.png)
The Hedgehog Dilemma at Work: How Close to Get to the Person You Support, and How to Find the Right Distance at Speed.
Updated: Sep 28
It is twenty to eight and you are walking beside your executive down a corridor that is slightly too long, reading the day off your phone. Somewhere between the lift and the boardroom they tell you, quietly, that the dental appointment on Thursday is really a scan, and could you keep that between the two of you. You say of course. Forty minutes later, in front of six people, they ask with an edge in their voice why the pre-read went out in the wrong order. You say you will sort it. Both of those moments happened before nine, between the same two people, and nobody watching the second would have guessed the first.
That is the hedgehog dilemma, at the speed most of you work. I wrote about it in 2025 as a leader's problem, the way pace turns capable people a little spiky, and after years of working as the right hand to executives I have come to think the person standing closest to the leader lives inside it more than anyone. What the research adds is how the dilemma actually works: why closeness and distance pull at the same time, what a close working relationship gives and quietly takes, what to share and when it backfires, and how trust gets built when there is no time to build it.
What you will get out of this
Why the fear of getting too close is usually bigger than the prick itself, from experiments on commuters, strangers and people asked to show a weakness
What happens to distance at the top of an organisation, and why the person beside the leader ends up holding it
The study of 350 employees that found workplace friendships raise performance and drain energy at the same time, and the five tensions that come with being friends with someone you work for
What to share at work and what to keep, including the finding that the same confession can make a leader seem more authentic or undermine them, depending on who is listening
The meta-analysis of 95 studies on emotional labour, and the one kind of it that costs far less than the other
Whose information is it, anyway? How close pairs set the rules for what the other one knows
How trust forms in a hurry, what it rests on in the first weeks, and how to repair it after a sharp word
Three dials for finding the right distance with the person you support, and a check that reads where yours are set
The Distance Pages: three pages for mapping, adjusting and agreeing the distance in a close working pair
Why closeness and distance pull at the same time
The parable is Arthur Schopenhauer's, from 1851. A company of porcupines crowd together on a cold winter's day to keep warm, prick each other with their quills, and move apart; the cold drives them together again, and they go back and forth until they find a moderate distance at which they can bear each other. Schopenhauer said that the distance people settle on is what we call politeness and good manners. Freud borrowed the image seventy years later to describe the mixed feelings in almost every close relationship, and the name stuck.
It sticks because it describes something real. The case that belonging is a fundamental human need, made in a review of the evidence in 1995, found that people form attachments readily under most conditions, resist breaking them, and suffer measurably without them. The need for warmth is built in, and so, it seems, is the fear of the quills.
The most direct test of the dilemma is a set of six experiments published in 2007, whose authors called it the porcupine problem. When people were made to feel excluded, did they withdraw to protect themselves, or reach out? They reached out. Excluded people showed more interest in making new friends, more desire to work with others, warmer first impressions of new people and more generosity towards new partners. But there were limits, and they are the useful part. People did not try to reconnect with whoever had excluded them. They did not reach towards people they did not expect to meet face to face. And people with a strong fear of being judged negatively did not reach out at all.
That last group matters at work. A review of close relationships describes a risk regulation system in which confidence in the other person's regard is what lets people risk depending on them; low confidence makes people protect themselves first and connect second. Research on rejection sensitivity found that people who anxiously expect rejection readily perceive it in other people's ambiguous behaviour, and a study applying attachment theory at work found that the anxious and avoidant ends of attachment each predicted how people regulated their emotions, whether they meant to leave, and how much they helped beyond their role. A curt two-word email means very different things depending on which of those people opens it.
Now the finding that I think should change how you read your own hesitation. In experiments on trains and buses, commuters asked to talk to a stranger had a more positive journey, and no less productive a one, than commuters asked to keep to themselves. Other commuters, asked to predict, expected the opposite. People underestimated how interested the stranger would be. A 2022 series of experiments found the same miscalibration for deeper conversations: people expected them to be more awkward than they were, felt more connected afterwards than they had predicted, and systematically underestimated how much the other person would care about what they shared. And across seven studies of what researchers named the beautiful mess effect, people judged their own moments of showing vulnerability, asking for help or admitting a mistake far more harshly than they judged the same moments in someone else.
The quills are real, and they are usually shorter than they look from the inside.
One more thing before we get to the office. A study of 8,943 people in 42 countries asked how close they liked a stranger, an acquaintance and someone intimate to stand, and found that the answers varied with age, gender and even the local climate. There is no correct distance, which is the whole point of the parable. The distance has to be found, by two particular people, and then found again when things change.
Try this. Next time you hold back from saying something warmer or more personal to the person you support, write down what you expect to happen if you say it. Then say it, lightly, and write down what did happen. After three or four of these, you will have your own evidence on how far your forecast and your reality are apart, and most people find the gap runs in their favour.
The research behind this section
Baumeister and Leary (1995) reviewed the evidence for a need to form and maintain strong, stable interpersonal relationships and found that people form social attachments readily under most conditions, resist the dissolution of existing bonds, and suffer ill effects on health, adjustment and well-being when attachments are lacking. Maner and colleagues (2007) ran six experiments on the porcupine problem and found that the threat of social exclusion led people to express more interest in making new friends, more desire to work with others, more positive impressions of novel social targets and greater rewards for new interaction partners, but not towards the perpetrators of exclusion or towards partners with no anticipated face-to-face interaction, and that people high in fear of negative evaluation did not respond in an affiliative way. Murray, Holmes and Collins (2006) proposed a risk regulation model in which confidence in a partner's positive regard allows people to risk seeking dependence and connectedness. Downey and Feldman (1996) found that people who anxiously expect rejection readily perceive intentional rejection in the ambiguous behaviour of others. Richards and Schat (2011) found in two studies that attachment anxiety and avoidance predicted emotion regulation behaviours, turnover intentions and supervisor-rated organisational citizenship behaviour, after controlling for other individual differences and organisational commitment. Harms (2011) reviewed adult attachment styles in the workplace. Epley and Schroeder (2014) found that commuters instructed to connect with a stranger reported a more positive and no less productive experience than those who remained disconnected, while separate participants predicted the opposite, partly because they underestimated others' interest in connecting. Kardas, Kumar and Epley (2022) found across a series of experiments that conversations between strangers felt less awkward and created more connectedness than participants expected, that people especially overestimated the awkwardness of deep conversations, and that they expected others to care less about their disclosures than others did. Bruk, Scholl and Bless (2018) found across seven studies that people evaluated their own shows of vulnerability, such as asking for help or revealing imperfections, more negatively than the same situations in others. Sorokowska and colleagues (2017) measured preferred social, personal and intimate distances in 8,943 participants from 42 countries and found that age, gender and regional temperature influenced interpersonal space preferences. Hall (1963) set out a system for the notation of proxemic behaviour, the study of how people use the space between them.
Distance at the top, and the person who stands next to it
If the dilemma were only personal, everybody would sort it out in their own way. It is not only personal, because organisations build distance into the job of leading them.
Researchers have been mapping this for more than thirty years. A 1993 review of distance in organisations treated it as something with several separate parts, and a 2002 review of leader distance pulled it into three that are easy to recognise from any office: how physically far the leader is from the people they lead, how socially distant they feel in status and rank, and how often they actually interact. A senior executive tends to be high on all three, and every promotion tends to push them further out.
Power itself seems to do some of the pushing. The social distance theory of power, set out in 2013, proposes that because the less powerful person depends on the more powerful one more than the other way round, the more powerful person feels more distant, and that this distance changes how they think: more abstractly, with less attention to other people's inner states. Separately, a pair of experiments found that simply putting people in a work frame of mind made some of them less attentive to the emotional tone in someone's voice and less likely to mirror another person's small movements. The office, in other words, can switch down the part of us that notices feelings, before the pace has done anything.
Here is the complication, and it is worth holding onto. Distance is not all loss. A study of 268 people in 50 teams found that when senior leaders were socially distant, their followers were less likely to copy their behaviour, which is a real cost; the same distance strengthened the link between the leader's inspiring leadership and the team's positive emotional climate and collective confidence. An earlier exploratory study found that people valued different things in close and distant leaders. A leader seen from a distance can hold a vision in a way that a leader seen every day at the photocopier sometimes cannot.
Which brings us to you. The executive assistant and the Chief of Staff are almost unique in organisational life: structurally close to someone the organisation has made structurally distant. A survey of 1,011 secretaries in the UK found that the role had moved on, with better qualifications, better pay and a minority now doing complex managerial work, while still being treated in many places as a support role. That is a precise description of the hedgehog's position. You are close enough to see the scan, and still expected to take the edge in the voice in front of six people.
The four protective spines I see most often in this work fit the research well: hyper-independence ("I will handle it"), ruthless prioritisation, being unflappable as a mantra, and laser focus on the agenda. The research above says they are rarely character flaws. They are what distance and pace do to capable people, on both sides of the desk.
Try this. Draw two quick lines on a piece of paper: how close your executive is to the rest of the organisation, and how close you are to your executive. If the first line is long and the second is short, you are the bridge, and some of the distance the organisation has built is being carried by you. Name one piece of it you could hand back: a message they could deliver themselves, or a room they could walk into more often.
The research behind this section
Napier and Ferris (1993) reviewed distance in organisations as a multidimensional construct. Antonakis and Atwater (2002) reviewed leader distance and proposed a theory built on physical distance, perceived social distance and perceived frequency of interaction between leaders and followers. Shamir (1995) set out theoretical notes and an exploratory study on social distance and charisma, finding that close and distant leaders were described in different terms. Popper (2013) examined leaders perceived as distant and close and the implications for psychological theory on leadership. Cole, Bruch and Shamir (2009) studied 268 individuals in 50 leader-follower groups and found that high social distance reduced or neutralised the association between senior managers' transformational leadership and followers' emulation of leader behaviour, while enhancing its association with perceptions of a positive emotional climate and collective efficacy. Magee and Smith (2013) proposed that asymmetric dependence produces asymmetric social distance, with high-power individuals feeling more distant than low-power individuals and engaging in more abstract mental representation. Sanchez-Burks (2002) found in two experiments that a work context reduced attention to emotional tone in spoken words and reduced unconscious mimicry among some participants, consistent with a relational ideology that treats relational concerns as out of place at work. Truss and colleagues (2013) surveyed 1,011 secretaries in the UK and found them better qualified and generally well paid, with a minority undertaking complex managerial tasks, while most still performed traditional tasks and career prospects remained poor.
What a friendship at work gives, and what it costs
Most of the advice on this subject is written as though closeness at work were simply good, and a good deal of the evidence agrees. A 2026 set of three studies with 1,305 employees found workplace friendships associated with better support, mood, health and performance. A meta-analysis of 161 samples and nearly 78,000 employees found that support from colleagues went with clearer roles, better attitudes to work, less withdrawal and better performance, and that antagonism from colleagues went the other way. A study of what positive relationships actually do for people found they served six functions, and that each one fed something different: help with the task was most linked to job satisfaction, friendship to positive emotions at work, personal growth to satisfaction with life, and the chance to give to others to finding the work meaningful.
That last one is worth pausing on for anyone in a support role. Giving is one of the things the relationship gives you, as well as a cost you pay for it.
So far, so warm. Here are the quills.
The best single study is a pair published in 2016, with 168 employees of an insurance company and 182 restaurant and retail staff. People with more friendships woven into their working relationships, which researchers call multiplex ties, were rated better performers by their supervisors. The friendships helped through trust and through positive mood. They also cost, through emotional exhaustion and through what the researchers called maintenance difficulty, the effort of keeping a friendship and a working relationship going in the same pair of people. Both paths ran at once. A mixed blessing, as the title put it.
A 2023 study of 429 employees found that friendship at work could act as a stressor because of its conflicting dual roles, leading to negative emotions and withdrawal, and that this was worse for people who react strongly to emotional events and in jobs where people depend heavily on each other to get the work done. Which is a fair description of the working pair at the top of most organisations.
Communication researchers have been mapping these tensions since 1992, when a survey of 162 adults with a close friend at work identified five of them: autonomy against connection, equality against inequality, impartiality against favouritism, judgement against acceptance, and openness against closedness. Tension was higher in more formal organisations and lower when the friendship was closer. And friends of unequal status tended to manage the tension by choosing one role over the other depending on the situation, rather than blending them. If you have ever gone from "how was your weekend" to "the board needs this by four" in the space of a sentence, you have done exactly that.
When these friendships go wrong, they tend to go wrong for recognisable reasons. Accounts of workplace friendships that deteriorated pointed to five causes: personality, distracting life events, conflicting expectations, promotion and betrayal. People mostly pulled back indirectly, by avoiding non-work topics and socialising less, and the consequences included stress, difficulty doing the job, and people leaving. A study of 269 people who described a work spouse, meanwhile, suggests the closest version of this relationship is common enough to have acquired its own name and its own unwritten rules.
There is a real disagreement in this literature and you deserve both sides of it. One camp, represented by the 2026 studies, argues that friendships are a basic part of human life that work cannot be separated from, and that the evidence for their benefits is strong. The other, set out most fully in a 2018 theory paper on the dark sides of workplace friendship, argues that friendship and work run on different rules, one on care and the other on exchange, and that holding both creates tensions the research has underplayed. My reading is that the two camps are describing the same relationships from different ends. A close working friendship is usually worth having, and it is worth having with its costs in view rather than out of sight.
A friendship at work is two relationships sharing a desk, and each of them occasionally needs the whole desk.
Try this. Look at the five tensions and pick the one that is loudest in your closest working relationship. For most assistants it is equality against inequality or openness against closedness. Then decide, in advance, which way you will lean when it comes up, so that you are not deciding it on the spot with a deadline in the room.
The research behind this section
Mirando and colleagues (2026) found across three studies with a combined 1,305 employed people that workplace friendships were positively associated with supportive, affective, health and performance outcomes. Chiaburu and Harrison (2008) meta-analysed 161 independent samples and 77,954 employees and found coworker support and antagonism linked to role perceptions, work attitudes, withdrawal and effectiveness, with results holding when controlling for leader influences. Colbert, Bono and Purvanova (2016) developed the Relationship Functions Inventory and found that task assistance was most strongly associated with job satisfaction, giving to others with meaningful work, friendship with positive emotions at work and personal growth with life satisfaction. Methot and colleagues (2016) found in 168 insurance employees that the number of multiplex workplace friendships was positively associated with supervisor-rated job performance, with a negative indirect effect through emotional exhaustion partly offset by positive affect, and in 182 restaurant and retail employees that these relationships enhanced performance through trust and detracted from it through maintenance difficulty. Wang and colleagues (2023) found in 429 respondents that workplace friendship could act as a stressor through its conflicting dual roles, triggering negative emotions and withdrawal, with emotional reactivity and task interdependence strengthening the effect. Bridge and Baxter (1992) surveyed 162 adults with a close friend as a work associate and identified five dialectical tensions, found that dual-role tension rose with organisational formalisation and fell with relationship closeness, and found that status-unequal friends were more likely to manage tension through selection between roles. Sias and Cahill (1998) interviewed 38 people in 19 peer friendships and traced the transitions from coworker to friend to close friend. Sias and colleagues (2004) found five causes of workplace friendship deterioration, namely personality, distracting life events, conflicting expectations, promotion and betrayal, mostly indirect disengagement tactics, and consequences including emotional stress, reduced ability to perform tasks and turnover. McBride and Bergen (2015) analysed 269 open-ended accounts to define the work-spouse relationship. Pillemer and Rothbard (2018) set out the dark sides of workplace friendship. Ingram and Zou (2008) reviewed business friendships. Methot, Melwani and Rothman (2017) reviewed ambivalent and indifferent workplace relationships. Uchino and colleagues (2001) found in 133 young and older adults that the number of ambivalent ties in a person's network predicted age-related differences in depression and in heart-rate reactivity during stress.
What to share, and when it backfires
At the heart of every close relationship is a series of small decisions about what to tell the other person, and the research on them is some of the oldest in social psychology.
The core finding is reassuringly simple. A 1994 meta-analysis found three separate links between sharing and liking: people who share more personal things tend to be liked more, people share more with those they already like, and people come to like the person they have shared with. The effects feed each other. A famous procedure published in 1997 built closeness between strangers in forty-five minutes, using sets of questions that escalated gradually from light to personal, and it produced more closeness than the same time spent on small talk.
Three details make this usable at work. First, reciprocity. In a 2013 experiment, the person who listened while the other shared came away liking their partner more, and the difference disappeared once they swapped roles; going first creates a small debt that the other person's turn repays. Second, what kind of thing is shared. A diary study of real conversations found that sharing feelings predicted a sense of intimacy more than sharing facts, and that the other person's responsiveness carried much of the effect. Third, the listener. Feeling understood turns out to be only modestly related to actually being understood, and it matters a great deal to both the relationship and the person.
Then the workplace adds the one variable the laboratory mostly leaves out, which is status, and the evidence splits.
On one side, a 2022 series of studies found that leaders who voluntarily revealed a weakness came across as more authentic, because observers concluded they were not managing their image, and that employees were then keener to work with them. The effect was stronger for higher-status people. A field experiment with 304 people found that a leader's transparency during a downsizing raised followers' trust. And a study of new starters in a large outsourcing company found that inviting them to bring their authentic best selves into their first days led to better customer satisfaction and retention six months later than the usual approach.
On the other side, a 2018 set of studies titled "When sharing hurts" found that self-disclosing a weakness undermined the task-oriented working relationships of higher-status disclosers. A theory paper on status distance argued that personal disclosure between people of different status can widen the gap it was meant to close. A study of company parties and other chances to mix found that these brought colleagues closer when they were demographically similar and did not do so across racial difference, with comfort during the mixing explaining much of the gap. And a study of more than 1,200 employees and survey participants found that owning up to an error lowered a leader's view of the employee's ability and raised their view of the employee's integrity, in the same moment.
My reading is that these findings are less contradictory than they look. The authenticity studies are mostly about how a disclosure looks from a distance: to followers, observers and new starters. The costly findings are mostly about what a disclosure does inside a working relationship that runs on the task, where the listener has to rely on the discloser tomorrow. Revealing a weakness that has nothing to do with what the other person needs from you tends to read as openness. Revealing one that sits right on top of it can read as a warning.
Share the weather, not the structural survey, until you know the other person can carry both.
This applies in both directions across the desk. When your executive tells you about the scan, they are disclosing up to the one person whose job it is to keep the day running around it, which is why it lands as trust. When you share something about yourself, the question to ask is whether it helps the other person understand you or asks them to manage you.
Try this. Before you share something personal with the person you support, run it through two questions. Does it touch what they rely on me for? And am I going first, or matching something they shared? If the answer is no and matching, it will almost certainly bring you closer. If it is yes and going first, share it with a plan attached.
The research behind this section
Collins and Miller (1994) meta-analysed the self-disclosure and liking literature and found significant effects for three links: people who make intimate disclosures are liked more, people disclose more to those they initially like, and people like others as a result of having disclosed to them. Aron and colleagues (1997) presented a 45-minute procedure of escalating self-disclosure and relationship-building tasks that produced greater closeness than comparable small-talk tasks. Aron, Aron and Smollan (1992) developed the Inclusion of Other in the Self Scale, a single pictorial measure of closeness. Sprecher, Treger and Wondra (2013) found in pairs of strangers that listeners reported more liking than disclosers after a first interaction, and that the difference disappeared after roles were switched. Laurenceau, Barrett and Pietromonaco (1998) found in two event-contingent diary studies that intimacy arose from self-disclosure and partner disclosure with perceived partner responsiveness as a partial mediator, and that disclosure of emotion predicted intimacy more than disclosure of facts. Reis, Lemay and Finkenauer (2017) reviewed research showing that feeling understood is only modestly related to being understood and matters for relationship and personal well-being. Jiang and colleagues (2022) found that sensitive self-disclosure of weaknesses made leaders seem more authentic because observers inferred they were not engaging in strategic self-presentation, that this increased employees' desire to work with them, that the benefits emerged when the disclosure was voluntary, and that they were more pronounced for higher-status disclosers. Gibson, Harari and Marr (2018) studied how and why self-disclosing weakness undermines the task-oriented relationships of higher-status disclosers. Phillips, Rothbard and Dumas (2009) theorised that disclosure of personal information across status differences may increase status distance rather than bring people closer. Dumas, Phillips and Rothbard (2013) found in two samples that integration behaviours such as attending company parties were associated with closer coworker relationships for demographically similar colleagues but not for racially dissimilar ones, mediated by comfort and enjoyment during integration. Zhang, Zhao and Yin (2023) studied 560 employees in 71 teams, 359 employees in 104 teams and 356 online participants, and found that error sharing lowered leader trust through evaluations of the employee's ability and raised it through evaluations of integrity. Norman, Avolio and Luthans (2010) randomly assigned 304 participants to conditions of leader positivity and transparency in a downsizing scenario and found both raised followers' trust and ratings of leader effectiveness. Cable, Gino and Staats (2013) found in a field experiment in an outsourcing company that initial socialisation focused on newcomers' authentic best selves led to greater customer satisfaction and retention after six months, and replicated the effect in a laboratory experiment.
The feelings you carry for two
Every close working relationship involves a certain amount of managing feelings, and the person in the support role usually does more than their share: absorbing the edge in the voice, smoothing the room afterwards, looking calm on the executive's behalf. The research name is emotional labour, and since a sociologist coined it in 1983 the psychology of it has been mapped in detail.
The most important distinction is between two ways of doing it. Surface acting is showing a feeling you do not have, the professional smile over a clenched jaw. Deep acting is changing what you feel, for example by trying to see the situation from the other person's side, so that what you show is close to what you feel. A meta-analysis of 95 studies and nearly 500 correlations put numbers on the difference. Surface acting, and the gap between felt and required emotion, were substantially linked to poorer well-being, with correlations between .39 and .48, and to worse attitudes to work. Deep acting was only weakly linked to poorer well-being and was positively linked to how well the emotional side of the job was done and to the satisfaction of the people being served.
A week-long study of 848 work conversations adds something important for the office. Faking a feeling was more common with customers, while the reverse, letting a real feeling show when the rules said it should not, was more common with colleagues. Both lowered well-being in the moment. The close colleague relationship is where people most often let their feelings out, and pay for it.
A 2022 study tracked the emotional journeys people made through the working day. Even when people ended up in a better mood, the steeper the climb to get there, the more it depleted them, and the depletion showed up later as sharper behaviour towards other people. Getting yourself from "I am overwhelmed" to "I am fine" has a cost, even when you arrive. A study that measured feelings every 200 milliseconds during a difficult phone call found that people surface acted and deep acted at the same time, switching between them moment by moment.
Much of what assistants do is managing someone else's feelings, and there is research on that too. A classification of the deliberate ways people try to change how others feel sorted 378 strategies into those that engage the other person with the situation or their feelings and those that work through the relationship itself. It matters because leaders' moods spread: in one study, groups whose leader was in a good mood caught the mood, coordinated better and needed less effort, and in 156 teams across two studies, leaders who tended to leave people feeling positive got more information shared in their teams, which in turn predicted innovation. When you calm your executive before the board meeting, you are managing the mood of the board meeting.
The helping research offers the most practical finding in this section. A daily study over three working weeks found that responding to requests for help drained people's self-control at an increasing rate through the day, and that feeling the help had made a difference to someone replenished it. The people most motivated to help were drained more and replenished less. And a ten-day study found that going beyond the job lifted people's mood and got in the way of their own goals, both on the same days.
There is one uncomfortable counterweight, and it concerns naming the cost. In three experiments, people who heard about the emotional toll of service work were less willing to reward lower-level workers, not more; reducing the sense of power distance between them reversed some of that. Telling a powerful person how hard you are working at their feelings does not automatically earn you credit. The route to recognition seems to run through the relationship first.
Try this. For one week, keep a tally at the end of each day of the moments you showed a feeling you did not have. Then pick the most frequent situation and try deep acting it next time: before you walk in, ask yourself what the other person is most worried about. And ask your executive, once, which bit of your support made the biggest difference that week. The answer is a top-up, and the research says you need it.
The research behind this section
Grandey (2000) set out emotional labour as emotion regulation, integrating earlier perspectives and proposing a model including individual and organisational factors; Grandey and Melloy (2017) reviewed and revised that model. Ashforth and Humphrey (1993) examined emotional labour in service roles and the influence of identity. Brotheridge and Grandey (2002) compared two perspectives on emotional labour and burnout in people work. Hülsheger and Schewe (2011) meta-analysed 494 correlations from 95 independent studies and found substantial relationships of emotion-rule dissonance and surface acting with impaired well-being (ρ between .39 and .48) and job attitudes (ρ between −.24 and −.40), a small negative relationship with performance (ρ between −.20 and −.05), and for deep acting weak relationships with impaired well-being but positive relationships with emotional performance (ρ = .18) and customer satisfaction (ρ = .37). Tschan, Rochat and Zapf (2005) studied 78 employees reporting 848 interactions over a week and found emotional dissonance more likely with customers and emotional deviance more likely with coworkers, with well-being lower in both. Frank and colleagues (2022) used experience sampling and found that steeper daily emotional trajectories towards positive affect led to psychological depletion and interpersonal counterproductive behaviour. Gabriel and Diefendorff (2015) captured emotions, surface acting, deep acting and vocal tone every 200 milliseconds in a call-centre simulation and found customer incivility causally shifted all of them, with surface and deep acting used simultaneously. Kim, Bhave and Glomb (2013) found in 2,072 employees in 274 workgroups that age diversity was positively related to emotion regulation, mainly among younger employees working with older coworkers. Niven, Totterdell and Holman (2009) classified 378 controlled interpersonal affect regulation strategies. Zaki and Williams (2013) proposed a framework for interpersonal emotion regulation. Sy, Côté and Saavedra (2005) found that leaders' positive mood spread to group members and improved coordination while reducing effort expended. Madrid and colleagues (2016) found in 87 teams and 69 teams that leaders' positive affective presence was related to team information sharing, which predicted team innovation. Frost (2004) wrote on handling toxic emotions as a challenge for leaders and organisations. Lanaj, Johnson and Wang (2016) found in a three-week daily diary study that responding to help requests depleted regulatory resources at an increasing rate, that perceived prosocial impact replenished them, and that prosocially motivated employees were depleted more and replenished less. Koopman, Lanaj and Scott (2016) found across ten workdays with 82 employees that daily citizenship behaviour was associated with positive affect and interfered with work goal progress. Moran and Ronen (2021) found in three experiments that hearing about the emotional costs of service labour reduced willingness to tip lower-level workers, and that reducing power distance between customers and workers increased it. Halbesleben (2006) meta-analysed social support and burnout and found work sources of support more closely associated with emotional exhaustion than non-work sources.
Whose information is it? Boundaries in a close pair
The scan on Thursday belongs to your executive, who has given you a share of it. That idea, that private information has owners and that telling someone makes them a co-owner, is the centre of a communication theory developed over thirty-five years, known as communication privacy management. The theory's useful move is to say that once information is shared, both owners need rules about who else may know it, and that most of the trouble in close relationships comes when those rules were never agreed, or were agreed differently in two heads. The theory calls that turbulence, which is a good word for the afternoon when your executive discovers that the finance director knew about the scan too.
Assistants spend their working lives as co-owners of other people's information, and very few have ever had the conversation about the rules. It is worth having, because the research on boundaries between work and home shows how much depends on the fit between what people want and what they get.
A study of 460 employees found that people who preferred to keep work and home separate were less satisfied and less committed when their employer offered policies that blurred the two, and more committed when it offered ones that helped them keep the line. A study of 325 people in a wide range of jobs found that the fit between a person's preference and what their workplace allowed affected conflict between work and home, stress and job satisfaction, and challenged the common assumption that blending the two is always better. A review in 2015 offered a reason both camps keep finding support: keeping the personal and professional apart helps people manage their responsibilities, while letting them overlap helps people build relationships and a sense of who they are at work. You need both, for different purposes, which is why the answer keeps changing.
The finding closest to the assistant's desk comes from an in-depth study of 70 lawyers in a large firm, published in 2013. It found that boundary work, the negotiating and adjusting of the line between work and the rest of life, could not be understood apart from the relationship it happened in. People set boundaries differently with different colleagues, depending on whether the relationship made them want to move towards the other person or protect themselves from them, and the success of the boundary and its effect on the relationship depended on both. The line you draw with your executive is a relationship act, and it lands like one.
Try this. Pick one category of private information you already co-own with the person you support: health, family, a sensitive negotiation, a board matter. Ask them, in one sentence, who else may know it and what you should say if asked. "If anyone asks why Thursday is blocked, I will say it is a personal appointment and leave it there. Is that right?" The rule takes ten seconds to agree and can save a very long afternoon.
The research behind this section
Petronio (2013) summarised the development and testing of communication privacy management theory, which treats people as owners of their private information, makes recipients co-owners when information is shared, and describes the boundary rules and turbulence that follow. Rothbard, Phillips and Dumas (2005) surveyed 460 employees and found that desire for segmentation moderated the relationship between access to organisational work-family policies and satisfaction and commitment, with people wanting more segmentation less satisfied and committed when they had greater access to integrating policies. Kreiner (2006) found in 325 employees that the fit between work-home segmentation preference and the segmentation provided by the workplace affected work-home conflict, stress and job satisfaction, with asymmetric effects that challenge the assumption that integration reduces conflict. Dumas and Sanchez-Burks (2015) reviewed personal and professional boundary management and posited that segmentation facilitates the management of role responsibilities while integration is more beneficial for managing workplace identity and relationships. Trefalt (2013) conducted an inductive qualitative study of 70 attorneys in a large US law firm and found that boundary work could not be fully understood without the relationships within which it happened, with different types of boundary work in different relationships activating approach or avoidance motivation and producing predictable outcomes for boundary success and for the relationship.
Trust at ninety miles an hour
Most of the classic trust research assumes time: two people watch each other, the evidence accumulates, and trust grows. At the top of an organisation there is rarely that kind of time. A new executive arrives, a Chief of Staff is appointed in the middle of a restructure, an assistant is covering for a colleague on leave, and trust is needed on day one.
Start with what trust is made of. The standard model, from 1995, says people judge three things about the person they are trusting: whether they are able, whether they mean well towards them, and whether they have integrity. A meta-analysis of 132 samples confirmed that each of those judgements predicted how people behaved at work, over and above trust itself. A 1995 study of 194 managers and professionals separated two kinds of trust, one resting on the head, from evidence of reliability and competence, and one resting on the heart, from care and emotional investment. And a study of managers' professional networks found that the head kind went with ties for task advice and resources, while the heart kind went with friendship, and was actually lower in relationships that were mainly about resources.
So what forms first, when there is no time? The research on teams thrown together at speed calls it swift trust. In a study of 75 teams working together across countries for eight weeks, early trust was predicted most strongly by how people judged each other's integrity and least by how they judged each other's goodwill, and the weight on ability faded as time went on. A follow-up study of teams that never met described swift trust as real but fragile and temporary. And a meta-analysis of 53 studies, with nearly 10,000 people in teams that had to start fast, found that trust fed into better team processes and performance, and that good processes and performance then fed trust.
This is where the researchers disagree, and it is a disagreement worth knowing about. A study of 68 temporary teams found that swift trust was sustained by what people did: setting and checking norms for how the work would be done, early. A survey of 172 project managers found the opposite emphasis: the strongest bases of swift trust were about who was on the team, reputation mattered less, and early actions mattered least of all. I think both are right about different moments. Who you are and what role you hold buys you the first week. What you do in that week decides whether you get the second.
There is also a warning, and a counterweight. A study of 285 project teams found that time pressure made knowledge transfer less effective and made trust harder to build, at the very moment trust was most needed. And an experiment with 40 teams found that teams given reason for low initial trust processed information more carefully, and did better on unusual problems as a result. A little healthy suspicion early on is not a failure of the relationship. It may be part of how a new pair gets things right.
At speed, warmth follows reliability; it rarely arrives first.
Then the sharp word in front of six people. The research on repairing trust is specific. In two studies, apologising repaired trust best after a failure of competence, while for a question of integrity, denying responsibility worked better when the person was later shown to be innocent, and apologising worked better when they were later shown to be at fault. A later study found that owning a competence failure as your own did more for repair than blaming circumstances. For the working pair, this means that when either of you gets it wrong on the task, the fastest route back is to say so plainly and fix it. A lengthy explanation of why it was not really your fault is the slower road.
Try this. In the first fortnight with a new executive, or after a rough patch with a familiar one, choose reliability over warmth on purpose. Make three small promises a day and keep every one of them, visibly. Save the personal conversation for when the head kind of trust is in place. It will come sooner than you expect, and it will have somewhere to land.
The research behind this section
Mayer, Davis and Schoorman (1995) set out an integrative model of organisational trust built on perceived ability, benevolence and integrity. Colquitt, Scott and LePine (2007) meta-analysed 132 independent samples and found that trustworthiness and trust propensity explained incremental variance in risk taking and job performance when trust was controlled. McAllister (1995) studied 194 managers and professionals and distinguished affect-based from cognition-based trust as foundations for interpersonal cooperation. Chua, Ingram and Morris (2008) found that cognition-based trust was associated with economic resource, task advice and career guidance ties, and affect-based trust with friendship and career guidance ties but negatively with economic resource ties. McKnight, Cummings and Chervany (1998) theorised initial trust formation in new organisational relationships. Lewicki, Tomlinson and Gillespie (2006) reviewed models of interpersonal trust development over time. Jarvenpaa, Knoll and Leidner (1998) studied 75 global virtual teams over eight weeks and found early team trust predicted most strongly by perceptions of integrity and least by perceptions of benevolence, with the effect of perceived ability decreasing over time. Jarvenpaa and Leidner (1999) found that global virtual teams may experience swift trust that is fragile and temporal. Crisp and Jarvenpaa (2013) found in a longitudinal quasi-experimental study of 68 temporary virtual teams that normative actions of setting and monitoring performance norms, supported by early trusting beliefs, increased later trusting beliefs and team performance. Kroeger, Racko and Burchell (2021) surveyed 172 project managers and found that bases related to team composition had the strongest effects on swift trust, reputational and institutional information much more limited effects, and action-related bases were largely irrelevant. Lines and colleagues (2022) meta-analysed 53 studies of swift-starting action teams (199 effects, 2,380 teams, 9,975 individuals) and found trust positively related to emotional reactions, team processes and performance, with processes and performance in turn related to later trust. Barrett (2026) reviewed swift trust in temporary systems. Bjorvatn and Wald (2020) studied 285 project teams and found time pressure negatively associated with knowledge transfer effectiveness, with trust a complementary mediator that time pressure also reduced. Thommes and Uitdewilligen (2019) found in 40 teams that a low swift trust manipulation led to more elaborate information processing and better performance on nonroutine tasks. Kim and colleagues (2004) found in two studies that trust was repaired more successfully when mistrusted parties apologised for competence-based violations and denied culpability for integrity-based violations, and when they apologised where there was subsequent evidence of guilt and denied where there was evidence of innocence. Kim and colleagues (2006) examined internal versus external attributions for the repair of trust after competence and integrity violations. Schweitzer, Hershey and Bradlow (2006) examined the restoration of trust after untrustworthy actions and deception. Dirks, Lewicki and Zaheer (2009) and Tomlinson and Mayer (2009) set out conceptual foundations for relationship and trust repair.
Three dials for finding the right distance
Put all of that together and the right distance stops being one thing. It is at least three, and they can be set independently of each other, which is the practical good news in this whole literature.
The first dial is what you share. It runs from strictly professional to fully personal. The research says it should move gradually and in turns: small disclosures, matched rather than leading, more about feelings than facts once the relationship can take it, and with care about anything that sits on top of what the other person relies on you for.
The second dial is what you carry. It runs from carrying none of the other person's feelings to carrying all of them. The research says the cost depends less on how much you carry than on how you carry it: deep acting costs less than surface acting, being able to see the difference you made refills the tank, and carrying without any recognition drains it. This is the dial that most often ends up in the wrong place without anyone noticing, because the person turning it up is usually the one doing the carrying.
The third dial is what you hold. It runs from knowing nothing private to knowing everything. The research says the problem is rarely how much you hold and almost always whether the rules for holding it were agreed. A lot of private information with clear rules is a sign of trust. A little with no rules is a turbulence waiting to happen.
Across all three, pace changes the order of operations. When the relationship is new or the week is frantic, the head kind of trust comes first and the warmth follows it. Call it sequencing rather than coldness.
And one thing sits outside the pair entirely. The person who holds the dials for someone else needs somewhere to set their own down. A study that followed low-income mothers in New Orleans from before Hurricane Katrina to five years after it found that those whose support networks had been scattered geographically were more likely to show post-traumatic stress, even after accounting for how much support they felt they had. A study of workplace loneliness linked it to how people performed their jobs. And the research on high-quality connections at work has found that the energy people get from the colleagues they work with most predicted how engaged they were a month later, over and above how supported they felt, and that good relationships built the psychological safety people needed to learn. Your own huddle is not a luxury; the evidence treats it as infrastructure.
Try this. Rate each of the three dials for your closest working relationship, from one to five, as it is today. Then rate where you think it should be. If any dial is two or more points out, that is the one to work on this month, and the check below will help you think through how.
The research behind this section
Owens and colleagues (2016) developed and validated a measure of relational energy across four studies and found that the relational energy employees experienced with their leaders predicted job engagement a month later, controlling for perceived social support, and job performance through engagement. Carmeli, Brueller and Dutton (2009) found in 212 employed part-time students that high-quality relationships were positively associated with psychological safety and, through it, with learning behaviours. Heaphy and Dutton (2008) reviewed how positive social interactions at work affect human physiology. Dutton, Workman and Hardin (2014) reviewed compassion at work as a process of noticing, feeling, sensemaking and acting. Kahn (1990) set out the psychological conditions under which people bring more or less of themselves to their work roles. Morris and Deterding (2016) used longitudinal mixed-methods data from low-income mothers before and five years after Hurricane Katrina and found that geographic dispersion of social networks was positively associated with the likelihood of post-traumatic stress, controlling for socio-demographic characteristics, hurricane-related trauma and perceived social support. Ozcelik and Barsade (2018) investigated the link between workplace loneliness and job performance. Jolly, Kong and Kim (2021) reviewed the social support at work literature and found its findings on moderation decidedly mixed. Rosenbruch and colleagues (2023) developed a taxonomy of leader trusting behaviours from 29 interviews, 224 qualitative accounts and a two-wave survey of 415 employees, identifying reliance, disclosure, acknowledgement and guardianship, with mostly positive but some negative follower reactions such as strain. Cropanzano and Mitchell (2005) reviewed social exchange theory, including the norms and resources of exchange relationships.
The distance check
Every idea in this post comes down to a moment you will recognise: something is about to be shared, carried or held in a close working relationship, and you are not sure whether it moves you closer or too close. So here is a tool for that moment.
Tell it who the relationship is with, how long you have worked together, what you are weighing up and how the relationship feels at the moment. It will read where your three dials are set, suggest which way to move, give you steps and what to check afterwards, flag what the research would warn you about, and write a short note you can copy.
Where to start
If you take one thing from all of this, make it the idea that the right distance is three distances. Most of the discomfort in close working pairs comes from treating closeness as a single setting, so that getting warmer in one way feels as though it must mean getting warmer in all of them. In practice you can share a great deal, carry a sensible amount and hold private information under clear rules, all at once, and the research suggests that is roughly what the healthiest pairs do.
After that, there are three moves worth making.
Agree one rule for the private information you already hold, this week. Swap one piece of surface acting for deep acting, by asking yourself what the other person is most worried about before you walk in. And when there is a sharp word, on either side of the desk, repair it quickly and plainly: name it, own your part, fix what needs fixing.
I have put the three pages I would use into a download: a map of the three dials for your closest working relationship, a planner for what to share and a one-week log of what you carry, and a one-page working agreement for the conversation about rules, pace and repair.
Schopenhauer's porcupines found their distance by trial and error, one cold night after another. You do not have that many nights, and you have something they did not, which is evidence about where the quills are and how long they really are. The distance you and the person you support settle on will be yours, and it will move as you both do. That is the relationship working, rather than the dilemma failing to resolve.
The part of this that shows up in how you are seen at work sits inside what the Connected Leader Assessment calls Influence and Stakeholder Trust: how quickly people come to rely on you, how well you hold what they tell you, and whether the relationships you build survive a hard week. It is the dimension that decides whether you are close enough to be useful and trusted enough to be told.
The assessment is free, takes about twelve minutes and gives you a read across all five dimensions. If you want to go further on this one, the Skill Sprints are short, focused programmes on a single dimension, and the Irreplaceable Audit is the place to look at which parts of your working relationship only you could hold, which is usually the part it took the right distance to build.
Stay savvy.
Meg ✌️
Free tools
The distance check, the interactive tool in this post, free
The Distance Pages, the download in this post, free
The Connected Leader Assessment, twelve minutes, free, with a read on Influence and Stakeholder Trust, the dimension this post sits in
P.S. Want to Go Deeper on the Hedgehog Dilemma?
If this resonated and you want to go further, here is the evidence the piece is built on. These are the papers that shaped my thinking on closeness and distance in a fast-moving working partnership, and how to find the right space, and they are well worth your time.
Every DOI below has been checked. Where I could not confirm a figure from the paper itself, I have left the figure out rather than guess.
📃 Maner and colleagues (2007) Six experiments on the porcupine problem. Exclusion made people reach out to new partners, though not to whoever excluded them. https://doi.org/10.1037/0022-3514.92.1.42
📃 Murray, Holmes and Collins (2006) The risk regulation model: confidence in the other person's regard is what lets people risk depending on them. https://doi.org/10.1037/0033-2909.132.5.641
📃 Baumeister and Leary (1995) The need to belong as a fundamental human motivation, tested against the evidence. https://doi.org/10.1037/0033-2909.117.3.497
📃 Aron, Aron and Smollan (1992) A single pictorial measure of closeness: two circles and how far they overlap. https://doi.org/10.1037/0022-3514.63.4.596
📃 Antonakis and Atwater (2002) Leader distance reviewed: physical distance, social distance and how often leader and follower interact. https://doi.org/10.1016/s1048-9843(02)00155-8
📃 Shamir (1995) Close and distant leaders were described as charismatic in different terms. https://doi.org/10.1016/1048-9843(95)90003-9
📃 Napier and Ferris (1993) Distance in organisations as something with several separate parts. https://doi.org/10.1016/1053-4822(93)90004-n
📃 Methot and colleagues (2016) 350 employees in two studies. Friendships at work raised performance through trust and cost it through exhaustion. https://doi.org/10.1111/peps.12109
📃 Pillemer and Rothbard (2018) The dark sides of workplace friendship, set out as theory. https://doi.org/10.5465/amr.2016.0309
📃 Ingram and Zou (2008) A review of friendships in business and what they do. https://doi.org/10.1016/j.riob.2008.04.006
📃 Colbert, Bono and Purvanova (2016) Six functions of positive work relationships, each tied to a different kind of flourishing. https://doi.org/10.5465/amj.2014.0506
📃 Collins and Miller (1994) Meta-analysis of self-disclosure and liking. Sharing and liking feed each other three ways. https://doi.org/10.1037/0033-2909.116.3.457
📃 Aron and colleagues (1997) Forty-five minutes of escalating questions built more closeness between strangers than small talk. https://doi.org/10.1177/0146167297234003
📃 Gibson, Harari and Marr (2018) When sharing hurts: disclosing weakness undermined the working relationships of higher-status people. https://doi.org/10.1016/j.obhdp.2017.09.001
📃 Phillips, Rothbard and Dumas (2009) Personal disclosure across a status gap can widen the gap it was meant to close. https://doi.org/10.5465/amr.2009.44886051
📃 Dumas, Phillips and Rothbard (2013) Company parties brought similar colleagues closer, and did not do so across racial difference. https://doi.org/10.1287/orsc.1120.0808
📃 Bruk, Scholl and Bless (2018) Seven studies. People judged their own vulnerability more harshly than the same moment in others. https://doi.org/10.1037/pspa0000120
📃 Trefalt (2013) 70 lawyers. How people set boundaries depended on the relationship they were setting them in. https://doi.org/10.5465/amj.2011.0298
📃 Sanchez-Burks (2002) A work frame of mind reduced attention to emotional tone and to mimicry in some people. https://doi.org/10.1037/0022-3514.83.4.919
📃 Jarvenpaa and Leidner (1999) Teams that never met formed a swift trust that was real, fragile and temporary. https://doi.org/10.1287/orsc.10.6.791
📃 McKnight, Cummings and Chervany (1998) How trust forms at the very start of a new working relationship. https://doi.org/10.2307/259290
📃 Kim and colleagues (2004) Apology repaired trust best after a competence failure; for integrity, it depended on later evidence. https://doi.org/10.1037/0021-9010.89.1.104
📃 Schweitzer, Hershey and Bradlow (2006) Promises, apologies and the restoration of trust after untrustworthy actions and deception. https://doi.org/10.1016/j.obhdp.2006.05.005
📃 Owens and colleagues (2016) Energy from interactions with a leader predicted engagement a month later, beyond social support. https://doi.org/10.1037/apl0000032
📃 Carmeli, Brueller and Dutton (2009) 212 employees. High-quality relationships built psychological safety, and safety built learning. https://doi.org/10.1002/sres.932
📃 Truss and colleagues (2013) 1,011 UK secretaries. Better qualified and paid, some doing managerial work, prospects still poor. https://doi.org/10.1111/j.1468-0432.2012.00587.x
📃 McBride and Bergen (2015) 269 accounts of the work spouse, and how the relationship is defined and managed. https://doi.org/10.1080/10510974.2015.1029640
📃 Grandey (2000) Emotional labour recast as emotion regulation, with a model of what shapes it. https://doi.org/10.1037/1076-8998.5.1.95
📃 Hülsheger and Schewe (2011) Meta-analysis of 95 studies. Surface acting went with poorer well-being; deep acting cost far less. https://doi.org/10.1037/a0022876
📃 Niven, Totterdell and Holman (2009) 378 ways people try to change how someone else feels, sorted into a classification. https://doi.org/10.1037/a0015962
📃 Madrid and colleagues (2016) 156 teams. Leaders who left people feeling positive got more information shared, and more innovation. https://doi.org/10.1037/apl0000078
📃 Frank and colleagues (2022) The steeper the climb to a better mood during the day, the more it depleted people. https://doi.org/10.1037/apl0000908
📃 Sorokowska and colleagues (2017) 8,943 people in 42 countries. Preferred distances varied with age, gender and climate. https://doi.org/10.1177/0022022117698039
📃 Sias and Cahill (1998) 19 peer friendships traced from coworker to friend to almost best friend. https://doi.org/10.1080/10570319809374611
📃 Crisp and Jarvenpaa (2013) 68 temporary teams. Setting and checking norms early sustained swift trust and performance. https://doi.org/10.1027/1866-5888/a000075
📃 Colquitt, Scott and LePine (2007) Meta-analysis of 132 samples. Ability, benevolence and integrity each predicted behaviour beyond trust. https://doi.org/10.1037/0021-9010.92.4.909
📃 Mayer, Davis and Schoorman (1995) The integrative model of trust: ability, benevolence and integrity. https://doi.org/10.2307/258792
📃 Methot, Melwani and Rothman (2017) A review of ambivalent and indifferent relationships at work. https://doi.org/10.1177/0149206316685853
📃 Uchino and colleagues (2001) 133 adults. Ambivalent ties predicted depression differences and heart-rate reactivity under stress. https://doi.org/10.1023/a:1010634902498
📃 Sprecher, Treger and Wondra (2013) The listener liked the sharer more, until they swapped roles. https://doi.org/10.1177/0265407512459033
📃 Kardas, Kumar and Epley (2022) Deep conversations were less awkward and more connecting than people expected. https://doi.org/10.1037/pspa0000281
📃 Laurenceau, Barrett and Pietromonaco (1998) Diary studies. Sharing feelings built intimacy more than sharing facts; responsiveness carried it. https://doi.org/10.1037/0022-3514.74.5.1238
📃 Brotheridge and Grandey (2002) Two perspectives on emotional labour and burnout in people work, compared. https://doi.org/10.1006/jvbe.2001.1815
📃 Tschan, Rochat and Zapf (2005) 848 conversations in a week. Faking feelings was commoner with customers, letting them out with colleagues. https://doi.org/10.1348/096317905x39666
📃 Kim, Bhave and Glomb (2013) 2,072 employees in 274 groups. Younger people regulated their emotions more around older colleagues. https://doi.org/10.1111/peps.12028
📃 Cable, Gino and Staats (2013) Inviting new starters to bring their best selves raised retention and customer satisfaction. https://doi.org/10.1177/0001839213477098
📃 Zaki and Williams (2013) A framework for how people regulate emotions through, and for, each other. https://doi.org/10.1037/a0033839
📃 Rothbard, Phillips and Dumas (2005) 460 employees. The fit between wanting separation and getting it shaped satisfaction and commitment. https://doi.org/10.1287/orsc.1050.0124
📃 Kreiner (2006) 325 people. Fit between preferred and provided work-home separation shaped conflict and stress. https://doi.org/10.1002/job.386
📃 Dumas and Sanchez-Burks (2015) Separate the personal and professional for responsibilities; blend them for relationships and identity. https://doi.org/10.5465/19416520.2015.1028810
📃 Petronio (2013) Communication privacy management: shared information has co-owners who need rules. https://doi.org/10.1080/15267431.2013.743426
📃 McAllister (1995) 194 managers. Trust from the head and trust from the heart as separate foundations. https://doi.org/10.2307/256727
📃 Chua, Ingram and Morris (2008) Head-based trust went with task ties; heart-based trust with friendship, and less with resource ties. https://doi.org/10.5465/amj.2008.32625956
📃 Lewicki, Tomlinson and Gillespie (2006) How trust between two people develops over time, reviewed. https://doi.org/10.1177/0149206306294405
📃 Dirks, Lewicki and Zaheer (2009) A conceptual foundation for repairing damaged relationships in and between organisations. https://doi.org/10.5465/amr.2009.35713285
📃 Popper (2013) Leaders perceived as distant and close, and what that means for leadership theory. https://doi.org/10.1016/j.leaqua.2012.06.008
📃 Cole, Bruch and Shamir (2009) 50 teams. Distance cut copying of the leader and strengthened the team's emotional climate. https://doi.org/10.1177/0018726709346377
📃 Magee and Smith (2013) Power creates felt distance, and distance changes how the powerful think. https://doi.org/10.1177/1088868312472732
📃 Ozcelik and Barsade (2018) Workplace loneliness and how it shows up in job performance. https://doi.org/10.5465/amj.2015.1066
📃 Harms (2011) A review of adult attachment styles in the workplace. https://doi.org/10.1016/j.hrmr.2010.10.006
📃 Richards and Schat (2011) Attachment anxiety and avoidance predicted emotion regulation, intention to leave and helping at work. https://doi.org/10.1037/a0020372
📃 Downey and Feldman (1996) People who anxiously expect rejection read it into ambiguous behaviour. https://doi.org/10.1037/0022-3514.70.6.1327
📃 Lanaj, Johnson and Wang (2016) Three weeks of diaries. Helping drained people; seeing the difference it made refilled them. https://doi.org/10.1037/apl0000118
📃 Koopman, Lanaj and Scott (2016) 82 employees, ten days. Going beyond the job lifted mood and slowed people's own goals. https://doi.org/10.5465/amj.2014.0262
📃 Frost (2004) Handling toxic emotions as a challenge for leaders and organisations. https://doi.org/10.1016/j.orgdyn.2004.01.001
📃 Sy, Côté and Saavedra (2005) A leader's good mood spread to the group and improved coordination. https://doi.org/10.1037/0021-9010.90.2.295
📃 Epley and Schroeder (2014) Commuters who talked to a stranger had better journeys than they, or others, predicted. https://doi.org/10.1037/a0037323
📃 Reis, Lemay and Finkenauer (2017) Feeling understood is only modestly related to being understood, and matters a great deal. https://doi.org/10.1111/spc3.12308
📃 Dutton, Workman and Hardin (2014) Compassion at work as noticing, feeling, sensemaking and acting. https://doi.org/10.1146/annurev-orgpsych-031413-091221
📃 Hall (1963) The founding system for recording how people use the space between them. https://doi.org/10.1525/aa.1963.65.5.02a00020
📃 Sias and colleagues (2004) Why workplace friendships deteriorate: five causes, including promotion and betrayal. https://doi.org/10.1177/0265407504042835
📃 Bridge and Baxter (1992) 162 adults with a close friend at work. Five tensions between friendship and the job. https://doi.org/10.1080/10570319209374414
📃 Mirando and colleagues (2026) Three studies, 1,305 employees. Friendships at work went with support, health and performance. https://doi.org/10.1177/00218863261433643
📃 Wang and colleagues (2023) 429 employees. Friendship's dual roles acted as a stressor where people depended on each other. https://doi.org/10.1002/ijop.12925
📃 Jiang and colleagues (2022) Leaders who volunteered a weakness seemed more authentic, especially higher-status ones. https://doi.org/10.1037/xap0000453
📃 Zhang, Zhao and Yin (2023) Owning up to an error lowered judged ability and raised judged integrity, at once. https://doi.org/10.1111/joms.13003
📃 Rosenbruch and colleagues (2023) Four ways followers read a leader's trust, including disclosure, and the strain it can bring. https://doi.org/10.1080/1359432x.2023.2250087
📃 Kroeger, Racko and Burchell (2021) 172 project managers. Who was on the team mattered most to swift trust; early actions least. https://doi.org/10.1093/cje/beaa041
📃 Thommes and Uitdewilligen (2019) 40 teams. Low early trust made teams think harder, and better on unusual problems. https://doi.org/10.1037/gdn0000102
📃 Bjorvatn and Wald (2020) 285 teams. Time pressure hurt knowledge transfer and made trust harder to build. https://doi.org/10.1108/jkm-05-2020-0379
📃 Lines and colleagues (2022) Meta-analysis of 53 studies of fast-starting teams. Trust fed performance, and performance fed trust. https://doi.org/10.1037/spy0000298
📃 Morris and Deterding (2016) After Hurricane Katrina, scattered support networks went with more post-traumatic stress. https://doi.org/10.1016/j.socscimed.2016.07.034
📃 Moran and Ronen (2021) Hearing about emotional costs made people less willing to reward lower-level workers. https://doi.org/10.1108/jsm-03-2020-0085
📃 Jolly, Kong and Kim (2021) An integrative review of social support at work and its mixed findings. https://doi.org/10.1002/job.2485
📃 Norman, Avolio and Luthans (2010) 304 people. Leader transparency and positivity raised trust during a downsizing. https://doi.org/10.1016/j.leaqua.2010.03.002
📃 Tomlinson and Mayer (2009) How the causes people see behind a failure shape whether trust can be repaired. https://doi.org/10.5465/amr.2009.35713291
📃 Kim and colleagues (2006) Taking internal blame for a competence failure repaired trust better than blaming circumstances. https://doi.org/10.1016/j.obhdp.2005.07.002
📃 Ashforth and Humphrey (1993) Emotional labour in service roles and the influence of identity. https://doi.org/10.2307/258824
📃 Gabriel and Diefendorff (2015) Emotions measured every 200 milliseconds. Surface and deep acting ran side by side. https://doi.org/10.5465/amj.2013.1135
📃 Grandey and Melloy (2017) Emotional labour as emotion regulation, reviewed and revised. https://doi.org/10.1037/ocp0000067
📃 Kahn (1990) The conditions under which people bring more or less of themselves to work. https://doi.org/10.2307/256287
📃 Halbesleben (2006) Meta-analysis. Support from work was most closely tied to exhaustion. https://doi.org/10.1037/0021-9010.91.5.1134
📃 Chiaburu and Harrison (2008) 161 samples, 77,954 employees. Coworker support and antagonism both shaped outcomes. https://doi.org/10.1037/0021-9010.93.5.1082
📃 Heaphy and Dutton (2008) How positive interactions at work show up in the body. https://doi.org/10.5465/amr.2008.27749365
📃 Jarvenpaa, Knoll and Leidner (1998) 75 teams, eight weeks. Early trust rested most on integrity and least on goodwill. https://doi.org/10.1080/07421222.1998.11518185
📃 Barrett (2026) A systematic review of swift trust in temporary teams. https://doi.org/10.1177/10464964251348901
📃 Cropanzano and Mitchell (2005) A review of social exchange theory and the rules of exchange relationships. https://doi.org/10.1177/0149206305279602
P.P.S. Why Do I Even Have the Nerve to Write This?
Fair question.
I am someone who has spent years working alongside remarkable executives, navigating chaos, translating vision into structure and figuring things out, sometimes beautifully and sometimes the very hard way. I have had the privilege of being the steady right hand to leaders who move fast, take risks and expect a lot. That environment has shaped me more than any classroom ever could.
I also happen to have an MBA, and I am studying psychology because people fascinate me. How we work. Why we disconnect. What actually holds us together when the pressure rises.
But really, none of that is the point.
I am here because I have lived the erosion and rebuilt from it. I have seen the cost of carrying too much and the relief that arrives when you finally stop disappearing inside your own competence.
So I share what I have learned in case it helps someone else. Take what is useful. Leave the rest. And remember this one truth that professionals like you often forget:
You are probably doing far better than you give yourself credit for.
Comments