Knowledge fieldProfessional practice · long-form study

Next useful moveContinue: When the birth clock cannot bear the question, ask the moment.
← The Grahvani Journal

A rectified time must predict events it was not tuned on.

Given enough freedom, any birth time can be adjusted until known events appear to fit. Rectification becomes evidence only when the candidate time is fixed first and then tested on events that played no part in choosing it.

01प्रश्न

Why rectification is needed

Hospital clocks drift, the recorded minute may be the cry, the delivery or the paperwork, and family memory blurs over decades. Meanwhile the finest techniques move fast: the ṣaṣṭyaṃśa lagna changes every one to three minutes, depending on the latitude and the sign rising, and KP sub-lords on the cusps change within minutes.

Rectification, adjusting a recorded time to the time the chart actually behaves as, is therefore part of advanced practice. It is also one of the easiest places to fool oneself.

02प्राण · विनाडी

Classical time and its units

Indian astronomy measures fine time in prāṇas, the time of a breath. In the Sūrya Siddhānta, six prāṇas make a vināḍī, sixty vināḍīs a nāḍī and sixty nāḍīs a sidereal day, so a prāṇa is very nearly four seconds, a vināḍī twenty-four seconds and a nāḍī a little under twenty-four minutes.

Traditional rectification checks, such as those based on prāṇapada, on divisional-chart boundaries or on the agreement of the lagna with the Moon, work at this scale. Their rules differ between texts, and many of them repeat at regular intervals across a window of uncertainty.

03पूर्वाग्रह

The confirmation-bias problem

If the same five events are used both to adjust the time and to prove that the adjusted time is right, the proof is circular. With divisional charts and five daśā levels to choose from, some configuration will fit almost any event at almost any candidate time.

The symptom is familiar: a time that explains the past perfectly and predicts the next event no better than chance.

04परीक्षा

A held-out protocol

A defensible procedure separates the events. Collect several objectively dated, significant events. Use some of them to narrow the candidate window, with traditional checks treated as generators of candidates rather than as proof. Fix the candidate time and write it down. Only then examine the held-out events and ask whether the fixed time accounts for them better than the alternatives.

If the fixed time fails the held-out events, it is rejected, not adjusted again on those same events. Once they have been used to adjust the time, they are no longer held out.

  • Several dated, significant events
  • Calibrate on some
  • Fix the candidate time
  • Test on the held-out rest
  • Reject, do not re-tune, on failure
05अभिलेख

Keeping both times

A rectified time is a working hypothesis with a confidence, not a correction of the record. Good practice keeps the originally recorded time, the rectified time, the uncertainty and the events used for calibration and testing, so that later work can revisit the decision.

That record also tells a later reader which fine techniques the chart can bear. A time rectified to within a minute supports more than one known only to the half hour, and the reading should say so.

THE GRAHVANI PRINCIPLE

Given enough freedom, any birth time can be adjusted until known events appear to fit. Rectification becomes evidence only when the candidate time is fixed first and then tested on events that played no part in choosing it.