00:00:00
Imagine an Excel spreadsheet
containing loads of data on things
00:00:03
like population numbers, taxes,
and even major conquests.
00:00:07
Except what you're looking at
is a bunch of knots and string…
00:00:11
and it's the year 1532.
00:00:12
That's basically what happened
when Spanish colonizers
00:00:15
encountered the Inca Empire along
the western South American coast
00:00:19
– a massive territory they’d been
growing for almost a century.
00:00:23
The Inca had pretty much everything
you'd expect in an empire,
00:00:26
like big cities and a developed road system.
00:00:28
But they were missing one thing virtually
00:00:30
every other world empire has had: writing.
00:00:33
Although, they did have all those
knotted cords, called khipu.
00:00:37
And it seems like these bundles
of strings helped the Inca
00:00:40
keep records and run that massive empire… Somehow.
00:00:44
So, what are all these knots about?
00:00:46
How can they be used to contain information?
00:00:49
And are they one of the world’s most
unique forms of writing… or knot?
00:00:53
[♪INTRO]
00:00:57
The word khipu comes from
the Quechuan word for "knot,"
00:01:00
and they look kind of like
a really intricate necklace.
00:01:03
But they're part of a debate about
what actually counts as writing,
00:01:07
and also what makes a successful civilization.
00:01:09
See, from Ancient Rome to the Abbasid Caliphate,
00:01:12
empires throughout history used
writing to pass on information…
00:01:16
and /also/ for fun administrative
tasks like taxes and census data.
00:01:20
You know, all the stuff that helps you
00:01:22
oversee a massive territory with a lot of people.
00:01:25
Defining writing is tricky.
00:01:27
But in most of the ways we’d recognize,
00:01:30
it seems like the Inca didn’t have it.
00:01:32
The Inca did just fine
running their sprawling empire
00:01:35
like drill sergeants
overseeing a training exercise.
00:01:38
But there's no evidence they
ever had a written language;
00:01:41
just that they had khipu.
00:01:43
It's hard to say how old khipu really are.
00:01:45
We do know for sure they were
used during the Wari Empire,
00:01:48
which ran from around the 6th to
11th centuries in modern-day Peru.
00:01:53
The Wari empire collapsed suddenly,
and we still haven't figured
00:01:56
out exactly what their khipu were
used for or how to read them.
00:02:00
But after the Wari came the Inca,
00:02:02
and with them, drumroll, the golden age of khipu.
00:02:06
Incan khipu were made from cloth, usually cotton,
00:02:09
and fibers from animals like alpacas and llamas.
00:02:12
Khipu typically included a
primary cord with pendant cords
00:02:16
hanging off it, as well as subsidiary
cords hanging off of those.
00:02:21
And they were structured a
bit like a fabric abacus,
00:02:23
with knots placed in specific areas.
00:02:26
They could be relatively small or really big —
00:02:29
the longest one ever found has over 1800 strings
00:02:33
and is around five meters in length.
00:02:35
Khipu also had plenty of other
attributes, like different colors,
00:02:38
and types of knots, and even different ways
00:02:40
the strings were twisted together, called the ply.
00:02:43
Historical records show
that Incan messengers called
00:02:46
chasqui delivered khipu using a relay road system.
00:02:50
There's no indicator that chasqui
could understand khipu, though.
00:02:53
For that role, the Inca used
specialists to interpret
00:02:56
khipu and share their contents with others.
00:02:59
These experts were called khipukamayuq,
00:03:02
and they were trained from a young age
00:03:04
to become savvy state recordkeepers.
00:03:06
Now, it's possible the
khipukamayuq were just uniquely
00:03:09
chosen to become the only literate
people in their civilization —
00:03:12
think of Egyptian scribes being specially
00:03:15
trained to read and write hieroglyphs.
00:03:17
But it's also possible they
were more like musicians.
00:03:20
Like, in theory anyone could
learn to read sheet music for
00:03:23
a viola part and understand what
the notes individually mean.
00:03:26
But to actually convey what the music is saying,
00:03:29
you'd have to know a lot more —
00:03:31
like how to move from a C-sharp on the
00:03:33
G string to an E on the A string and so on.
00:03:36
In other words, it’s possible that the
00:03:38
khipukamayuq had specialized knowledge,
00:03:40
with the khipu serving as
something like a memory aid,
00:03:43
rather than containing all
the information within itself.
00:03:46
We know the Inca used khipu up to
and through Spanish colonization.
00:03:51
And sometimes the Spanish encouraged their use,
00:03:53
for religious or recordkeeping purposes.
00:03:56
But, typical Spanish conquerors,
00:03:58
for the most part they
condemned khipu as idolatrous
00:04:01
and destroyed most of the
ones they got their hands on.
00:04:03
Over time, khipu use died out. Some
Andean villages kept using khipu,
00:04:08
but otherwise… they pretty much
went the way of the dinosaur.
00:04:11
And so that we don’t do the same,
let’s take a quick break for an ad.
00:04:16
Thanks to Brilliant for
supporting this SciShow video!
00:04:18
Brilliant is an online learning platform where you
00:04:21
can develop your skills and deepen your knowledge.
00:04:23
And you can do all of that for a month for free!
00:04:25
Which is something we can
get behind here at SciShow.
00:04:28
They’ve got courses on math, data analysis,
00:04:30
and more, so you won’t run
out of topics to explore.
00:04:32
Whether you’re just starting
or you’re already a pro,
00:04:35
Brilliant has classes designed to build on
00:04:37
your skills and give you an interactive challenge.
00:04:40
You’ll get to solve puzzles
and work with real data,
00:04:42
so you stay engaged while you learn.
00:04:44
To try Brilliant, visit
brilliant.org/scishow, scan the QR code
00:04:48
onscreen, or click on the
link in the video description.
00:04:51
You’ll also get 20% off an
annual premium subscription.
00:04:55
So how do we know so much about khipu?
00:04:58
For one thing, we have the
records of Spanish colonizers.
00:05:01
But we also have a mathematician with
00:05:03
the delightful name of
Leslie Leland Locke to thank.
00:05:06
A little over a hundred years ago,
our guy Leslie figured out that
00:05:10
a lot of khipu are numerical
and use a base-10 number system.
00:05:14
Decades later, other researchers
found that khipu typically
00:05:17
consisted of cords that had powers
of ten, with different types
00:05:20
and placements of knots coming
together to convey a number.
00:05:23
Each chord on a khipu was arranged like this,
00:05:26
with knots in a series of rows.
00:05:29
Those knots typically made up a total number,
00:05:31
with rows divided into sections
— a ones row, a tens row,
00:05:35
a hundreds row, and on and on
for as long as they wanted to go!
00:05:39
In the ones row, you’d find
a long knot with twists in it
00:05:43
representing what the number should be.
00:05:44
So, if a khipu cord wanted to
convey the number 432, the ones row,
00:05:49
representing 2, would have
a long knot with two twists.
00:05:53
But, at least in some cases,
the rest of the cord would
00:05:56
show single knots representing multiples of 10.
00:05:59
So for 432, above the long
knot with two twists there
00:06:02
would be a cluster of three knots
together in the tens place for 30,
00:06:06
and then above that would be a cluster of
00:06:08
four knots together in the hundreds place for 400.
00:06:11
You’d be reading these top-down,
which is how you’d get to 432!
00:06:15
Things got a little spicier
if you wanted to represent
00:06:18
the number one, which used a figure eight knot.
00:06:20
Meanwhile, no knot at all, meant zero.
00:06:23
That would all be on one cord,
sometimes with many more to follow.
00:06:27
Not all numerical khipu are the
same; some had knots that were
00:06:31
more like codes, standing in for
things like places, or individuals.
00:06:35
Like, khipu from one archaeological
site near Lima appear
00:06:38
to use a set of knots representing
the area of Puruchuco…
00:06:42
kind of like an ancient zip code.
00:06:44
About two-thirds of the khipu
we've ever found are numerical.
00:06:47
But the rest are a big mystery.
00:06:49
We tend to call those narrative
khipu, and we think they
00:06:53
contain a lot of information
about battles, history,
00:06:56
and even things like poetry and astrology.
00:06:58
After all, khipu are really complex,
potentially way more so than they'd
00:07:02
need to be if they were just meant
to jog a khipukamayuq's memory.
00:07:06
Plus, there are a lot of
Spanish records claiming that
00:07:09
khipu held information that
went far beyond numbers.
00:07:12
They wrote about how the
chasqui runners carried khipu
00:07:14
containing things like administrative
tasks or descriptions of conflict.
00:07:18
But how was that information recorded?
00:07:21
Was it like written Quechua: Knot Edition,
00:07:24
or more like conveying pictures through string?
00:07:27
Well, welcome to a big dispute among experts.
00:07:30
One theory is that narrative
khipu are semasiographic,
00:07:34
which would mean they're a system
00:07:35
of symbols that contain information.
00:07:37
Think about road signs; we all kinda know that
00:07:40
an image of a moose means to
be on the lookout for gigantic
00:07:43
mammals with antlers that can
roll through your windshield.
00:07:46
Some supporters of this theory argue that
00:07:48
khipu sort of form a seven-bit binary code,
00:07:51
using elements like fiber type and the ply and
00:07:54
spin of a knot to create
distinctions between symbols.
00:07:57
Researchers supporting this theory say khipu could
00:08:00
have stored more than 1,500
units of information this way —
00:08:03
beyond what ancient writing systems
like Egyptian hieroglyphs could.
00:08:07
Another big one argues that
khipu are logosyllabic,
00:08:10
meaning they use symbols to
stand in for words and concepts,
00:08:13
but also sounds — like Chinese does.
00:08:16
That theory rests heavily on an
18th century rebellion effort
00:08:19
against Spain by leaders from the small village of
00:08:22
San Juan de Collata in the Peruvian Andes.
00:08:25
Their communications were
recorded on khipu and kept for
00:08:28
centuries by villagers, until
anthropologist Sabine Hyland arrived
00:08:32
and was invited to view this
Holy Grail of khipu research.
00:08:36
Hyland had only 48 hours to examine the khipu…
00:08:39
and it was a full sensory experience.
00:08:42
The khipu had 14 colors and hair
from six different types of animals,
00:08:47
some of which could only be identified by touch.
00:08:50
While using touch and texture
to guide her inspection,
00:08:53
Hyland and her husband cross-referenced the khipu
00:08:56
with more than 1,000 pages of
historical written testimony.
00:09:00
She finally determined that the khipu
relied on a system of 95 symbols.
00:09:04
Some were phonetic, or represented sounds;
00:09:07
the rest were ideographic, meaning they
00:09:09
represented an idea or a concept.
00:09:12
Hyland got to that conclusion
by parsing out what seem to be
00:09:15
the signatures of former village
leaders from the rebellion plot,
00:09:19
who used a combination of color, fiber,
00:09:21
and knots to form syllables and concepts.
00:09:24
Signatures make sense at the end of
a communication like a letter, and
00:09:28
Hyland already knew she was
looking at messages between people.
00:09:31
That made it a lot easier for her to catch on
00:09:33
quickly when she began to
notice phonetic patterns.
00:09:36
In one instance, a final blue pendant seemed
00:09:39
to signal the local word for the color, "ankas,"
00:09:42
providing the letters K and A
for the end of the name "Alluka."
00:09:45
Alas for Collata's scrappy resistance fighters,
00:09:49
records show their rebellion wound up crushed.
00:09:51
And while decoding parts of the Collata khipu was
00:09:54
a watershed moment in khipu
research and understanding,
00:09:58
here are still some major caveats.
00:10:00
For one thing, they're the only
phonetic khipu ever discovered.
00:10:04
They could be a one-off
regional variation that doesn't
00:10:06
share much in common at all
with other khipu from elsewhere.
00:10:10
Plus, they date to a time
when the Inca had already
00:10:13
interacted with the Spanish
and writing as a concept,
00:10:16
which could have introduced
changes in the way khipu were used.
00:10:19
So, are khipu a kind of writing?
00:10:22
An unsatisfying answer is
that we won't know until we've
00:10:25
actually cracked the code;
for now, those narrative khipu
00:10:29
lying around are keeping
their secrets to themselves.
00:10:32
But a better question might be
whether it actually matters.
00:10:35
Instead of wondering how an empire
could thrive without writing,
00:10:38
maybe we need to reconsider what it
00:10:40
means to administer one in the first place.
00:10:42
Writing is an incredible tool.
00:10:44
But it's not the only way to communicate
00:10:46
or effectively perform administrative tasks.
00:10:49
The Inca are a great example —
00:10:51
they used khipu to do so much
and to do it really well.
00:10:54
Maybe we'll find that there
are many types of khipu,
00:10:56
and some are writing and some aren't,
00:10:58
and some are a cool hybrid of writing and math
00:11:01
unlike anything we've ever encountered.
00:11:03
And maybe they used khipu for telling stories,
00:11:05
and maybe a khipukamayuq had to get involved,
00:11:08
and maybe we'll never know.
00:11:10
But what we do know is that they
came up with a really complex
00:11:13
and fascinating system unlike
anything we've ever seen.
00:11:16
That was enough to help them
run a successful empire,
00:11:19
whether it was writing, or… knot.
00:11:22
[♪OUTRO]
