The Chart Is Holding You Back

There is a chart on this site with every character in the International code laid out in dots and dashes. It is accurate, it is useful, and it is the first thing almost everyone reaches for. It is also, used the obvious way, the most expensive mistake available to a beginner — not because the information is wrong, but because studying it trains the wrong mapping, in the wrong direction, and does so efficiently enough that you will not notice for two months.

The habit you are actually building

Copying Morse is one operation: sound in, letter out. When you learn the code from a table, what you build is something else entirely. You hear the signal. You form a mental picture of the dots and dashes. You search for that picture in a remembered table. You read off the letter. Four steps wearing a coat and pretending to be one, and three of them are conscious, serial and slow.

The trap is that it works. At five or six words per minute there is ample time for all four steps, progress is immediate and measurable, and you can have the whole alphabet apparently learned in a weekend. Nothing about the experience warns you. The bill arrives three weeks later, at about ten words per minute, when the time available per character drops below the time the lookup needs. That is the wall described in the plateau dispatch, and this is where it is built.

Mnemonics make it worse, not better

The aids invented to make the table easier to memorise all share the same defect, usually in a stronger form. Word mnemonics — a phrase whose stressed syllables match the rhythm — install a second intermediate on purpose: sound, then phrase, then letter. Charts that draw each letter out of its own dots and dashes install a visual one. The binary decision tree, dot left and dash right, is an elegant piece of structure and a hopeless decoding procedure: traversing it is a conscious serial operation four or five levels deep, and at 20 WPM the entire budget for a character is about six hundred milliseconds.

Every one of these is a deliberately rehearsed intermediate step. That is not a side effect. It is what a mnemonic is, and it is a precise description of a habit you will have to demolish.

The honest qualification: they are not useless. A mnemonic is a fine way to get unstuck on the one or two characters you keep dropping. The rule is dosage — three or four characters, for a few days, then gone. If you still have a mnemonic for the whole alphabet after a month, you do not have a memory aid, you have a decoding system.

Why it is so hard to undo

You cannot decide to stop doing something automatic. Telling yourself to stop visualising the dots is about as effective as telling yourself to stop hearing an accent. Automatic processes are not removed by instruction; they are outcompeted.

Which gives the one lever that moves: make the old route unusable by starving it of time. Run the characters fast enough that the lookup cannot complete, and it stops getting rehearsed, and the direct sound-to-letter association — forming quietly underneath all along — becomes the only route that delivers an answer. That is the entire mechanism, and both standard teaching methods are built on it.

What Koch does about it

Ludwig Koch was a German psychologist working on Morse instruction in the 1930s, and his result was that students taught slowly and then sped up plateaued, while students started at the final speed with a tiny alphabet did not. The method that came out of it is deliberately primitive: two characters, at full target speed from the first minute, copied until you are hitting 90 percent, then a third character at the same speed, then a fourth. The speed never changes. Only the alphabet grows.

Read against the problem above, what Koch is doing is obvious. At 20 words per minute there is no time to picture anything, so the visual route never forms and there is nothing to unlearn later. Koch does not teach you to hear characters. It prevents you from learning to do anything else, then gets out of the way while the hearing builds itself.

The 90 percent threshold is not a grading standard either. It is a check that the direct association has formed before you add load. Advance at 70 percent and you carry the gaps forward; the ones you half-know become the ones you stall on four characters later, and you will not be able to tell which they are. The correct response to a failed level is to repeat the level — emphatically not to slow the characters down, which undoes the only thing the method is doing for you.

What Farnsworth does about it

Koch leaves one practical problem: full speed from minute one is brutal, and a beginner needs a way to make the task easier that does not touch the character speed.

Farnsworth spacing is that knob. Characters go out at 18 or 20 words per minute with every internal ratio intact, and the silences between them are stretched until the message as a whole crawls along at eight. You hear the real sound of the character with a long pause after it. The pause is thinking time, thinking time comes down as you improve, and when it reaches zero the characters are already where they need to be. Nothing gets rebuilt.

What matters is that it separates difficulty from speed, which are otherwise the same dial. Watch the two numbers independently. A practice tool offering one speed control and nothing else is lowering your character speed to make things easier, and teaching you to count.

A progression that does not have to be undone

Once, at the start. Open the chart, go down it, and say each character aloud as a rhythm — di-dah, dah-di-di-dit. Ten minutes. The chart's job is to tell you what the sounds are called. Then close it and do not study it again.

Two characters, full character speed, wide spacing. Somewhere around 18 to 20 WPM characters and 8 WPM effective. Five-minute runs. Add a character at 90 percent. Fifteen to twenty minutes a day, most days, beats an hour on a Sunday, because this consolidates between sessions rather than during them.

Say them, do not write them. For as long as you can keep up, speak the character aloud instead of writing it. A pencil adds a lag that hides recognition failures inside handwriting, and it lets you cheat with a half-formed letter you fix a second later.

Random groups before plain text. English is redundant and will quietly repair your copy, which feels excellent and teaches nothing. Once the alphabet is complete, alternate: random five-character groups to test recognition, real sentences to build rhythm.

Then the rest of the character set. Digits, the four punctuation marks that actually get sent — full stop, comma, question mark, slash — and the prosigns, which are single run-together characters and must be learned as single sounds rather than two letters jammed together.

Close the spacing. One word per minute at a time, several sessions per step, until effective speed meets character speed. Purely mechanical, no shortcut in it.

Then callsigns, then no pencil at all. Callsigns because they have no redundancy to lean on. Head copy because that is the actual skill, and because the first time somebody answers your call you will be writing with one hand and reaching for the key with the other.

How to tell if you are still looking things up

Four tells, and any one of them is enough. You can write the letter but you have to stop and think to say its rhythm aloud. Some characters arrive instantly and others need a visible beat — the slow ones are almost always the ones you learned from the table. Your copy collapses when you put the pencil down. And the diagnostic one: you can produce the code for a letter faster than you can recognise the letter from the code.

That last asymmetry is the whole point. Sending and receiving are two mappings running in opposite directions, and a chart only trains one of them. For sending it is exactly the right tool. For receiving it is the thing standing between you and 20 words per minute. Use the trainer for the ear and the chart for the hand, and nothing at all for the step in between.