Imagine an Excel spreadsheet  containing loads of data on things like population numbers, taxes,  and even major conquests.

Except what you're looking at  is a bunch of knots and string… and it's the year 1532. That's basically what happened  when Spanish colonizers encountered the Inca Empire along  the western South American coast – a massive territory they’d been  growing for almost a century.

The Inca had pretty much everything  you'd expect in an empire, like big cities and a developed road system. But they were missing one thing virtually every other world empire has had: writing. Although, they did have all those  knotted cords, called khipu.

And it seems like these bundles  of strings helped the Inca keep records and run that massive empire… Somehow. So, what are all these knots about? How can they be used to contain information?

And are they one of the world’s most  unique forms of writing… or knot? [♪INTRO] The word khipu comes from  the Quechuan word for "knot," and they look kind of like  a really intricate necklace. But they're part of a debate about  what actually counts as writing, and also what makes a successful civilization. See, from Ancient Rome to the Abbasid Caliphate, empires throughout history used  writing to pass on information… and /also/ for fun administrative  tasks like taxes and census data.

You know, all the stuff that helps you oversee a massive territory with a lot of people. Defining writing is tricky. But in most of the ways we’d recognize, it seems like the Inca didn’t have it.

The Inca did just fine  running their sprawling empire like drill sergeants  overseeing a training exercise. But there's no evidence they  ever had a written language; just that they had khipu. It's hard to say how old khipu really are.

We do know for sure they were  used during the Wari Empire, which ran from around the 6th to  11th centuries in modern-day Peru. The Wari empire collapsed suddenly,  and we still haven't figured out exactly what their khipu were  used for or how to read them. But after the Wari came the Inca, and with them, drumroll, the golden age of khipu.

Incan khipu were made from cloth, usually cotton, and fibers from animals like alpacas and llamas. Khipu typically included a  primary cord with pendant cords hanging off it, as well as subsidiary  cords hanging off of those. And they were structured a  bit like a fabric abacus, with knots placed in specific areas.

They could be relatively small or really big — the longest one ever found has over 1800 strings and is around five meters in length. Khipu also had plenty of other  attributes, like different colors, and types of knots, and even different ways the strings were twisted together, called the ply. Historical records show  that Incan messengers called chasqui delivered khipu using a relay road system.

There's no indicator that chasqui  could understand khipu, though. For that role, the Inca used  specialists to interpret khipu and share their contents with others. These experts were called khipukamayuq, and they were trained from a young age to become savvy state recordkeepers.

Now, it's possible the  khipukamayuq were just uniquely chosen to become the only literate  people in their civilization — think of Egyptian scribes being specially trained to read and write hieroglyphs. But it's also possible they  were more like musicians. Like, in theory anyone could  learn to read sheet music for a viola part and understand what  the notes individually mean.

But to actually convey what the music is saying, you'd have to know a lot more — like how to move from a C-sharp on the G string to an E on the A string and so on. In other words, it’s possible that the khipukamayuq had specialized knowledge, with the khipu serving as  something like a memory aid, rather than containing all  the information within itself. We know the Inca used khipu up to  and through Spanish colonization.

And sometimes the Spanish encouraged their use, for religious or recordkeeping purposes. But, typical Spanish conquerors, for the most part they  condemned khipu as idolatrous and destroyed most of the  ones they got their hands on. Over time, khipu use died out.

Some  Andean villages kept using khipu, but otherwise… they pretty much  went the way of the dinosaur. And so that we don’t do the same,  let’s take a quick break for an ad. Thanks to Brilliant for  supporting this SciShow video!

Brilliant is an online learning platform where you can develop your skills and deepen your knowledge. And you can do all of that for a month for free! Which is something we can  get behind here at SciShow.

They’ve got courses on math, data analysis, and more, so you won’t run  out of topics to explore. Whether you’re just starting  or you’re already a pro, Brilliant has classes designed to build on your skills and give you an interactive challenge. You’ll get to solve puzzles  and work with real data, so you stay engaged while you learn.

To try Brilliant, visit  brilliant.org/scishow, scan the QR code onscreen, or click on the  link in the video description. You’ll also get 20% off an  annual premium subscription. So how do we know so much about khipu?

For one thing, we have the  records of Spanish colonizers. But we also have a mathematician with the delightful name of  Leslie Leland Locke to thank. A little over a hundred years ago,  our guy Leslie figured out that a lot of khipu are numerical  and use a base-10 number system.

Decades later, other researchers  found that khipu typically consisted of cords that had powers  of ten, with different types and placements of knots coming  together to convey a number. Each chord on a khipu was arranged like this, with knots in a series of rows. Those knots typically made up a total number, with rows divided into sections  — a ones row, a tens row, a hundreds row, and on and on  for as long as they wanted to go!

In the ones row, you’d find  a long knot with twists in it representing what the number should be. So, if a khipu cord wanted to  convey the number 432, the ones row, representing 2, would have  a long knot with two twists. But, at least in some cases,  the rest of the cord would show single knots representing multiples of 10.

So for 432, above the long  knot with two twists there would be a cluster of three knots  together in the tens place for 30, and then above that would be a cluster of four knots together in the hundreds place for 400. You’d be reading these top-down,  which is how you’d get to 432! Things got a little spicier  if you wanted to represent the number one, which used a figure eight knot.

Meanwhile, no knot at all, meant zero. That would all be on one cord,  sometimes with many more to follow. Not all numerical khipu are the  same; some had knots that were more like codes, standing in for  things like places, or individuals.

Like, khipu from one archaeological  site near Lima appear to use a set of knots representing  the area of Puruchuco… kind of like an ancient zip code. About two-thirds of the khipu  we've ever found are numerical. But the rest are a big mystery.

We tend to call those narrative  khipu, and we think they contain a lot of information  about battles, history, and even things like poetry and astrology. After all, khipu are really complex,  potentially way more so than they'd need to be if they were just meant  to jog a khipukamayuq's memory. Plus, there are a lot of  Spanish records claiming that khipu held information that  went far beyond numbers.

They wrote about how the  chasqui runners carried khipu containing things like administrative  tasks or descriptions of conflict. But how was that information recorded? Was it like written Quechua: Knot Edition, or more like conveying pictures through string?

Well, welcome to a big dispute among experts. One theory is that narrative  khipu are semasiographic, which would mean they're a system of symbols that contain information. Think about road signs; we all kinda know that an image of a moose means to  be on the lookout for gigantic mammals with antlers that can  roll through your windshield.

Some supporters of this theory argue that khipu sort of form a seven-bit binary code, using elements like fiber type and the ply and spin of a knot to create  distinctions between symbols. Researchers supporting this theory say khipu could have stored more than 1,500  units of information this way — beyond what ancient writing systems  like Egyptian hieroglyphs could. Another big one argues that  khipu are logosyllabic, meaning they use symbols to  stand in for words and concepts, but also sounds — like Chinese does.

That theory rests heavily on an  18th century rebellion effort against Spain by leaders from the small village of San Juan de Collata in the Peruvian Andes. Their communications were  recorded on khipu and kept for centuries by villagers, until  anthropologist Sabine Hyland arrived and was invited to view this  Holy Grail of khipu research. Hyland had only 48 hours to examine the khipu… and it was a full sensory experience.

The khipu had 14 colors and hair  from six different types of animals, some of which could only be identified by touch. While using touch and texture  to guide her inspection, Hyland and her husband cross-referenced the khipu with more than 1,000 pages of  historical written testimony. She finally determined that the khipu  relied on a system of 95 symbols.

Some were phonetic, or represented sounds; the rest were ideographic, meaning they represented an idea or a concept. Hyland got to that conclusion  by parsing out what seem to be the signatures of former village  leaders from the rebellion plot, who used a combination of color, fiber, and knots to form syllables and concepts. Signatures make sense at the end of  a communication like a letter, and Hyland already knew she was  looking at messages between people.

That made it a lot easier for her to catch on quickly when she began to  notice phonetic patterns. In one instance, a final blue pendant seemed to signal the local word for the color, "ankas," providing the letters K and A  for the end of the name "Alluka." Alas for Collata's scrappy resistance fighters, records show their rebellion wound up crushed. And while decoding parts of the Collata khipu was a watershed moment in khipu  research and understanding, here are still some major caveats.

For one thing, they're the only  phonetic khipu ever discovered. They could be a one-off  regional variation that doesn't share much in common at all  with other khipu from elsewhere. Plus, they date to a time  when the Inca had already interacted with the Spanish  and writing as a concept, which could have introduced  changes in the way khipu were used.

So, are khipu a kind of writing? An unsatisfying answer is  that we won't know until we've actually cracked the code;  for now, those narrative khipu lying around are keeping  their secrets to themselves. But a better question might be  whether it actually matters.

Instead of wondering how an empire  could thrive without writing, maybe we need to reconsider what it means to administer one in the first place. Writing is an incredible tool. But it's not the only way to communicate or effectively perform administrative tasks.

The Inca are a great example — they used khipu to do so much  and to do it really well. Maybe we'll find that there  are many types of khipu, and some are writing and some aren't, and some are a cool hybrid of writing and math unlike anything we've ever encountered. And maybe they used khipu for telling stories, and maybe a khipukamayuq had to get involved, and maybe we'll never know.

But what we do know is that they  came up with a really complex and fascinating system unlike  anything we've ever seen. That was enough to help them  run a successful empire, whether it was writing, or… knot. [♪OUTRO]