, at least, always arrives.
The second is that the cue can arrive at the wrong moment. A 2006 study gave twenty people a location-based reminder system on their phones for four weeks. When a reminder arrived in the right place, the task was more likely to get done. But a fixed radius around an address turned out to be crude: someone on a motorway needed the alert two exits earlier; someone at home got nagged about a nearby place they had no intention of going to; sometimes the alert came after the person had already driven past.
A good cue is not just any cue. It is one that is certain to happen, at a moment when you can act.
Which one to use when
Peter Gollwitzer's implementation intentions, plans in the form "when X happens, I will do Y", help explain the criterion: the plan works because it hands control of the action to a concrete situation. So the real question is which situation will hold your intention, and sometimes the best situation available really is a time.
| Time-based reminder | Event-based reminder | |
|---|---|---|
| Who does the remembering | You, by checking the clock | The environment, together with you |
| Typical failure | Fires when you cannot act | The event never happens or arrives at a bad moment |
| Best for | Fixed deadlines and scheduled commitments | Actions that depend on a place or situation |
| Evidence | Weaker performance in studies, especially in older adults | Stronger performance in the lab and the badge study |
The honest limits
It is worth being clear about the size of the evidence. Einstein and McDaniel's studies were run in the lab, with simple tasks, and the event-based advantage showed up mostly in older adults. The badge experiment analysed twelve people. The location study had twenty, with no comparison group. I could not find a controlled trial comparing time and location reminders on ordinary people's phones over weeks.
What can be said with confidence is the mechanism: an intention tied to the clock depends on you remembering to look; an intention tied to a cue shares that work with the world. The rest is judgement, not law.
What I took from this
I used to write almost every reminder with a time, because time is the mandatory field in the app. Now I think first about the moment when I will be able to act, and only then about the clock.
That is the behaviour I wanted in ASB: a reminder that lands in the conversation you already open, asks whether you managed it, and if you did not, comes back at a better moment instead of disappearing.