This morning, I encountered a viral post on a social media site claiming that ancient Indian philosophers knew about gravity before Isaac Newton. As evidence, the author cites a translation of the Vaiśeṣika Sūtras (VS), a very old work of metaphysics. (It goes back before the start of the Common Era, though just how far back is controversial.) The meme-writer goes on to state that Isaac Newton didn’t discover the laws of nature, but the author of the VS did.
Shocking if true, but the translation is misleading, and the post is simply wrong. The full story brings up some questions about translation and some worries about fact-checking in the age of AI.
Newton and gravity
The concept of gravity, as Newton defined it, is not something that existed in premodern Sanskrit literature. Newton’s Philosophiæ Naturalis Principia Mathematica set out (as its title suggests) mathematical principles for things such as how bodies can act upon each other at a distance. He distinguished between weight and mass, arguing that gravity is a force that acts “in proportion to the quantity of solid matter, and whose action is extended everywhere to immense distances, always decreasing as the squares of the distances” (The Principia: Mathematical Principles of Natural Philosophy (1687), 3rd edition (1726), trans. I. Bernard Cohen and Anne Whitman (1999), General Scholium, 943.)
The term “gravity,” for Newton, became a technical term, embedded within mathematical conceptions of the laws of nature. There is no evidence that philosophers on the Indian subcontinent had this precise conception of gravity.
Premodern India and gurutva
However, they did have theories about the lawlike regularities of nature, of how movement operates, how fundamental particles combined to make wholes, and so on. And there is discussion in these very old texts, the VS, of how heavy, or weighty, objects fall when there is nothing obstructing them. Holding a stick, I can use it to spin a potter’s wheel. Were my hand not in contact with the stick, it would fall.
The word for “heavy” in Sanskrit, guru, is a word that has a range of meanings, not unlike “heavy” in English. A person can be “a heavy” in English slang, meaning that they are weighty or important. Likewise, a guru can be a weighty or important person. In the sūtra text, the word that was translated as “gravity” is related to this word: gurutva means “being guru.” Essentially, this word means “heaviness” or “weightiness.”
However, there are no discussions of mathematical relations between force, mass, and weight in the VS. These premodern philosophers observed objects falling—like sticks, water, arrows, pots, etc.—and inferred that there were regularities in when they fell or moved. For instance, when there is no conjunction (saṃyoga) between the earthen parts of an embankment holding water back, water would flow.
However, although it’s tempting to translate gurutva, “heaviness” as “gravity” (and indeed, many early translators and esteemed Sanskritists have done so), in a modern context, this might suggest more—and a different—theoretical apparatus than the Vaiśeṣika Sūtras has.
(If, at this point, you think I’m channeling modern Western sensibilities, see an article by Swaraj making a similar point—a magazine that can hardly be charged with being leftwing or Western!)
What does this have to do with AI translation?
So, we have a misleading translation of an old Sanskrit text and a misuse of it for popular and political purposes. My concern is that the rise of machine translation and ChatGPT-like writing bots could make these kinds of errors increasingly difficult to identify and remove.
Machine translation, like Google Translate, uses a bunch of existing texts, known as a a corpus, and text-predictive capabilities to produce translations from a source to a target language. If previous translations in the corpus have used “gravity” for gurutva, it’s likely that this translation will be reproduced. Further, translation bots often ask users to judge whether a translation was useful or accurate. If users rank “gravity” highly, this will be more likely to be employed in the future. And ChatGPT bots might deploy the term “gravity” as a synonym for gurutva without nuance. As these bots become more common producers of Internet content, could there be a self-perpetuating cycle in which bot content is part of the new corpus scraped for further bot content?
As of today, ChatGPT answers my question, “Did ancient Indians know about gravity?” similarly to how I have above:
The concept of gravity, as understood in modern physics, was not explicitly known to ancient Indians in the same way it is understood today. However, ancient Indian scholars made significant contributions to the understanding of the natural world, including discussions related to the nature of matter, motion, and the Earth's gravitational force.
However, ChatGPT then goes on to state that a seventh-century CE Sanskrit mathematician and astronomer, Brahmagupta, did have an “explicit explanation of gravity as a force of attraction between objects,” and that he “stated that all objects are attracted towards the Earth due to the force of gravity.” Without a source, this claim is difficult to verify. In fact, Brahmagupta’s Brāhmasphuṭasiddhānta is a sophisticated work of astronomy, containing advanced mathematical concepts. However, the term translated as “gravity” is gurutva. Just what did Brahmagupta think gurutva was? How did he think it functioned in relationship to force and mass? (Did he distinguish mass and weight? Did he apply mathematical concepts to them?)
I would give you a link to an English translation of his text, but no complete translation exists. Likewise, scholarly work on his massive treatise is scant, much of it focusing on his mathematical theories (he is credited as the first mathematician to work with zero, for instance). You can look at the first volume of his text in Sanskrit on archive.org, with a thorough introduction, here.
From some scattered translations of his text, though, we can see that Brahmagupta’s cosmology is not precisely Newtonian: “The earth's sphere is surrounded by the orbits of the Moon, Mercury, Venus, the Sun, Mars, Jupiter, and Saturn, and (the universe) ends with the orbit of the constellations; it is shaped by the good and bad deeds of creatures” (David Pingree, “Brahmagupta, Balabhadra, Pṛthūdaka and Al-Bīrūnī,” , Journal of the American Oriental Society 103 (2) (Apr. - Jun., 1983), 357.)
Without more information, I’m skeptical (as usual) that ChatGPT has gotten things right, or that I want to use the word “gravity” for what Brahmagupta is talking about. I want to wait until I know more to decide.
“Gravity” then and now
Maybe, though, Brahmagupta did have some more robust conception of a gravitational force than the VS. In that case, perhaps the term gurutva should be translated as “gravity,” with a little note to say that it’s not exactly Newtonian gravity.
After all, we do use the term “gravity” differently than Newton did—modern physics has made advances since his time and we haven’t coined a new term for the phenomenon modern science describes. That’s because there’s enough in common for us to call both things “gravity.”
So, back to the original problem I began with: does VS 5.1.7 refer to Newton’s Laws of Motion when it says saṃyogābhāve gurutvāt patanam?
Definitely not! This short sūtra simply says “When there is no conjunction, there is falling, because of weight.”
In context, the conjunction is between someone’s hand and a stick. When the hand doesn’t support the stick, it will fall because it is heavy. The VS does not have a robustly mathematical conception of regularities due to force and mass, although it certainly does employ methods of observation and reasoning to consider how movement works.
Did ancient thinkers in the Indian subcontinent discover the Laws of Motion before Isaac Newton? Again, strictly speaking, no.
However, whether we should use the word “gravity” to translate gurutva in some texts is a different question. If readers can distinguish between “gravity” in Newton and “gravity” in Einstein, then perhaps they could distinguish between “gravity” in Brahmagupta and Newton. That may be an increasingly big “if,” these days, though.