The Most Expensive Pew in New York

The German word is Balkentreter. A Balken is a beam, and a Treter is one who treads, so the word names the man, more often the boy, who stood behind the organ and walked the bellows-beams up and down with his feet so that the instrument had wind. English called him the organ blower, or the bellows-blower, and paid him by the Sunday. He is the humblest wage on a list that a St. Louis journal printed in May of 1856, and he is the reason I noticed the list at all, because he stands there next to a pastor who was paid four thousand dollars a year.

What everyone already knew

That the fashionable churches of New York were expensive was not news in 1856. It was one of the standing jokes of the American religious press, and the German papers took it up gladly. A church on Fifth Avenue was a brownstone monument with a spire, a pulpit filled by a man whose name was in the newspapers, a choir hired from the opera, and pews that were not seats but property, bought and sold and rented by the year at prices that a farmer in Illinois could not have believed. The received story was that such congregations spent on themselves what would have kept the Gospel going across a whole territory of the West, and the received story was, as far as it goes, true.

Lehre und Wehre was two years old that spring. Its back pages carried a department it then called Vermischte kirchliche Nachrichten, miscellaneous church news, and into the May number, after a set of Tuscan prayers to the patriarch Noah for the health of the vines and a note on the Lutheran Church of France, the editors dropped an item under the letterspaced heading Enormer Aufwand einiger Modekirchen in New-York, "enormous expenditure of some fashionable churches in New York." A Modekirche is a fashion-church, a church in the mode. The item runs eight lines on page 158 and four on page 159, and it is a budget.

The budget, as printed

Here it is, in the journal's order. In one New York congregation the pastor's salary is $4,000, and his assistant's is $500. The gifts made to the pastor in the course of a year average $500; those made to his assistant, about $5. The choir, the organist's salary, repairs to the organ, and the bellows-treader's wage together come to about $1,600. Cleaning, together with the sexton's salary, averages $900. The annual loss in the value of the building and its furnishings through use and time may be reckoned at $1,000. "The costs of the annual upkeep of the church therefore amount to $22,505," the item concludes, "which is as much as $432.78 for a single Sunday. This sum could comfortably maintain 22 country churches or 8 city churches." Pew rents paid for it. That is what a pew in such a church was, not a seat but a share in a twenty-two-thousand-dollar subscription, and the most expensive pews in New York were the ones nearest this pulpit.

Read it once for what it says and it says what the editors wanted it to say. Four hundred and thirty-two dollars every Sunday, for one congregation, in a year when a pastor of the Missouri Synod on the frontier counted himself fortunate to be paid at all. The five-dollar gift to the assistant is the kind of detail a good editor leaves in without comment, and these editors left it in without comment. The bellows-treader, lumped with the choir and the organ repairs, is the other. Somebody in that church was paid four thousand dollars to speak, and somebody else was paid a few dollars to make the wind for the music, and the journal set them in the same paragraph in the same type and let the reader draw the line.

Read it a second time

Then read it again with a pencil, which is the habit this archive has forced on me, because a page that prints a sum is a page that can be checked. Add the figures the item actually lists. Four thousand, and five hundred, and five hundred, and five, and sixteen hundred, and nine hundred, and one thousand. The total is $8,505.

The printed total is $22,505. The difference is exactly $14,000, and it is not on the page.

I went to the scan to make sure the transcription had not dropped a line, because that is the first thing to suspect when a column fails to add. It had not. The page sets the seven items and then the word daher, "therefore," and then the total, with nothing between. The division, on the other hand, is honest: twenty-two thousand five hundred and five dollars over fifty-two Sundays is four hundred and thirty-two dollars and seventy-nine cents, which the compositor set as seventy-eight, a penny lost in the rounding. Whoever worked the sum worked it from a total of $22,505. The seven items in front of the reader do not produce that total. Something was counted that was not printed.

I think I know what it was, and I will mark this as a guess so that it stays one. Fourteen thousand dollars is seven per cent of two hundred thousand, and seven per cent was the lawful rate of interest in New York in 1856, and two hundred thousand dollars is the order of what the grandest of the new churches had cost to build. The one item a budget of this kind always carries, and the one item a congregation never sees in the collection plate, is the interest on the money sunk in the building. The American paper from which this was almost certainly clipped, and which the journal does not name, very likely had a line for it. The St. Louis compositor, or the St. Louis editor, or the German paper that stood between them, let it fall, and kept the total. The result is a paragraph whose parts say one thing and whose sum says another, printed in a journal edited by a man who would have lectured you for an hour on the difference between a premise and a conclusion.

Why it matters that it does not add

There is a temptation to fix it. A tidy transcription would quietly supply the missing line, or quietly change the total, and the page would come out balanced. We do not do that in this archive, and the reason is the reason the mistake is interesting. The page as printed is the evidence. What it shows is not that a New York church spent $22,505 a year, which may or may not be true, but that a St. Louis theological journal in 1856 was willing to print a figure it had not checked, in the service of a point it already held. The editors were not deceived and were not deceiving anyone. They were clipping. The item served, and it went in.

I have found the same thing in the same journal in the other direction, where a table of theology students at twenty German universities came through the scanner with nine numbers wrong and every one of them was recoverable from the row and column totals, because that table had been set from a source that added. The rule I have come to is simple and I apply it to my own work as much as to Walther's: a column that sums to its printed total is a proof, and a column that does not is a finding. It is not to be corrected. It is to be reported, and then explained if it can be, and left standing if it cannot.

What the bellows-treader was paid

The item does not say. He is inside the $1,600, with the choir, the organist, and the organ repairs, and a hired choir in a church of that kind would have taken the greater part of it. Whatever was left for the boy at the bellows, it was a small figure set beside a large one, in a paragraph where the pastor made four thousand and the pastor's assistant made five hundred and five. The gift of $5 to the assistant sits in the middle of it like a joke nobody told on purpose.

He has no name in the record and never will. What he has is a job that no longer exists. The electric blower reached American organ lofts around the turn of the century and within a generation had put the bellows-treaders out of work, and the word went with the job, so that today you have to explain it, as I have just done. But he was there every Sunday in 1856, out of sight, making the wind, and the journal in St. Louis put him on the page, which is more than the church in New York seems to have done.

Now

Every congregation I have served has known, to the dollar, what it costs to open the doors on a Sunday. The figure is discussed at voters' meetings with the same mixture of pride and alarm that the St. Louis editors were inviting in 1856, and it is always compared, in the same breath, to what the money could do somewhere else. The comparison is not wrong. It is just older than we think, and the arithmetic behind it has been shaky for a hundred and seventy years.

The two pages are in the archive, German and English side by side, at page 158 and page 159 of the 1856 volume, in the May number. Add the column yourself. Then go and find out what your own church pays, per Sunday, for the wind.

Note: the journal prints the per-Sunday figure with a German decimal comma, $432,78, and I have given it the American point. The scans were made from a library copy; the transcriptions and translations are machine work corrected against the page, not yet human-reviewed.