We restore books nobody has reprinted in a century. We rebuild languages that are down to their last speakers. We put myths back into print in the languages they were told in.
So when we shipped a dream dictionary, a few people asked what that had to do with anything.
It is the same work. Here is why.
What we actually do
Strip the projects back and there is one job underneath all of them.
Somebody scanned a book in 1903 and the scan is a mess. The typeface is unfamiliar, the OCR software has never seen it, and what comes out the other end is thousands of words of wreckage: an `n` read as an `x`, a long `s` read as an `f`, whole pages of a language the machine does not know it is reading. Fixing that is slow, unglamorous, and mostly invisible when it is done well.
Underneath the fixing is the reason for it. These texts carry things that exist nowhere else. A grammar written by somebody who could still hear the language spoken. A folk tale collected from a person whose grandchildren did not learn the words. Once the last copy rots and the last scan stays broken, that is genuinely gone, and no amount of goodwill brings it back.
So: restore the text, so the language survives. Restore the language, so the stories survive. Restore the stories, because they are where a culture kept what it knew.
That is one chain, and every project we run is a link in it.
Where dreams come in
Here is the part that surprised us.
Work on this material long enough and the same images keep arriving. A serpent. A well. A threshold you should not cross. A bird that knows something and will not say. They turn up in Irish, in Norse, in Greek, in Egyptian, in stories told by people who never met, never traded, and could not have copied each other.
That is strange enough to be worth taking seriously.
Jung had a name for it and called it the collective unconscious, and you do not have to accept his machinery to notice the thing he was pointing at. You can explain the overlap several ways. Shared inheritance from further back than any of these traditions. Shared human circumstance, since everyone has a body, a night, a mother, and a death. Shared borrowing across a longer contact history than anyone has mapped. Take your pick, or hold all three.
What is not in dispute is the practical part: you did not invent the images you dream in.
Nobody does. You dreamt a snake, and the snake arrived carrying several thousand years of freight you never agreed to. Some of it came from a book you have not read, through a story you were told by somebody who half remembered it, in a language that may not even be the one you speak. The vocabulary was handed to you. You just supplied the sentence.
So a dream is a private event made of public material. It is genuinely yours, and almost none of the parts are.
That is what a dream is a window into. Not the future, which it does not know. Not a diagnosis, which it cannot give. It is a window into the stock of images a culture handed you without asking, and which you are now using to think with at three in the morning.
Which means the dictionary and the restoration work are the same enquiry. Restoring the texts is restoring the vocabulary people dream in.
Why nobody else builds it this way
Look at the dream sites that exist. The biggest has not published anything since January 2020. The next one down has thousands of entries and not one citation. They tell you a snake means transformation. They do not tell you who decided that, or when, or whether anyone disagreed.
You are asked to take it on faith from an anonymous voice.
That is the same disease we deal with in language work. Somebody types a word into a box and gets an answer with no source behind it. No dialect noted. No attestation. No indication of whether one person said it once in 1840 or whether half a county says it now.
We do not think that is good enough for a language. It is not good enough for a symbol either.
Four readings, not one verdict
So our dream dictionary does not tell you what your dream means.
It reads every symbol four ways. Freud, who thought a snake was about sex and anxiety. Jung, who thought it was about transformation and the part of the mind that is not human. Modern dream research, which notes that snakes appear in dreams worldwide and has a fairly unromantic explanation for why. And myth, where the serpent is medicine in Greece, wisdom in the north, and a warning in a garden.
Four traditions, disagreeing, on the same page. Each one named. Each one carrying the text it came from, so you can go and check.
Then it gets out of the way. You collect the symbols from your own dream and it shows you where the readings overlap, and asks you a couple of questions. It never hands down a verdict. It never predicts. It does not diagnose.
We built it that way because the honest position is that nobody knows what your dream meant. Everyone in that list was guessing. Some were guessing carefully, from a great deal of evidence, which is worth something. None of them were guessing about you.
The same discipline, four times
The text restoration says: here is a page the machine read wrongly, here is what it actually says, here is the scan so you can check us.
The language work says: here is a word, here is where it is attested, here is who recorded it and when.
The myth work says: here is a story people stopped printing, restored, with the translator named and the edition stated.
The dream dictionary says: here is a symbol, here are four traditions that read it differently, here is the source behind each one, and the meaning is yours to settle.
The subject changes. The discipline does not. Cite it, or do not publish it.
What it is not
It is not an oracle. It will not tell you that a black dog means a death in the family. If you want that, it is easy to find and it is free, and you will get it delivered with total confidence by somebody who has no idea who you are.
It is also turning out to be useful for something we did not plan. We are adding a second section to each entry, written for writers: what a symbol does in a story, how genres use it differently, and what happens when you turn it inside out. Every common dream has one now and we are working through the rest. That came straight out of our own work. When you are restoring a text, or writing a new one, you need to know what a raven has meant to readers for a thousand years before you decide what it will mean in your chapter.
Go and break it
It is live. Nine hundred and eighty seven entries, every one carrying four sourced readings.
Search a symbol from something you dreamt. Read all four. Notice which one you want to be true, and then notice that you wanted it.
That gap is the interesting part, and it is the same gap you meet when a language turns out not to have the word you assumed it must have. Both are moments where you find out that the equipment you were thinking with came from somewhere, and that somewhere had its own ideas.
We are still adding. If we have got something wrong, or missed a tradition that belongs in there, tell us. That is how the rest of the work goes too.
Try it: the Dream Dictionary, the full A to Z of symbols, and the sources behind every reading are all in the navigation.