Gematria | Hacker News Reader

Gematria

en.wikipedia.org

46 pointsuserbinator47 comments

Gematria is featured a lot in scott alexander’s Unsong (serialized fiction that was published in via blog) http://unsongbook.com/

Very quick summary is that Kabbalah is real, and corporations set up content farms to brute force search for names of god so they can copyright and sell them.

It’s a really fun read (author probably also has a hacker news account, so if you see this comment thank you for writing such a fun book).

This sounds a lot like Neal Stephenson's Snow Crash, in that they're a mash-up of (1) taking an ancient tradition literally and (2) a critique of capitalism and modern culture.

Throw in a bit of the movie Pi for good measure :)

With The Bible Code in 1997, Pi in 1998, The Matrix in 1999, Waking Life in 2001, the turn of the century was a trip.

So a few things to think about with regard to this and related concepts (numerology in general, and really with different arguments about almost any other kind of divination):

Understanding _what math is_ is a very deep concept that we didn't really have a firm foundation of until the 19th century.

Numbers in many languages were represented by letters. They understood that they could find true things about the world by manipulating those symbols and letters using rules (ie, doing math). It was not at all clear then (and perhaps not even now) why that should be the case, and it's not especially obvious on its face how to know which manipulations will yield useful results and which won't. A lot of _real_ math they were doing seemed like magic to them, and indeed were sometimes used in the context of magic or in theurgical rituals. And there's nothing wrong with numerological methods as mathematical manipulations necessarily, it's interpretations of those results that are where they go wrong. These people were not _stupid_, in a lot of cases they make quite sophisticated arguments to justify the validity of it, and invalidating those arguments required invalidating many of the assumptions they held about the nature of reality (many of which are still held by many people today, intuitively, if they haven't thought deeply about it).

It's also the case that divination had a practical value in terms of being a way to get out of analysis paralysis and bike shedding. You can't decide what to do, you go to some theoretically neutral third party and he does some random process that produces an answer and you just go with that. It doesn't really matter what the reasoning is behind it or if it "works", the important thing is that it's a decision that stakeholders respect and nobody gets blamed for following it. It's no different than flipping a coin, really.

Also, if you squint, word2vec and related concepts are sort of a very sophisticated version of this.

Bingo - also in the case of astrology, you had a way for someone who’s smart enough to do astronomical and calendrical calculations for practical purposes to advise in a low-personal-stakes way.

I was told members of my prior family were Cabalistically fascinated by the number 3, looked into almost everything in life for patterns of 3, lived in a menage a trois.. Maybe this 'assign a number to a word' thing is more pervasive than people think?

Q-Anon is into gematria: https://podcasts.apple.com/us/podcast/episode-168-the-mutant...

It's kind of interesting that there's no mention of this on Wikipedia and your comment is downvoted, there's been a lot of media coverage of gematria in q-anon. Even if it's a misuse of gematria (which it probably is), it's one of the main things people will commonly use the word for these days. Many will have heard of this use and not know about the ancient traditions.

Yeah, I'm not sure why this got downvoted. Are people upset that I mentioned Q-Anon? That Q-Anon has adopted gematria? That gematria is on HN? We'll never know.

I dunno, but it's not even that q-anon adopted gematria, i think there is some gematria or gematria-like stuff literally in some of the early posts from q which probably sent them down that particular rabbit hole.

I was really confused about why such a generally anti-semetric group was so into kabbalah. I think it might be antisemetic reasoning as well, this is how the ((globalist)) conspiracy signals each other so they need to understand it or some shit.

Sounds like you’re guessing while assuming negative intent. Is there any practical way to attach your assumptions to cold, hard, touch-grass reality? If not, why make these comments?

I mean, despite any antisemitism, ancient Hebrew carries weight to a Christian audience, since there is overlap in scripture and tradition.

I think they're not thinking of it really hard.

Your comment comes off as negativity toward the concept via tying it to a detestable group

I've got nothing against Judaism, Kabbalah, gematria, etc. I just pointed out the simple, documented fact that a bunch of Q-Anon kooks are into... their version of gematria, where JFK + NWO + WHO = 17 + 45 + 420 + 69 or whatever. They even have a mobile app for this!

Have they tried decompressing the main shuttle bay?

I propose the following challenge: find the smallest integer number that, when written in English words, equals the sum of it's letters (following gematria rules i.e. A=1, B=2 etc).

   T  W  O  H  U  N  D  R  E  D  A  N  D  F  I  F  T  Y  O  N  E  
  20+23+15+08+21+14+04+18+05+04+01+14+04+06+09+06+20+25+15+14+05 = 251

If you don't include "and" it's more interesting. I don't think there is an answer because the value of a number grows much faster (exponentially) than the value of its gematric sum (linearly), so once you've checked there's no example below some limit (one million is enough), there can't be any examples larger than that.

Maybe you'll find some with the scheme "x score and y" =)

I don't want to write out all of the written-out numbers, but I wrote some code:

    letterToValue = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
        .split('')
        .reduce( (acc, letter) => { acc[letter] = letter.charCodeAt()-64; return acc; }, {}) 

    getValue = (word) => word.split('')
        .map(letter => letter.toUpperCase())
        .filter(letter => Object.keys(letterToValue).includes(letter))
        .map(letter => letterToValue[letter])
        .reduce((acc, val) => acc + val, 0);


    ['ZERO', 'ONE', 'TWO', 'THREE', 'FOUR', 'FIVE', 'SIX', 'SEVEN', 'EIGHT', 'NINE', 'TEN']
        .reduce((acc, word) => { acc[word] = getValue(word); return acc; }, {});

    ['ZERO', 'ONE', 'TWO', 'THREE', 'FOUR', 'FIVE', 'SIX', 'SEVEN', 'EIGHT', 'NINE',
    'TEN', 'ELEVEN', 'TWELVE', 'THIRTEEN', 'FOURTEEN', 'FIFTEEN', 'SIXTEEN', 'SEVENTEEN', 'EIGHTEEN', 'NINETEEN',
    'TWENTY', 'TWENTY ONE', 'TWENTY TWO', 'TWENTY THREE', 'TWENTY FOUR', 'TWENTY FIVE', 'TWENTY SIX', 'TWENTY SEVEN', 'TWENTY EIGHT', 'TWENTY NINE',
    'THIRTY', 'THIRTY ONE', 'THIRTY TWO', 'THIRTY THREE', 'THIRTY FOUR', 'THIRTY FIVE', 'THIRTY SIX', 'THIRTY SEVEN', 'THIRTY EIGHT', 'THIRTY NINE',
    ]
        .map(word => getValue(word));

I really wanted to find an answer for your challenge but it seems there aren't any answers using the regular 1-26 gematria and numbers written in their ordinary English form (without "and"). However, I did find some answers using a "zero-indexed" gematria ranging from 0-25, and with this other English gematria described here on Wikipedia: https://en.wikipedia.org/wiki/English_Qabalah#R._Leo_Gillis'...

Python 3:

>>> num_strings = ['Zero', 'One', 'Two', ... 'Seven Hundred Thirty Two' ... ] # I tried 0 to 1000. num_strings generation script sold separately

>>> def find_gematria_matches(values): return [i for i,ns in enumerate(num_strings) if i == sum(values[ord(c)-65] for c in ns.upper() if c.isalpha())]

>>> find_gematria_matches([0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25]) # "zero-indexed" gematria

[213]

>>> find_gematria_matches([5,20,2,23,13,12,11,3,0,7,17,1,21,24,10,4,16,14,15,9,25,22,8,6,18,19]) # R. Leo Gillis' Trigrammaton Qabalahn

[232, 242]

No number does, at least by the usual way of writing them.

https://gist.github.com/mminer237/926bc453a9e09b2a2803305c5e...

Even in the millions, few gematria sums get above 800, so it's not as hard to figure out as it seems.

Two numbers are one off, though I won't spoil them here.

There is a match if you include "and" though.

I was bored and played around with various letter numbering schemes, like reversed order, or digits of Pi (or E), or the value of the letter multiplied by how many times it appears in the string and so on, and so forth. I got bored eventually, but it is a great fun (for certain minds, I guess, but still).

Word embedding used in NLP feels similar to this (https://en.wikipedia.org/wiki/Word_embedding)

See also, The Number of the Beast: https://en.wikipedia.org/wiki/Number_of_the_beast

An interesting synchronicity, because I just ordered Roots of the Bible[1] by Friedrich Weinreb[2] which teaches exactly this.

[1]: https://www.amazon.com/gp/aw/d/1621388034/

[2]: https://en.m.wikipedia.org/wiki/Friedrich_Weinreb

There’s a similar concept in the Islamic world called Abjad (https://en.m.wikipedia.org/wiki/Abjad_numerals), ebced in Ottoman. The building dates of most public buildings (mosque, fountain, etc) are embedded in the text written on them. It was heavily used for numerology and divination, too.

I'm surprised there isn't a section on Chinese - I thought there was similar number-meaning conventions there too: https://www.fluencypending.com/learn-with-me/introduction-to...

It's a fascinating psychological technology for saturating your brain with associations between anything and anything else. Eventually you can start to read numerical coincidences or synchronicities into anything - bus route numbers, time of day, order numbers... strangely reminiscent of Arronofsky's Pi, where a mathematician straddles the boundary of number theory and numerology [0]. All of it tied back to some notion of divinity, or whatever notions of "the most important things" exist in your head.

I've sometimes wondered if occult instruction in gematria, numerology, tables of correspondences, relating names, colours, symbols, scents, etc... to mythological figures is intended to produce this hyper-associative, nonlinear form of thinking - a different mode of cognition than the typical linearized, rational sort of thought, that is able to leverage our ability to quickly and subconsciously make inferences, associations, and snap judgments about phenomena, instead of using slow, deliberate, deductive forms of reasoning. Kahneman's system 1 vs. system 2 maybe approaches what I am getting at here.

The associations that surface may give insight, if only dimly, into one's psychological state - how they feel at the moment, or what meaning a particular event has for them. A sort of looking glass for the mind, similar to the way one may lay out a tarot spread and not seek to prophesy the future, but speak some symbolic language understood by this hyper-associative faculty to see what meanings the practitioner projects from their subconscious and out onto random phenomena; or the way one may consult the I Ching to partake in an "experimental dream" - in which, instead of trying to isolate and remove the subjectivity of the observer, the experiment embraces the fact that observer and observed are inextricably linked in this game of trying to ascribe meaning to phenomena. Where "synchronicity takes the coincidence of events in space and time as meaning something more than mere chance, namely, a peculiar interdependence of objective events among themselves as well as with the subjective (psychic) states of the observer or observers." [1]

If you find yourself bereft of meaning in this life, a practice of gematria may provide you with the semantic infusion you seek - though it may not be exactly what you bargained for. You can really blow yourself up if you get too deeply into this.

[0] https://www.youtube.com/watch?v=d1IzNKIHhp0

[1] http://the-iching.com/extra/iching_jung_foreword

> a different mode of cognition than the typical linearized, rational sort of thought, that is able to leverage our ability to quickly and subconsciously make inferences, associations, and snap judgments about phenomena, instead of using slow, deliberate, deductive forms of reasoning.

I mentioned this in my other comment, but you don't even really need to get this deep about it -- it gives people a justification for making essentially _random_ decisions. It's not always the best thing to do in many contexts to make the "logically correct" decision. There are many cases where being unpredictable is a benefit, or people doing new or unusual things can lead to progress, or just getting over the hump of not being able to decide what to do.

Yes, Julian Jaynes also expands on this point in his book, "The Origin of Consciousness in the Breakdown of the Bicameral Mind" and the section on "Divination." He posits that forms of divination, augury, sortilege, etc... were invented as attempts by newly "post-bicameral" man to once more hear the "voices of the gods" that would give them authorization or assurance that their course of action was correct; an attempt to recapture the bicameral era, when "free will" as a concept did not even exist, and all of one's actions were in fact orchestrated by the gods (where Jaynes presupposes that what were referred to as "gods" were actually psychological phenomena).

Is it possible to use this technique for compression? For example, if we created a dictionary of the most commonly used words / phrases, then we could drastically reduce space. You would still need to store the dictionary but that's a one-time cost.

I mean, this is how a lot of compression works already. It's just that there isn't generally a global dictionary, the items to be compressed are scanned for large common sequences and then those are tokenized into a dictionary.

https://en.wikibooks.org/wiki/Data_Compression/Dictionary_co...

Do you mean LZW?

https://en.wikipedia.org/wiki/Lempel–Ziv–Welch

Yeah that sounds like what I was thinking of, thanks!

Without code examples, it's hard to follow.

Is it just

   AutoTokenizer.from_pretrained('mispar/hechrechi')

or am I missing something?

In the method described, it could also be:

'mispar/{gadol|katan|siduri|bone'eh|kidmi|p'rati|ha-merubah ha-klali|meshulash|ha-akhor|mispari|ne'elam|katan mispari|musafi}' as well as 'otiyot/be-milui', and 'kolel'

You can then also have #includes or pre-processor directives such as:

#include Atbash (opposites), Albam and Achbi (two 11 letter alphabets), Akhas beta (three groups [778]), Ayak bakar (letter*10), Ofanim (last letter replace), Avgad (letter++)

And if you really want to, you can do all the other parts of 'math' like polynomials, curve fits, blah, blah, just with letters.

Why is this a HN post?

And why did someone ask for "code examples"?

I am so confused.

HN is for anything intellectually interesting, especially if it's uncorrelated with other topics. This seems to me to fit fine.

https://news.ycombinator.com/newsguidelines.html

Fair enough, I certainly find it interesting. I guess I have no complaint, it was just a bit out-of-place seeming.

HN didn't seem like the kind of crowd to be into superstitous practices, but I obviously read the room wrong. Third or fourth occult-related posted I've seen this week. Stuff is interesting but also really popular with paranoid schitzophrenics, q-anons, and such, so I wonder if it's a bit unhealthy for the mind.

> Fair enough, I certainly find it interesting

I rest my case! :) - it's actually surprisingly common for people to see things that they personally find interesting but which they feel don't 'belong' on HN - but if it's personally genuinely interesting, that's the essence of what belongs on HN! Of course tastes vary dramatically from person to person, but hopefully there's a wide and unpredictable range of something for everybody.

> Third or fourth occult-related posted I've seen this week

That's almost certainly one of those random subsequences that doesn't feel random - they happen all the time. There has always been an occasional trickle of such posts to HN, and that's fine. More than that would not be fine, of course.

Does it not remind you of a certain computer science concept that maps strings to (possibly identical) numbers?

I'm not a computer scientist or a cryptographer but I assume you mean a hash table? Enlighten me :)

edit- or something related to LLM tokens?

oh was it compression? i think thats the most technically apt comparison?

Why is this nonsense in HN? Don't people know about it?

It's this sort of scam where people would find numeric significance between a word/name to try to convince you of some power or god. Since they have a number to represent pretty much anything there's always something. It's a scam and a bad one.

Please let's not go into flamewar about this stuff - there's no curiosity value in that.

I had a relative reveal that the DNA proves God the creator because the four letters used in DNA coding (ACTG) somehow transform into YHWH or something like that.

I wasn't convinced...

Reply on news.ycombinator.com