Why CW Refuses to Die

On 31 January 1997 French naval coast stations sent their last message in Morse, ending with a line that has been quoted ever since: a last cry before eternal silence. On 1 February 1999 the satellite-based Global Maritime Distress and Safety System became fully effective and the century-old obligation to keep a Morse watch at sea was gone. On 12 July 1999 the last commercial Morse message in the United States went out from a California coast station, closing with the text Samuel Morse had sent in 1844 and the prosign SK. That was the end of the job. Morse had been the way the world talked to ships for ninety years and then it was not the way anything talked to anything.

The requirement to know it went next. The international obligation on countries to test amateur licence applicants for Morse proficiency was removed at the World Radiocommunication Conference in 2003, and national administrations dropped their tests over the following few years — the United States ended the last of them in February 2007. So within a decade the mode lost both the thing that paid it and the thing that compelled it.

It should have gone quiet. The odd thing is that more people appear to be learning it now than were learning it when it was still compulsory.

The awkward part: measuring it

Nobody counts Morse operators, so the honest position is that there are no clean numbers and anyone quoting you one has made it up. What there is instead is a set of indicators that all lean the same way. Free, structured, multi-week CW courses run by volunteers are routinely oversubscribed. Weekly short CW contest sessions built specifically for people practising get turnouts that did not exist twenty years ago. The portable award programmes that have grown enormously over the last decade are disproportionately CW, because CW is what works on five watts from a hillside. And there is now a small manufacturing sector making keys and low-power CW transceivers that would have had no market at all in 1999.

That is circumstantial, and it is the kind of circumstantial that is hard to explain away. Something is drawing people to a skill with a hundred-hour entry cost and no external reward whatsoever.

The bandwidth argument, stated fairly

The technical case for CW is real and it is arithmetic. A Morse signal at conversational speed occupies well under a hundred hertz of spectrum. A receiver can therefore filter it at 250 or 500 hertz and throw away everything else. A single-sideband voice signal needs around 2.4 kilohertz to be intelligible, and noise power scales with bandwidth, so narrowing from 2400 down to 250 discards about ten decibels of noise while keeping all of the signal. On top of that, a human ear tracking a steady tone works at a lower signal-to-noise ratio than a human ear trying to reassemble speech.

The result is what operators have always claimed: CW gets through conditions that break voice, on the same radio, into the same antenna, at the same power. That has not stopped being true.

But it has stopped being the whole story, and pretending otherwise is how the argument gets lost. The modern weak-signal digital modes decode far below where a human ear gives up — the most popular of them works down around twenty-one decibels below the noise in a 2.5 kilohertz reference bandwidth, which is a long way past anything a person can copy. If the only objective is to complete a contact in the worst possible conditions, the machine wins and it is not close.

So the claim has to be stated more carefully than it usually is. CW is the best mode in which a human being is in the loop. That is a narrower statement and it is one you can actually defend.

The simplicity argument, which is underrated

A CW transmitter is close to the simplest thing that can cross an ocean. An oscillator, a stage that the key switches on and off, an amplifier and a filter to keep the harmonics at home. People build them on a Saturday from parts that cost less than dinner, and those radios work — not as a stunt, but as a station.

A voice transmitter needs modulation and linearity and an order of magnitude more of everything. A digital mode needs a computer, a sound interface and a clock accurate to within a second or two. Both are excellent and neither fits in a coat pocket alongside a battery that will last all day.

There is a second-order version of the same point. CW degrades gracefully. When the antenna is a compromise, the band is marginal, the battery is low and the power amplifier has stopped working properly, Morse is what is left. That is not a romantic claim about the spirit of the operator; it is a consequence of needing so little bandwidth and so little signal. It is also why it keeps reappearing in emergency communication long after anyone planned for it to.

The craft argument, which is the real one

Morse is a motor skill and a perceptual skill, and skills of that kind have a particular shape: a steep and frankly unpleasant beginning, then years of slow, visible improvement with no obvious ceiling. You can be better at this in five years than you are now. Most technologies do not offer that, because most technologies are designed so that competence is immediate and then flat.

It is also individual in a way that no other mode is. Two operators sending identical text sound different — the weighting, the spacing, the small hesitations — and people who spend time on the bands recognise each other by it. That is why the wartime services talked about a fist as a form of identification, and it is still true on any evening on 40 metres. There is a person at the other end and you can hear which person it is.

And it is live. Two people making sound at each other in real time, at a pace set by their hands, able to interrupt, to slow down, to make a joke, to ask again. An automatic mode exchanging structured reports is a different and entirely legitimate activity, but it is not a conversation, and conversations are what most people are actually after.

There is a case that the automatic modes have helped CW rather than hurt it. When something arrived that produced contacts more efficiently than Morse ever could, the people who mainly wanted contacts went and used it. What was left on the CW portions of the band was people who wanted CW, which is a considerably better band to be on.

And the nostalgia, honestly

A good deal of the attraction is romance and it would be dishonest to pretend otherwise. Brass and bakelite. Ships calling on 500 kilohertz and the silent periods twice an hour when everyone stopped to listen. Operators copying under fire. Volunteers still fire up the original transmitters at one of those closed American coast stations every year on the anniversary of the last message, and people tune in from around the world to hear a commercial Morse station that has had no commercial purpose for a quarter of a century.

Two caveats worth holding onto. Nostalgia is not a technical argument and should never be dressed up as one — the dressed-up version is easy to knock down and it takes the good arguments with it when it falls. And nostalgia does not explain persistence. It will make somebody buy a straight key. It will not carry them through the 10 WPM plateau, which is months of unglamorous daily work with no scenery in it at all. Nobody grinds through that out of sentiment. They do it because the thing on the other side turns out to be worth having.

That is the whole answer, really. A technology outlived its job by decades because it stopped being a technology and became a skill, and skills are kept alive by the people who enjoy practising them. The history is worth knowing, the timing takes an afternoon to understand, and the first two characters take about ten minutes. The rest of it takes as long as you want to give it.