Shalabhasana

The Gheranda Samhita describes shalabhasana as lying face down with the chest and belly against the ground, the palms planted, and the legs lifted behind. Shalabha is the locust or grasshopper, and the raised legs behind a body pressed to the earth are the insect's hind legs.

The posture is not in the Hatha Yoga Pradipika. It belongs to the set of four prone forms that the later manual adds, with the crocodile, the camel and the serpent, and those four together do more than any other group to distinguish the two texts. The earlier manual's fifteen postures are seats with two arm balances and a reclining rest; the later manual's thirty two open the body out onto its front and add kneeling and standing work as well.

That widening is a fact about the texts rather than an explained development. Whether the Gheranda Samhita records practice that already existed when the Pradipika was compiled, or practice that had grown up in the interval, is not settled by anything in the manuscripts, and no evidence has been identified that decides it. The atlas states the distribution and marks the question as open rather than offering a narrative to cover it. Mallinson and Singleton's Roots of Yoga presents the primary material in translation and is the basis on which the comparison can be made.

The classical posture vocabulary is populated almost entirely by large animals: the eagle, the cow, the bull, the camel, the lion, the peacock, the crocodile, the tortoise. The locust is the one insect, and its presence says something about how the naming worked. What the compiler needed was a creature whose characteristic silhouette a reader would recognise instantly, and a grasshopper at rest, with its hind legs folded high behind a low body, is as distinctive a shape as any of the larger animals offer.

The locust also carries an association the manuals do not use. In agricultural South Asia, as elsewhere, locusts are the swarm that destroys a harvest, and they appear in that role in literature and in ritual concerned with averting calamity. Nothing of that reaches the posture entry, which names the shape and stops. The atlas records the resemblance the text implies and does not import a symbolism the text does not claim.

Reading the animal names as a group rather than one at a time is the useful move here. They are drawn from what a person in medieval North India would see in a day: livestock, birds, temple carvings, the creatures of a river bank and a field. That is a vocabulary of resemblance built from ordinary life, which is exactly what an oral technical tradition needs if a list of thirty two shapes is to be carried in memory.