Abstract
The question this article answers: when an idea offered in a meeting is later credited to somebody else, what usually happens in the minds of the people in the room, why does it keep happening to the same people, what part does personality play on both sides, and what can the person who lost it, the people who saw it, and the manager do differently?
It follows a single composite meeting in which an analyst’s idea comes back ten minutes later in a manager’s words, and everyone present leaves with a sincere, different memory of whose it was. It explains why the theft story is the one people reach for, then describes four ordinary failures of memory that move ideas between people without anybody intending it, and the newest one, ideas drafted with AI.
It then shows why the losses are not random: the gap in memory is filled by status, gender, and fit, and modest people hand over credit. A separate part weighs the deliberate credit-taker, the evidence on narcissism and knowledge theft, and what losing an idea does to the people it happens to. The closing part covers the people most often mislabelled, including the reader, and what each person in the room can do, saying plainly where no trial has been run.
Keywords: credit for ideas; idea theft; knowledge theft; office politics; meetings; source monitoring; cryptomnesia; unconscious plagiarism; next-in-line effect; overclaiming; self-serving bias; Matthew effect; status; gender; narcissism; amplification; AI and memory.
PART I. TEN MINUTES LATER
Chapter 1: The Problem in Every Meeting
1.1 Single Ownership
This scene is a composite, drawn from recurring patterns in research and advisory practice, and it describes no real person or organisation. It will follow us through the article.
The weekly review of a regional insurance business had been stuck for twenty minutes on the same problem: claims from one product line were taking twice as long to settle as everything else, nobody could agree why, and the director in the chair had already asked twice for a suggestion that was not more staff. A claims analyst who had been with the business for three years and rarely spoke in the review said the delay seemed to start when a claim passed between two teams, and suggested that one person should own each claim from the first call to the final payment. There was a short silence; the director said it was worth thinking about, and the meeting moved on to staffing.
About ten minutes later, an operations manager, who sat at the director’s end of the table and spoke in most reviews, came back to the delays. What the business really needed, he said, was single ownership, one person carrying each claim end to end, and he had been thinking about it for a while. The director said that this was exactly the kind of thinking the business needed, asked him to draw up a plan, and wrote his name beside the action on the whiteboard.
1.2 Three Memories of One Meeting
The analyst said nothing in the room, but afterwards she told a colleague that her idea had been stolen, and the colleague, who had been sitting opposite her, said he was not sure it had been: he clearly remembered the operations manager raising it, and could not quite remember whether anybody had said it earlier. The operations manager, asked about it weeks later, was certain the plan was his. He remembered working it through in his head during the meeting, and he had added two details that made it workable: how claims would be allocated and what would happen when somebody was on leave.
The plan worked. Settlement times on the product line fell within a quarter, the operations manager presented the results at the next regional meeting, and the analyst stopped offering suggestions in the weekly review. Nobody noticed that she had stopped. Asked separately, each of the three would have said that they remembered the meeting clearly, and each would have been telling the truth as they knew it.
1.3 Why It Matters More Than It Looks
A single idea changing hands looks like a small injustice, the kind of thing people are told to shrug off. It is not small, and it is not rare. In one of a series of studies with more than 1,500 workers, 91 per cent said they had had an idea or piece of work taken, had taken one themselves, or had watched it happen to somebody else (Zweig, Damp & Scott, 2025).
People who had lost credit this way reported stronger intentions to hide what they knew, stay quiet to protect themselves, and work against the organisation in other ways. The second of those is exactly what the analyst did.
The costs run well beyond the person who loses the idea. Organisations live on the suggestions people choose to offer, and when managers took up an employee’s suggestion and then claimed the credit for it, the employee felt less respected, identified less with the group and offered fewer suggestions afterwards, even though the idea itself had been adopted (Johnson, Wu, Zhang & Lyu, 2024). The business in Chapter 1 gained a plan and lost its source of the next one.
Because the people who lose credit most often are, as later chapters show, the people who were already least likely to be heard, the loss is not spread evenly. It falls on the quiet, the junior and the people whose faces do not fit the room’s picture of who has ideas.
The problem is also growing a new form. Ideas are now drafted with software as routinely as they are discussed in meetings, and a recent experiment found that a week after people had written up an idea that a chatbot had suggested, most of them could no longer say whose idea it was (Zindulka, Goller, Fernandes, Welsch & Buschek, 2026). Credit was already hard to track between two people. It is harder still between a person and a machine.
1.4 Why the Usual Responses Fail
The advice people usually get is to speak up, confront the person who took the idea, or keep quiet and play the long game. Each of these assumes someone in the room knew exactly what they were doing, and each tends to fail for that reason, because telling the analyst to speak up ignores that she did speak up, clearly, and the room still lost her name. Confronting the operations manager would have meant meeting a man who sincerely believed the plan was his and would have heard the accusation as an attack on his character, which, as Chapter 2 shows, is how people hear it.
1.5 What This Article Argues
Most organisations treat a lost idea as a matter of character, a thief and a victim, and respond with accusation, advice to be more assertive, or silence. Yet people remember ideas far better than they remember who said them; the gap is filled by whoever fits the room’s expectations, and personality decides who takes credit and who hands it over. Therefore, most lost credit is a predictable failure of memory, steered by status and personality, and the most effective remedy is not an accusation but a single sentence, said at the time, that names the source.
The article follows the questions somebody in that room would ask: Part II asks how an idea can lose its owner without anybody stealing it, Part III asks why it keeps happening to the same people, and Part IV turns to the colleague who does it on purpose and to what losing credit does to the people it happens to. Part V asks what the analyst, her colleague, and the director could each have done differently, and what each can do on Monday.
Chapter 2: Theft, or Memory
2.1 The Story We Tell
When an idea changes hands, most of us reach for an explanation about character. The person who repeated it is ambitious, careless, or dishonest, and the person who lost it was too quiet or too junior to defend it; the story is attractive because it is simple and because it puts the whole problem inside one person rather than in the room. It is often wrong.
We also judge this kind of loss more harshly than almost any other. In a set of experiments, people judged somebody who had stolen a colleague’s idea to have worse character than somebody who had stolen money, were less willing to help them at work, and, when forced to choose, preferred to work alongside the money thief (Ellis, 2022). The harsher judgement came from the belief that taking an idea says more about who a person really is, and stealing a creative idea was judged more severely than stealing a practical one. An accusation of idea theft is therefore one of the most damaging things one colleague can say about another, which is a good reason to be sure before saying it.
2.2 Why Everybody Is Sure
There is an ordinary reason why each person in the review left certain. Across 266 studies, people took more credit for success and less blame for failure than an even-handed account would give them, a tendency psychologists call the self-serving bias, and it appeared in nearly every sample studied, across ages and cultures, although it was considerably smaller in Asian samples (Mezulis, Abramson, Hyde & Hankin, 2004). The operations manager’s certainty that the plan was his is the state of mind most people would be in, and on its own it is not evidence of dishonesty.
The analyst’s certainty is just as understandable, and on the central fact she is very likely right: she said it first. What she cannot know is what the others heard, and the evidence in Part II suggests they heard less than she said and remembered even less about who said it. Both memories can be sincere. Both can be partly wrong.
2.3 What the Evidence Says Instead
The research on memory offers a different account from the theft story, and it is not a kinder one. People remember what was said far better than who said it; the moment somebody is preparing to speak is when they take in the least of what is being said, and improving an idea is the thing most likely to make a person remember it as their own. When the memory of the source fails, credit then flows towards whoever fits the room’s picture of somebody with good ideas.
None of that excuses deliberate theft, which happens and which Chapter 11 treats on its own terms. What it changes is the remedy. A problem of character is answered by confronting a person, while a problem of memory is answered by changing what gets recorded, who names the source, and when. Those are things anybody in the room can do.
PART II. HOW CAN AN IDEA LOSE ITS OWNER WITHOUT ANYBODY STEALING IT?
Chapter 3: Memory Keeps the Idea and Drops the Name
3.1 How the Mind Files a Source
The mind does not store where a memory came from the way a computer stores a file's author. Psychologists call the process of working out where a memory came from source monitoring, and the review that set out the idea argued that the source is rebuilt each time from the feel of the memory: how vivid it is, how much of the setting comes with it, and how much effort seemed to go into it (Johnson, Hashtroudi & Lindsay, 1993). Because the source is inferred rather than recorded, it can be inferred wrongly, which is why a memory of having had an idea can be vivid, detailed and mistaken all at once.
3.2 Told Not To, and Doing It Anyway
One error this predicts is cryptomnesia, or unconscious plagiarism: producing an idea one has encountered before while believing it to be new. The first systematic test seated people in groups of four and had them take turns producing examples of a category; in two of the three experiments, the other three were live participants taking turns alongside them (Brown & Murphy, 1989). Later, each person was asked to recall the items they had produced and to come up with new ones, and in every task and every experiment people repeated other members’ items as their own, although they had been told explicitly to avoid exactly that, and the items they most often took were those of the person who had spoken immediately before them.
The tasks were simple word and category exercises rather than business problems, and the step from there to a meeting is an inference, though a short one. A meeting is, after all, a group of people taking turns to produce ideas, each of them busy with their own.
3.3 What That Means for the Review
The common belief is that somebody who repeats your idea must have heard it and decided to take it. The evidence offers a more ordinary possibility: the operations manager heard it, kept the idea, and lost the memory of who said it; when it came back to him ten minutes later, it arrived with no source attached, and the most available source was himself. Nobody lied.
In my experience, the phrase I’ve been thinking about for a while is often true in a way the speaker does not intend. The idea really has been in their head for a while, put there by somebody else ten minutes earlier.
The uncomfortable part is that this works in every direction. Anybody who has sat in enough meetings has almost certainly done it. They would have no way of knowing, because an idea that has lost its source feels exactly like an idea of one’s own. The feeling of having thought of something is not evidence of having thought of it first.
Chapter 4: The Ten Seconds Nobody Hears
4.1 The Next Person to Speak
People hear least of what is said just before they speak, because they are busy getting ready to speak. The original study of this is short and striking: 88 men sat round a table in groups, took turns reading words aloud, and then tried to recall everything that had been read, and they remembered 87 per cent of the words they had read themselves and 37 per cent of the words read by the person beside them (Brenner, 1973). Recall dipped for anything read in the few turns either side of a person’s own, about nine seconds in each direction, and it dipped further when a distracting noise made the task harder. Psychologists call it the next-in-line effect.
A more recent set of experiments traced the cause to anticipation itself. Words studied silently just before a person had to read something aloud were recognised less well than words studied before another silent item, and the cost largely disappeared when there was nobody in the room to perform for (Forrin, Ralph, Dhaliwal, Smilek & MacLeod, 2019). Getting ready to perform for other people pulls attention away from whatever is happening at the time, and in a meeting, what is happening at the time is usually somebody else’s idea.
4.2 Why the Quiet Pay Most
Both studies used word lists rather than conversation, and I couldn’t find any study that measured the effect on memory for ideas in a working meeting. Nothing in the evidence, though, suggests that meetings are kinder to memory than word lists, and the implication is uncomfortable: an idea offered into a pause that several people are waiting to fill is offered at the moment it is least likely to be remembered, and least likely to be remembered as yours.
That cost does not fall evenly, because people who speak often tend to speak when they have a point ready, while people who speak rarely tend to wait for a gap, and a gap after twenty minutes of disagreement is, in my experience, exactly when several people have just decided to say something. The analyst in Chapter 1 spoke into that gap. Her suggestion was heard as a suggestion, and not as hers.
Chapter 5: Let Me Build on That
5.1 Improving an Idea Rewrites Its Owner
The person who adds to an idea is the person most likely to remember it later as their own. A study designed to test this used a creativity task in which people produced unusual uses for everyday objects, and then had participants work on ideas that somebody else had produced, either imagining them or improving them (Stark & Perfect, 2006). When people were later asked to recall their own ideas, those who had improved somebody else’s idea claimed it as one of their own at high rates, and the authors concluded that it is the generative work done on an idea, the act of adding to it, that decides whether it is later appropriated.
There was a twist in the same study, and it explains why the operations manager could be so sure. Working on somebody else’s idea, by improving it or simply imagining it, made people less likely to offer it again when they were asked for something brand new, so they knew perfectly well that the idea already existed. What improving it changed was who they remembered it to be.
5.2 Why the Additions Matter
This fits the account in Chapter 3, because source monitoring relies partly on how much effort a memory seems to carry, and an idea one has worked on carries the effort of one’s own thinking, so the memory comes back feeling home-grown. The improvement is a genuine contribution, and it is worth crediting. It also overwrites the trace of where the idea started.
The operations manager added two details that made single ownership workable, and those additions are the most likely reason he came to remember the whole plan as his, which means that the moment an idea is most at risk is not when it is ignored but when somebody more senior starts to improve it, because that is when the memory of its source is being written over in the head of the person whose memory will decide the credit. In the reviews I have sat in, the sentence " Let me build on that” is often followed, within a few minutes, by an idea with a new owner.
Chapter 6: Everybody Did Most of the Work
6.1 The Sums That Do Not Add Up
Ask each member of a team what share of the work they did, and the answers will add up to more than the whole. The pattern was first documented in married couples, who each claimed more than half the responsibility for household tasks and, more surprisingly, for starting arguments, so that their claims together slightly exceeded the total, a pattern the authors traced to people simply remembering their own contributions more easily than their partner’s (Ross & Sicoly, 1979). Psychologists call this overclaiming, and in groups it is much larger.
The clearest figures come from a set of studies that followed it as groups grew (Schroeder, Caruso & Epley, 2016). Among 699 MBA students in study groups, the shares people claimed implied totals of more than 140 per cent in groups of eight or more, although nothing was offered to anybody for claiming more, and in a national sample of 1,962 adults recalling a real group they had worked in, the share each person claimed, multiplied by the size of the group, came to about 141 per cent for small groups and 171 per cent for large ones. A large team whose members all claimed what its average member claimed would have accounted for well over one and a half times the work it actually did.
6.2 The Handgrip Test
The most persuasive of these studies took place at a science museum, where 339 visitors squeezed a handgrip together in teams of three or six: the six-person teams claimed, between them, an average of 114 per cent of the effort, against 104 per cent for teams of three, although their actual grips did not differ (Schroeder, Caruso & Epley, 2016). Nobody worked harder in the bigger teams. The extra claims came from memory, not from effort.
6.3 One Question That Shrinks It
The same studies found what reduces it, and it is a single question. Asking people to think about what each of the others had contributed before estimating their own share brought the national sample’s figure down from about 204 per cent to between 118 and 145 per cent, and the lowest figure came when people considered every other member before turning to themselves (Schroeder, Caruso & Epley, 2016). Changing what people remembered first changed what they claimed.
The question has a cost worth knowing before asking it. In four experiments, asking group members to consider others’ contributions reduced overclaiming, but the members who had contributed most became less satisfied with the group and less keen to work with it again (Caruso, Epley & Bazerman, 2006). Seeing the whole picture is fairer, and it is not always more comfortable.
6.4 What That Means for the Review
The common belief is that a colleague who claims too much is unusual. The evidence says almost everybody claims too much, sincerely, and that a team in which every member feels under-credited may simply be a team of people behaving normally. The operations manager remembered his own thinking in detail and the analyst’s sentence only faintly, while the analyst remembered her own sentence in detail and his thinking not at all; both were overclaiming in the ordinary way.
Chapter 7: The Machine in the Meeting
7.1 A Week Later, Whose Idea?
The newest version of the problem has a different contributor. In a study published in 2026, 184 people produced and wrote up ideas for a set of problems, sometimes alone and sometimes with an AI chatbot, each person working through every combination of help and no help, and about a week later they were asked where each idea had come from (Zindulka, Goller, Fernandes, Welsch & Buschek, 2026). When they had done all the work themselves, they named the source correctly about 92 per cent of the time, but when the chatbot had produced the idea, and they had written it up, they were right about 38 per cent of the time, and their confidence did not fall in step with their accuracy.
This is Chapter 5 again, with software in the role of the colleague. Writing an idea up is work, and that work makes it feel owned, so the chatbot becomes one more source for memory to drop.
7.2 What Changes When Drafts Are Shared
The study is one experiment with an online panel, and the step from it to shared work in teams is an inference, but the implication is still worth taking seriously, because a person who drafts with AI may sincerely present as their own an idea a tool generated, and a team that circulates drafts produced partly by software has a further source in the room that nobody’s memory will track reliably. In my experience, the teams that handle this best keep a short note inside the draft about where it came from, which nobody has yet tested but which costs a line.
7.3 Where Part II Leaves Us
An idea can lose its owner without anybody stealing it, in at least four ordinary ways: memory keeps the idea and drops the name, the people about to speak hear it least, improving an idea makes it feel home-grown, and everybody remembers their own part best. None of this is random, though. When memory for the source fails, something fills the gap, and Part III is about what fills it.
PART III. WHY DOES IT KEEP HAPPENING TO THE SAME PEOPLE?
Chapter 8: The Idea Goes to the Person Who Fits
8.1 Memory Keeps the Kind of Person
When nobody is quite sure who said something, the idea tends to be credited to the person who looks like somebody who has ideas. A classic study showed this basic error: people watched a group discussion and were later asked who had said what (Taylor, Fiske, Etcoff & Ruderman, 1978). They often remembered that a woman or a black member of the group had made a point but confused which one, so remarks were swapped between people who shared a visible category far more often than between people who did not. Memory had kept the kind of person and lost the individual.
8.2 Status Fills the Gap
Status fills the gap in a similar way, and sociologist Robert Merton described what he called the Matthew effect: the tendency for credit for shared or simultaneous work to flow to whoever is already most eminent, so those who have already been recognised receive more recognition (Merton, 1968). For decades it was hard to separate that from the possibility that eminent people simply do better work.
A natural experiment, a change made for other reasons that happened to work like a controlled test, separated the two: in an internet standards body, a surge of submissions led the announcement emails to replace authors’ names with et al., and when a high-status name was hidden in this way, the proposal received less attention; the authors estimated that signals carried by names could explain up to three-quarters of the gap in publication between high-status and low-status authors (Simcoe & Waguespack, 2011). The effect vanished for proposals that were screened more carefully, which suggests status matters most when attention is short. A weekly review, with twelve items on the agenda and an hour to cover them, is a place where attention is short.
8.3 What That Means for the Review
The previous article in this series was about the loudest voice in the room, and why rooms so often hand it their trust. This is the same room seen from the side of the idea. The operations manager sat at the director’s end of the table, spoke in most reviews and already owned several of the business’s improvements, and the analyst did none of those things, so when the room’s memory of the source began to fade, it faded towards him.
The common belief is that a good idea will be recognised whoever says it. The evidence says that when memory for the source is weak, which Part II suggests is most of the time, credit follows the person rather than the idea, and it moves towards whoever is senior, visible or matches the picture the listeners hold of somebody with ideas. Merit still matters. It matters most when somebody can remember whose merit it was.
Chapter 9: When Nobody Is Sure Who Did What
9.1 Joint Work, Unequal Credit
Gender shapes the gap too, and the clearest evidence comes from joint work: in three experiments, when a man and a woman were described as having succeeded together, and it was unclear who had done what, the woman was rated as less competent, less influential and less likely to have taken a leading role, unless each person’s contribution was made explicit or there was clear evidence of her competence (Heilman & Haynes, 2005). The authors called the process attributional rationalisation: when the facts were ambiguous, people resolved them in line with what they already expected.
The same pattern shows up in careers: among academic economists, a paper written alone raised a woman’s chance of tenure more than a man’s, while a paper written with others raised a man’s chance by about 8 percentage points and a woman’s by about 6, and women who had co-written all their papers were tenured at about 37 per cent, against about 72 per cent for men (Sarsons, Gërxhani, Reuben & Schram, 2021). Two follow-up experiments, one with 506 online participants and one with 479 working recruiters, found the same pattern in settings where nothing but the assignment of credit could explain it. In economics, co-authors are conventionally listed alphabetically, which tells a reader nothing about who did what, and that is exactly the ambiguity the experiments describe.
9.2 The Experiment That Replays the Review
The most direct test of the review in Chapter 1 is new. In five preregistered experiments, meaning that the methods and predictions were lodged publicly before any data were collected, 2,264 people watched a video of an engineering team at work, in which a woman offered an idea and a male colleague later repeated it (Dennehy, Engstrom, Hernanto & Schmader, 2026). Across the three experiments that measured it, only about half the observers, between 48 and 57 per cent, correctly said the idea was hers.
Then a colleague said, in a single sentence, that it had been her idea, and correct attribution rose to around four in five. Chapter 15 returns to that sentence, because it is the best-evidenced step in this article.
For now the point is the first figure. Half the people who watched a short exchange on video, with no stake in the outcome, could not say whose idea it was. A real meeting, in which people are also following the argument and waiting to speak, is unlikely to be easier.
9.3 What That Means for the Review
This is why the problem is not spread evenly. Everybody loses ideas to failures of memory, but the failures are not random: a junior analyst who speaks rarely, a woman in a mostly male team, anybody whose face does not fit the room’s picture of who has good ideas, loses more of them, and the people who gain them do not experience it as gaining anything. The experiments also point to the remedy. In the joint-work experiments at 9.1, the gap largely disappeared once it was made clear who had contributed what (Heilman & Haynes, 2005), which is a matter of record rather than of anybody’s personality.
Chapter 10: The Colleague Who Gives It Away
10.1 Credit That Is Handed Over
Some credit is not taken at all. It is handed over, sincerely, by people who doubt their own part, and the evidence on this comes from an unexpected place. In a study of 160 students who took a creativity test in pairs and were told they had done very well or very badly, people low in narcissism did not simply claim less: they gave their partner more credit for a shared success than they took themselves, a pattern the authors described as other-serving (Campbell, Reeder, Sedikides & Elliot, 2000). In a pair made up of one high scorer and one low scorer, both would have agreed that the high scorer deserved the credit.
The same thing has been shown for women working with men: in four studies, women who had succeeded on a task with a male partner gave him more of the credit and took less for themselves, unless their own role was made clear or they were given evidence of their competence at the task, and they did not do so when the partner was a woman (Haynes & Heilman, 2013). When it was unclear who had done what, they resolved the uncertainty against themselves.
10.2 Speaking Up Is Not the Same as Being Credited
How a person speaks matters as much as how much: in a three-wave field study and an experiment, people who spoke up to suggest improvements gained status with their peers and were more likely to emerge as leaders later, but the gain went mostly to men, and women who made the same kind of suggestions gained less from them (McClean, Martin, Emich & Woodruff, 2018). Talking was not enough to be credited, because who was talking changed what the talking earned.
10.3 What That Means for the Review
The common belief is that people who lose credit have it taken from them. The evidence says much of it is given away by people who are uncertain of their contribution to people who are certain of theirs, and the analyst, who did not object, would have been read by the room as agreeing that the plan was the manager’s. Nobody had to take anything from her for the credit to move.
In my experience, the phrase people use most often about their own lost ideas is it was a team effort. It is generous, and it is usually believed. The answer is not for modest people to become somebody else, because the studies that tested it found the pattern weakened or disappeared once each person’s contribution was made clear (Heilman & Haynes, 2005; Haynes & Heilman, 2013). The fix sits in the record, not in the personality.
PART IV. WHAT ABOUT THE COLLEAGUE WHO DOES IT ON PURPOSE?
Chapter 11: The Credit-Taker
11.1 Narcissism and the Shared Success
Some people do take credit deliberately, and the trait most clearly linked to it is narcissism. Here, that means the everyday personality trait, a mix of grandiosity, a need for admiration and a sense of entitlement that everybody has to some degree, and not a clinical diagnosis. Memory explains most lost ideas, but not all of them, and the evidence on deliberate credit-taking is worth knowing precisely because it is narrower than the theft story suggests.
The study in Chapter 10 is also the clearest test of how narcissism shapes credit, and it found that when the pair succeeded, people higher in narcissism took more of the credit, and when it failed, they took less of the blame (Campbell, Reeder, Sedikides & Elliot, 2000). When people worked alone, with no partner to take credit from, the difference largely disappeared, so the trait did not make people think more of their work in general. It showed up when credit could be taken from somebody else, which is exactly the situation of a shared idea in a meeting.
11.2 Knowledge Theft
Outside the laboratory, the deliberate version has a name, and the researchers who developed a way of measuring knowledge theft, a colleague intentionally taking credit for somebody else’s work or ideas, tested it across seven samples and more than 1,500 workers, and found it to be distinct from other kinds of hostile behaviour at work, such as undermining or aggression (Zweig, Damp & Scott, 2025). It is its own kind of harm, and people recognise it as such.
Managers are not exempt. A field study in China and an experiment in the United States found that when managers endorsed an employee’s suggestion and then claimed the credit for it, much of the value of the endorsement was lost: the employee felt less respected, identified less with the group and offered fewer suggestions later (Johnson, Wu, Zhang & Lyu, 2024). The research team behind the knowledge-theft measure was also funded to study whether psychopathy and Machiavellianism predict it, and I could find no published result from that work. That is an absence of evidence, and it should be read as one rather than as evidence that those traits play no part.
11.3 Why It Is Harder to See Than It Sounds
The common belief is that the credit-taker is a type, and that the type is easy to spot. The evidence supports something narrower and harder to see. A tendency to take credit when it can be taken from somebody else goes with a measurable trait. The same person may be perfectly fair about their own work when nobody else is involved, so the behaviour shows up only in the moments that matter.
It also means that deliberate theft and sincere misremembering can look identical from the outside, and the operations manager may have been either.
The evidence in Part II says that the second is more likely in any single case, while the evidence here says that the first is real and concentrated in some people. Neither licenses a judgement about a particular colleague from one meeting. The pattern worth watching for, in my experience, is not a single borrowed idea, which happens to everybody, but a colleague whose name is attached to the team’s successes and whose team is attached to the failures, over a year.
Chapter 12: What Losing an Idea Does to People
12.1 It Is Felt as a Loss
Losing credit for an idea changes how a person behaves afterwards: they share less, say less and guard more, and the knowledge-theft studies found that workers who had experienced it reported stronger intentions to hide what they knew, to stay silent to protect themselves and to behave in other ways that harm their workplace (Zweig, Damp & Scott, 2025). Two follow-up experiments, with more than 1,600 participants between them, found that what drove the anger was a sense of loss, of ownership, recognition and opportunity (Zweig, Scott, Damp & Paquin, 2026). That is why being told to let it go so rarely helps.
The same experiments found what eased it: when a leader or a colleague stepped in publicly at the time to acknowledge the original creator, or restored the credit by pointing to a record in which ideas had been logged under their originators’ names, people reported significantly less loss and less anger (Zweig, Scott, Damp & Paquin, 2026). Both steps worked by restoring what had been lost, rather than by asking the person to feel differently about losing it.
12.2 The Idea Survives, the Suggestions Do Not
The effect on future suggestions appears in the study of managers too. When a manager endorsed an idea and took the credit, the employee offered fewer ideas later, although the idea itself had been adopted (Johnson, Wu, Zhang & Lyu, 2024). The idea survived. The suggestions did not.
Most of these measures are reported intentions rather than observed behaviour, and the workplaces in the experiments were simulated. They describe the analyst all the same. She stopped offering suggestions, and nobody in the review noticed. An absence is the hardest thing to see in a meeting, and repeated a few times it can turn a contributor into a spectator without anybody deciding that it should.
12.3 The Arithmetic for Whoever Runs the Team
For anybody running a team, this changes the sums. The person who repeated the idea gained a little credit, and the team lost the next suggestion from the person who did not get it, and the one after that. Nobody records that loss. When a previously active contributor goes quiet in meetings, the first question I ask is whether something of theirs recently appeared under another name, because in my experience it is one of the most common causes and one of the least often checked.
PART V. WHAT CAN I DO ABOUT IT?
Chapter 13: The People You Will Mislabel, Including Yourself
Read this before acting on anything above. Every pattern in this article describes what tends to happen across many people and many meetings, and each person below can be misread by someone who has just learned it.
13.1 The Colleague Who Repeated Your Idea
Most people who repeat an idea are not stealing it. The memory evidence in Part II says that hearing an idea, losing its source and rediscovering it is ordinary, and that improving it makes the rediscovery more likely. Accusing somebody of theft based on one meeting will often be wrong, and it will be heard as an attack on their character, which Chapter 2 shows is how people judge idea theft. The test is a pattern over time, not an incident.
13.2 The Colleague Who Seems to Take Credit Every Time
A colleague who keeps appearing as the source of ideas may simply be the source of them, because some people do generate more, and the overclaiming evidence applies to the people complaining about them as much as to them, so before concluding that a colleague is a credit-taker, it is worth checking a written record, if there is one, against your memory. Memory, as Part II shows, is a poor witness in its own cause.
13.3 The Quiet Colleague
A quiet colleague who rarely gets credit may also rarely have contributed. Chapter 10 says that modest people give credit away when it is unclear who did what, not that every quiet person is an unrecognised source of good ideas. The correction is a record that shows who contributed what, not an assumption running the other way.
13.4 The Manager Whose Name Is on Everything
A manager’s name often appears on a team’s work because the manager is accountable for it, and that is not theft. The study in Chapter 11 concerns managers who endorse an idea and then present it as their own, which is different from presenting the team’s work while naming the team. Accountability is not authorship.
13.5 You
If you have read this as a story about other people, it is also about you. The overclaiming studies found the pattern in almost everybody, and the memory studies found it in people who had been told to avoid it. You have almost certainly presented as yours an idea that was not, with no way of knowing which. The habit worth keeping is to ask, before claiming anything, who else was in the room when it formed.
Chapter 14: If Your Ideas Keep Going Missing
14.1 Write It Down Before You Say It
Memory for who said what is weak, and memory for what is written is much less so. Sending a short note of a suggestion before a meeting, or straight after it, gives everybody’s memory a source to attach to, and the joint-work experiments in Chapter 9 suggest why it should help: once it was clear who had contributed what, most of the gap in credit disappeared (Heilman & Haynes, 2005), although no study I could find has tested the note itself in a real workplace, so this is practice built on that evidence rather than a tested remedy.
14.2 Choose When the Room Is Listening
The next-in-line evidence in Chapter 4 suggests that an idea offered into a pause that several people are waiting to fill is the idea least likely to be remembered, so offering it when you have been asked, or tying it by name to the previous speaker’s point so that the room has to stop and connect the two, may make it more likely to be heard as yours. It has not been tested in meetings, but it costs nothing to try.
14.3 Claim It Plainly When It Comes Back
When your idea returns in somebody else’s words, the evidence in Chapter 9 says that naming the source at the time restores most of the credit, and Chapter 10 says that modest people tend to let it go instead, so a short, neutral sentence- that’s the point I raised earlier, and I’d be glad to work on it with you- puts your name back without accusing anybody of anything. The experiments tested the sentence from a colleague rather than from the person whose idea it was, so the version you say yourself rests on practice and on the logic of those experiments rather than on a trial.
14.4 Keep a Record That Speaks Over Time
Keeping a dated note of what you proposed and what came of it is the individual version of making each person’s part clear, which is the step that removed most of the gap in the joint-work experiments (Heilman & Haynes, 2005; Haynes & Heilman, 2013). It also gives you something better than memory to consult before deciding that a pattern exists, which Chapter 13 says you should do first.
Chapter 15: If You Hear It Happen
15.1 Name the Source, Out Loud
This is the best-evidenced step in the article, and it belongs to the colleague sitting opposite. In the experiments in Chapter 9, when a colleague simply said whose idea it was, correct attribution rose from around half of observers to around four in five. Naming the source after the idea had already been repeated by somebody else restored the credit almost entirely, and it worked as well when a man said it as when a woman did (Dennehy, Engstrom, Hernanto & Schmader, 2026).
The same studies compared simply naming the source with a more elaborate version, known as amplification, in which the colleague also invites the originator to say more. Naming the source worked as well. That matters because it means the remedy does not have to fall on the people already losing credit, and it does not require a speech. That was her idea a few minutes ago.
15.2 Do It at the Time
The knowledge-theft experiments found that a leader or colleague publicly acknowledging the original creator, at the time, reduced the sense of loss and the anger that losing credit produces (Zweig, Scott, Damp & Paquin, 2026). In my experience, an acknowledgement made the following week does much less, because by then the name on the whiteboard has already become the memory. The colleague opposite the analyst had one sentence available to him in that meeting and didn't know he needed it.
15.3 Ask What Each Person Brought
When a discussion turns to who deserves credit for a piece of work, asking each person to say what the others contributed before their own share reduced overclaiming in every study that tested it (Schroeder, Caruso & Epley, 2016), although the people who did most should expect to find it uncomfortable, because the same question made high contributors less satisfied with their group (Caruso, Epley & Bazerman, 2006). It is fairer, and it is worth the discomfort.
Chapter 16: If You Run the Team
16.1 What the Director Controlled
A director cannot fix anybody’s memory, and trying to make people more honest would miss the point, since most of them already are. What the director in Chapter 1 controlled was the record, when sources were named, and whether her own behaviour taught the team that ideas were safe to offer. Each of those has evidence behind it.
Making each person’s contribution to joint work clear removed most of the devaluation of women’s contributions in mixed teams (Heilman & Haynes, 2005), and naming the originator of an idea restored correct attribution (Dennehy, Engstrom, Hernanto & Schmader, 2026). Pointing to a record of ideas logged under their originators’ names reduced the anger of people whose ideas had been taken (Zweig, Scott, Damp & Paquin, 2026), while managers who endorsed an idea and then claimed the credit for it received fewer suggestions afterwards (Johnson, Wu, Zhang & Lyu, 2024).
16.2 Five Habits That Follow
Write the originator’s name beside an action on the whiteboard or in the minutes, alongside the name of the person who will carry it out, because in my experience the written name becomes the room’s memory within days. Had the director written analyst’s proposal, operations manager to plan, the review would have remembered both, and both would have been right.
When you take up an idea yourself, name its source before you build on it, in words as plain as building on what you said earlier, which is the step the attribution experiments tested, and when your own name goes on the team’s work, say who did it, since the study of managers shows what not doing so costs. Keep a simple log of proposals with their originators and dates, which reduced anger after theft in the experiments, although nobody has tested whether it prevents the taking. And when an active contributor goes quiet, ask whether something of theirs has recently been credited elsewhere.
The knowledge-theft researchers also suggest rewarding teams rather than individuals, to reduce the motive for taking credit in the first place (Zweig, Damp & Scott, 2025). That is the authors’ suggestion, and it has not been tested. No trial, as far as I could find, has measured whether minutes, idea logs or a habit of naming sources change who gets credited or promoted in a real organisation over a year, so the experiments show that each step works in a single moment and leave the long run open.
Chapter 17: Back to the Review
17.1 What Happened in That Room
Run the meeting again with what the evidence describes: the analyst offered her idea into a pause several people were waiting to fill, and the operations manager kept the idea and lost its source, then improved it, writing his own effort over its origin.
The colleague opposite remembered the senior voice, the director wrote down the last name she heard, and the analyst, unsure whether objecting would sound petty, let it go. Nobody in that room had to be dishonest for the credit to move, and if one of them was, nothing about the meeting would have looked any different.
The cost was not the plan, which worked. It was every suggestion the analyst did not make in the year that followed, which nobody counted because nobody saw them not being made.
17.2 What I Still Do Not Know
Four things remain open. Most of the memory evidence comes from word lists and creativity tasks, and no study I could find counts how often ideas change hands in real meetings, so the claim that memory explains most cases rests on the laboratory evidence and on practice rather than on a count.
The evidence on deliberate credit-taking covers narcissism and leaves the other dark traits untested; the remedies have been tested in single moments rather than over time in real teams, and the study of ideas produced with AI is one experiment, published this year. Any of those could change what this article recommends.
17.3 Questions for Monday Morning
The temptation is to tell this as a story about credit thieves and their victims, and sometimes that is the right story. For most cases, the evidence supports a less dramatic and more useful one: memory keeps ideas and drops their sources, fills the gap with whoever fits, and does so in everybody, including the people who complain about it. Four questions follow from it.
- In your last three meetings, whose name went beside the actions, and was it the person who first raised them?
- Is there anybody on your team who used to make suggestions and has stopped?
- When you last said Let me build on that, did you say whose idea you were building on?
- What did each of the others bring to the last piece of work you were proud of?
What I would most like to have made harder is assuming that a repeated idea is a stolen one. What I would most like to have made easier is the sentence that fixes most of it, said at the time: that was her idea. It costs nothing to say, and in the experiments it worked whoever said it.



