You wrote four words because you were between meetings. "Can you resend this." No full stop, no preamble, nothing hostile intended.
The person who received it read it three times, forwarded it to a friend, and spent an hour wondering what they had done.
You cannot hear your own tone
Justin Kruger, Nicholas Epley, Jason Parker and Zhi-Wen Ng (2005) ran the study that should be printed above every inbox. People writing emails were confident their intended tone, sarcasm, seriousness, humour, would be understood. It was not, at rates barely better than chance, and the writers' confidence did not move with their accuracy.
Boaz Keysar and Anne Henly (2002) found the same pattern in speech: speakers systematically overestimate how effectively their intended meaning is being conveyed.
Your clarity is illusory. The message is clear to you because you already know what you meant.
Kristin Byron (2008) added the direction of the error. Email does not just strip emotion, it skews it negatively: neutral messages tend to be read as slightly negative, and mildly negative ones as more hostile than intended. Alan Sillars and Theodore Zorn (2021) documented the same hypernegative interpretation of workplace email directly.
So the bias is not random. It runs one way, and you have to correct for it deliberately.
The move: Write for the Reader's Default
Method at a glance
Write for the Reader's Default
- Assume the negative read, and warm the opening. The first line sets the frame everything after is read through.
- Put the tone in words. Whatever you hope comes across on its own, will not.
- Cut the ambiguity you cannot see. Do not ask if it is clear. Ask what it could be read as.
- Match the channel to the weight. Some things should not be sent as text.
One and two, in practice
This does not mean a paragraph of pleasantries. It means an early, specific signal that gives the reader something other than the default to build on.
Seven extra words. "No fire here" is move two: the reassurance you were assuming would be inferred, put on the page instead.
Three. The two gaps that cause most of it
The chapter names two specifically. The pronoun with no clear referent, the "that" or "this" that is obvious to you and ambiguous to them. And the implied deadline, the "soon" or "when you get a chance" that you privately mean as today and the reader reasonably hears as sometime.
When you reread, do not ask whether it is clear. From inside your own head the answer is always yes. Ask what someone could read this as if they do not already know what you meant.
Four. Channel
The rough test: if the message is mostly information, write it, because writing is precise and creates a record. If it is mostly relationship, disappointment, praise that matters, anything about someone's standing, move it to a channel with a voice attached. Those are what curdle when the tone is stripped out.
Susanne Braun and colleagues (2019) found that a leader's choice of channel is itself read as a signal, and factors into how the leader is evaluated. The medium is part of the message whether you intend it or not.
Where this shows up outside work
Group chats, especially with family. A one-word reply that meant "agreed" and read as "fine, whatever." The mechanism is identical and the stakes are sometimes higher.
Where this tips into something worse
Two ways.
Warming every message into padding is its own failure. If every note opens with three lines of throat-clearing, the signal stops working and you have made everything longer for nothing. The correction is specific and small, aimed at messages that carry tone.
The other is using written warmth as cover for avoiding a conversation. A carefully warmed email about a serious performance issue is still an email about a serious performance issue, and choosing the channel because it is easier for you is the move move four exists to prevent.
Two things to change today
Your one-line replies
The shortest messages carry the most tone risk, because there is nothing else for the reader to weigh. Those are the ones that need the four extra words.
Every "soon" and "when you get a chance"
Replace with a date. You mean one, and they cannot hear it.
Where to start
Take the last terse message you sent and reread it as if you disliked the sender. That reading is closer to how it landed than yours is.
Then add one sentence at the top of the next one. Not an apology. A signal.
Common questions
Why do my emails come across as rude when I do not mean them to?
Because intended tone does not transmit in writing and the error runs negative. People are confident their tone will be understood by email, and it is understood at rates close to chance (Kruger et al., 2005), while neutral messages tend to be read as slightly negative (Byron, 2008).
How do I make a short email sound less cold?
Put the tone in words rather than trusting it to come through. A few words at the top that name the register, such as “quick one, no fire here,” do more than any amount of softened phrasing, because the opening sets the frame the rest is read through.
Is it better to call instead?
For anything relationally heavy, yes. The rough test is whether the message is mostly information or mostly relationship. Information belongs in writing, which is precise and creates a record. Disappointment, correction, or anything touching someone's standing loses the tone that would have softened it.
Why can I not tell that my own message is ambiguous?
Because you already know what you meant, so rereading confirms rather than tests it. The useful question is not “is this clear” but “what could this be read as by someone who does not already know.” The two usual culprits are an unclear “this” and an implied deadline.
Sources
Kruger, J., Epley, N., Parker, J., & Ng, Z.-W. (2005). Egocentrism over e-mail: Can we communicate as well as we think? Journal of Personality and Social Psychology, 89(6). · Keysar, B., & Henly, A. S. (2002). Speakers' overestimation of their effectiveness. Psychological Science, 13(3), 207–212. · Byron, K. (2008). Carrying too heavy a load? The communication and miscommunication of emotion by email. Academy of Management Review, 33(2), 309–327. · Sillars, A., & Zorn, T. E. (2021). Hypernegative interpretation of negatively perceived email at work. Management Communication Quarterly. · Braun, S., Hernandez Bark, A., Kirchner, A., Stegmann, S., & van Dick, R. (2019). Emails from the boss: Curse or blessing? Management Communication Quarterly.