Why Sanskrit is commonly referred to as Devabhāsā?
Sanskrit is the only language originating from the Indian sub-continent that carries the moniker "The Language of the Devas". Why is it called that?
नृत्तावसाने नटराजराजो ननाद ढक्कां नवपञ्चवारम् ।
उद्धर्तुकामः सनकादिसिद्धान् एतद्विमर्शे शिवसूत्रजालम् ॥
When you talk to most people from Bhārat about why Sanskrit is called devabhāṣā, they’ll simply say that the Vedas are written in Sanskrit, and the āstikas will say they’re apaurusheya (not of man made origin), hence its cosmic origins make it the “language of the devas”.
Now, it is true that an overwhelming majority of spiritual and related works from Bhārat are written in Sanskrit. But, attributing just this aspect to it isolates the scientific beings who want to look at and analyze data.
Essentially folks like me … or perhaps even you.
Now, data as it is requires preservation and the Vedic tradition is responsible for data preservation. If you’re interested in a quick cursory view of that tradition of data preservation over time, perhaps you’d be interested in this piece.
Anyway, I’ll get back to the Vedic tradition later, but for now, analyzing scientifically we have to also start looking at contemporary works as well. Those are that are outside of the spiritual tradition. Those that fall under what is called ‘Laukika Bhāṣā’ a.k.a Classical Sanskrit.
And, this is where Pāṇini comes in.
His dates are under flux. Most mainstream historians whom I pretty much consider inept (if it isn’t obvious by now) consider him to be around circa 600 BCE, and some attribute circa 300 BCE to his brother Piṇgaḷa, the originator of the chhandas shāstra.
But, some Sanskrit scholars push them both as far back as 2850 BCE (1) as shown here:
But, it doesn’t actually matter from the point of what we’re discussing today. Even 300 BCE (2300 years ago) is just as impressive, although I feel unlikely to be accurate.
Now, Pāṇini as we know is certainly post-Vedic and he devised what is called the Aṣṭādhyāyi or ‘Eight Chapters’. Each of these chapters is divided into 4 pādas or parts making a total of 32 parts.
Pāṇini’s work is considered by many to be the first work which has a ton of similarity to computing as he devised a meta-language to describe an existing language and gave a framework to create new words using meta rules.
Before we get into this, let’s understand how the Aṣṭādhāyi works as there are …
Several Parts Needed to Use It
Akṣara is not the same as an English letter from the alphabet, as it is a sound representation first and a written representation only secondarily. It is the smallest unit of sound that cannot be divided further. The meaning of kṣara is to destroy, break, divide etc. Akṣara is ‘That which cannot be divided’.
The script used to represent akṣaras is called a lipi. Sanskrit has been written in many different lipis including Brāhmi, Shārada, Thai, Tibetan etc besides Devanāgarī which has become the most common.
It has also been written with all the other alphabets of India, except Gurmukhi and Tamil, and with other alphabets such as Thai and Tibetan. The Bhaiksuki, Grantha, Sharda and Siddham alphabets are used only for Sanskrit (2)
When these akṣaras are put together, it becomes an …
The Akṣaramālā - This is primary arrangement and is based on the origination of sounds from the mouth and the various parts used for it’s production. There is a default arrangement by origination.
This arrangement is described rather well in this article by
This default arrangement is very scientific and is also used in the Aṣṭādhāyi. There is also an important alternative arrangement called …
The Māheshwara Sūtras - This is an alternative arrangement of the Akṣaramālā to enable creation of the sūtras. Essentially it’s an encoding mechanism to group related letters into what is called a pratyāhāra.
Therefore these are also known as Pratyāhāra sūtras (3) and there are 14 of these.
१ अइउण्
२ ऋऌक्
३ एओङ्
४ ऐऔच्
५ हयवरट्
६ लण्
७ ञमङणनम्
८ झभञ्
९ घढधष्
१० जबगडदश्
११ खफछठथचटतव्
१२ कपय्
१३ शषसर्
१४ हल्
These 14 are the default arrangements and are also pratyāhāras (encodings) by themselves. Each pratyahārā starts with an akṣara and ends with a delimiting consonant (which is not a part of the letters used). Just like the delimiter is never an actual part of data.
The first sutra is अइउण् and therefore is represented as अण् which means any of the akṣaras अ, इ, उ when used elsewhere. Similarly, ऋऌक् are the ऋ, ऌ vowels represented as ऋक्. When you combine these two, you start with अ and end with क् so it is अक् and represents the 5 short base vowels अ, इ, उ, ऋ, ऌ.
Note: That last vowel ऌ is not to be written or pronounced as लृ which is ल + ऋ and often mispronounced by a majority of people, even Sanskritists.
Using this method of the consonant delimiters as line ends and the first letter using as the beginning point several pratyāhārās can be constructed from the Māheshwarāṇi Sūtrāṇi.
So, you can see that the अच् pratyāhārā combines the first 4 sūtras to make an encoding representation of all the vowels, and the rest of the letters or consonants can be described by the हल् pratyāhārā. So, words ending in vowels or अच् are अच् + अन्त (ending) are called अजन्त (ajanta) words and the ones ending in consonants are called हलन्त (halanta) words.
The arrangement is not nonsensical either. There’s logic behind it.
ञम् are all the nasal consonants ञ, म, ङ, ण, and न and are a vertical representation of nasal across the original horizontal category.
As you can see, many encodings or pratyāhārās can be created simply by using the first akṣara and the line ending delimiting consonant.
But, it doesn’t stop there.
While the line ending consonant delimiters remain the same, the actual start can be from any akṣara.
So while अच् is all the vowels, if you start with the second akṣara इ and make इच्, then it would represent all the vowels but अ. Similarly, ऋच् would start from ऋ and therefore ignore the first 3 vowels.
It’s incredible, how many pratyāhārās can be made. However, there’s a limited number of use cases, and therefore the aṣṭādhyāyi only uses a little over 40 of them.
You have to remember that the language already existed when Pāṇini composed these marvels. He could be considered not just an important grammarian, but possibly the first linguist.
Only after people understood Pāṇinian grammar, did they understand many things in western languages, even Latin. That’s not to say that Pāṇini was the first, as there may have been folks before him. The Vedas also have several Vedāṇgas which include things like sound and etymology.
But, because Pāṇini’s grammar was discovered by westerners when they came to Bhārat, and it provide them the key to several ancient language grammars, he can possibly be hailed as the ‘father of modern linguistics’.(4)
The next and extremely important portion of what it takes to put it all together is …
The Dhātupāṭha - The dhātus are essentially approximately 2000 verb roots which are split into groups called gaṇas based on their common behaviour when used in grammar. There are essentially 10 core gaṇas and they’re named by the first dhātu in the set.
The first set or first gaṇa starts from the verb root भू (bhū)and is therefore भू आदि (bhū and more) and is called भ्वादिगण (bhvādigaṇa).
When you apply meta rules as a series on these verb roots then you get both verbal and nounal forms of words. The root of the verb भवति (bhavati) or happens and the noun भूमि (bhūmi) or ground, earth is the same verb root भू.
Using these ~2000 verb roots and running a few of the ~4000 meta rules on them, one can create billions, even trillions of words. It’s really unlimited. While most languages borrow words from other languages, Sanskrit for the most part, doesn’t need to do so, as a majority of the nouns are actually descriptions.
Let’s take a few simple examples:
खग (khaga) which is the commonly used word for bird is essentially खे गच्छति (khe gacchati) or that which goes in the sky. It can just as easily be used for a plane or a helicopter or anything that flies. Actually even the word for sky which is खम् (kham) actually represents a space and is context dependent too. It can be used for sky, space or even zero in the Mathematical context.
नग (naga) which is commonly used for mountain is न गच्छति (na gacchati) or it doesn’t go/move.
अनन्तं (anantam) which essentially directly translates to endless is used for sky, space, infinity and even Shiva who is Anādi (no beginning) and Ananta (no end) so can be a representation of the universe or kāla (time) which is why Shiva is also called Mahākāla.
Anyway, this leads to a different aspect called Vyuttapatti which loosely translates to what is called etymology but is different because there’s actual descriptive meaning attached to the word in the given context. Although that is a rather interesting property of the language and results in works like sahasranāmas (1000 names).
So, after you’ve understood both arrangements of the varṇamālā, the pratyāhārās of the Māheshwara sūtras, you can then go ahead start looking at …
The Aṣtādhyāyi - These are the meta rules that are put together in an arrangement by Pāṇini of 8 chapters into 4 sub-sections per chapter, making 32 sections in total. The total sūtras are approximately 4000.
There is a tremendous amount of compression in this work because of how it is all arranged. Before we get to that, it’s important to understand the classifications of the sūtras (5) :
Adhikāra or Anuvṛtti Sūtras - These are considered governing sūtras and are headings or topics. These have their adhikāra from where they are mentioned all the way to a certain point and can run through 100s of rules at a time. They’re only mentioned once, but must be mentioned every time they have adhikāra or authority. There’s a sub-section called anuvṛtti which only pushes a portion of the sūtra (one or more words) down.
Saṅjñā Sūtras - These are definition or naming sūtras. They are the simplest part and are in the front portions of the work. The very first sūtra is:
वृद्धिरादैच् which is compressed and expands to वृद्धिः आत्-ऐच्. ऐच् is essentially ऐऔच् which are the ऐ and औ akṣaras and आत् is just आ. These 3 akṣaras namely आ, ऐ and औ are being named as वृद्धि (Vriddhi). Another example is
मुखनासिकावचनोऽनुनासिकः which expands to मुख नासिका वचनः अनुनासिकः which defines अनुनासिक or nasal consonants as anything said with both the mouth and nose together.
Paribhāṣā Sūtras - These are interpreting sūtras or metarules.
Vidhi Sūtras - These are operational and can be called prescribing rules.
Niṣedha Sūtras - These are negations or exceptions.
Niyama Sūtras - These are restrictive
Atidesha Sūtras - These are extension sūtras which extend some property or operation.
The traditional method of learning the Aśṭādhyāyi is:
कण्ठस्कथिकरणम् or Memorization of the text
व्यवस्थाबोधः or Understanding the structure
सूत्र निर्मिति or Building the formulae using the rules
अवगमनम् or Understanding
प्रयोगः or Usage
These are achieved through the following:
पदच्छेद -Splitting into separate words
विभक्ति वचन - Understanding the grammatical cases of the split words and verbal forms.
समास - Compounded words often accompanied by change in meaning
अधिकार and अनुवृत्ति - Already explained earlier
अर्थ or understanding the meaning
उदाहरण or ability to give examples
सिद्धि or understanding of the procedure of sūtra application. Usually this is a confirmation of the understanding
Let’s take some examples to understand the structure.
एचोऽयवायावः which breaks into एच: अयवायावः and further into एचः अय् अव् आय् आवः(6)
The pratyāhārās are:
एच् = ए, ओ, ऐ, औ ।
अच् = अ, आ, इ, ई, उ, ऊ, ऋ, ॠ, ऌ, ए, ऐ, ओ, औ ।
The sūtra states that
When followed by a स्वर, the letters of the एच् प्रत्याहार are respectively converted as
ए → अय्, ओ → अव्, ऐ → आय् and औ → आव्
वने + इति = वन् ए इति = वन् अय् इति = वनयिति
साधो + इति = साध् ओ इति = साध् अव् इति = साधविति
Let’s now take something with adhikāra.
3.1.1 प्रत्ययः has adhikāra from 3.1.1 to 5.4.160
3.1.2 परश्च has adhikāra from 3.1.2 to 5.4.160
So when constructing the metarule, even though it’s only mentioned once at 3.1.1, it will apply to every line where it’s adhikāra is valid.
3.1.91 धातोः has adhikāra from 3.1.91 to 3.4.117
So, what does this mean. Let’s take
3.2.123 वर्तमाने लट्
To build the rule, the adhikāra sūtras will take position as follows:
प्रत्ययः परश्च धातोः वर्तमाने लट्
With understanding this is re-arranged and then a verbal form is affixed to it to complete the sentence, like so:
धातोः वर्तमाने लट् प्रत्ययः परश्च भवति (5)
This is the complete construction of one metarule.
When there’s an anuvṛtti, then only one or two words of the sūtra are taken into the metarule.
As you can see, there is a huge amount of compression going on.
But, why devabhāṣā?
Pāṇini essentially created a language construct which makes it impossible for the language to die, so it ends up being immortal by its very nature with the ability to produce new words.
Just like you can use a combination of a limited set of keywords and data to produce entire programs in programming languages, you can use a combination of 2000 verb roots and 4000 metarules to produce new words in Sanskrit.
When people ask me, if Sanskrit is suitable for programming, I tell them that, it’s never really been investigated despite all we see on the internet, but on the other hand, Sanskrit a la Pāṇini is programming.
Now, let’s get back to our original claim.
In the Amarakosha we have several synonyms for devās if which some are:
अमरा निर्जरा देवास्त्रिदशा विबुधाः सुराः (7)
The first one is अमरा which means without death or mortality. Sanskrit has the same properties due to Pāṇinis codification and therefore devabhāṣā.
And, finally getting back to Vedic tradition, it also is focused on data preservation via oral transfer of knowledge. There were several serious measures made to ensure there’s no corruption of the dataset. And, the walking talking databases were people who took the pain to learn all the mnemonic devices and checks and balances and memorized large portions of the corpus.
That is another way Sanskrit was also kept alive and not permitted to die, again making it immortal.
References
Ancient Indian Leaps into Mathematics - Yadav and Mohan
https://omniglot.com/writing/sanskrit.htm
https://ashtadhyayi.com/shivasutra
https://en.wikipedia.org/wiki/P%C4%81%E1%B9%87ini
The Tested Easiest Method for Learning and Teaching Sanskrit - Pt Brahmadatt Ji Jijnasu
https://ashtadhyayi.com/sutraani/6/1/78
https://ashtadhyayi.com/amara/1.1