The Translation Problem: Why Being Understood Is Harder Than You Think

At 11:42 p.m., your phone lights up.

You’re half-asleep, scrolling nothing in particular, when the message arrives from someone who matters more than you’d like to admit.

“ok”

Lowercase. No punctuation. No emoji.

You stare at it for a full five seconds, which is roughly four seconds longer than it takes to actually read.

What does “ok” mean?

Is it annoyed “ok”? Resigned “ok”? Tired “ok”? The “ok” that secretly means “I am not ok, but I don’t have the energy to elaborate”? Or just… regular ok?

Your brain begins its late-night Netflix binge of catastrophic possibilities. You replay the previous conversation, now under suspicious lighting. Was that joke actually passive-aggressive? Did they misread your earlier message as cold? You hover over the keyboard, drafting and deleting escalations disguised as clarifications.

“Did I say something wrong?”

“Are you mad?”

“Sorry if that didn’t come across right—”

You put the phone down. Pick it back up. Check if they’re still online. You reread “ok” as if more pixels will appear.

Meanwhile, across town, the sender has already brushed their teeth, turned off the light, and fallen asleep in the warm glow of believing they’ve replied promptly and politely.

They think the matter is settled.

You think the relationship is on fire.

Welcome to the translation problem of being human.

We speak as if words were wires: I plug a sentence into my mind, it carries meaning across, and plugs into yours. If something goes wrong, we assume someone must be careless, cruel, or stupid.

But what if the more honest metaphor isn’t a wire at all?

What if every conversation you have is closer to passing a handwritten note through a storm, hoping most of the ink survives the rain?

This sounds depressing, but it’s actually liberating. Because when you start to see communication as translation between private worlds rather than the transfer of a single shared world, a lot of things that used to feel personal turn out to be structural.

You’re not bad at explaining. They’re not uniquely dense. The channel itself is noisy.

And once you accept that the channel is noisy, you can get much better at designing messages that survive the weather.

Claude Shannon, the engineer who basically invented information theory in 1948, opened his landmark paper with a deceptively simple sentence: “The fundamental problem of communication is that of reproducing at one point a message selected at another point.” (en.wikipedia.org)

He wasn’t thinking about late-night texts; he was thinking about telephone lines. But the logic is eerily applicable to ordinary conversations.

In Shannon’s world, you have a source (you), a transmitter (your mouth or your thumbs), a channel (air, fiber-optic cable, Slack), a receiver (their ears or screen), and a destination (their mind). In between, there’s noise: static on the line, compression artifacts, autocorrect, your friend’s stress from a meeting you know nothing about, the fact that they’re also half-watching a show while you talk. (en.wikipedia.org)

Engineers responded to this problem with two big ideas.

One is compression: packing a message efficiently into fewer bits. A photo becomes a JPEG. A long thought becomes a tweet.

The other is redundancy: repeating the critical parts of a message in different ways so that even if some bits are lost, the core survives. This is why digital signals include error-correcting codes. It’s also why when you dictate directions, you instinctively say, “Take the second left—by the blue house—right after the gas station.” You’re saying the same thing three times, in different forms.

Every time you talk, text, email, or write a spec, you are doing some version of both.

Compressing, because you don’t have infinite time or attention.

Adding redundancy, because you hope your meaning can be reconstructed even if half your nuance dies en route.

Most of the time, you do this intuitively. When you say, “It’s fine,” but add a certain tone, a shrug, a tiny pause before the “fine,” you’re adding analog error-correcting codes. The words carry one layer of meaning; the nonverbal cues carry another.

Modern life, unfortunately, has turned a terrifying amount of our communication into pure, stripped-down compression. Text. Email. Slack. Comments on Figma files. Twitter replies.

We’ve kept the compression.

We’ve discarded much of the redundancy.

No wonder so many of us feel like we’re constantly being misread—and constantly misreading others.

Psychologists Nicholas Epley and Justin Kruger once ran an experiment that reads like a horror story for anyone whose job depends on email. Pairs of participants sent each other short messages that were either sincere or sarcastic. Half recorded the messages out loud; half sent them in email. Everyone predicted their partner would correctly detect the intended tone about 78% of the time. In reality, listeners hearing vocal tone were right roughly three-quarters of the time. Email readers? Barely better than a coin flip, correctly guessing tone only about 56% of the time. (en.wikinews.org)

Worse, both senders and receivers remained convinced they were doing great. People thought they’d correctly interpreted tone about 90% of the time.

That gap—between how clear we feel we’re being and how clear our communication actually lands—is everywhere once you start looking.

Deborah Tannen, the linguist whose work on conversational style became unexpectedly famous in the 1990s, has spent decades showing how people can leave the same conversation with entirely different ideas of what was said. She describes friends who interrupt one another as a show of enthusiasm, talking over each other like jazz musicians riffing on a theme. Put them in a room with someone whose style is to pause, reflect, and finish every sentence, and suddenly the same interruptions feel like aggression, domination, or insult. (deborahtannen.com)

Nobody changed their intentions.

They changed only the interpretation rules.

It’s as if two people are playing different games with the same deck of cards. One is quietly building houses out of them. The other is shuffling for poker.

They’re both using the cards “correctly” inside their own rules. But when the structures collapse, each blames the other’s clumsiness.

Underlying a lot of this is something cognitive scientists call the curse of knowledge.

The simplest way to experience it is with a game that started as a Stanford dissertation experiment in 1990. One person, the “tapper,” is given a list of well-known songs like “Happy Birthday” or a national anthem. Their job is to tap out the rhythm on a table. The other person, the “listener,” tries to guess the song.

Before the listener guesses, the tapper is asked: how likely is it that your partner will identify the tune?

Tappers tend to predict that listeners will get it right about half the time.

In reality, listeners guess correctly about once in forty. (en.wikipedia.org)

From the tapper’s perspective, the result is baffling. They can hear the melody. It’s obvious. How can the listener not hear “Happy Birthday” in what to them feels like nearly the entire song?

From the listener’s perspective, all they have is a series of dull, repetitive knocks.

Once you know the tune, it’s almost impossible to imagine what it’s like to hear only the taps.

That’s the curse of knowledge: once we know something, we find it hard to remember what it felt like not to know it. And because we find it hard to reconstruct the other person’s ignorance, we routinely underestimate how much we need to explain, or how ambiguous our words really are.

If you’ve ever watched a programmer attempt to explain a bug to a non-technical colleague and say, “It’s just a race condition with the API call,” then be mystified when the other person still looks confused, you’ve seen this. If you’ve ever been in the other chair, you’ve felt what it’s like to be the listener hearing only taps while someone else hears a full symphony.

The curse of knowledge doesn’t just apply to facts. It applies to feelings.

You know how ashamed, or worried, or grateful you feel inside. You hear the full orchestra of your internal monologue. You know the hours of backstory behind an “ok.”

To the other person, all that arrives is a small constellation of words and gestures. A handful of taps on a table.

And yet we’re constantly surprised when they mishear the song.

There’s another illusion that makes this worse, especially in high-stakes conversations.

Kenneth Savitsky and Thomas Gilovich studied what’s called the illusion of transparency: our tendency to overestimate how much other people can see our internal states. In a series of public-speaking experiments, they had students give short speeches while both they and observers rated how anxious the speakers appeared. Predictably, speakers believed their nervousness was glaringly obvious. Observers, meanwhile, rated them as far calmer than the speakers did themselves.

When another group of speakers was told about this illusion beforehand—essentially, “You feel more exposed than you actually are”—they not only felt better about their performance, but independent judges rated their speeches as more effective. (sciencedirect.com)

Just knowing that others can’t see everything inside you, that your internal chaos is largely invisible, made people more relaxed and therefore clearer.

The same phenomenon shows up in negotiations: people think that their preferences and private calculations are more legible to the other side than they actually are, which can lead them to overcorrect or under-communicate strategically important information. (scholars.northwestern.edu)

Put the curse of knowledge and the illusion of transparency together and you get a strange, slightly tragic creature: a person who thinks their inner world is both obvious and shared.

Of course they understood I was joking.

Of course they could tell I was uncomfortable.

Of course they knew I was being generous.

Maybe not.

From the outside, there’s just a pattern of taps, and a face that looks about as calm as anyone else’s.

Now add culture to the mix, and things become even more interesting.

The anthropologist Edward T. Hall famously described a spectrum of “high-context” and “low-context” cultures. At one end, people rely heavily on shared background, implicit understanding, and nonverbal cues; much of the meaning lives in what goes unsaid. At the other end, people expect communication to be explicit, direct, and self-contained; most of the meaning lives in the words themselves. (en.wikipedia.org)

Think of a family that’s lived together for decades, where a raised eyebrow can carry an entire paragraph. That’s a micro high-context culture. Then think of a cross-functional Slack channel with a rotating cast of people from three time zones and four companies. That’s about as low-context as it gets.

Neither style is inherently better. High-context communication can be incredibly efficient and emotionally rich among people who share a lot of background. Low-context communication can be wonderfully clear and inclusive when you’re working across differences.

The trouble starts when the context expectations don’t match.

Put a high-context speaker in a low-context environment—say, someone who grew up in a culture where reading the air is essential, now trying to give feedback in a blunt, individualistic workplace. They may hint instead of state. They may soften criticism so thoroughly that to them it feels appropriately diplomatic, but to the recipient it feels like vague niceness with no actionable information.

Reverse the pairing, and you get someone whose idea of “being honest” lands as rude or brutal, because they’ve packed all the meaning into the explicit words and neglected the atmosphere around them.

It gets even weirder online.

Hall’s original work predated emojis by decades, but his ideas resonate in the way different people use digital cues. Research on computer-mediated communication finds that some users lean heavily on punctuation, emoji, and timing to carry nuance—essentially rebuilding a high-context layer out of symbols—while others fire off stark, almost context-free text messages and expect that to suffice. (sciencedirect.com)

You’ve probably experienced a micro version of this:

One person sends: “Sounds good.”

Another sends: “Sounds good!!! 😊”

To the first person, adding anything more would be unnecessary; the words already convey agreement. To the second, the bare phrase feels chilly, even passive-aggressive, like a waiter intoning “enjoy” without smiling.

They’re not just using different punctuation. They’re playing by different context rules.

And then there are emojis themselves: our collective attempt to bolt some of the lost nonverbal redundancy back onto text.

For all their cuteness, emojis are remarkably unruly creatures.

In one study, researchers showed people various emoji and asked them to rate their meaning and emotional tone. Even when participants were looking at the same rendering of the same emoji, they disagreed about whether it was positive, neutral, or negative about a quarter of the time. When you consider that the same emoji looks different on different platforms—a grinning face on one device might look more like a grimace on another—interpretation disagreements only grew. (scholars.northwestern.edu)

Other research suggests that different generations and cultures use emoji differently. A simple “thumbs up,” which older adults often interpret as a straightforward “okay,” can be read by some younger users as sarcastic, dismissive, or a sign that the conversation is over and they are done with you. (pubmed.ncbi.nlm.nih.gov)

We invented a new layer of redundancy, then promptly made it ambiguous again.

That doesn’t mean emojis are useless. They often do soften messages, make intentions clearer, and help people convey tone they’d otherwise struggle to express in text. But they’re not a magic fix.

They’re another set of taps on the table, and we’re still arguing over what tune they represent.

If this all sounds like a mess, that’s because it is.

Yet somehow, despite all these glitches, we build marriages, companies, social movements, and software systems out of nothing but these fragile signals. That alone is a minor miracle.

The point, then, isn’t that communication is hopeless.

It’s that clarity is an achievement, not a default.

It’s what happens when you treat every conversation as a small engineering and storytelling project: How do I get this living, breathing, multi-dimensional experience inside my head compressed into a form that can survive the trip into yours?

Some people do this instinctively. Others learn it the hard way, through painful misunderstandings and hard-won repairs.

Either way, the skill comes down to acting less like someone broadcasting obvious truth and more like a translator working between two different, opaque languages.

Imagine for a moment that everyone you talk to speaks a slightly different dialect of “Human” than you do.

People from your hometown or subculture share many of your idioms, your sense of timing, your idea of what counts as polite. Over years of interaction, you’ve built up a massive, invisible phrasebook together: the private jokes, the short-hand references, the look that means “we should leave this party soon.” Conversations with them feel silky because your translation dictionaries overlap so much.

Now picture meeting someone entirely new: a colleague in another country, an investor thirty years older than you, a partner who grew up with different family rules about conflict. You have no shared dictionary yet. You are, quite literally, translating as you go.

Seen through this lens, a lot of interpersonal friction stops looking like malice or incompetence and more like a lack of shared phrasebook.

They say, “We should talk,” intending “I’d like to understand what’s going on.”

You hear “You’re in trouble.”

They decline your invitation with “Maybe another time,” intending “My week is genuinely slammed.”

You hear “This relationship matters less to me than it does to you.”

They give you feedback in a brisk, declarative way, because to them that’s respectful—getting straight to the point instead of dancing around it.

You hear “You’re a disappointment.”

One of Deborah Tannen’s key insights is that people’s conversational styles are like cultural dialects: timing, overlap, indirectness, teasing, formality, volume. These aren’t random quirks; they’re learned repertoires. Put two different repertoires together with no translation effort, and each side will interpret the other’s style through their own norms. (deborahtannen.com)

A high-involvement speaker who finishes your sentences to show connection may come off as rude to someone who equates not interrupting with respect.

A reflective speaker who leaves long pauses may come off as cold or disengaged to someone who expects quick reassurance.

In both cases, the words might be fine. It’s the wrapper that goes misread.

Howard Giles’ communication accommodation theory adds another useful layer here. He observed that people often unconsciously adjust their speech patterns—accent, tempo, vocabulary—to match those of their conversation partners, a process called convergence. When done skillfully, this feels like rapport: “We’re on the same wavelength.” When styles clash and neither party adjusts, you often get divergence: we remain apart, each talking in our own key. (en.wikipedia.org)

Think of a time you met someone for the first time and, without planning, you both settled into the same pace and slang, and the conversation just flowed. That’s mutual micro-translation at work.

Think of a time you left a meeting fuming that “they just don’t get it.” Often, that’s translation failure.

If being understood is so tricky, why does it matter so much?

Because feeling understood isn’t just a nice-to-have social bonus. It’s one of the basic ways humans register belonging and safety. Experiments on social support and communication consistently find that the sense of “this person gets me” predicts emotional well-being more than the sheer number of interactions or the amount of advice offered. (arxiv.org)

Interestingly, this shows up even in conversations with machines. In one study, people who talked with a version of Amazon’s Alexa that engaged in simple active listening—nodding equivalents, little backchannel noises like “uh-huh” and “I see”—reported feeling more heard and ended up sharing more about themselves than people talking to a version that just waited in silence between prompts. (arxiv.org) A more recent experiment with a social robot that offered contextually appropriate affirmations and reflections found that participants disclosed more, and felt the robot was a better listener, than those talking to a version without these cues. (arxiv.org)

These systems didn’t actually understand anything.

They mimicked the signals of understanding—and our brains filled in the rest.

If a metal cylinder with a pleasant voice can make us feel heard just by saying “mm-hmm” at the right times, imagine how much power you have, as an actual human, to make others feel the same way with a handful of intentional signals.

So how do you become a better translator in a noisy world?

You could, of course, memorize an arsenal of conversational techniques: reflect back what you heard, label emotions, choose the right medium for the message. Those are useful.

But under all the tactics is a more fundamental shift: treating meaning as a joint construction, not a solo performance.

Instead of assuming, “I said the words; my job is done,” you begin to assume, “I sent a first draft of meaning; now we’ll edit it together.”

This mindset changes small behaviors in subtle ways.

You start to show more of your work. Instead of dropping conclusions out of the blue, you include a sentence or two of the reasoning that got you there. People see the path, not just the destination, which reduces the gap for the curse of knowledge to hide in.

You add a bit of redundancy when it matters. Not repeating yourself like a malfunctioning robot, but reinforcing key points through different channels: written and verbal, abstract explanation and concrete example, data and story. Shannon’s original model treated redundancy as something that could help overcome noise; in human conversations, restating something in a new form is often the difference between “huh?” and “ohhhh.” (en.wikipedia.org)

You become more explicit about context. Before delivering a short, blunt message—especially in text—you might add a line of framing that says what game you think you’re playing. “Flagging this early so we can fix it together,” before criticism. “This isn’t urgent; just a thought for next week,” before a request. Tiny preambles that act like legends on a map.

You check your assumptions in small, low-drama ways. Instead of “You’re clearly upset,” you might try, “I’m reading you as frustrated; is that right?” Instead of “We all agree on this plan,” something like, “Let me say back what I think we’re doing; tell me what I miss.” This isn’t therapy-speak; it’s error correction. It’s sending a test signal through the channel and seeing how it comes back.

You pay attention to style as well as content. If you notice someone consistently answering your long paragraphs with short bullet points, or vice versa, that’s a clue about their preferred dialect. You can stay yourself while swiveling a few degrees toward their language—adjusting your tempo, your level of directness, how much you wrap things in story versus summary—so your messages don’t have to fight through as much stylistic friction. That’s communication accommodation in action. (en.wikipedia.org)

Most of all, you become more generous in your interpretations.

Not in a naïve, “everybody means well all the time” way, but in a practical, probabilistic way.

Given what we know—that people regularly overestimate how clear their emails are, that emojis and punctuation can be read wildly differently across groups, that our own internal states are far less visible to others than we think—it’s simply more accurate to treat many small hurts as likely translation problems first. (en.wikinews.org)

The colleague who didn’t respond to your carefully nuanced message with equal nuance might not be dismissing you; they might just be a low-context communicator pinging off short replies between meetings.

The friend who wrote “ok” at 11:42 p.m. and then vanished might not be seething; they might genuinely have meant “ok” and then fallen asleep.

The parent whose texts feel like terse status reports might be playing by a different rulebook about emotional display, not withholding affection on purpose.

You still draw boundaries. You still notice patterns. Chronic disregard for impact is different from occasional misfires. But at the level of any given “ok,” any given awkward pause or stilted response, the most statistically sensible first move is often: “Translation error; request clarification,” rather than, “Character defect; prepare retaliation.”

There’s an odd freedom in accepting that you will never be perfectly understood.

Your internal universe is too detailed; their decoding system is too different; the channels you use are too lossy. A certain percentage of your meaning will simply not make it across.

This could make you feel lonely.

Or it can make you feel oddly relieved.

Because if total mutual understanding is impossible, then the goal shifts from “say it once, perfectly” to “keep iterating together.”

The novelist is misread by some portion of their audience no matter how carefully they craft each sentence. The founder sees their product used in ways they never anticipated. The partner in a relationship discovers, ten years in, that a phrase that always seemed harmless to them has been stinging the other person the entire time.

Of course.

We’re all tappers, and listeners, trying to explain songs to each other with nothing but knocks and gestures and little yellow faces with ambiguous smiles.

The work, then, is to build, over time, richer private dictionaries with the people who matter most to you.

Every friction-laced conversation that doesn’t end in flames but in, “Wait, what did you hear me say just now?” adds a few entries to that shared lexicon.

Every time you adjust your style a bit to meet someone else where they are, you broaden the bridge between your dialects of “Human.”

Every time you decide not to catastrophize an “ok,” but to ask, “Hey, just checking, are we good?” and discover that yes, actually, you are, you move one small pebble from the pile labeled “They’re impossible” to the one labeled “This channel is improvable.”

Over years, if you’re lucky, you end up with a handful of people with whom communication feels almost telepathic. Not because it is, but because the two of you have done so much back-and-forth translation that your error rates have dropped dramatically.

Among strangers and weaker ties, the error rates will always be higher. That’s fine. You design messages accordingly. More redundancy here, more explicitness there, more humility about how your words might land.

You will still send late-night texts that implode.

You will still write emails that made perfect sense in your head and leave someone else bewildered or bruised.

You will still be the listener sometimes, hearing only taps and trying to guess the song.

But if you can hold, in the back of your mind, that every act of communication is a small experiment in lossy translation rather than a verdict on anyone’s worth, you’ll move through all of it with less defensiveness and more curiosity.

And on some quiet future night, when your phone lights up with a bare “ok,” you might feel the familiar spike of confusion rise in your chest—and then soften into something else.

Not certainty.

But the willingness to ask, gently:

“Just to be sure I’m getting you… what does ‘ok’ mean here?”

That single question is a translator’s move.

It’s also a declaration of something deeper:

I know there is more in your head than fits in that word.

I am willing to build a better dictionary with you.