Transactive memory: you outsourced facts, not your days
Try to recall a phone number. Not your own, and not an emergency line. A friend’s, a parent’s, the number of the place you worked five years ago. Most people can produce one, maybe two, and then stop. Twenty years ago the same person carried a dozen in their head without effort.
Nothing has gone wrong with anyone’s memory. The numbers left the head because a better place to keep them appeared. The phone holds them now, and the head, sensibly, stopped duplicating the work. This is such an ordinary trade that it barely registers as one. It also has a precise name in psychology, and a body of research that explains both what the trade is good for and where it quietly fails.
A partner outside the skull
The trade has a name: transactive memory. The psychologist Daniel Wegner introduced the term in the 1980s to describe something he had noticed in long-married couples. Neither partner remembered everything. One held the finances, the birthdays, the route to the in-laws. The other held the names of the neighbors, the history of the car, the contents of the freezer. Asked a question alone, either could come up short. Together they could answer almost anything, because each had become the other’s external store, and each knew which questions belonged to whom.
Wegner’s point was that the edge of a person’s memory is not the edge of their head. A transactive memory system is built out of two things: information held somewhere, and a reliable sense of where it is held. You do not need to carry the fact if you can carry the address of the fact. Carrying addresses instead of facts is not a weaker kind of memory. It is how every group of people who have lived or worked together for any length of time actually operates.
The internet joined the system
For most of human history the partners in these systems were other people. Then they were also notebooks, address books, the relative who “remembers everything.” In 2011 a paper put numbers on what happens when the partner is a search engine.
Betsy Sparrow, Jenny Liu, and Daniel Wegner published “Google Effects on Memory” in Science. Across four experiments, they tracked what people’s minds do when they expect a machine to hold the information.
In the first, participants answered a run of hard trivia questions and then took a Stroop test, naming the ink colors of words while ignoring the words themselves. After the difficult questions, people were measurably slower on computer-related words like Google and screen. Being stumped had pulled their attention toward the machine. The reflex to look it up fired before any decision to look it up.
In the second, participants typed forty trivia statements into a computer. Half were told the file would be saved; half were told it would be erased. The ones who believed the statements were saved recalled far fewer of them afterward, even when they had been explicitly asked to remember. The expectation of future access was enough, on its own, to switch off the effort of internal storage.
The remaining experiments showed the other side of the exchange. When statements were filed into named folders, people remembered the folder a statement had gone into better than they remembered the statement itself. Memory for where had quietly replaced memory for what. Sparrow and her colleagues concluded that the internet had become “a primary form of external or transactive memory.” The search box had taken a seat that used to belong to a spouse or a colleague.
The trade is sound, where it works
The popular version of this finding is that Google is making people forgetful, or worse. That is not what the data shows. Nobody in those experiments lost the ability to remember. They reallocated it. The memory budget that used to go to holding facts went to holding addresses instead, exactly as it does in a long marriage.
That reallocation is a good deal when two conditions hold. The partner has to actually keep the information, and you have to be able to get it back. For facts, both are true. The capital of a country, the year a film came out, the formula you half-learned in school: these sit somewhere addressable, and the address is stable. Offloading them frees the mind for things it is better at and costs you nothing, because the fact is always one query away.
The Google effect, in other words, is not a malfunction. It is a working transactive memory system, scaled up to the size of the internet.
The memory with no partner
There is one category of memory, though, for which neither condition holds.
Think about offloading last Tuesday. Not a fact you could have looked up: the actual texture of that specific day, the order things happened in, who you spoke to, what the afternoon felt like. When that memory leaves your head, where does it go? There is no search box for your Tuesday. It was never typed into a file, never filed into a folder, never given an address. The only partner it was ever handed to is the one labeled “I’ll remember it,” and that partner is the least reliable member of the entire system. It does not keep what it is given. It compresses, it rewrites, and within a week it has thinned an ordinary day down to almost nothing.
This is the gap the Google effect hides. The effect works because people offload things that are stored somewhere retrievable. It feels safe to forget a fact, because forgetting a fact is not really forgetting, it is delegating. But the same instinct, applied to your own days, delegates them to nobody. There is no folder name to fall back on.
We have built an extraordinary transactive memory partner for the world’s facts and never built one for our own lives. The drift of a lived day into a flat summary is something we have written about before; transactive memory explains why it goes unnoticed. The brain hands the day off with the same ease it hands off a phone number, and never learns that this time, nothing caught it.
A partner for the days
deariary builds each day’s entry from the traces your tools already leave on their own: calendar events, messages, commits, completed tasks, check-ins, the songs that played. It is not a prompt to reflect and not a box to fill in at night. It is a record that exists whether or not you think to make one.
In the language of this research, deariary is the transactive memory partner the system was missing. It takes the seat for autobiographical memory that Google took for facts. The arrangement is one you already trust completely: you do not carry a friend’s phone number because you know the phone has it; you can stop carrying the details of Tuesday because the diary has them. The relief Sparrow measured in the lab, the mind letting go once it trusts the store, becomes available for the part of memory that needed it most and never had it.
What changes is not how much you remember. It is what remembering your own life now means. It comes to mean the same thing remembering a fact already means: knowing, with justified confidence, where to find it.
We outsourced memory to our devices a long time ago, and on the whole it was the right call. The mistake was assuming the call covered everything. It only ever covered the facts. The ordinary days, the ones you would actually want back, were left inside, with the one partner that does not keep its files.
That is the gap deariary was built to close. It gives your own life the same deal your phone number already has.
