Quick quiz: how do you write “x² + y³” in plain text, no HTML, no LaTeX?
Easy, right? `²` and `³` are just characters. Now write "E = mcᵠ" — wait, that should be a superscript "q". Try to find it. I'll wait.
You won't, because **Unicode has no superscript q**. When I built a [free superscript converter](https://literalkit.com/superscript-generator), this turned out to be the tip of
a very weird iceberg. Here’s what I learned.
Superscripts aren’t formatting — they’re separate characters
Most people assume superscript works like bold: take a character, apply a style. That's how `` works in HTML, and how rich text editors work.
Unicode plain text doesn't have styles. Instead, every superscript character is its own pre-composed codepoint, donated to the standard by different communities at different
times:
- `¹²³` live in **Latin-1 Supplement** (U+00B9, U+00B2, U+00B3) — they've been there since the 1980s.
- `⁰ ⁴-⁹ ⁺ ⁻ ⁽ ⁾` live in the **Superscripts and Subscripts** block (U+2070–U+209F).
- Superscript *letters* are scattered across **Phonetic Extensions** (U+1D2C–U+1D7F), because linguists needed them for phonetic transcription: `ᵃ ᵇ ᶜ ᵈ ᵉ ᶠ ᵍ ʰ ...`
- And a few stragglers like `ˣ` (U+02E3) hide in **Spacing Modifier Letters**.
So "convert text to superscript" really means "look up each character in a hand-built mapping table someone assembled from four different corners of the standard." There's no
algorithm. It’s a scavenger hunt.
The honest-output problem
Here's where it gets philosophically interesting. The mapping has holes:
- **No superscript `q`.** It's the famous gap. Proposals exist, but as of today there's no official codepoint.
- **Uppercase letters are incomplete** — you get `ᴬ ᴮ ᴰ ᴱ ᴳ ᴴ ᴵ ᴶ ᴷ ᴸ ᴹ ᴺ ᴼ ᴾ ᴿ ᵀ ᵁ ⱽ ᵂ`, but try spelling a full sentence in superscript caps and you'll hit walls fast.
- Subscripts are even sparser: `ₐ ₑ ₕ ᵢ ⱼ ₖ ₗ ₘ ₙ ₒ ₚ ᵣ ₛ ₜ ᵤ ᵥ ₓ` — that's most of what exists.
Most converters on the web quietly patch the holes with **lookalike characters** — a Greek letter here, a math alphanumeric symbol there. The output *looks* fine until someone
copies it into a screen reader, a search box, or a font that renders the substitute as tofu (□).
I made a different call: when a character has no real superscript, the tool says so instead of faking it. It feels worse in the demo and better in real life — what you copy is
what you meant.
