First, last, and forgotten

Think back to the last handful of new words you tried to learn in one sitting — a dozen of them, say, met one after another. If you test yourself now, the result is oddly lopsided. The first two or three come back readily. The last two or three are still warm, right where you left them. And the ones in the middle? A blank, or a vague sense that they were there, once. You didn't study the middle words any less. They simply drew the worst seat in the house.

This is not a personal failing or a bad day for your memory. It is one of the most reliable shapes in all of psychology, so dependable it has a name and a curve to go with it: the serial position effect. Where a word falls in a sequence changes how likely you are to keep it — and once you can see the curve, you can stop putting your hardest words in the exact spot where memory is weakest.

A curve with two humps

The clearest early portrait of it comes from Bennet Murdock, in a 1962 paper in the Journal of Experimental Psychology with the flatly descriptive title "The serial position effect of free recall." His method was as plain as the name. People heard a list of ordinary words, one per second, and then — after a short pause — wrote down as many as they could, in any order they liked.

Plot how often each word was remembered against its position in the list, and you don't get a flat line or a gentle slope. You get a lopsided U: a rise at the beginning, a long sag through the middle, and a steep climb at the very end. The two ends of the list are remembered well; the belly of it is remembered badly. Those two humps have their own names. The advantage for the earliest items is the primacy effect; the advantage for the last items is the recency effect. Between them lies the trough where most of the forgetting happens.

Murdock didn't invent the observation — Hermann Ebbinghaus, testing his own memory on nonsense syllables in the 1880s, had already noticed that position mattered. But Murdock's curves made it unmissable, and the shape has since turned up in study after study, across words, pictures, and even events, in laboratories and out in the wild.

Two humps, two different reasons

The tidy thing about the two humps is that they don't share a cause. They are produced by two different parts of memory doing two different jobs — which is exactly why they behave so differently once you leave the room.

The primacy advantage is thought to come from rehearsal. The first words arrive into an empty mind with nothing competing for attention, so they get repeated, turned over, and quietly copied into durable, longer-term storage. By the time the middle words show up, your attention is already crowded and each new one gets a thinner slice of it. The early words, in other words, are remembered well because they were genuinely learned — encoded deeply enough to last.

The recency advantage is a different animal. The last words are remembered not because they were learned well but because they are still, in a sense, echoing — sitting in the front of your mind at the very moment you go to recall them. That makes recency feel just as strong as primacy in the instant of the test. It is not.

The words at the end of a session feel learned. Often they are merely still in the room.

The catch: recency is borrowed, not banked

Here is the finding that turns a curiosity into practical advice. In 1966, Murray Glanzer and Anita Cunitz ran a simple, decisive experiment, published in the Journal of Verbal Learning and Verbal Behavior. After the last word of a list, they had some people wait — counting backwards, or filling a brief span with an unrelated task — before recalling.

A delay of just a few seconds of distraction did something remarkable: it flattened the recency hump almost completely, while leaving the primacy hump and the middle untouched. The end-of-list words, so vivid a moment earlier, evaporated the instant something else needed the front of your mind. The early words, already written to deeper storage, sat there unbothered.

The lesson is quietly deflating for anyone who studies in a rush. The comfortable feeling that the last few words of a session are safely learned is, to a large degree, a mirage — a loan from short-term memory that gets called in the moment you look at your phone, start the car, or begin the next task. Primacy reflects learning that will still be there tomorrow. Recency mostly reflects learning that won't survive the walk to the kitchen.

What this means for a study session

Put the two findings together and the shape of the problem is clear. In any single run of new words, the middle is a graveyard, and the end is a trap that feels like a triumph. Only the beginning is doing reliable work. So the goal is simple to state: have fewer words stuck in the middle, and don't be fooled by the ones at the end.

The most useful move is almost boringly practical — make your sessions shorter. A long list has one beginning, one end, and a vast forgettable middle. Two shorter lists have two beginnings and two ends, and a middle that is half the size. Break twenty new words into two sittings and you have doubled the number of strong positions and shrunk the weak one. This is the same reasoning that spacing runs on, arriving from a different direction: shorter, more frequent bouts beat one long slog, and the serial position curve is part of why.

A second move costs nothing: don't study words in the same order every time. If your list never changes, the word marooned in position seven is marooned there forever — always rehearsed least, always recalled worst. Shuffle the order between sessions and every word eventually gets its turn at the favourable ends. This is a close cousin of why mixing up your practice works, and it's one of the few study habits that improves your results by doing less arranging, not more.

And there is an escape hatch for any word that must live in the middle: make it stand out. A single distinctive item planted in an otherwise uniform list gets remembered far better than its bland neighbours — a companion phenomenon known since Hedwig von Restorff described it in 1933 as the isolation effect. A middle word that you say aloud, tie to a vivid mental image, or fold into a sentence about your own life stops being interchangeable. Distinctiveness is the thing that lets a word climb out of the trough on its own.

A couple of honest limits

Two cautions, so the curve isn't asked to carry more than it can. First, the serial position effect was mapped mostly on lists recalled minutes later — a snapshot of a single session, not a full theory of how vocabulary is built over weeks. It tells you how to arrange one bout of study; it does not replace the slower machinery of the forgetting curve and repeated exposure. Even a perfectly ordered session leaves its words exposed to ordinary forgetting if you never return to them.

Second, none of this is a reason to stop meeting words in the middle — it's a reason to stop relying on a single long list to do the teaching. The fix isn't to study less. It's to break study into shapes with more edges and fewer soft middles, and to lean, as always, on bringing each word back from memory across days. Position sets up the first encounter. Spacing and retrieval are what keep it.

Where vocabla fits

A fair amount of this is built into how vocabla already works. Practice arrives in short, self-contained bouts rather than one endless scroll, so there are more beginnings and more ends and far less forgettable middle. The words you review don't march past in a fixed order — they're reshuffled and spaced to return just as each one begins to fade, which means no word gets permanently stranded in position seven. And because every word is met inside a real sentence, even a middle-of-the-list word arrives with something distinctive attached, rather than as one more interchangeable pair.

The next time you sit down to learn a batch of words, don't reach for the longest list you can stand. Reach for a short one, meet it, and come back to a different short one later. Give more of your words a front row or a back row — and, for the few that end up in the middle, give them something to make them impossible to confuse with the word beside them. The curve is old and stubborn. You don't have to fight it; you only have to stop feeding it your best words.