Colleague Feels Dismissed by Your Message: Text or Call?

Acknowledge the impact before choosing the channel

If a colleague says your message felt dismissive, the first decision is not whether a call is more emotionally intelligent. It is whether you can show that you heard the concern without turning the reply into a defense of your character. A short written response often gives both people room to slow down and creates a shared reference to the exact exchange. For example: “Thanks for telling me. I can see that my message came across as dismissive. Which part would you like me to address first?” That sentence acknowledges the reported impact, points toward the specific wording, and leaves room to learn what the colleague means.

Do not begin with “That wasn’t my intention,” even if it is true. Intention may matter later, especially if the colleague believes you meant to belittle or exclude them. At the start, though, an immediate explanation can sound like a request that they withdraw the concern before you understand it. Nor do you have to confess to facts you do not yet believe. “I’m sorry you feel that way” can sound distancing because it places the entire issue in the other person’s reaction. A more useful opening names what you know: they experienced the message as dismissive, and you are willing to understand what landed that way.

Research on workplace apologies, including [“When apologizing hurts: Felt transgression and restoration efforts”](https://onlinelibrary.wiley.com/doi/full/10.1002/job.2672), helps explain why an automatic apology is not always repair. In three studies, a 2023 paper compared apologies made when people felt responsible for a transgression with apologies when they did not. Its first study collected workplace recollections from 251 full-time UK employees; the second tested a workplace scenario; and the third analyzed recent workplace apologies recalled by 187 working professionals. In the scenario experiment, participants assigned to apologize without feeling they had transgressed reported lower restoration efforts than those in the felt-transgression condition. In the first study, 67.21% of people recalling an apology without felt transgression reported only negative feelings toward the recipient, compared with 33.33% of those who felt they had transgressed. This was a recollection and self-report, not an observed measure of later relationship repair. It does not mean a colleague should withhold every apology until an investigation is complete; it suggests that an apology used only to end discomfort may leave the apologizer less engaged in putting the relationship right.

You can recognize an impact without settling every disputed interpretation in the first reply. “I’m sorry that my wording shut down your point. I should have asked about the risk you raised instead of calling it a distraction” accepts a concrete communication failure if you now see it. If the facts remain unclear, acknowledge the experience and ask for the phrase or moment that mattered. The distinction is useful because repair has two parts: understanding what happened and deciding what to do next. A formulaic apology can skip both; a careful acknowledgment begins them.

Keep the first note modest. A long message listing your deadlines, prior contributions, and reasons for urgency will likely introduce issues that the colleague has not raised. It can also make the apology feel conditional: you will recognize the impact only after they accept your explanation. A brief acknowledgment creates a narrow opening. If the reply identifies a specific sentence, you can decide whether a text exchange can resolve it. If it brings up repeated incidents, unequal authority, or a larger unresolved work concern, a call may be useful, provided the other person is comfortable with that format.

The proposed sentence is a practical application, not a script validated in an experiment. The apology research concerns responsibility, emotion, and restoration intentions; it did not compare written and spoken apologies or test exact wording. Use its finding to avoid an apology that merely quiets the moment. Then move to the concrete message, the work issue it carried, and the channel that lets both colleagues discuss those things clearly.

Sources: When apologizing hurts: Felt transgression and restoration efforts

Why your intended tone is hard to recover from a message

A message can feel neutral to its sender and dismissive to its recipient because each person is reading different evidence. The sender remembers the meeting, deadline, prior agreement, and intended emphasis. The recipient sees a sequence of words, punctuation, timing, and whatever history they bring to the exchange. A line such as “We already covered this” may mean “I think we have the answer in the notes” to the writer. A colleague who raised a new concern may read it as “Your contribution is not worth considering.” Neither interpretation is automatically proved by the sentence alone.

A 2005 set of five email experiments, [“Egocentrism over e-mail: Can we communicate as well as we think?”](https://web-docs.stern.nyu.edu/pa/kruger_email_ego.pdf), found that senders overestimated their ability to convey tone. The accessible abstract describes tasks involving emotion, humor, and sarcasm, and reports that senders’ confidence partly reflected difficulty setting aside their own perspective. When people compose a line while hearing the intended humor or emphasis in their heads, they can overestimate how much of that tone the reader will recover. In one lab comparison, 154 pairs of university students tried to convey scripted emotions through email, voice-only, or face-to-face. Predicted accuracy averaged 88.8% against 70.4% actual; email senders were more overconfident, and email recipients identified the intended emotion correctly 62.8% of the time versus 73.3% for voice-only listeners. This controlled task was not a real workplace conflict. It does not show that all email is confusing, that chat fails, or that a call secures agreement.

A workplace study, [“Hypernegative Interpretation of Negatively Perceived Email at Work”](https://journals.sagepub.com/doi/10.1177/0893318920979828), gives the receiver’s side a more complicated shape. In the workplace study [“Hypernegative Interpretation of Negatively Perceived Email at Work”](https://journals.sagepub.com/doi/10.1177/0893318920979828), participants supplied emails from coworkers that they had perceived negatively and described their reactions. Receivers rated the emails more negatively than independent observers did, while receiver ratings had only weak relationships with message features such as negative wording or flaming. The study called this pattern negative intensification: in this selected set of upsetting workplace emails, recipients’ interpretations were more negative than observers’ assessments. The authors also reported stronger intensification in poor communication climates and among subordinates.

That finding is not a license to tell someone their reading is wrong. The study recruited emails already perceived negatively; it does not establish that the receiver misread the sender’s intention, that an observer is a better judge of impact, or that recipients are generally over-sensitive. A recipient may be noticing a real pattern that an observer cannot see from one message. The weak relationship between message features and receiver judgments also means a writer cannot inspect a sentence mechanically and infer exactly what it will mean to a coworker.

The sender has a narrower and more useful task: compare the actual words with the work point they meant to convey. If a message says “Stop reopening this,” but the intended request was “Please put new concerns in the risk log so we can assess them,” the gap is concrete. The sender can acknowledge that the first version shut down discussion and restate the second request. If the colleague says the problem is not one phrase but a history of being cut off, editing the message alone will not answer the concern. Ask what they want addressed rather than guessing.

An interpretation can be understandable without being the only possible one. That distinction matters when a person wants to clear up facts: they can clarify which decision was already made without arguing that the colleague had no reason to feel dismissed. Conversely, hearing that the message hurt does not require accepting an accusation of intentional disrespect before you know what happened. A useful reply separates three statements: what the text literally said, what the sender meant to do at work, and what the recipient says it communicated. Keeping them separate allows a correction without pretending the other two statements are irrelevant.

The evidence has a boundary. The five experiments concern email tone in controlled tasks, and the workplace study examined negatively perceived emails selected by participants. Neither compares a current charged message across chat, phone, and video. Their value is more modest: they show why a sender’s private sense of what they meant is insufficient evidence about what a colleague could reasonably infer. That is a reason to ask and inspect, not a reason to presume that all written conversation is defective.

When reviewing your own line, read it without supplying the context you had while typing. Does it name the task or only judge the person? Does it answer the concern or tell the colleague to stop raising it? Does “obviously,” “again,” or “as I said” add useful information, or mainly communicate impatience? These questions cannot score a message or determine how it landed. They can reveal one observable thing the sender can change before sending a clarification. If the colleague then says the remaining issue is an unresolved decision, move the discussion to that decision rather than continuing to debate the tone of a sentence.

Sources: Egocentrism over e-mail: Can we communicate as well as we think?; Hypernegative Interpretation of Negatively Perceived Email at Work

What should the first written reply accomplish?

The first reply should create a common object to discuss. Name the message or phrase accurately, state the work point you intended to communicate in one sentence, and ask one open question about the impact or next step. A reply might read: “I meant that the release date had already been agreed, but I can see that ‘we already covered this’ dismissed the risk you were raising. What part of the risk do you think still needs a decision?” The first clause clarifies the decision; the second acknowledges the phrase; the question invites the work substance back into view.

This is a process of grounding: checking that people are referring to the same point and understand the exchange in roughly the same way. The term comes from communication research, including a systematic review of communication media in negotiation. In everyday use, grounding can be as simple as asking, “Do I have your concern right?” and restating the answer before replying. It is not a demand that the colleague agree with your summary. If they correct it, you have learned something that a more polished defense would have missed.

The [systematic review of communication media in negotiation](https://link.springer.com/article/10.1007/s10726-020-09662-6) examined 98 empirical studies of communication media and negotiation. The studies varied in their medium, negotiation task, and outcomes. The review distinguishes conveying information from converging on a shared interpretation, and discusses how real-time feedback can support convergence while written or asynchronous exchange lets people consider information and respond later. Negotiation is not the same as repairing a one-to-one work message, and the review does not validate a particular number of text turns before switching. It offers a useful lens: determine whether you still need to convey a fact, or whether you need rapid feedback to find out whether you share its meaning.

A written reply is doing useful work when it makes the disputed phrase visible, allows a colleague to correct your paraphrase, or establishes a precise work fact that should not be lost. It is also useful when either person needs time to think before answering. Those are real communication tasks. A message should not be treated as merely a defective version of a conversation. It can let a person read carefully, avoid being interrupted, compose a measured answer, and refer back to a task detail later.

The first written exchange should not try to settle every question about intent, fairness, and the working relationship. That is too much to ask of a single message, especially when the original line was emotionally charged. Keep the scope visible: “I want to understand what sounded dismissive and make sure we have the risk decision right.” If the colleague wants to discuss a wider pattern, acknowledge that this is larger than the single sentence and ask whether they would prefer to talk about it now, at a set time, or in writing.

A practical test is whether the next message adds information. If the colleague names a phrase and you can clarify it without repeating yourself, the written exchange may be working. If each reply adds another accusation or defense, if you are debating what a sentence “really meant,” or if several separate decisions are entangled, another paragraph is unlikely to make the exchange clearer. That is not a word-count rule. It is an observable change in the conversation: are the two of you becoming more specific, or are you sending increasingly long versions of the same position?

Do not ask the colleague to prove an emotion. “Can you show me why that was dismissive?” can sound like a cross-examination, especially when they have already identified a phrase. Ask for the detail you need to act: “Was it the wording, the fact that I closed the discussion, or something else in the exchange?” That gives options without insisting that the impact be objectively measurable. If the answer points to a concrete behavior, respond to it. If it stays broad, ask what change would make the next work exchange feel more open or useful.

There are times to stop the text exchange even before it becomes heated. One person may need time to regulate; a decision may require careful documentation; the chat may be public or include people who should not be drawn into a personal repair. In those cases, a short holding reply can protect both the relationship and the work: “I want to answer carefully. I’ll come back to this after I review the change request.” A pause should include a realistic time or next step where possible. Silence after acknowledging a concern can feel like withdrawal, especially when the person has taken a risk by raising it.

The point of this first reply is not to make the colleague reassure you that you are a good person. It is to reopen the issue in a way that makes the work question answerable. A message can begin that process when both people are still able to read and respond with care. If the exchange produces no shared account after a small number of clear turns, switch the task rather than escalating the prose: ask whether live conversation would help you test a paraphrase and agree on a next step.

Sources: Egocentrism over e-mail: Can we communicate as well as we think?; From Letter to Twitter: A Systematic Review of Communication Media in Negotiation

Text and calls solve different parts of the problem

Text is strongest when precision, reflection, access, or a written record is central. A call is useful when the people need to ask and answer questions quickly, hear vocal emphasis, and test a summary before either side drafts another long explanation. These are task differences, not moral rankings. A calm written clarification can be more respectful than an unexpected call. A brief call can be more considerate than making someone parse five paragraphs about a disputed phrase. The medium should fit what remains unresolved.

Communication researchers use the word synchronicity for how closely a medium supports people responding at the same time. A live call is more synchronous than a text exchange that can pause for minutes or hours. The Geiger review describes both conveyance, or sharing information for later consideration, and convergence, or working toward a shared interpretation. The distinction helps with this decision. When you have a factual correction to send, asynchronous writing may be enough. When each person’s interpretation needs to be checked against the other’s in several quick turns, a call may help.

A systematic review can sound as though it should produce a simple ranking: richer media for emotional ambiguity, leaner media for routine facts. The empirical picture is more mixed. The review covers 98 studies in negotiation rather than everyday workplace repair, and the author describes competing results across different channels, tasks, and outcomes. Some theoretical approaches predict that more feedback and cues will help with ambiguous matters. Other research and the review’s synthesis show that technology, communicator preferences, the precise task, and combinations of channels can change the result. Do not turn a theory about media into the rule that a call is always best whenever feelings appear.

A 2025 theoretical review, [“Re-thinking email richness and impact in institutional communication from Media Richness Theory”](https://www.scielo.org.pe/scielo.php?lng=en&nrm=iso&pid=S2304-22652025000100187&script=sci_abstract&tlng=en), makes a related point from the written side. It argues that modern email can carry more than the low richness assigned to it in early versions of media richness theory: it can include varied content, arrive quickly, be personalized, and be archived. It also notes the risks of overload and misunderstanding. The article is a comparative theoretical analysis, not a trial showing that email resolves conflict better than a call. It supports a limited conclusion: written channels have functions that are easy to miss if one thinks of them only as the absence of voice.

Suppose the dispute concerns whether a review deadline changed. Text can preserve the date, the earlier agreement, and a link to the current plan. A call may help the colleagues discover why one remembers the conversation differently, but the date itself still belongs in a written record. Suppose instead the disagreement concerns whether “I need this fixed today” was a firm request, criticism, or an attempt to assign blame. A quick live exchange may expose those competing meanings more efficiently than successive text replies, though it cannot guarantee that the colleague will feel heard.

A call does not supply facial expression unless it is a video call, and it does not erase past experience. Voice offers timing, pauses, volume, and immediate questions. Those cues may clarify the speaker’s present tone, but they cannot prove the intent behind an earlier message or establish that its impact was harmless. In some conversations, the added immediacy helps; in others, it makes it harder for someone to think, take notes, or hold a boundary. The relevant comparison is not “text lacks cues, call has cues.” It is which medium lets these two people address this particular point with enough control and feedback.

The staged approach uses both channels where each does a distinct job. A short message acknowledges the impact and asks what needs attention. If clarification stalls, a call tests understanding and settles an immediate work question. A follow-up note records the agreed action. The sequence is a practical inference from research on conveyance, convergence, and written records. It has not been tested as a single intervention, so it should remain flexible. If the colleague prefers a written exchange or needs time to formulate responses, keep using text unless a concrete operational constraint makes that impossible.

Nor does this mean that every emotional conversation requires two channel changes. If one clear written exchange resolves the issue, finish there. If the call itself is the accessible and comfortable first choice, you can agree to one. If a matter needs an accurate record throughout, keep it in an agreed written channel and use shorter, structured messages. The method is not valuable for its own sake. It is valuable only when each change makes the next task easier: understanding impact, resolving meaning, or preserving a decision.

Sources: From Letter to Twitter: A Systematic Review of Communication Media in Negotiation; Re-thinking email richness and impact in institutional communication from Media Richness Theory

When is it time to switch to a call?

Suggest a call when the exchange has stopped producing new understanding. Observable signs include the same phrase being interpreted differently after you have named it, a work decision entangled with several relationship concerns, repeated long explanations that neither person can summarize, or questions that would be answered faster through turn-taking. The point is not that emotion has crossed some invisible threshold. It is that the current channel is no longer helping the two people do the next piece of work.

The distinction between sharing information and converging on a meaning offers a rationale for switching. Text gives people time to inspect exact words and plan a reply. But when you need to ask, “Do you mean the decision itself felt dismissive, or the way I announced it?” and hear the answer before responding, the delay between written turns can make it difficult to test understanding. A call permits quick paraphrase and correction. The negotiation studies reviewed by Geiger show why feedback and task requirements matter; they do not directly prove that calls resolve colleague conflict more often.

A useful invitation lowers the cost of saying no and states what the call is for: “Would a short call help us separate the deadline decision from how my message came across? I’d like to understand the impact and agree on what happens next. If writing works better for you, we can stay here.” This frames a call as an option, not a summons or a test of the colleague’s willingness to repair. It also tells them that the conversation has a bounded purpose, which may make a difficult exchange feel less open-ended.

Avoid asking for an immediate call simply because you are anxious to end the discomfort. If you are activated and likely to interrupt, explain, or pressure the colleague to accept your view, a call can amplify the problem. Give yourself time to prepare one question and a concise account of the work facts. You might say, “I can see that my reply was abrupt. I need to check the project note before discussing the date. Can we speak at 14:00, or would you rather continue in writing?” This recognizes impact, avoids an unverified claim, and gives both people control over timing.

A colleague may prefer text for reasons you cannot see. They may need to process language slowly, use assistive technology, work across time zones, take notes, or avoid a live conversation that feels pressured. The channel they ask for is relevant evidence about what would help them participate. Ask once, accommodate where feasible, and do not label that preference as avoidance. A person’s wish for a written record can also be reasonable when earlier decisions were contested or when the exchange concerns responsibilities with consequences.

There are conditions where a direct call may not be appropriate. If the other person has asked you not to contact them outside a work channel, follow that boundary. If there is a power imbalance, an allegation of repeated mistreatment, or concern about retaliation, an informal call may not be a fair setting. Keep the immediate work issue clear and consider an appropriate neutral process. This recommendation does not assume that every conflict is symmetrical or that two colleagues have equal freedom to speak spontaneously.

When the colleague accepts, agree on a length and a purpose. “Ten minutes to understand the wording and decide how to handle the issue tracker” is more useful than “We need to talk.” Send any essential material in advance if the call depends on a particular message or decision. This gives both people access to the same facts and reduces the chance that the first half of the conversation will be spent searching through old threads.

If they decline, do not keep trying to move them to voice. Reply with one specific question and a proposed written next step. If writing is becoming unwieldy, ask whether they would rather have a third person facilitate or schedule more time. A channel change should reduce friction, not transfer control to the person who proposed it. The conversation succeeds only if both people can take part in a way that allows them to say what happened and what needs to change.

Sources: From Letter to Twitter: A Systematic Review of Communication Media in Negotiation; What the differences in conflict between online and face-to-face work groups mean for hybrid groups: A state-of-the-art review

How to clarify without turning the call into a defense

A call is not helpful merely because it is live. Its value comes from listening, checking your understanding, and responding to what the colleague actually says. Begin by restating the concern in ordinary language: “You read ‘we already covered this’ as me closing down your concern about the release risk. Have I got that right?” Then pause. A paraphrase is a question, not a clever way to put words in someone else’s mouth. If they correct it, use the correction instead of defending the first version.

After you understand the impact, clarify the work fact without treating it as a rebuttal. “I was trying to say that the review date had not changed. I should not have used that as a reason to close the risk question.” That order keeps both issues present: the date and the way the message shut down discussion. If the deadline really is fixed, state why and what choices remain. Acknowledging poor wording does not require pretending an immovable deadline is flexible. Equally, a fixed deadline does not make dismissive wording useful.

Keep your turns short enough that the colleague can respond. Long monologues often contain several points that the other person must either accept or interrupt. Ask a question that could change your view: “What would you need me to do differently when a risk is raised after a decision?” If their answer is actionable, name the action you can take. If you disagree about an event, separate the fact you can verify from the interpretation you cannot settle immediately. A useful call may end with some difference still present but a clearer plan for work.

Research on communication channels supports the importance of feedback in a qualified sense. The negotiation review considers how people establish shared meaning and how media permit responses. A 2025 qualitative study, [“Understanding workplace incivility dynamics from the perspective of conflict mediators: a qualitative study”](https://www.frontiersin.org/journals/organizational-psychology/articles/10.3389/forgp.2025.1683369/full), describes a different setting: conflict mediators recounted difficult employee relationships and the discussions they facilitated. The study’s interviews show that people could become more open to another interpretation through facilitated perspective-taking, while not every case developed trust. Because the evidence comes from mediator accounts and not a controlled comparison of call and text, it cannot establish that a live exchange itself causes understanding. It points to what the conversation needs to contain: reciprocal perspective-taking, acknowledgment, and space for the other person to answer.

That mediator study also describes a risk in centering intent too early. In the accounts, participants often interpreted behavior through personal norms and suspected disrespect even when the initial actor had not intended harm. In one mediator account, the repair involved recognizing how direct wording might have sounded and then explaining the intended work point. This is an illustrative research account from a small qualitative dataset, not proof that intent never matters. It shows why “I didn’t mean it that way” is incomplete by itself: it tells the colleague about the sender’s intention but not whether the sender understands the effect.

If you realize that the wording was dismissive, make a specific repair. “I cut you off when you raised the risk. I’ll reopen that item and ask you to walk me through the evidence before I close it” links accountability to an observable action. If you still see the underlying decision differently, say so plainly and explain the reason. Acknowledgment does not require agreement on every conclusion. It does require that the person’s concern be allowed into the decision process if that is what the conversation is meant to resolve.

If the call gets louder, more defensive, or less specific, stop trying to force a resolution. You can say, “I want to understand this, and I’m noticing we’re repeating our positions. Let’s pause and return after we’ve both had time to review the notes.” If the colleague asks to stop, stop. If they are being interrupted, leave room. A pause is useful when it includes a clear next step and does not quietly abandon the issue. It can be another scheduled call, a written exchange, or a facilitated meeting, depending on what the unresolved concern requires.

The goal is neither a perfect emotional performance nor a rapid agreement. It is to learn enough to decide whether the message needs correction, the work decision needs reconsideration, or both. A call can support that sequence by letting each person check what they have heard while the other is present. It cannot replace evidence about the work, guarantee that the colleague feels safe, or make a power imbalance disappear. Treat it as one tool for clarifying the exchange and make sure the colleague has room to use it too.

Sources: From Letter to Twitter: A Systematic Review of Communication Media in Negotiation; Understanding workplace incivility dynamics from the perspective of conflict mediators: a qualitative study

The record can protect clarity, but it cannot settle tone

Use writing to preserve work facts and decisions, not to prove who was the more reasonable person. A written record can show which date, owner, task, or decision was agreed. It cannot establish that a colleague felt respected or that a message had only one plausible meaning. A call can help people repair a misunderstanding in real time, but it may leave a task owner or deadline fuzzy unless someone records it. Keep each channel responsible for the work it does well.

The 2025 theoretical review of institutional email names flexibility, archiving, and low cost among email’s advantages while also noting information overload and possible misunderstanding. It argues that email’s present capabilities complicate a simple classification as a low-richness medium. The article does not demonstrate that a saved thread is neutral or that a record resolves interpersonal conflict. It supports using writing for reference while recognizing that the artifact captures words, not the whole relationship or the recipient’s inner experience.

After a call, a proportionate summary might say: “Thanks for talking. I’ll reopen the release-risk item today and add your evidence before we confirm the date. We agreed that the date remains Friday unless the review identifies a blocker. Please correct me if I missed something.” This records actions and leaves space for correction. It avoids adding a judgment such as “We agreed you misunderstood my tone” or a psychological claim such as “You were upset because you dislike direct feedback.” Neither statement is a work decision, and neither follows from the call merely because it ended politely.

Keep the recap brief enough that it remains usable. Name who will do what and by when only if the people actually agreed to those details. If no deadline was settled, write that it remains open. If the colleague disputes the summary, correct it together rather than treating your version as official. The same standard applies when the original issue came from your message: do not use documentation to freeze your own interpretation into the record.

There are good reasons to stay in writing throughout. A colleague may need a transcript-like trail for memory, accessibility, language processing, or organizational accountability. The work may involve a decision that has to be auditable. Text can also give each person time to choose words more carefully. In such cases, make the exchange easier to follow: discuss one point at a time, quote the precise work item, distinguish a question from a decision, and state when you will respond. If the thread becomes tangled, propose a written structure instead of insisting that a call is the only route to clarity.

A record can also create a different kind of pressure. Copying an audience into a repair message, attaching unrelated old exchanges, or drafting a long chronology can make a colleague feel that a private concern is now a case against them. If others need to know a task decision, record that decision in the normal project channel. Keep personal repair between the relevant people unless a workplace process or safety need calls for someone else to be involved. The article by itself cannot determine which internal process applies; follow the organization’s actual responsibilities and the colleague’s stated boundaries.

When the issue concerns potential repeated harm, discrimination, retaliation, or a serious breach of policy, a private call followed by a short note may not be sufficient. A record should not be used to discourage a colleague from accessing an appropriate reporting or support process. Be clear about what you can address directly and what needs a fairer structure. The immediate conversation can still acknowledge impact and clarify the work issue, but it should not be framed as the only acceptable way for the other person to raise a concern.

The useful record is therefore intentionally narrow. It helps the two people return to the same practical agreement later. It does not determine whether the original message was kind, whether the colleague’s reaction was proportionate, or whether the relationship is repaired. Those questions depend on what each person says and does over time. If you want to know whether the repair worked, look for observable follow-through: did the issue get reconsidered where appropriate, did each person know the next step, and did later work exchanges allow concerns to be raised without being shut down?

Sources: From Letter to Twitter: A Systematic Review of Communication Media in Negotiation; Re-thinking email richness and impact in institutional communication from Media Richness Theory

The same channel can feel different across workplaces

A communication channel does not have the same meaning in every workplace. One team may routinely resolve small issues by chat and schedule calls for complex decisions. Another may use calls to make fast decisions but keep all task changes in writing. A colleague who has experienced repeated interruptions may find a surprise call less open than a written question. A person who has been left out of decisions may regard a request to “take it offline” as an attempt to remove the conversation from view. The channel’s effect depends partly on norms and history, not only on its technical features.

The 2020 review [“What the differences in conflict between online and face-to-face work groups mean for hybrid groups”](https://www.rcommunicationr.org/index.php/rcr/article/view/24) examined 68 articles on online, hybrid, and in-person work groups. The reviewed literature suggested that conflict was more common in online groups than in face-to-face groups, while strong norms and leadership behaviors that built trust and cohesion appeared to reduce conflict and its effects. The authors emphasized that evidence about hybrid groups remained incomplete. This is group-level research, not a study of two colleagues choosing text or phone after a charged message. It makes a workplace-level point: the same medium operates inside shared expectations, leadership behavior, and team history.

The workplace email study by the workplace email study adds a power concern. Its negative intensification pattern was stronger in poorer communication climates and among subordinates in the sample described by the authors. That does not mean a subordinate is mistaken when reading a manager’s email negatively. It means the writer should not assume that a call gives each person equal freedom to disagree. The person with less authority may face more consequences for challenging an interpretation in a live exchange, or may value a written record precisely because it permits them to formulate a response without immediate pressure.

The 2025 qualitative study by workplace mediators supplies concrete but bounded examples. The researchers analyzed mediator accounts of conflicts in professional relationships, including employees who described direct messages as dismissive, blaming, or belittling. The accounts showed that participants sometimes attributed harmful intentions to behavior that the mediator understood as unintended; they also described situations in which perspective-taking and acknowledgment helped reduce defensiveness. Because the source is qualitative and based on mediator accounts, it can show how conflicts unfold in those cases, not estimate how often a channel choice produces repair in the wider workforce.

The study does not support a tidy instruction to “just talk face to face.” One account described disagreement about whether a person should pick up the phone rather than rely on email, with the channel preference itself entangled in different expectations and work styles. Other accounts described efforts to understand a colleague’s perspective, including the effect of direct wording. The relevant implication is that the channel can itself become part of the dispute. Ask what would make the discussion workable instead of assuming that your preferred medium is neutral.

Consider accessibility before proposing a call. A colleague may use captions, need an interpreter, require more time to process spoken language, be working in another time zone, or need to consult a document while responding. A voice-only call may remove useful visual information or create practical barriers. If a call seems helpful, ask what format and timing would work. If text is preferred, break a difficult issue into clear questions and allow a reasonable time to answer. Do not infer unwillingness to collaborate from a format choice.

Power also affects how an invitation lands. A manager asking a direct report to call immediately has more force than a peer asking another peer. If you supervise the colleague, state that they may choose a written reply or request time, and follow through without penalizing that choice. If you are the person with less authority, you may ask to keep the work decision in writing, request a neutral facilitator, or bring a support person where the policy allows. These are not signs of poor emotional skill. They are ways to preserve a fair opportunity to speak.

Urgency can separate the work task from the repair. If a production issue needs an immediate answer, send the operational facts and ask for the decision the team needs now. The conversation about how your earlier message came across can happen later when both people can participate. Do not use a deadline to force the colleague to abandon the concern, but do not leave a time-sensitive work question unanswered while waiting for an ideal emotional conversation. Identify what must be decided now and what can be discussed at an agreed time.

These conditions do not make channel selection unknowable. They make it relational and practical. Ask the colleague which format would help them discuss the issue; state the purpose and limits of any proposed call; preserve the work record when needed; and watch whether the chosen format lets both people contribute. If the channel repeatedly prevents one person from expressing a concern, revisit it. Trust and communication norms develop through repeated choices, not one technically optimal medium.

Sources: Egocentrism over e-mail: Can we communicate as well as we think?; What the differences in conflict between online and face-to-face work groups mean for hybrid groups: A state-of-the-art review; Understanding workplace incivility dynamics from the perspective of conflict mediators: a qualitative study

What emotional skill does this exchange let you practice?

The relevant emotional skill is a sequence of observable actions: notice your own urge to rebut, pause long enough to choose a response, identify what the colleague says they experienced, and keep the work question clear. This is more useful than asking whether you are an empathetic person. A person can care about colleagues and still send a sharp message under deadline pressure. The question is what they do after they learn that the message felt dismissive.

Start with awareness. Before replying, identify what you are feeling in plain terms: embarrassment, irritation, worry about the deadline, or fear that your competence is being questioned. Naming an emotion is not a diagnosis and does not prove its cause. It can help you notice the difference between the colleague’s complaint and the response you are tempted to send. If you are writing to show that the colleague is unfair, wait. If you are writing to understand the specific point and decide the next work action, proceed.

Regulation does not mean suppressing every reaction or pretending to be calm. It means choosing when and how to express a reaction so it does not take over the task. You might draft a reply, remove the sentence that argues about the colleague’s character, and keep the line that addresses the work fact. Or you may send a holding message and return after reviewing the issue. The goal is not a polished tone at any cost; it is making a deliberate response possible while preserving the substance that needs to be discussed.

Empathy here means trying to understand the colleague’s account without claiming that you know it better than they do. Ask what felt dismissive, paraphrase the answer, and let them correct you. The 2025 mediator study describes perspective-taking as one route by which people in some cases reconsidered their assumptions about a colleague’s intent. It also reports cases where trust did not develop. Perspective-taking can make another view more intelligible; it does not guarantee agreement, forgiveness, or a repaired relationship.

Boundaries still matter. Listening does not require accepting an insult or abandoning a necessary work decision. If the colleague uses personal attacks, you can name the limit and return to the issue: “I want to discuss the message and the deadline. I can’t do that productively while we’re calling each other incompetent. Let’s pause and continue at 15:00.” If a boundary is needed around after-hours messages, state the response window. Emotional intelligence does not mean unlimited availability or taking responsibility for every feeling another person has.

The repair should include behavior after the conversation. If you agree to reopen a risk review, do it. If you agree to ask a clarifying question before closing a concern, use that question next time. If your message was accurate but abrupt, you can acknowledge the abruptness and still maintain the decision with a clear explanation. This follow-through helps distinguish a practical apology from a social gesture made only to smooth the moment. The workplace apology studies link felt responsibility with restoration efforts, though their designs do not establish which follow-through will repair a particular colleague relationship.

A single exchange cannot establish someone’s stable emotional skill. The same person may respond thoughtfully in one situation and react defensively in another, especially under different pressure, authority, or team norms. Treat the episode as information about a behavior you can examine: Did I ask what the colleague meant? Did I make the work decision understandable? Did I follow through on an agreed action? Avoid turning it into a verdict such as “I am bad at empathy” or “They are impossible to talk to.” Those labels obscure the next observable choice.

For anyone who wants a more structured reflection after the issue is settled, the Emotional Skills Profile is a private educational tool for reviewing recent behavior and scenario judgments across recognition, understanding, regulation, and responding to others. It is not a hiring score, diagnosis, fixed ability measure, or normed test, and its results do not establish how either person behaved in this exchange. Its useful role is narrower: help you select a behavior to practice and reflect on it. The article’s example offers one such practice without requiring a score to decide whether a colleague deserves to be heard.

Sources: Egocentrism over e-mail: Can we communicate as well as we think?; From Letter to Twitter: A Systematic Review of Communication Media in Negotiation; Understanding workplace incivility dynamics from the perspective of conflict mediators: a qualitative study

Send one clear invitation

The decision rule is straightforward. Acknowledge the impact and clarify in text when the colleague has raised the issue there, the work point is narrow, and each reply is making the exchange more specific. Invite a call when meaning or next steps remain tangled and immediate back-and-forth could help. Keep it optional, bounded, and accessible. After a call, record the task decision in a neutral note. This sequence respects text’s precision and a call’s feedback without assuming either one will repair the relationship by itself.

There are important exceptions. A call may be the wrong choice when the colleague requests writing, a power imbalance makes live speech risky, the decision needs an accurate record, or the issue belongs in a formal process. Text may be the wrong choice when a long thread is multiplying interpretations and the other person welcomes a live discussion. If a call is unsuitable but the written thread is failing, ask about a different format, a scheduled conversation, or a neutral facilitator. The medium should make it easier for both people to be heard and to work with the facts.

The strongest caution against the staged recommendation is that the evidence does not directly test the stages. The email experiments examine tone judgments; the workplace study, selected negatively perceived messages; the negotiation review, varied media and tasks; the group review, group conflict; and the apology studies, felt responsibility and restoration. These sources support relevant mechanisms and boundaries, not a claim that the sequence guarantees repair. The sequence is the publication’s synthesis: begin with acknowledgment, use the channel that fits the next communication task, and preserve only the agreed work actions.

If you need to send a message now, make it specific enough to begin a conversation and short enough to leave space for the colleague’s answer: “I’ve reread my message about the review. ‘We already covered this’ shut down your concern, and I’m sorry I used it that way. I meant that the date was unchanged, not that the risk was irrelevant. Would you tell me what still needs consideration? If it would be easier, I can make time for a short call; writing is fine too.” Adjust the apology to what you genuinely accept, and do not include a promise you cannot keep.

If the colleague replies with the issue in writing, stay with that issue. If they accept the call, listen first, clarify the work decision, and agree on a next step. If they do not want to continue, respect that boundary and leave a clear path for necessary work information. Repair is not measured by whether the other person forgives you on demand. It is visible in whether you understood what they raised, addressed the part you could change, and kept the work moving without repeating the same dismissal.

Sources: When apologizing hurts: Felt transgression and restoration efforts; Egocentrism over e-mail: Can we communicate as well as we think?; What the differences in conflict between online and face-to-face work groups mean for hybrid groups: A state-of-the-art review

Questions readers ask