The Medium Is Always Dying: America's 180-Year Habit of Predicting Communication Collapse

In 1858, when the first transatlantic telegraph cable briefly connected the US and Britain, observers on both sides of the Atlantic declared that letter-writing was finished. Why would anyone commit words to paper and wait weeks for a reply when a message could cross an ocean in minutes? The letter, they said, was an artifact of slower times. It would not survive the decade.

American letter-writing volume increased every year for the next forty years.

This should be the end of the story. It isn't, because we have repeated the exact same prediction cycle — with the exact same outcome — approximately every forty years since, with a reliability that suggests something important about how human beings actually use communication tools versus how we think we use them.

The Prediction Always Looks Reasonable

To be fair to the letter-writing doomsayers of 1858: their logic was not stupid. The telegraph was faster, cheaper per unit of information, and more dramatic. It was hard to imagine a scenario in which, given the choice between instant and delayed, people would consistently choose delayed.

The same logic animated every subsequent prediction. When the telephone became widely available in the early 20th century, observers noted that real-time voice communication made writing look redundant. When email arrived in the 1990s, a thousand op-eds declared the phone call an endangered species. When SMS texting took off in the 2000s, email was pronounced dead. When Snapchat and Instagram emerged, texting was supposedly on its way out. The logic in each case was identical: why use the slower, more effortful medium when the faster, easier one exists?

The logic keeps being wrong for the same reason. It assumes communication is a single activity with a single goal, and that people will naturally route all of it through whichever channel is most efficient. Neither assumption is true.

What Communication Is Actually For

Here's what the data, gathered over nearly two centuries of postal records, telephone logs, and now digital analytics, consistently shows: new communication technologies don't replace old ones. They specialize them.

When the telegraph arrived, letters didn't disappear — they changed. The telegraph handled urgent, transactional communication. Letters absorbed everything else: the reflective, the relational, the things that benefited from time and care. The telegraph made letters better by relieving them of the burden of urgency. Letter volume went up.

The same thing happened with the phone. It didn't kill letters; it took over the conversational, spontaneous exchanges that letters had never been ideal for anyway. Letter volume didn't collapse. It shifted toward the more deliberate, more permanent communications that voice couldn't capture well.

Email didn't kill phone calls. It absorbed the asynchronous, documentable, moderately formal exchanges that phone calls handled awkwardly. Phone call volume kept climbing for years after email became ubiquitous, because it turned out that some things are genuinely better handled by voice — quick clarifications, emotional conversations, anything that benefits from tone.

Texting didn't kill email. It took over the brief, informal, low-stakes exchanges that email had never been great at. Anyone who has received a three-word email saying "sounds good!" understands intuitively that this was not what email was designed for.

The pattern is consistent enough that it has predictive power. Every new communication medium absorbs the use cases it's genuinely better suited for, which frees the older medium to concentrate on what it was always better at. Both survive. Both often grow.

Why We Keep Getting This Wrong

The prediction failure is interesting because it persists even among people who watched the previous prediction fail. Baby Boomers who saw email survive the rise of texting still confidently predicted that texting would kill email. Millennials who watched texting survive Snapchat are now confidently predicting that something will kill texting.

The error isn't ignorance. It's a consistent cognitive bias toward what researchers call "tool substitution" — the intuitive assumption that new tools replace old ones rather than joining them. We think about communication the way we think about transportation: when cars arrived, horses largely went away. The substitution was real and visible.

But communication isn't transportation. A horse and a car do the same job — moving a person from one place to another — and the car does it better in almost every dimension. Communication tools don't do the same job. They do adjacent jobs, with different tradeoffs, and humans are remarkably good at sorting those jobs to the right tools once the sorting period is over.

The sorting period is what gets misread as collapse. When a new medium arrives, usage patterns are genuinely chaotic. People are figuring out what it's good for. Older mediums see weird usage dips as some people over-migrate to the new thing before discovering its limitations. Observers watching the dip declare the older medium dead. Then the sorting completes, usage patterns stabilize, and both mediums grow — usually together.

The Anxiety Layer

There's a second thing happening underneath the prediction cycle, and it's more psychologically interesting than the predictions themselves.

Every new communication medium doesn't just change how we communicate — it changes who can communicate, at what cost, and with what degree of permanence. Each of those changes creates genuine social anxiety, and the death-of-communication narrative is partly a way of processing that anxiety.

The telegraph made communication faster, but it also made it more surveillable and more expensive. The phone made communication more intimate and immediate, but it also made it harder to think before speaking and impossible to revise. Email made communication cheaper and more permanent, but it also created an expectation of constant availability that the letter never carried. Texting made communication ambient and low-stakes, but it fragmented attention in ways that longer-form communication hadn't.

Each new medium creates genuine losses alongside its gains. The death-of-communication narrative is the culturally legible way of naming those losses. When someone says "texting is killing real conversation," they're not actually making a prediction about medium survival rates. They're expressing something real about what ambient, low-stakes communication costs in terms of depth and attention. The prediction is wrong, but the anxiety is legitimate.

The Layer Keeps Stacking

What we have now is not a simpler communication landscape than the one our grandparents navigated. It's a dramatically more complex one, with more mediums, more specialized norms for each, more anxiety about which to use when, and more social cost attached to choosing wrong.

The letter is still here. So is the phone call. So is email, texting, video chat, voice memo, and approximately fifteen messaging apps with subtly different social contexts. None of them died. They all specialized, found their niches, and stuck around.

The next medium will not kill any of them either. It will join the stack, absorb the use cases it handles best, and generate a fresh round of predictions about what it's replacing.

We will be wrong again. We always are. Human communication needs are not getting simpler, and the tools we use to meet them are not getting fewer. They're getting more numerous, more specialized, and more layered — which is exactly what five thousand years of evidence suggests happens when you give people more ways to talk to each other.