# How to share knowledge to a future human civilization who doesn't know our languages?

Consider that I want to let an object to share knowledge to a civilization living in a far future (~10.000 years after today). It can be a book or any object which can store information, must be functional in this future and must be autonomous (it requires no other devices to work or to be read).

This civilization has the same habitabilities as us (vision, touch, same IQ, etc) and the same knowledge as us at the beginning of Writing (they understand that the object gives information and must be decrypted).

Every present languages will have disappeared and unfortunately there is no Rosetta Stone or similar object to help to translation.

Can we make this object and what would it look like?

Edit: This question has been initially asked for the language and communication aspects, but responses concerning the object itself and the way to preserve it are also welcomed.

• I recommend reviewing our written-material and preservation tags. You may find those questions to be of interest. – a CVn Sep 6 '16 at 15:46
• You say "unfortunately there is no Rosetta Stone or similar object to help to translation", but that's exactly what you're going to need to provide in order for the future people to be able to read your content. Give them as much help with understanding the message as you possibly can. You'll probably need more content for that than for the message itself. – Simba Sep 7 '16 at 10:17
• some answers from that question Million-year-old records address problem of reading information after long time of storage, in therms of overcoming comprehensive barrier – MolbOrg Sep 7 '16 at 12:29
• Many posts repeat or re-invent things explained in detail in Benford’s Deep Time book. You really should read that. Here is an early report, now used as an introduction, that made him want to write a whole book, free to read. – JDługosz Sep 7 '16 at 13:42
• On a much smaller scale, this very problem exists with storing nuclear waste – BruceWayne Sep 8 '16 at 4:44

A Visual Dictionary

A visual dictionary would be fantastic to find. Under nearly every word describing a real-world object a picture would be found, along with a written description. The written descriptions would take awhile to work out, but once you have catalogued enough correlations between bolded words and images, you could start to read the descriptions. Then you could begin to read the descriptions for more abstract entries and from there you're off to the races. Including a few children's stories after the dictionary probably be beneficial as well, as they feature simple sentences good for those learning the language.

As a physical relic the book would be ridiculous in size, but luckily it's not likely paper could survive for 10,000 years anyway. You would almost require it to be stored in metal, possibly on a set of one of those archival/permanent CD's (M-Disc's). How people would read the CD's in 10,000 years of course is another problem, but if you left them a few old CD readers they could probably backwards-engineer the whole thing.

Note: Even M-Disc's have a limited lifespan of around 1,000 years, but if the goal really was to preserve use I'm sure that could be significantly expanded with correct packaging (hermetic sealing, possibly with a vacuum).

• I don't have enough reputation yet to upvote your response but I like it. :-) – roipoussiere Sep 6 '16 at 14:18
• Hard to be sure. The Domesday Book is now almost 1000 years old and has survived most of its life without serious modern archiving methods or technology. Not sure what kind of "paper" was used, but likely a good "rag" paper rather than anything more modern. Real quality paper properly preserved can last a very long time. – user2338816 Sep 6 '16 at 22:09
• I'm sceptical of M-Disc's claimed 1000-year lifespan, and even if true there's no guarantee that they'll be using the same data storage techniques or file formats in 1000 years. I have enough trouble trying to convert my grandfather's files from 1980s wordprocessors into something that current versions of MS Word can open. The future engineers would need a complete computer to reverse-engineer how the discs worked (good luck getting that to survive) - and a disaster may have knocked society back to pre-computer tech by then. A physical book has the advantage that anyone with eyes can use it. – anaximander Sep 7 '16 at 10:51
• Children's stories are terrible for learning a language. It turns out that children actually know the language, they just cannot read. Additionally the stories are typically written in a language different to what people normally use, adding an additional layer of difficulty. When you try teaching a foreigner your language using children's stories you will notice that they are way too complicated to be useful. Fortunately there already are texts specifically made for people who don't know the language well, use those instead. – nwp Sep 7 '16 at 11:09
• I have not mentioned that the manual should be compact. A huge encyclopedia in thousand of volumes put in a big box should be technically possible: for instance a paper version of Wikipedia takes 7,473 books (printwikipedia.com/#/about), which is fine. For my opinion clarity beats density in this case. – roipoussiere Sep 7 '16 at 21:25

What you want is the Rosetta Disk which is part of the Long Now Foundation's 10,000-year library. This disk holds 13,000 pages of language documentation; is a physical, durable object; and requires only optical magnification to read.

The goal of the Rosetta Disk project is to hold a record of human languages, and holds 1,500 languages and descriptions and usages of each, as well as maps for the usage of the languages. Instead of that, you could instead hold two or three languages, and a pictographic bootstrapping to teach those languages. Why not just one language? Depending on how their language is structured, it may be much more difficult for them to learn English than, say, Japanese. So I'd pick 3 very different (structurally) languages and everything completely redundant.

After the pictographic bootstrap that teaches the basic uses of the language and provides a simple dictionary, you could even then strictly replicate the Rosetta Disk data for the language, since it's meant to be a reasonably complete record of the use of that language. After that, you should still have thousands of pages worth of space to transmit whatever data you wanted to this civilization.

• Good idea, but the object must be autonomous (it requires no other devices to work or to be read). I edited my question to define this word more precisely. – roipoussiere Sep 6 '16 at 16:47
• Oh and I'm not sure the Rosetta Disk is understandable by people who don't know any of the languages written on the disk. – roipoussiere Sep 6 '16 at 19:04
• @roipoussiere It wouldn't be inherently understandable, which is why you need the pictographic bootstrap. The bootstrap will have to teach them enough of the language that the rest can be taught using the language itself. As for the "no other devices", it wouldn't be difficult to scale up the disk. The Long Now disc is 3" and requires 650x magnification; to get the same amount of writing space without magnification, the disc still only needs to be 162.5 feet across (or a 10 foot high wall for 2,000 feet). Figure out the level of magnification these people have access to, and size accordingly. – Azuaron Sep 6 '16 at 19:33

Preservation: Send capsules in space

Content: very small pictures in hard material or a Rosetta disk

An Earth-style planetary environment is pretty corrosive, so if you want to send something 10,000 years into the future it might be best if it didn’t spend that intervening time on the planet at all. Shoot off your hardened message on a rocket that will send it looping through the solar system a few times before it decays and comes back home.

Going to space solves a lot of problems you will otherwise have. Sure you could make a time capsule and fill it with a non-reactive substance and bury that, but your still on the planet. You still are going to have corrosive gasses and liquids hitting the outside of the capsule. You still will have plants and microbes and animals pulling and prying at the outside of the capsule. The capsule might end up in the middle of a lava flow and melted or sucked down a mid sea trench and folded. Being on the planet is dangerous.

10,000 years is, as far as stellar mechanics go, not that long of a time. You just need to make sure your initial setup and orbit is correct to eventually decay at the right time. As for powering it or steering it... you shouldn't need to; if the message is a physical object and the orbit is set correctly, it will drift on its merry way till it has its too-close encounter with the atmosphere again. If your afraid of loosing the one capsule... send ten...or a thousand. You could time them so one attempts to make landfall every hundred or so years.

Your message might get some nicks and dings from micro-meteors and be bathed in some stronger radiation then would be normally comfortable, but besides that there’s not much that’s actually going to corrode your message, so you could send almost anything. If your medium of transition is also hardy, you could expose it to the near hard-vacuum to remove any remaining corrosive gasses or freeze any active chemicals that are accompanying it. You could also send it in a capsule packed with non-reactive gas (such as Argon).

A big enough capsule with a light enough payload could probably aero-break enough to survive reentry even if your parachute is in a sorry state (or maybe it’s designed to smack into the ground with only aero-breaking…to create a sort of a “I’m over here” crater).

How the message makes it down Without burning to a crisp? the same way most other stuff makes it back down; Just think about the shape of the actual pod (probably a sphere [imagine one of those man sized inflatable gerbil balls]) and how much that sphere will weigh (not much.) The upper atmosphere might ablate some or most of its surface as it enters and slows down, but because there is not much mass there, it will slow down fairly quickly.

As for the message itself, you could literally just use pictures; really small ones in monochrome on some sort of material that will retain them; work up to a Latin style (readable, but not really spoken) language from there. Or as per the suggestion by Azuaron; a Rosetta disk [I had forgotten they were a thing].

• How do you steer the thing? Are you expecting your rockets to remain functional after 10,000 years? Do you send it on a trajectory so that it returns in 10,000 years without any steering(sounds like some risk of getting totally obliterated on re-entry, unless you have parachutes that you expect to work in 10,000 years). And why not just bury the capsule on earth, if the capsule will prevent against corrosion? – Salmoncrusher Sep 6 '16 at 15:48
• @Salmoncrusher Updated, I think I address those points but would be happy to expand on it again. – Marky Sep 6 '16 at 16:04
• I like the idea because it works even for 100.000 or more years, and the landing by itself is an information which says "Hey, look at this object" and it's easier to find than a Rosetta disk on earth, among the billions of tons of objects we have today. – roipoussiere Sep 6 '16 at 16:21
• @Marky, it's an interesting idea, and it certainly works for the purposes of worldbuilding. It seems to me that surviving re-entry is a much greater engineering challenge than surviving in the ground for a while, but space does have the advantage that it's out of reach of future humans. Is the capsule radiation-proofed? Even visible light degrades paintings over just a few hundred years, and there's a lot more ionizing radiation in space. – Salmoncrusher Sep 6 '16 at 19:38
• @Salmoncrusher that's partly why you used an embossed surface; even if it 'bleaches' its still there physically. Re-entry is a problem mostly because objects are traveling bleeding fast compared to the planet/atmosphere. With a low enough mass object and much higher shock tolerances you can slow down much much faster and avoid a lot of the re-entry burn. This IS an engineering challenge no doubt, but one I believe is well within our technology horizon, and easier then dealing with the corrosive nature of being on/around a planet. – Marky Sep 6 '16 at 20:54

In my story idea (never written fully) it is a sentient robot. It is dorment in the library cache which was burried for thousands of years (see Deep Time: How Humanity Communicates Across Millennia by Gregory Benford for the burried vault I have in mind).

Once found, the robot follows the guy back to his village and lives as a child, learning the culture and language. The robot develops rapidly though and “grows up” in a month or two, and then is ready to teach people about their history and serve as a translator for all the knowledge in the vault.

A distinct feature of my story is that the robot was very grassile, looking like a sticks coming out of a block for the torso. The idea is that it is deliberatly weak looking so will not be used for labor; and is a “cute” doll toy readily accepted as friendly.

Of course, for complex machinery such as working AI computers and library storage, not to mention the robot body, to function after being left for 10,000 years the storage is not purly passive. The nanotechnology self-repairs like living tissue, and needs a power source. The mothballed machines need to fight back entropy and a radiothermal generator provides a trickle of electricity. (Though how to preserve whatever is not part of the question, I mention it for completeness since cybernetics is more delicate than passive artifacts.)

I really reccomend Benford’s book as preliminary research if you are interested in this subject.

• I really like the idea of this robot! Do you think future people will feel cuteness in the same way as us? – roipoussiere Sep 6 '16 at 19:43
• Yes, because cuteness is instictive so we take care of our young, and domesticated animals have evolved to be cute so the standard has been the same for the 50Kyears we’ve had puppies, and traits are common among all mammals not just humans. – JDługosz Sep 6 '16 at 20:17
• Unfortunately with no current technology(and I'd doubt that close future will bring us much better one), no robot would survive too long; I'd set a safe 1000 years limit, but I guess that it would break much sooner. – MatthewRock Sep 7 '16 at 12:59
• @MatthewRock self-repairing nanotechnology. – JDługosz Sep 7 '16 at 13:27
• You should post that in an answer. – JDługosz Sep 7 '16 at 13:34

Real-world scientists and linguists have previously debated this.

Pictograms and reference to common physics/maths and elements would likely be a universal place to start, as used aboard the Pioneer space probes.

Whilst this is by no means a complete communication method, it is designed to show Humanity, our location, etc.

A Pictographic method using mathematics would, most likely, be the most language-independent method.

• How future people could understand mathematics pictograms? I read some articles about the Pioner Plaque and it seems this was a bad idea, since 99.99% of the present population don't understand it without explanation. – roipoussiere Sep 6 '16 at 14:07
• I believe the assumption is that the basic concepts of maths SHOULD, in theory, be universal in how they apply to Physics etc. If a civilisation is capable of the Space flight needed to retrieve the Pioneer, they should, again in theory, have enough understanding to interpret them with some thought and provide some common ground (however basic) for further communication. – sequoiad Sep 6 '16 at 14:19
• Remember though, this is designed for communication with aliens, rather than humanity in the far future as your question stipulates, but it is definitely possible as a method somehow for storytelling. – sequoiad Sep 6 '16 at 14:22
• If we found an alien artefact out there in the solar system, we would spend a LOT of time trying to decode it. The pioneer plaque contains a lot of information in a way that a scientifically advanced species of alien would be able to figure out, in time. The ability of a human who doesn't have a lot of knowledge of astrophysics to determine the plaque's meaning within a minute or so of seeing it really isn't the yardstick the efficacy of the plaque should be measured by. – A Simmons Sep 6 '16 at 14:34
• Interesting that the illustrations are meant for anonymous strangers but are not suitable for public viewing here? – JDługosz Sep 6 '16 at 18:07

Some sorts of "Rosetta Stone" cannot go out of date for lack of any known language.

The most obvious is an annotated picture, as used for teaching children in elementary schools. All that is needed is for the recipient to form correct associations between the strings of symbols and the entities in the picture. If you are going to be sending to assumed human beings a long way in the future, you could do worse than dump the content of some learn-to-read books for young children into the package, as a backup to a carefully constructed pictorial dictionary for adults.

As for the medium that might survive 10,000 years: the simplest and cheapest is glazed clay. Visit any museum and you'll see that writing and drawing on glazed clay pots and plates can still be read milennia later. In fact, the only degradation is corners where the items have been broken, but if the pieces are not scattered, a future archaeologist will reassemble the pieces and it's unlikely that much will have been lost. If you are trying to send information forwards in time, you'll include multiple copies.

If you have the money the 21st century can do a lot better than simple clay in terms of resilience. But this might backfire. If you create something that becomes unique and "impossible", a future barbarian may destroy it as sorcery, or scatter the found pieces to the winds as gifts. If it's "just" commonplace old broken pottery, he'll likely leave it in the hole in the ground where he found it. Maybe even pre-break your plates?

You can also get a much higher information density stored on a clay plate than can be read with an unaided eye. Think of those perfectly preserved potters' fingerprints, and they were not even trying. If your target is a future that has magnifying glasses or microscopes, then you can put a lot of data into what a primitive will think is a patch of random texture in a picture, by a simple process of embossing. This can also create a lot of identical copies of the data very easily.

We know more about Sumeria than many (most?) civilisations since, because they wrote on clay tablets which they then threw away. Even some of the un-fired tablets (their ephemera, private letters) have survived the milennia.

• In Benford’s Deep Time he points out that you want the monuments carted off to put into museums. So the Rosetta stone or key or whatever was small enough, not cemented down, and present in several copies. – JDługosz Sep 6 '16 at 18:04
• Old civilizations have written on clay because they did not have got better. A book written in metal will not be treated as sorcery, because our current civilization uses it a lot. I think archeology in 10.000 years will be very difficult since we will let after us a big very durable trash, so get useful data will be a big challenge. – roipoussiere Sep 6 '16 at 19:24
• @roipoussiere you are assuming that in 10,000 years, civilisation will be maintained. If there is a collapse, a metal book may have greater value to an illiterate as a raw material than as an incomprehensible text. Broken crockery will not (or if recycled as hardcore, it may still preserve its information content). – nigel222 Sep 7 '16 at 10:09
• @JDlugosz have the exhibits that were in museums in Bagdhad, Mosul and Damascus fared well of late? If you want information to carry across one or more dark ages, I am convinced that buried pottery is the best bet. It won't be removed from holes in the ground until archaeology is once again a science. Mildly radioactive location markers might be a good addition. – nigel222 Sep 7 '16 at 10:14
• @immibis I don't know if archaeologists do dig in places with mildly elevated radiation levels (probably not since that's not a thing for past civilizations), but as soon as the civilization is aware of radiation, someone will dig up this places to see what's going on, and when they find the pottery, they will call the archaeologists. By the 3rd time this happens, they will pick up the pattern and start actively searching for other such locations. – xDaizu Sep 8 '16 at 7:01

This is a real and legitimate problem, which those disposing of radioactive waste have been wrestling with for a long time.

Their problem is a lot simpler, and yet harder: how do you make something look to future archaeologists like they really, REALLY don't want to go digging in there?

You could just try to hide it. But that's not gonna work. People will always find it. And if they think something's buried there, they're gonna want to know what. But the radioactivity symbol won't mean anything to them.

So you need to signpost it. So they spent a million dollars trying to figure out good ways of doing that, and came up with a 351 page report: Expert Judgement on Markers to Deter Inadvertent Human Intrusion into the Waste Isolation Pilot Plant (Sandia National Laboratories report SAND92-1382 / UC-721, p. F-49) http://downlode.org/Etext/WIPP/

Basically, your problem is an architectural one. Only something on that scale will have a good chance of lasting thousands of years.

It took a million dollars to create a message as simple as "don't dig here", but given enough funds, I suspect a simple dictionary is possible, with diagrams and a simple language.

To ensure survival, though, I'd go for spreading dictionary-plaques throughout the world as building tools, to the point where they become ubiquitous. So, for example, subsidize bricks and concrete slabs and formers to have at least one word or phrase on each, with a simple accompanying diagram to explain it. Subsidized, they will become the cheapest building materials, and every building becomes a dictionary clad in plaster. Hopefully the people will support the movement, and will be happy to do this, though many will of course subvert the idea and put other messages on their bricks, too. But without subsidizing, they won't be so cheap, so won't be so ubiquitous.

Building materials get reused a LOT, and some few constructions last a LONG time, so odds of survival of the message are pretty high.

The danger there is that some other entity with more money decides that their message is more important to preserve, and thus spends more money on subsidies. Say, a church. So your message could get muddied by other signals.

• Will the idea of easily understood signs by the masses scale to a rosetta stone meant to be deliberatly translated by future scholors? – JDługosz Sep 7 '16 at 13:36
• It took a million dollars to create a message as simple as "don't dig here". I suspect a simple dictionary is possible, with diagrams and a simple language. I'd go for spreading dictionary-plaques throughout the world as building tools, to the point where they become ubiquitous. So, for example, subsidize bricks and concrete slabs and formers to have a word or phrase and an accompanying diagram, so they become the cheapest building materials, and every building becomes a dictionary clad in plaster. – Dewi Morgan Sep 7 '16 at 14:46
• you should write that up as an answer! – JDługosz Sep 7 '16 at 19:32
• @JDługosz Yep, I plopped it up into my main body text of the answer :) It made it more on-topic, I think, so thank you for a good penetrating comment that forced me to improve my answer! – Dewi Morgan Sep 7 '16 at 20:27

## Language

Plenty of suggestions of Rosetta Stones. The problem with a Rosetta Stone is that it has to go from an unknown language to a known language, and we don't have a known language for the future people. Also we don't really care about sharing our language or languages with them. So previous answers suggesting saying the same thing in several languages are simply wrong.

What you want is a self-starter language books. Show pictures of a person, show them doing things, and apply words to describe the concept. Repeat the same words for different people doing the same action, and you start getting a noun-verb grammar established. Your "known language" is pictures, and we can (by the OP's question) assume they can understand pictures.

The important thing here is sharing information though, not sharing the language. The language is just the means of sharing information, and you don't really care how you do it. So you want to use a language which has no irregular elements whatsoever and keeps a strict order to its written grammar. English would be a spectacularly bad choice for this. Esperanto is probably your best bet.

You don't necessarily have to provide any context for how words sound, either. "This sequence of marks means 'run'" is enough to start with.

## Maths, physics, chemistry

Maths has been covered already by other answers. The basic concepts will be common to anyone who ever does maths.

Basic physics will generally follow from maths, with diagrams of forces, etc.. These are also universal concepts. Language to describe the concepts may help, but the basic equations and diagrams may be sufficient.

Basic chemistry will probably require a periodic table, diagrams of electron shells, and tables of valences. All universal concepts again.

## Biology and ecology

Teaching language will have to give some illustration of what our idea of a person looks like. Later on, you can start describing variations in what a human being looks like, how our limbs move, and our internal body structure.

This may also be a good place to start talking about the rest of the world, with pictures and descriptions of animals and plants.

## History

Ancient history is full of civilisations who created artifacts and structures but left no record of how they lived. If we've got language sorted, we may want to provide some details of events in human history and the development of our knowledge as a species. It might also be interesting to describe how we lived at various stages of history - aspirationally, this can describe "when you can do this, then you'll be able to live like this too".

History can also help with engineering too. Engineering needs iterative improvements. We can give them lists and descriptions/blueprints of various inventions, in approximate order of mastery, and the crucial concepts needed to master each one.

From human history, we can then step backwards to evolution of species, which is another essential concept. From there, backwards to how the Earth was created, and backwards from there to the start of the universe. Some science knowledge will be needed for this, and this will give a context for taking science a step further.

• Upvoted for Esperanto. – Murphy Sep 7 '16 at 12:58
• @Murphy would not recommend esperanto as it relays on prefixes and postfixes(?) in changing roots meaning. It could be second language or the third maybe. Not necessary to be a spoken language too, which might (or not) make easier to construct such. Layers of languages might be a good chose - as pyramid of languages. Or language or method of description of information which have pyramid base which is probably not the case with most existing languages. – MolbOrg Sep 7 '16 at 14:14
• @MolbOrg Esperanto does change the verb for tenses, but in predictable ways. As you say though, you can pick (or create) your constructed language for this. But your "second or third language" or "pyramid of languages" is flat wrong for the OP's purpose, because we're only aiming to convey information. We are talking to future engineers and scientists, not future linguists. Having to teach them one language is a necessary evil for conveying information. More than one language is flat wrong for this purpose. – Graham Sep 8 '16 at 9:43
• If there is one future guy, truing to decrypt information of entire civilization, he will have other problems besides languages. It will be no a book, but a big library of books(digital, paper, stone whatever). Starting from 0 common language, 0 cultural inheritance - if hundred languages is needed to overcome that problem let it be. If they do not have 100 men equivalent power - to learn those 100 languages to learn that one evil language to read the books, they do not have needed power to process that library any way. Natural languages have some of pyramid properties - kids learn mama papa – MolbOrg Sep 9 '16 at 0:29
• @MolbOrg Exactly one language is needed. 10 or 100 or 1000 or all the languages that ever existed in the world will not pass that information along any better - in fact it will make matters much worse because your future people don't know that they're all different languages. It will take a linguist to initially decode the language, sure, but after that the information will be useable by anyone. And you can easily build up a vocabulary by showing pictures of people doing things, or of the things themselves, next to the words. I already covered that in my answer to the OP. – Graham Sep 9 '16 at 8:42

Assume that our descendants are not stupid.

Provide them with lots of material to work with. Don't just give them 1 book, give them a library of information.

Give them basic, easy to decode info. Numbers are good to lead in with.

1
one
•

2
two
••

3
three
•••


Etc

once they have numbers include the info from lots of childrens books with pictures.

Gradually include more advanced info on things like how to make glass and microscopes.

Include in your archive much more dense storage medium that can be read with the instructions from the earlier ones, first microscopes which provide access to instructions on how to build basic computers and the protocols to read stored data in highly durable formats.

At that stage we can give them digital data. you could even include instructions on algorithms and our own cutting edge methods for machine translation. We could include vector maps of all known words in all known languages and their relationships which would immensely help future linguists. Once they have one language we don't need to just give them language, we can give them methods to translate any language.

There's the added advantage that at this point you could even include instructions on how to build a speaker and digital data to play the spoken language.

What would it look like? Like a vacuum sealed time capsule filled with layers of more and more dense and more and more complex information.

• @JDługosz words can be viewed as having a position relative to each other in an n-dimensional space. In different languages the words for the same things tend to have similar locations and be located similar distances from other words. "dog" is a similar distance and offset from "cat" as "perro" is from "gato" in spanish. the x/y graph shows a simplified 2-dimensional version of this for the numbers 1 to 5 in english and spanish. technologyreview.com/s/519581/… – Murphy Sep 8 '16 at 11:08
• That reference should be part of the answer. Introducing the idea though is probably distracting. Unless you want to cover it in detail, why not just show a normal translation dictionary? – JDługosz Sep 8 '16 at 14:21
• because they're not really the same. The idea is that you don't just have to give them a dictionary like they're used to. you could give them entirely different cutting edge ways of approaching translation that might even allow them to translate texts for languages not in the archive at all. – Murphy Sep 8 '16 at 15:08

Consider a smallish cube, self powered and capable of projecting interactive holograms. If your far-future race communicates verbally, the cube would be equipped with a language-learning routine, so it would learn the language of the future rather than the inverse. Holographic projection would first allow picto-graph type communication until the cube learned the language and then it could begin to communicate verbally.

Preserve it inside one of those molybdenum aluminum boride mononiths (thanks to Pommerbot for the science!) and make it a "quest" to get inside the monolith with all the associated rewards of solving the quest.

Pictures are good to get started with Math — you could show numbers from 0 to 100 with the appropriate number of blobs with each numbers. Then you can define +, − , =, etc. with examples.

For chemistry, you can use the periodic table. There's only one way to lay it out (not counting reflections, rotations, etc — but with numbers it's obvious). That gets you names for the elements; with a few pictures you can show what formulae like H₂O mean.

For (rather coarse) timestamps, you could use astronomical data — precession of the celestial pole, and distortion of constellations due to star movement.

For other stuff, the visual dictionary idea mentioned above sounds good.

As to format, check the museums. Ceramics seem to survive better than a lot of metals (although gold would probably work if no one melted it down in the interim). So you could either go for lots of redundant copies so that at least some survive, or go for a big stone monument. Or put a copy on the moon and spray big arrows pointing towards it over the lunar surface :).

The best way to make something survive around 10000 years would be to create a molybdenum aluminum boride monolith, or the most durable material available. You would need to create a modern (by our terms) Rosetta stone. You would need to carve what you wanted to say into the stone in as many different languages as possible. You might even want to utilize pictures to help explain what you are attempting to leave behind. If you used something technological than you assume that the future civilization would be able to use the technology. For example, in 100 years very few people would be able to listen to cassettes, because everything would have the newest way to play something. In short, there would BE no cassette players.

• One language would be sufficient, other languages would only bring confusion. Choose a rather simple language and provide information in such a way that someone without knowing the language could learn from it quite fast. – MatthewRock Sep 7 '16 at 13:31

The nuclear industry has done a lot of work for this.

I refer you to the OECD's Preservation of Records, Knowledge and Memory Across Generations bibliography. Which contains far more information than I can easily read or summarise.

I heard of an research that is a silicon disk with visible instructions on how to build a microscope, and then readable though a microscope the instructions on how to build a computer to read the digital data embedded in the disk.

I think we underestimate our power to understand things. We can't give up because we have limited understanding of what happened 1000 years ago. The records of that time was in paper and clay, no much information to work with and most deteriorated so much.

For example, a modern encyclopedia with our writing and pictures will be understandable by context. In the text below each figure there is a Fig X and in the text somewhere has an Fig X... it will not take long for a reader to understand that is a reference and by the text near the pictures get some words out. Coincidences between the text near the reference and the picture caption will help too. A page with lots of pictures of birs and a boldface bird write on it. It probably will take decades to decipher... but it will allow a better understanding of our language than the understanding we have of ancient languages.

Not counting videos. Videos made to this purpose can relate our words (in subtitles and by voice) with objects and actions.

The problem is if this time capsules would be open before its time, stolen or misplaced somehow. To make sure a future civilization reads them... they should be everywhere and therefore should be cheap to make.

Which I think is not a problem, we are entering in the age of small machines and 3d printers. Few decades from now this will be very cheap to do.

Prime numbers is a good way to make contact

Actually if you are ever put in this situation yourself, one of the quickest way to show that you're smart is to display prime numbers. Pick white and black stones, and put a white stone at every prime number :

.xx.x.x...x.x...x.x...x

prime numbers do not depend on the counting basis, if they can count, even if they have 23 fingers, they should understand instantly.