Memory Techniques from Vedic Learning [pdf] | Hacker News Reader

Memory Techniques from Vedic Learning [pdf]

artofmemory.com

111 pointskeiferski37 comments

Fascinating. As a child, I semi-reluctantly learned vedic chanting when staying with relatives over the summer. It was very interesting to understand the “algorithms” behind the techniques they used to have us memorize each “shlok”. As the author says, these served an important purpose to transmit knowledge to future generations when writing materials and books were scarce and delicate.

The idea that memorization was used due to a lack of writing material is just a modern and shallow interpretation.

The ancients considered written text as less authentic, and rightly so.

Oral knowledge is spread among the populance and can not be destroyed or distorted by few actors.

It seems increasingly obvious to me that memorization and learning play a really important role in maintaining basic intellectual standards. Everything can’t be googled and even if it could be, retrieval isn’t equivalent to knowledge. Having personally memorized a lot of things via Anki, there is a noticeable difference when you can “chew on” something mentally at a subconscious level.

Any tips for using anki?

I'd suggest spending the time needed to read through the documentation they have. It's not a lot of reading and it helps you understand how it works, what choices it makes and subsequently, it is easier to figure out for yourself how you want to use it.

https://apps.ankiweb.net/docs/manual20.html

Sure!

1. Use Cloze cards, not regular ones. Much more powerful. Also learn the shortcut for applying the Cloze label to a text selection.

2. Always create your own cards. I never download premade ones.

3. Break down things into small pieces. Cards should only be a sentence or two maximum.

4. Use audio, images, and animated gifs. Especially with languages, adding audio files to the cards is a game changer.

5. Create a deck for small random things and add them occasionally. For example, I can understand the alphabets in various foreign languages because I added the letters to Anki a few years ago.

6. Try adding stuff that you wouldn’t normally consider “necessary to memorize.” For example, a daily journal entry and notes from books you’ve read. Spaced Repetition guarantees that you will remember them over time. I try to do this with every book I read.

Is there a way to store the database on my own server instead of the anki servers?

If you never create an Anki web account, I don’t think your data ever hits their servers at all. The app itself is a downloadable. Unfortunately then you won’t be able to sync across devices.

If you google around, I’m sure there is a solution to this problem somewhere.

Now that I think of it, I suppose you could export your decks then import them manually on a separate desktop /laptop. Not sure if that would work on mobile.

I had not heard of that interpretation. Interesting.

> The ancients considered written text as less authentic, and rightly so.

I know some of the ancients were skeptical of the impact of writing on knowledge and culture, but others seem to have held rather the opposite view:

https://en.wikipedia.org/wiki/Verba_volant,_scripta_manent

If you have a long text and it is corrupted, it’s really hard to tell, particularly before the invention of checksums. Is this a hapax legomenon or did someone just misspell a word? Even things that should be clear like “is this part of the text or part of the commentary on the text” get murky: there are several parts of the Bible where scholars will say things like, we think this was meant to be an explanatory comment but it got put in the main text by mistake.

Oral tradition can make these problems better because if one person has a distorted view of the text, they won’t be able to be certified as a reciter by their master. The Indians in particular had various checksum methods, plus a tradition of teaching how to pronounce things, which all would have made the text drift much slower than “a scribe messed this up and the only way to tell is to go to another region and see if their copy has the same words or not.”

Chants were not random data though. They had meanings in Sanskrit, (very similar to poems) and there were grammatical rules (called Chhandas). The deformation happened mostly after the practicing class stopped learning Sanskrit language and blindly memorized them.

you've obviously never played the game of telephone.

If anything, purely relying on oral transmission means that the further that knowledge moves away from its source, the further variations and slight changes that will occur... intentionally or unintentionally. It is the reason why when you try to collect old folk tales or mythological stories, it is difficult to figure out the canonical version of it.

We live in a culture that makes zero effort towards teaching memorization techniques. I’m sure if you had a bunch of memory champions play telephone, the end result would be pretty accurate.

Memory isn't the faculty that makes errors in that game, the errors come from misunderstanding and that's true for any oral and written text, but in the case of writing the misunderstanding doesn't mutate the original.

But ironnically, chants themselves get slightly changed / deformed through generations. Especially because recent generations have not learned Sanskrit, and in fact many people who practiced them, while being Brahmins, had other occupations such as agriculture and accounting etc..

Few days ago, had to explain someone some word that they mispronounced was actually pronounced as something else and what it meant in Sanskrit. (Had to learn some Sanskrit in high school).

The practices & customs are good only as long as the reasons / meaning are known. Otherwise they become superstitions. How many of your Indian priests actually study Upanishads? For most of them it's a job that brings food to the ~~table~~ banana leaf.

> The oral tradition of vedic learning has preserved the entire vedic texts by purely human memory for several generations. This in an enormous accomplishment given the vast corpus of text preserved without any errors creeping in.This has been made possible by a diligently devised systematic scheme of chanting the vedic corpus.

A system of chanting for memory has been used since pre-history across many cultures. See for example Greek epic poetry, Hebrew Scripture, Arabic Koran, Anglo-Saxon Beowulf, Viking sagas.

In ptr-literate societies, bards were able to memorize huge amounts of information and recite to their listeners. Much of the poetry with its rhyme/meter/repetition was written that way to aid in memorization.

That sort of memorization got a lot less as literacy grew and people could just read stuff for themselves, but it was pretty widespread and certainly not something unique to the Vedic tradition.

> This in an enormous accomplishment given the vast corpus of text preserved without any errors creeping in.

This bit is particularly questionable. It's unproveable -- if a text was only transmitted orally, there's no way to compare its modern form to an older version -- and it's probably wrong. The short-range memorization techniques described in the article don't guard against larger-scale errors in memorization, like omitting a "chapter" of a text, nor do they prevent single-word changes, like a student mishearing a word for another similar-sounding word.

I disagree, since it was spread orally and these texts were widespread in the subcontinent you could measure the similarity in different current renditions. If you looked at enough people's renderings you could even get a sense of how the transmission corrupts over time.

apparently brahmins in different parts of India, speaking different languages, from different schools, separated for generations recite these texts identically. while it doesn't prove anything it's plausible that due to these techniques and processes that have been able to preserve these texts verbatim.

I disagree. The key difference is here is that language sanskrit is unlike any other languages in the world. Somewhere during 400 BC. a grammarian named Panini formalized Sanskrit grammar as precise, formally defined 4200 rules. This book called `Ashtadhyayi` or 18 chapters was an exhaustive book of Sanskrit grammar and remains so even today. Panini defined a meta language to define Sanskrit grammar.

What this meant was that language stopped evolving since last 2400 years ago. You can create new words but there are precise rules for it. All this means there is no scope for "dialects", words meanings changing with geography etc. Different vedic chants separated by thousands of miles and 1000 years in time are exactly the same and perfectly understandable in any part of the country for anyone who has studied Sanskrit.

This makes Vedic chants unlike anything English, Latin or Arabic has to offer.

While what you say is true for Sanskrit is true, that Sanskrit is in a way fixed, it doesn't mean Sanskrit stopped evolving, it has certainly evolved since the time of Panini, just evolved in a way that doesn't break the rules set up by Panini. That however is a topic for another discussion as the vedas aren't composed in Paninian Sanskrit.

As for the vedas, specifically Rigveda (which is the only one I can comment on with some expertise) we have what we can call different recensions of the text. These texts, independently transmitted, differ very little from each other. This fact alone strongly indicate that even while they've been orally transmitted for many generations, they are more or less uncorrupted by the act of transmission.

Much more could be written on this subject (and has been). However, english isn't my native language, it is late and this comment could go on and on…short story is: memorization and oral transmission seems to work!

> it has certainly evolved since the time of Panini, just evolved in a way that doesn't break the rules set up by Panini.

This is important. This constraint alone helps oral traditions a great deal.

How does a rigid, comprehensive, and prescriptive grammar ensure accurate oral transmission of linguistic works across generations? How does such a grammar prevent semantic shifts across time?

Grammar is not rigid, it is just well defined. You can write a sentence in a dozen different ways, it is just that the meaning can be derived by a formal process.

Imagine you had to pass down a Java program (through oral tradition) to sort numbers and you had a Java Language Specification that remains unchanged. Now compare that to say English language where and you write an essay on sorting numbers. If both things are passed down orally over few dozen generations which one is likely to keep its form ?

There is error correction built in to the chants. Really advanced stuff, but I expect nothing less from those who authored the vedas.

https://hinduism.stackexchange.com/questions/9517/procedures...

I'm dubious of the claims made on that page.

"Divisional protection" appears to simply be stating that the Vedas are divided into sections. This doesn't provide any form of error correction; it's simply a means of organization.

"Meter protection" states that the Vedas consist of metrical verse. This provides a limited level of protection against certain types of errors (like inserted or deleted words), but not against larger changes.

"Swara protection" and "Mudra protection" state that the chants are tonal, and that those tones are associated with particular gestures. I'm not sure what protection this is meant to provide.

"Patha protection" is a restatement of the same memorization technique described in the OP.

You compare different lineages and use standard linguistic techniques for detecting language change. It’s not perfect, but with long texts you end up doing the same thing because it’s rare for written things to last more than a couple of hundred years without extremely good luck and often it takes just that long for a text to become popular enough for there to be enough copies for you to find and compare.

do you think if, for say, 20 generations, your family practiced these mental exercises... would your DNA/RNA (more % turned 'on'[1]) be primed for increased memorization ?

[1] https://science.howstuffworks.com/life/genetic/genes-turned-...

From my understanding, this is essentially Lamarckism, which was not taken as fact for a long time but has (recently?) been investigated again.

https://en.m.wikipedia.org/wiki/Lamarckism

Counterpoint: Epigenetics is a thing, which is influenced by the parents life style.

https://en.wikipedia.org/wiki/Epigenetics_of_Human_Developme... https://sciencing.com/epigenetics-definition-how-it-works-ex...

>Other studies have looked at the effect of famine. When mothers were exposed to famine during pregnancy, as was the case in Holland in 1944 and 1945, their children had a higher incidence of obesity and coronary disease compared to mothers not exposed to famine. The higher risks were traced to reduced DNA methylation of a gene producing an insulin-like growth factor. Such epigenetic effects can be inherited over several generations. >Exposure of parents to cocaine may affect memory.

I can't find evidence of it, but it seems reasonable to extrapolate that long term life style choices will also affect epigenetics for the better.

There have been similar studies on male fitness level at the time of conception.

There is growing scientific evidence in human and animal studies that should inspire every father-to-be to start exercising more and eating healthier before conceiving a baby. A wide range of studies have found that fathers who are obese at the time of conception pass on a predisposition to their children and grandchildren to be overweight, to develop diabetes, or both.

https://www.psychologytoday.com/us/blog/the-athletes-way/201...

Most of those high-level epigenetics findings are on shaky ground.

Even if true that their children had higher rates of obesity this could easily be due to nurture - e.g. those Mothers could've been more likely to stock up on food at home and feed their children more due to the trauma, with the methylation explanation being post-hoc.

No.

Reply on news.ycombinator.com