There is a particular sound that no longer exists: the sound of a page of Raphael's being turned back and forth, twice, three times, because you have lost your place in the column of proportional log…

[Daniel Burkett]

There is a particular sound that no longer exists: the sound of a page of Raphael's being turned back and forth, twice, three times, because you have lost your place in the column of proportional logarithms and you are not going to get the Moon right until you find it again.

I want to describe the procedure, because a generation of practitioners has now come up without ever performing it, and the description alone does some of the work of the argument.

You began with time. Not the time on the birth certificate - that was raw material, not data. You converted it to Greenwich, which meant knowing the time zone, and knowing whether the jurisdiction was on daylight saving, and knowing that in 1943 the United States called it War Time and observed it year-round, and knowing that some counties in Indiana observed nothing at all. Half the errors in hand-cast charts entered here, in the first two minutes, before a single planet had been touched.

Then sidereal time. The ephemeris gave it to you at noon Greenwich for that date. You added the interval from noon to the birth, then a correction for the acceleration of sidereal on solar time, roughly ten seconds for every hour of that interval,then an adjustment for the longitude of the birthplace, four minutes of time for every degree east or west. What came out the other end was the local sidereal time, and with it you entered the Tables of Houses.

The Tables were printed for specific latitudes. Yours was never one of them. So you interpolated: between two sidereal times, and between two latitudes, and you did it for the midheaven and the ascendant and the intermediate cusps, and at high latitudes the ascendant would leap ten degrees between adjacent entries and you would sit there deciding how much you trusted the birth time.

Then the planets. The ephemeris gave a position at noon and another at noon the next day. The difference was the day's motion. You took the proportional logarithm of that motion, added the proportional logarithm of your interval from noon, looked up the antilog, and added the result to the noon position. Ten times. The Moon last, and slowest, because the Moon travels twelve to fifteen degrees a day and will not forgive you.

Forty-five minutes, if you were practiced. An hour and a half if you were not, or if the birth was near midnight, or if you had to correct for a birth south of the equator and your Tables assumed north.

Here is where this blog post is supposed to tell you that this was better. It was not better. It is important to say so before saying anything else, because an elegy that lies about its subject is just advertising.

Hand-calculated charts were wrong. Not occasionally, routinely, in small ways, and sometimes in large ones. A transposed digit in the sidereal time moved the ascendant by degrees. An arithmetic slip in the lunar interpolation put the Moon in the wrong sign. And these errors did not announce themselves. They went into the wheel, and the wheel went to the client, and the error was interpreted. Someone was told something true-sounding about a Mars that was not there. Whole consultations, whole self-understandings, were built on the wrong side of a rounding. We have no idea how many. The tradition has never seriously reckoned with this, and I think it should.

And the labor was a gate. Not a threshold, a gate. It meant that casting a chart required an ephemeris for the right year, a Tables of Houses, a table of logarithms, an afternoon, and a teacher to show you the sequence. That combination selects for leisure and access more reliably than it selects for aptitude. Every practitioner who ever gestured at the difficulty of the work as evidence of its seriousness was, whether they meant to be or not, defending a filter that had nothing to do with the sky.

So: not better. Slower, more expensive, less accurate, and less available. Anyone who wants it back as a requirement is asking for something worse in the name of something they have not correctly identified.

What was lost is narrower than nostalgia claims and harder to replace than nostalgia understands.

You used to have to be inside the chart before you could look at it. Forty-five minutes of arithmetic is forty-five minutes of contact with those specific degrees, in a state of attention that has no equivalent. By the time the wheel was drawn you had handled every position in it. You knew the Moon was at twenty-two Scorpio because you had made it be there, out of two numbers and a logarithm. That is a different relationship to a figure than receiving it rendered.

You also used to know why. Why the ascendant is unstable at high latitude. Why four minutes of clock time moves the midheaven exactly one degree while the ascendant moves anywhere from a fraction of that to several, depending on where you are and what sign is coming up. Why a rectified birth time is a claim about an interpolation and not a claim about a fact. None of this was taught as theory. It was extracted from you by the procedure, the way a language is extracted by living somewhere.

And you cast few charts. This is the part that gets least attention and matters most. When a chart costs an hour, you cast the ones that deserve an hour. Scarcity did the work that discernment is supposed to do. Now the marginal chart is free, and a practitioner can generate four hundred figures in an afternoon and think seriously about none of them. The software did not make us shallow. It removed the friction that had been doing our thinking for us, and most of us did not notice we had been outsourcing it.

I should be honest about my position. I build astronomical computation software. The engine I have spent years on exists specifically so that nobody ever has to open a table of proportional logarithms again, and it is accurate to a precision that Raphael's could not have represented, let alone achieved. I am not a bystander to this loss. I am one of the people making it permanent, and I would do it again, because the alternative is a craft that is beautiful and wrong.

But I would ask one thing of anyone who reads charts.

Do it once. Your own, by hand, with a paper ephemeris for your birth year and a Tables of Houses and an afternoon you are willing to spend badly. Not as practice, as an initiation, in the ordinary sense of the word: an act performed once that changes the standing of everything you do afterward. You will get a number slightly wrong. You will find it. And when you go back to the software the next day, the figure on the screen will have stopped being an image and become a result, something with a derivation behind it, produced by a mechanism you have touched with your hands.

That is the whole of what was lost. Not accuracy, which we have improved. Not rigor, which we have improved. The chart used to be something you made, and now it is something you receive, and the difference is legible in the interpretation every single time.