Elliott waves are a way of splitting price movement into parts: five waves in the direction of the trend, then three against it. The five are called an impulse, the three a correction. Each wave divides the same way at a smaller scale.
The idea dates from the 1930s and is essentially a grammar: rules by which a stretch of chart decomposes into legs.
What actually gets marked up
Markup answers one question: which wave the most recent significant extreme — the peak — is. There is usually more than one answer, and each is called a reading, or a count.
Each reading is a complete answer: where the preceding waves sit, which wave is running now, and the zone where it could end according to the geometry of its form.
The hard part is discipline
Markup is easy to fit to an opinion you already hold: nudge a wave boundary by a couple of candles, "fail to notice" a broken rule, pick a convenient scale. That is exactly why wave analysis has hard rules — they exist to rule out convenient but impermissible counts.
The flip side of an honest approach: sometimes no numbering passes the rules at all. "There is no markup" is a legitimate answer, not a reason to invent one.
Why there are several readings
Because the markup is ambiguous by nature. The same peak can be the end of a fifth wave, the end of a third, or the end of wave B inside a correction — and each version is self-consistent.
We do not pick one and call it correct. Everything that passed the rules is shown, and the differences between readings are visible. There is a separate write-up on that ambiguity.
What the markup does not do
It does not predict price. A wave's completion zone is the geometry of a form, not a promise that price will get there. Wave markup describes the structure of a stretch already travelled and the continuations the rules permit; it does not become a probability or an entry signal.