The many schools of jyotisha

A "school" in jyotisha, the Indian tradition of sky-reading, is not a building. It is a set of default moves: what you look at first, what counts as a connection, what you may say afterwards. Two people working from different defaults, neither careless, are speaking different grammars over the same sky.

The mainstream is Parashari, named after the big Parashara handbook but not limited to it. It reads each of twelve houses through its ruling graha, the planet said to own that house. Its timing comes from the Vimshottari dashas, stretches of time each handed to one graha. This is what most people in India mean by Vedic astrology, and what nearly all software speaks.

Jaimini is a different instrument. In Parashari, which graha stands for the self is fixed. In Jaimini it is assigned by measurement: whichever sits at the highest degree becomes the atmakaraka, so Saturn may hold a post in your chart that Venus holds in mine. Its text is in sutras too compressed to use without a commentary, and the commentaries disagree.

Tajika reads the year rather than the life. Its instrument is the varshaphala, a chart cast for the sun's yearly return to its birth position. It never pretended to be homegrown: its name reaches Sanskrit through a Persian word meaning Arab, and its vocabulary is Arabic in Sanskrit dress.

Nadi is two things sharing a name: a body of Sanskrit texts, and the Tamil palm-leaf practice where a seeker is said to be matched to a leaf inscribed long ago. Two different kinds of claim, not to be argued as one.

Underneath everything sit the computation traditions, where the argument is arithmetic rather than meaning. Hence two almanacs printing different festival dates.

If this were physics, one description would have won by now. It has not. The essay's conclusion: a system like this is a lens, not a measurement, each school defined by what it leaves out. A prophecy in four incompatible dialects would be worthless. A way of thinking that comes in several is just a field with more than one method.