Ah, the khipu. The way that it represents numeric info is somewhat well understood already:
- it’s all base 10, positional. The tens/hundreds/etc. of different strings in the same khipu are aligned.
- zero = no knot
- 1~9 in the tens, hundreds etc. are represented by 1~9 simple knots
- 1 in the units is represented by a figure 8 knot
- 2~9 in the units is represented by a long knot with 2~9 turns
This might sound complicated but it’s really elegant, and representing the units in a different way allow you to cram multiple numbers into the same string.
So for example. Let’s say that you want to record 234 and 506 into a string. You’d do the following:
- 2 simple knots
- 3 simple knots
- long knot with 4 turns
- 5 simple knots
- space
- long knot with 6 turns
In some cases there might be geographical info in the khipu too, with numbers representing localities. Kind of like postal codes. The material of the string and the colour likely encode some info too, but AFAIK nobody knows it any more.
I’m almost sure that it doesn’t contain any sort of textual info, though. Like, something you can read. Classical Quechua had at least 17 consonants, this would be impractical to represent through knots, specially as Quechua tends towards large words.
My bet on both “paired” khipukuna is that one encodes income, another outcome. Kind of like double bookkeeping but for material.
textual info, though. Like, something you can read.
I think they probably do, but like other early forms of record keeping, the focus definitely seems to be numbers. IIRC there are some knots that just don’t look like numbers.
The Chinese also used knots to record details of events, not just numbers, but the system that is used to decode the knots was complex enough that it functioned like a cipher - without already knowing how they did it, there’s just no way you’d be able to work it out.
I can’t recall the details but I think it’s covered in the Ancient Americas video on the khipu on youtube.
By “textual info” I mean plain language, like we’re using now. It’s theoretically possible to encode it in khipu, not just for Quechua but for any other language; but doing it in a practical way is another can of worms.
Instead what I think that they used is what the video calls a “semasiographic system” - there are standardised codes for almost everything worth registering (from a bureaucratic PoV), and the officer/kamayuq is expected to be able to decode it.
For a silly example using English, it would be a lot like writing “Jn Smth in ptt 20 mze 35” and then reading it as “John Smith stored 20kg of potatoes and 35kg of maize here”.
That hypothesis makes sense. Seems like we’ll never know for sure since no contextual clues survived.
