Kamesh Ramakrishna Aiyer
May 31, 2003
This is collection of speculations on the origin of Samskrtam, the perfected language of the civilization of Ancient India. There are three strands to these speculations:
a) Vedic Thinking
b) What’s “ir” got to do with it?
c) The origin of Constellations
I happened to read a book on “Hindu Mathematics” in the mid-1960s, courtesy of my uncle. It discussed a number of arithmetical and proto-algebraic techniques apparently from ancient Indian books, including the Lilavati of Baskaracharya and Aryabhatta’s “??”. It was quite fascinating and I was very intrigued by the shortcut arithmetical techniques that could be explained using modern notation.
I lost access to the book and was not able to find it for many
years. Around 2001, on a trip to
The earlier book had not claimed to be “Vedic”. By no stretch could Baskara and Aryabhatta, who have been placed in 200-600 CE be considered “Vedic” philosophers. But the preface set me straight.
The author of the preface explained that people who thought that Vedic as an adjective referred to work done by people who might have lived in the so-called Vedic age of Indian pre-history were simply mistaken. The Vedas were intended to be all of knowledge. Anything that was true knowledge was truly a part of the Vedas. He then explained that the Vedas followed a particular style of knowledge discovery, exposition, and explanation and that any work composed in that style should be considered Vedic.
Part of the Vedic technique of knowledge discovery was “revelation” that could come to appropriately prepared sages. Tapas, i.e., strenuous meditation, was a technique of such preparation. The author of Vedic Mathematics, the senior Jagadguru Shankaracharya of Puri, had spent many years in meditation, during which he had discovered how to express many modern discoveries in sutras that followed the style of Vedic sutras. This book was the fruit of those revelations.
My initial reaction to this was irritation, and when I found that many Indians accepted this explanation, the irritation changed to anger. To some extent, I continue to feel cheated by the way this shifts the sands of Vedic thought. But, as I continued to think about it, I came to the conclusion that there was something to be learned from it after all. And the arena that I applied my thinking to was the problem of Samskrtam, the perfected language.
Sanskrit is not a language like any other. Properly speaking, it is “Samskrtam,” the “perfected language”. Note, it is not the perfect language. Some people translate it as “orthodox language”, but it is not that either. Because it is difficult to understand the term, people have accepted it as meaning “formal” or “grammatically perfect” or something like that. And the usual explanation is to point to the number of cases in Sanskrit, the expressive power of the language, the fact that the first generative grammar for any language was written for Sanskrit by Panini and that the concept of generative grammars and the specification of a second language using a generative grammar had to await the research of twentieth century linguists.
As explanations go, this is good, but it does not explain enough. After all, Latin and Lithuanian have the multiple noun case structures like Sanskrit, and other languages have intricate webs of verb conjugations for past, present, and future, their continuous and perfect versions, the subjunctives thereof and so on. So does Sanskrit. But only Sanskrit calls itself the perfected language.
Vedic Sanskrit is different from classical Sanskrit and it is classical Sanskrit that is this perfected language. So what’s perfected about it?
Think “Vedi”-cally. The name did not come by because somebody decided that a new name was needed. It came about because someone decided that they had accomplished something that was significantly better, intellectually deeper and meaningful.
I imagine a linguistics committee rationalizing the language. They create sub-committees for various parts of their job – parts of speech, case structure, tenses, adjectival structure, sentence structure and so on. But this is the obvious stuff – every language learner has to internalize this. Language learning focuses on word-roots and how they are put together.
Human language skills are such that we do not need to go into the structure of word-roots in order to learn its meaning and usage. A vocabulary of between 1000 and 10000 word-roots can be built up without any understanding of word-root structure. Basic English uses about 800 of these; Esperanto uses about 5000 (from a wider, if still Romanic, base). Beyond that, as any language pedagogue will declaim, understanding Latin or Sanskrit or old English helps the learner.
The invention of writing brought about a big change. The rebus principle, which has been used to decode a number of ancient scripts, identifies how symbols representing concepts are used to represent sounds and sound-fragments. The written symbol by itself might represent the concept, but when combined with other similar symbols it represents some part of a sound element inherited from that concept. The rebus principle does not apply to just written language but also to sign languages – a concrete action is idealized into its component memes and is subsequently used to represent some subset of memes in a context. As the context becomes increasingly general, the subset of memes that this action can code becomes smaller until in the most general context, the action represents a single atomic meme. This rebus principle itself may have been a discovery (as has been its application to decoding ancient languages in the last 100 years), but it is possible that it is innate to human or animal development of communication, or possibly even innate to all developed coded communication. In other words, our code elements can carry their history even if that history is not used or not understood.
In the context of language, the single atomic meme can be almost anything. Some candidates have been the syllable, an extended sound, or a sub-syllabic modifier of a syllable. In the extreme, the sub-syllabic entity could be an element of an alphabet.
The committee of rsis charged with producing a language that was perfected would have looked at this as an opportunity for rationalization. They froze two elements in Sanskrit writing:
a) The use of single symbols for the consonantal part and the vowel part of a syllable
b) The systematic construction of the consonantal part of a syllable from spoken elements of speech.
But actually, via the process of Vedic thinking (what else), I conclude that they actually intended something much more. I believe that they applied what I term the “inverse rebus principle” to rationalize how consonants were built up into words. From their perspective, they had produced a language that was built up by mapping the minimal elements of spoken speech to minimal elements of meaning, which included BOTH grammar and lexicon. This was the “perfected language”.
The rebus principle says that sound elements (“phonemes”) are coded by icons/markings (“characters”) that represent entities (“memes”) whose spoken representation (“words”) contain the phoneme in a distinguished way (first or last sound).
We
have to ask ourselves what things differentiate Sanskrit from other languages
of its time. We list the obvious:
1)
The detailed elucidation of the structure of parts of
speech: this rationalizes what is
already visible on the surface. The
generative grammar of Panini lays this out.
2)
The use of a syllabic script.
Other written languages of that time had applied various forms of the
rebus principle to construct languages that had symbols for sounds and parts of
sounds such as consonants and vowels.
However, the Sanskritic poet-performer, paid
for his time by the number of syllables he uttered, needed a script that would
count syllable unambiguously.
We
claim that the following non-obvious aspect of Sanskrit lead to the name:
3)
The inverse rebus principle: writing/sound elements are
explicitly identified with key semantic components that can then be used to define
a grammar for constructing words from syllables. (For the record, let me state that I may be
the first person to have stated the inverse rebus principle). Thus a sound element (a “phoneme”) is used to
represent a root entity (a “meme”) and coded with an icon (a “character”) and
this is used to construct a spoken representation (a “word”) that represents
complex memes. As we discuss below under
“What’s ir doc”, the root syllable, “r”, meaning
“plough” or “law”, followed by “s”,
meaning “controls”, would be “rsi” – one who controls “r”, or the ploughman, while “sri” might mean one who gets the benefit of “r”, the patron.
They
did not create a new language – in that sense, they did not behave like the
Esperantists. Nor did they simplify the
existing language – in that sense, they did not behave like the inventors of
Basic English. Their goal was precisely
the scientific and engineering goals of modern-day computer linguists – they
were trying to precisely and accurately define the language using their
understanding of how it was to be used.
They were also psychologists – they knew that language constrained the
terms in which people thought. They
therefore tried to create a language that would not constrain thinking. They were liberals in spirit – they did not
make religion the basis of the language and Samskrtam
was not the language of the gods, even if that is what it became. And lastly, they based their application of
the inverse rebus principle on a task analysis of concrete human thinking.
They
were also a committee – this means that many poetic flights of fancy had to be
suppressed and pragmatic design decisions played a role. Not everything could be governed by a rule,
because every rule would have exceptions that constrain the applicability of
the rule. This also means that they did
not invent new meanings for existing syllables unless there was no
alternative. My current speculation is
that, fortunately, they did not have to invent very often, and, when they did,
the new word extinguished the old.
The
invention of a syllabic script and the inverse rebus principle both address the
issue of how to construct words from phonemes and what a minimal set of memes
might look like. This is different from
other language grammars that accept word-roots as given. The hypothetical Samskrtam
committee was the first and only language committee that constructed a grammar
for constructing words.
This
was what made Sanskrit the perfected language, deserving of the name Samskrtam. What then
of Panini, the grammarian? Possibly, he was the one
who, much later, systematized the result of what the committee did.
The component structure of Sanskrit syllables introduces a three-dimensional structure – consonants, vowels, modifiers, and extended consonants:
a) The base consonants are k, c, t, T, p
b) The base consonant modifiers – voiced/unvoiced, aspirated, nasalized, emphasized, etc.
c) The base vowels – a, A, i, I, u, U, e, E, ai, o, O, au, ir
d) The vowel modifiers – long/short, aspirated, nasalized
e) The extended consonants – y, r, l, v, s, S, Sh, rr, R
If the linguists committee had been trying to cover all
languages, they would not have limited themselves to the sounds of
Sanskrit. Clicks, guttural modifiers,
etc., are common in
This may be contrasted with the efforts of modern linguists who try to discern differences between languages within a family and attempt to reconstruct a non-living parent language (such as ur-Indo-European). They have come up with detailed analyses of components and how they have changed or evolved, what their meanings are, and how to arrive at the possible pronunciation. Modern day linguists are not trying to re-design an existing language, but that was the charter of my imagined Samskrtam committee. The work of the committee in this regard comes closest to the work of English lexicographers such as Samuel Johnson and Webster who regularized English spelling.
Basically, the Samskrtam committee went way beyond regularizing Sanskrit spelling. Since Sanskrit spelling was based on a syllabary rather than an alphabet, spelling was not the problem. This brings us up against a question – what exact problem with Sanskrit was this committee trying to solve? I will address this later.
That is why the inverse rebus principle is important – it was used to both analyze and rationalize Vedic Sanskrit to construct Samskrtam.
The inverse rebus principle requires that every one of these components be associated with a meme. The rsi committee associated the consonants with memes representing objects, events, actors, actions, and other entities that they appeared to correspond to in Vedic Sanskrit (as it then was). Vowels were associated with memes involved with describing the effects of these entities, i.e, the transitions caused by actors, actions, or events.
The result could not be perfect – there must have been words that did not fit. Syllables were agglutinated to create the right word while the pre-existing word would be tagged as archaic or otherwise inappropriate.
And this was my foray into Vedic thinking. The root “ir” or “ri” meaning “plough” or “furrow” lead the way. It was one of the most pervasive roots from ancient Sanskrit that illuminated the process.
I want to argue by analogy with two phenomena of modern
English spelling was a mess that was not resolved until two
lexicographers, Samuel Johnson in
The
I think both events may have had their analogy.
In the period following the initial invention of a syllabary, it is possible that many different syllabaries were invented and in use. After all, writing is code, and enables one to hide what one is saying as much as communicate what one wishes to say. Different symbols for the same syllable, different ways of showing the combination of syllables, different ways of representing the vowel as modifier would have created a mess, not very different from the spelling mess in pre-Victorian English. A genius, possibly Vasistha, might have lead the way by standardizing the syllables, their pronunciation, as well as their written form.
Indian myth-history speaks of the establishment of Academies of poets and literatteurs who had great moral and spiritual authority within their regions. It is possible that some such Academy was placed in charge of syllable standardization.
It seems clear that the Vedic civilization succumbed to
environmental stresses. It is unclear
what these might have been caused by, but a common element to all such stresses
is population movements, involving the mingling of peoples and their cultures. It is possible that droughts and famines
northwest of
It is possible that Vedic Sanskrit, the language of the elite of the Indus-Saraswati civilization might have been under stress from these immigrants who were adding new words or constructs to the spoken language of the people. The last years of this civilization might well have been spent in creating Samskrtam as one of its defenses against the on-rushing tide of immigrants. That would date this to around 2000 BC to 1500 BC.
The root “ri” or “ir” is sometimes termed a vowel, and is sometimes treated like a consonant. It appears to be associated with the plough in Vedic Sanskrit (i.e. pre-perfected Sanskrit). Its meaning may well be “furrow” or the straight line drawn by the plough on a field. To distinguish it from other forms of the “r” consonant, I will use the token “:r” to represent it.
:r has a long and distinguished use. It refers to a key artifact of settled, plough-based agriculture. That is, by the time :r was in use to describe the result of ploughing, it was a well-understood, important function. The size of the word attests to its importance.
Its generic meaning may have been the long, straight, and narrow furrow produced by the plough. Ideally, the seed must be thrown into the furrow. The ruler of the agricultural settlement might have created the first furrow as a way of emphasizing their primacy within the group. The first furrow of a season, the first use of the plough may have been shrouded with ritual significance. Doing it right was important.
I propose that :r meant “doing it right” or the “right action” or the “straight action”. The symbol for this as writing (or proto-writing) developed would have been one or more straight lines dug into the writing medium (clay or mud or sand) either vertically or horizontally. Another symbol could have been the cross-section of a furrow with two small hills on either side, or some combination of the two.
Based on this, I tried constructing a matrix of the roots I am familiar with and speculated on what such a committee might have produced. I realized as I started that identifying roots and their meaning was work that others have done for many centuries, so after a short foray I stopped. I plan to re-start this after I have reviewed enough existing work that I feel comfortable with the more extreme speculations, but my intent is speculation, not reconstruction. (Please contact me at kamesh@acm.org if the current state of this matrix would interest you).
For some time, I had been speculating on the origin of constellations, the fanciful construction of real and fantastic gods, humans, and animals that populate the night sky. How did they come about? Whose idea was it at first? Which was the first constellation to be named? And so on.
Clearly, there is no evidence, no Rosetta stone that exists that uniquely identifies the inventor of constellations. However, there is no shortage of modern-day historians or other scholars who are willing to claim the credit for this invention on behalf of the Ancient Greeks with a slight nod to the Babylonians and Chaldean astronomers. The arguments usually consist of denigrating anybody else’s possible claim and pointing to the apparently universal acceptance of the apparently Greek zodiac with its Greek names.
Being a materialist, I tend to proceed differently. The first and most important question to be answered whenever an invention succeeds is, “Who benefits?” For the Greek zodiac the beneficiaries are astrologers, i.e., the inventors themselves. This turns the concept of invention into a mockery – as a materialist, this is not acceptable.
Once a concept has been invented, it is often the case that an idealist component of that invention is used to organize the space delineated by the invention. That idealized space often has value beyond the limits of the original invention and can take off. In the case of constellations, once one or two constellations were invented, it was possible for somebody to see that these could be used to mark out the path of the Moon, and the planets, and ultimately, the Sun. Marking out the path of the Moon among the stars in conjunction with the season allows for better management of agriculture and rituals involved with agriculture. This is further improved when the path of the Sun among the stars can be marked out – harder to do initially because the Sun seems to drive the stars away.
The invention of the Zodiac and the other constellations describe the path of the Sun and the Moon and do not perform any other function. The first step in delineating a path consists of identifying a few landmarks – so we should expect that some one or two landmark constellations of stars came first. That is, the “Greek” constellations that fill the modern sky must have come about as a consequence of a prior successful understanding and invention of landmark constellations.
We know that individual stars have been recognized by many
civilizations. The Pole Star by virtue
of its apparent steadiness in the northern hemisphere, especially north of the
tropics (which includes all of
Unfortunately, identifying individual stars does not automatically lead to the identification of constellations. That’s because constellations of stars do not really exist out there – they are names that humans make up for these apparently associated stars. Some groups of stars form strongly suggestive groups – for example, the three stars of Orion’s belt are in a straight line and with Orion’s head and foot form a cross. Sirius is close by.
Another obvious collection of stars is the Milky Way – in a clear sky, this band of stars truly glistens like a path in the sky.
Orion’s belt is an example of a concrete constellation – our eyes compel us to accept the positioning of the belt and the cross. It does not follow that they are a belt or a cross to anybody. That’s what I am calling them. A bushman who makes fire with two sticks might see a fire-stick or a bundle of fire-sticks held together by a cord.
A concrete constellation is collection of stars that are in a pattern compelling to the human eye and brain (i.e., a variety of simple patterns). In addition, if this pattern matches some aspect of the watcher’s life, it has concrete existence for the watcher. It can be named by analogy. When these two conditions are met, we have a named, identifiable concrete constellation usable as a landmark.
My thesis is that the first named constellations must have been such concrete constellations. This brings us up against a real problem – none of the Greek constellations are concrete! Aries for example is not a simple pattern and it takes most people an active imagination to discover a ram in it. This is true of all the zodiacal constellations. We can therefore conclude that the current set of Greek zodiacal constellations are second or later generation names.
To support the thesis, we must be able to identify candidates for concrete constellations in the past. Some simple patterns:
a) Fixedness – if the position of a star does not change from night to night
b) Relative brightness
c) Simple arithmetic patterns such as two bright stars pointing to some other milestone or six bright stars all close together.
d) Simple geometrical patterns such as triangles, squares, crosses, zigzags, circles, clumps, … There aren’t too many of these as they are only a small number of brain-and-eye-compelling geometric patterns.
In addition, if the pattern matches some aspect of the watcher’s life, a memorable name can be created by analogy (as opposed to names of interest to technicians like “Cross-1” or “Triangle-5” and “square-3”!). If we believe that labeling constellations became necessary for planning the agricultural year, the watchers must have been farmers. So, at least some, if not all of the landmark constellations would have agricultural referents.
The Great Bear is also called the Plough. If you look at this collection of stars, the seven brightest ones, and the ones generally visible are shaped like a plough or like a pan with a handle. But this is not the significant aspect of the constellation.
Two of its stars point directly to the Pole Star. Since the Pole Star has a bunch of equally bright stars around it, the “pointers” of the Plough are critical to identifying the Pole Star. Why is it critical to identify the Pole Star?
The Pole Star identifies true north. Practically it is a navigation guide. This is important during group migration and is particularly important for nomads. However, when settled agriculture begins, it become even more important for traders. It is clear from both Paleolithic and Neolithic human remains that trading between groups even across great distances has always been an important feature of human societies.
So the Pole Star is important for both nomadic and agricultural societies. It is possible that it had a different name in a nomadic culture and another name in early settlements that did not use the plough, and that the name “The Plough” is a legacy of some agricultural society AFTER the plough was invented.
The Plough is called the Seven Rsis or Saptarishi in Sanskrit. This would argue that just like the Greeks, the ancient Indian name is not the concrete name either.
The syllable “ri” (sometimes called the vowel “r”) or some variation of it is the name for the plough in ancient Sanskrit. It is one of those unremarked curiosities that the sages of Indian mythology who produced the sacred books are called “rsis”, which could be interpreted as somebody having something to do with the plough.
It is a known fact that many settled agricultural communities had rituals associated with ploughing the land. The first furrow was often cut by somebody who had been ritually purified. Sometimes, sacrifices were offered at that spot, the ploughers, whether men or women would purify themselves before getting involved with the ritual. This was considered necessary for a successful crop.
I hypothesize that a culture that named the constellation the “Plough” might have chosen to call the individual stars after the ritual ploughmen. In that case, the constellation could also have been called the Seven Ploughmen.
Can the word “rsi” also be translated as “ploughman”? Was it a traditional meaning?
Suppose you accept that the traditional Indian name for the
Plough was “Sapta Rsi”. Also assume that when trade between regions
of
The sound “rsi” would have been pronounced as “ursus” or “ursi” by the Greeks (and others? What is the origin of this word?), that is “bear”. Assuming that “Sapta” was translated as “Seven”, this would have meant the “Seven Bears”.
Some clever Babylonian or Greek saw that this made no sense, but that it was possible to imagine a bear made up of the seven stars called the Bears. In particular, the Pole Star is itself part of a collection of seven stars that are also shaped like the Plough, and it was possible to construct a smaller bear around that collection. The concrete idea here is that it was possible to imagine creatures in the heavens marked out by the stars! This was the genesis of the idea of a constellation constructed by imagination as opposed to by concrete metaphor.