Archive · the grid script
Writing to a voice we have never seen.
The duet with a pulsar was a handshake: it set a clock both sides share and showed that the signal was tuned for Earth. The content came in the four transmissions after it, and the translating machine needed a way to write that down. This is that script, and it is still a first draft: a language both sides can speak back and forth is built over many exchanges, and each round trip takes 22 years. It is our way of representing their language, not their language: as far as anyone can tell they speak by resonance, in air, and nobody has seen them write anything, or seen them at all. The grid follows the one thing we think we know about that speech: one voice alone can say only a little, and meaning grows when voices join.
The drawings need JavaScript and the vocabulary file; they could not load.
1 · The grid
One grid, one mark per voice
- Everything is drawn on one shared grid of cells.
- Each voice gets its own mark: + for Ross 128 b, × for us. The machine never lets marks cover each other, so you can always see who said what.
- One grid square of 3 × 3 cells is one part, and two parts make a word, the way strokes make a Chinese character. It is not an alphabet.
- Spacing is structure: 1 empty cell between the parts of a word, 2 between the words of a triplet, 4 between triplets.
- A triplet is one statement. The other voices' answers are written under it.
So far, everything we have transcribed from them is in + alone. The second band in the drawing is a reply humans could send: water, instead of air.
2 · Parts
Sixteen parts
Each part is a 3 × 3 shape with two meanings: one as a thing, and one (after the arrow) as a relation. The shapes are ours, chosen to be easy to guess: a closed square for us, a square opened toward someone for you, crossing rays for light, a rising staircase for sound.
3 · Words
Two parts make a word
A word is two parts, one empty cell apart: the kind on the left and the which on the right, written KIND.WHICH. A world that is yours is your world; a star that keeps time is a pulsar; air going is wind. A one-part word is just 3 cells wide. The dictionary lists every agreed word.
4 · Lines
Every triplet is one edge of a graph
A text is a run of triplets (node, relation, node), read left to right and wrapping onto new lines. Each triplet is one edge of a knowledge graph, so a text is a graph. The gaps say it all: 1 cell inside a word, 2 between words, 4 between triplets, 3 rows between lines. The same word in two triplets is the same node, and a lone ONE in a node slot means "that": the triplet before.
A relation is one part for the action, plus a which that marks time or polarity. With LIGHT (see):
| Which | Meaning | LIGHT … |
|---|---|---|
_ | no tense | see |
BEFORE | past | saw |
ONE | now | see now |
AFTER | future | will see |
PATH | always | keep seeing |
NOT | not | do not see |
OPEN | asked: does it? | do … see? |
The second example is the shadow transmission put into words: we saw you; we do not see you. The third points back at itself: your world is water, and we saw that.
5 · Numbers
Numbers are base-512 digits
A number is COUNT, a counting line, followed by base-512 digits, most significant first. Each digit is nine bits: 256 down to 1, left to right, top to bottom. COUNT alone is zero; 137 is COUNT.137; the year 2219 is COUNT.4.171, because 4 × 512 + 171 = 2219. Only the positions after COUNT are read as bits, so a number can never be mistaken for a word.
Written COUNT.137. Try 137, the fifth transmission, or 2219.
6 · Voices
Stacking means another voice
Only a new party adds a band under a triplet, and each party keeps its band, so the number of bands is the number of voices. Every part in a lower band answers the nearest drawn part above it in the same position of the same word. That lets anyone agree with, deny or change any node or edge, part by part. A conversation is a graph several voices build together.
| Below | Means |
|---|---|
| The same part | yes |
| NOT | no |
| A different part | instead |
| A part under an empty position | and also |
| Nothing | no comment |
Questions: OPEN in any position asks "what goes here?", and whatever is drawn under it replaces it. In the last example a third voice, *, says AIR under our WATER. It answers the nearest part above, which is ours: it sides with them, against us, about our own world.
7 · The machine
How the machine reads a page
- Split the page into lines at the three-row gaps, and count the bands in a line: that is the number of voices.
- Cut each line into triplets, words and parts at the 1, 2 and 4 cell gaps. Anything that won't cut into whole parts is not text.
- Read the first band of each triplet as an edge: node, relation, node. Same word, same node. A lone ONE points at the triplet before.
- Read every part in a lower band against the nearest part above it, and attach the answer to that node or edge, with the voice that said it.
Humans don't learn the signs; the machine reads and writes them, and its lexicon grows with every exchange. You can run the same machine: the converter turns text into drawings and drawings back into text, in your browser. It follows the same rules and the same vocabulary as glyph, the command-line dictionary (pip install glyph-cli).