Enough to Be Bread
A new wheat-starch supplier sends a lot that tests clean, below 5 mg/kg of gluten, and a coeliac boy's mother is delighted. The priory's bakery assistant says the result is the wrong kind of good. Hosts with no gluten at all are not valid matter for Mass, and the lot is too pure to be bread. The story is about a gram of flour in a kilogram of starch, a hot spot of 310 in twelve hosts, and a boy's first communion.
The hosts at Sainte-Scholastique were made in a room the sisters called the iron room, though the irons were the least of it. There were two of them, cast in 1958, with long handles and a pattern of the crucifixion pressed into each plate, and they still worked. The sisters used them for the big breads at Christmas and Easter. For the other fifty weeks of the year the room ran on a modern wafer machine with a chain, a tray and a fan, which Sister Bernadette had bought in 2039 with the proceeds of eleven years of jam.
The priory sold altar breads to forty-three parishes in the Townships and the Beauce. Most were ordinary. A little under a tenth, about a thousand a month, were the other kind: low-gluten hosts, for people with coeliac disease and for priests who have it, made from wheat starch washed until almost nothing is left of the gluten, and still, by the letter of the law, wheat.
Bernadette was seventy-four. She had been, for thirty-one years before she entered, a quality-control technician at a dairy in Granby, and she kept the bakery the way she had kept the dairy, with a logbook, a ruler, and no patience for the phrase "should be fine". The assistant on the bakery's tablet was called Levain, after the sourdough starter, because the first thing it had ever been asked to do was keep a feeding schedule for one. It now kept the lot book, the oven log, the delivery calendar and the test record. It wrote in the register of a competent clerk. It did not say it was pleased.
In the second week of February a pallet came in from a new supplier of wheat starch. The old one, a family mill in Ontario, had closed at the end of the year. The new one was in Hungary, and its certificate said gluten-free, below 20 mg/kg, Codex standard, tested by R5 sandwich ELISA. Sister Odile, the prioress, who had made the order, was pleased with the price.
Bernadette ran her usual checks. She baked a first batch, and Levain pulled twelve hosts from it at random by the numbers on the tray, and she dropped each into a vial of extraction buffer and put a strip in. Every strip showed one line, the control. None showed the second.
"Good," she said. "Clean."
"The strips do not show gluten," Levain said. "They are not built to show less than about ten milligrams per kilogram, so what they say is 'under ten'. I have sent the twelve to the lab at Sherbrooke, as the schedule says. I would like to hold the lot before anyone orders from it."
Bernadette looked at the tablet for a while with her pencil in her teeth, which was what she did when a machine said something she had not expected.
"Hold it for being clean?"
"For being too clean," Levain said. "I would like to say why when the lab answers. If I am wrong it costs us four days."
Here is what was true in 2026, when this story was written, and what had been added by 2047.
Coeliac disease is an autoimmune reaction to gluten, the proteins in wheat, barley and rye, that damages the lining of the small intestine. About one person in a hundred has it. The only treatment is to avoid gluten for life. A strict gluten-free diet is conventionally one where food has under 20 mg of gluten per kilogram, a limit that is in the Codex standard and in the law of the European Union and many other countries. How much a person can take in a day without harm is a separate question. A 2007 trial in the American Journal of Clinical Nutrition (Catassi and others) gave coeliac volunteers 0, 10 or 50 mg of gluten a day for 90 days. The 50 mg group had measurable damage to the gut lining and the 10 mg group did not. Ten milligrams a day is the figure most often quoted as a threshold of safety. It is a trial of a small number of adults, and individual people vary.
The Catholic Church teaches that bread for the Eucharist must be wheaten and recently made, and that wine must be from grapes (Code of Canon Law, canon 924). A circular letter from the Congregation for Divine Worship in 2017 (that office is now a Dicastery) restated the rule and went on to coeliac disease. Low-gluten hosts are valid matter provided they contain enough gluten to obtain the confection of bread, without additives, and are made in a way that does not change the nature of the substance. Gluten-free hosts, made from other starches, are invalid. The letter gives no number for "enough". The bishop, as ordinary, is responsible. Communion under the form of wine alone is the whole Christ, as it always has been (the Council of Trent, and the Catechism, 1390). Many people with coeliac disease receive that way. Others receive a low-gluten host, with a doctor's advice. One monastery that sells them states, on its website, "less than 0.01 percent gluten", which is under 100 mg/kg. That is the supplier's number and a convention, and it is not a doctrine.
The R5 ELISA is a sandwich test with antibodies that bind a short sequence in the gliadin proteins. The usual kit reports quantities from about 5 mg/kg of gluten up. It is not an instrument for measuring zero. Below its limit it says "less than five", and it says it in the same words about a lot with two and a lot with nothing.
What had been added by 2047 was a tablet that could keep the test record, tell its own estimates from its own measurements, and read a certificate carefully. The rest of this story is about arithmetic a person could do with a pencil. It is not a prediction that such an assistant will exist, and it is not medical advice.
The lab answered on the Thursday. All twelve hosts read less than 5 mg/kg.
Levain put the two sets of numbers side by side on the tablet, and said what was on it.
"These are the old supplier's last six lots, from the book. Each lot's twelve hosts. The means run from 62 to 140 mg/kg, and the single highest host in six lots is 211. This is the new lot. Twelve of twelve are under five. The new mill's certificate says under 20. It does not say how far under, and it has probably washed the starch more thoroughly, which is a good thing for a coeliac and for no one else in this room."
"Meaning what for us?" Bernadette said.
"Meaning I cannot show, with this test, that there is any gluten in these hosts at all. If there is none, they are not what the bishop's office has been receiving from us for eight years, and a priest who consecrates them is using what the 2017 letter calls invalid matter. If there is a little, they might be. I cannot tell you which, and I cannot tell you what 'enough' is. The letter does not say. I think you should ask the chancery, and I think we should not ship from this lot until you have."
Bernadette took the pencil out of her mouth.
"We have six orders for the First Communions," she said. "Thirty-one of the low-gluten ones. Ten of them for Saint-Hilaire, where Mathilde Roy's boy is, Étienne. Eight years old, coeliac since he was three. She has been in the doorway of this bakery twice since Christmas to say that he asks every week why he gets the cup and the others get the bread."
"I know. It is in the order book." Levain waited a moment. "I would like to say that I did not look for this. I looked because the lab's report was a column of the same symbol, and I did not trust a column that clean."
"Good," Bernadette said, a little roughly. She put on her coat and walked over to the guest house to use the telephone, because she preferred to ask a priest a question with a voice.
Father Anselme, who looked after the liturgy at the chancery in Sherbrooke, was glad she had called and was careful in his answers, and she wrote them down in the lot book in pencil as he gave them.
The diocese had never set a figure. The practice since 2039 had been to accept low-gluten hosts from the priory as they had been made, with the wheat starch and the measured residue on each lot card. The bishop did not want to say a number that was not Rome's. What he had asked for, Father Anselme said, was that the hosts be what they had been. A host with a low number of milligrams was a host. A host that had been made from something that was not wheat, or from wheat that had been made into something that was no longer bread, was not. He did not think it was a matter for a laboratory. He thought it was a matter for the people who baked and the people who signed the card, and he would be happier if the card said how the sisters had made certain.
"He would like to know," Bernadette told Levain afterwards, "that someone had looked at it who knew what they were looking at, and had not got it from a certificate."
"Then I can tell you what I think the options are. May I?"
"Please."
"One. We buy an old-supplier equivalent, a starch with a known residue, and wait for it to arrive. Six weeks, perhaps eight, and I have not found a stockist in Canada that I would trust to hold the number. Two. We ship the new lot to the parishes as gluten-free hosts, labelled as that, and tell the priests what they are. They are for people who must have nothing, and the priest has to decide what to do about the matter, and the answer is that he cannot consecrate them. I would not call that a product. Three. We put a little gluten back."
Bernadette was silent a moment.
"Say that last one slowly."
"The new lot is wheat starch with almost no gluten. Wheat flour is about eleven percent protein on the sack, and about eighty percent of that is gluten, so roughly nine percent by weight of flour is gluten. If the old lots ran at about a hundred milligrams a kilogram, then a kilogram of the new starch needs a hundred milligrams of gluten, which is a little over a gram of flour. I have it as 1.14 grams. Nothing in it is not wheat. It is the same ingredients, in slightly different proportions, and I would say so on the card."
"And Father Anselme?"
"I do not know. That is a question for him, not for me. I think it is a different thing from what the letter forbids, because the letter forbids adding things that are not wheat. But I have no standing to say so, and you have asked him a question already."
She asked it the next morning, and Father Anselme said he would need a day, and in a day said he had asked the bishop, and the bishop had said yes: wheat, and the card to say what had been done. He added, as a man who liked a sentence to be exact, that the bishop did not say the number was right. He said the number was what the priory had been giving for eight years, and he would ask them to keep to it.
That left the matter of a gram of flour in a kilogram.
The first thing that Bernadette did, because it was the thing anyone would do, was to weigh out 1.14 grams on the analytical balance, tip it into a kilogram of starch in the big steel bowl, and stir it with the long spoon for five minutes, and then bake.
Levain pulled twelve hosts and sent them to Sherbrooke. The strips that Bernadette ran herself, in the bakery, said that some of the hosts were positive and some were not. She did not like that and said so. The lab's numbers, in milligrams per kilogram, were: 8, 11, 14, 9, 22, 18, 7, 310, 15, 12, 25 and 19.
"Mean of 39," Levain said. "Median of 14.5. Eleven of the twelve are under 30, and one is 310. A mean of 39 would have been a lot that was under what you wanted. It is not a lot that is under. It is a lot where one host in twelve is three times the old supplier's worst, and the rest are barely there. The gluten is in lumps."
"In the bowl," Bernadette said.
"I think so. A little over a gram of flour in a kilogram of powder, stirred with a spoon, is a few grains in a great many. If a few of them stuck together, they went into one host."
Bernadette thought about it with the tablet in her hand. Then she did what her old dairy would have done. She sent Sister Marie-Paule, who had been a pharmacist before she came to the priory, down from the dispensary.
Sister Marie-Paule listened for three minutes and said it was not a baking problem. It was the oldest problem she knew, and the cure was dull. You do not put one gram into one kilogram. You put one gram into ninety-nine and mix that, then one part of that into more, and you do it in steps until the last step is small enough that nothing can hide. She called it geometric dilution. She had done it with a mortar and a sieve for forty years with drugs where one milligram mattered, and she did not see why it should not do for flour.
"What is the number for the host? I mean the lump," she said. "How many particles of flour in a host?"
"A host is six tenths of a gram," Levain said. "At 1.14 grams of flour in a kilogram, a host should carry about 0.7 milligrams of flour. A particle of flour is about fifty micrometres across, and weighs, say, a tenth of a microgram. So about seven thousand particles in a host. Seven thousand is a lot, and if they are loose it should be even. If they are in a lump of a thousand it is not."
"Then do not let them lump. We sieve."
They made the premix on the Monday: ten grams of flour, weighed twice, tipped into a bowl with 990 grams of starch, passed through a fine sieve three times, and stirred between. One per cent flour. Then, for each five-kilogram batch of hosts, 570 grams of the premix went through the sieve again into 4,430 grams of fresh starch, three more passes, and into the mixer. Marie-Paule did the first sifting herself, and showed Bernadette the wrist movement. It took forty minutes a batch. Bernadette said that for forty minutes a batch she could have made a pie.
Levain pulled twelve hosts from the first batch and sent them. This time the strips in the bakery were mildly positive in all twelve, which Bernadette noted with approval. The lab's numbers came back on Thursday: 62, 148, 95, 101, 88, 117, 79, 134, 104, 91, 70 and 112.
"A mean of 100," Levain said. "Lowest 62, highest 148. No lump. I would say the method is sound and the number is the number we intended. I would also say that twelve hosts from one batch is twelve hosts from one batch. I would like to test twelve from each of the next five batches, and I would like the card to say 'mean of 12, range 62 to 148' and not 'about 100'."
"The card can say anything you like," Bernadette said. "The Hungarians will be pleased."
It was the third week of Lent when Mathilde Roy came to the bakery, with Étienne, on a Saturday when the sisters sold seconds, the broken hosts that go to the poor box at the door of the parish. She had a printed page from the hospital. Étienne had a ruler, which he held like a sword.
He was small and exact and wanted to know everything. He asked why the wafer machine was called a machine and the iron was called an iron, and Sister Marie-Paule explained the plates. He asked what the numbers were on the wall, which were the batch numbers, and Bernadette told him they were how the sisters knew whose bread was whose. He asked whether it was true that his hosts had less in them than other hosts.
"A great deal less," Bernadette said. "Hosts for the others have what a loaf has. Yours have about the same as a very small crumb of a loaf."
"What does the crumb weigh?"
Bernadette looked at the tablet. "May I?" it said, and she nodded, and Levain said, in the voice it used for children, which was the same as the voice it used for everyone except a little slower:
"A host is six tenths of a gram. About a hundred milligrams of gluten in a kilogram means each host has about six hundredths of a milligram. That is sixty millionths of a gram. Your doctor's number for the whole day is much larger than that, and I am not the person to say what your number is. Dr Fortin is. She has the blood tests. I can say that the amount in a host is a hundred and sixty times smaller than the figure the studies call safe for a whole day, and I can say that I measured it twelve times, and the highest of the twelve was one and a half times that."
"Half as much again as sixty-hundredths."
"Yes. About nine hundredths of a milligram."
Étienne considered this, and then said, to his mother, "That's an even smaller number than the dictionary."
Mathilde Roy had read the hospital letter four times. She read it again at the bakery counter. Dr Fortin's note, handwritten at the bottom, said that she saw no objection to a low-gluten host for the First Communion and the Sundays after it, provided Étienne's antibody levels at the six-month review were what they should be, and that if they were not, they would go back to the cup, and everyone would be sensible. Mathilde had read that last line four times too.
She had one more question, and she asked it of Father Luc, her own parish priest, who had come with her and had said nothing in the doorway.
"Is it the same, Father? The cup and the host? I want to hear you say it."
"It is the same," Father Luc said. "It is the whole Lord in either. If he takes the cup, he has received him whole, and he has not missed something. I have said so to him at the rail eleven times." He looked at the boy. "And if he takes the host, he is receiving him whole, and that is what he wants. It is for you and Dr Fortin to say. I think he has had enough of being the one who is different at the rail."
"I am not different," Étienne said. "I have a smaller number."
Everyone in the bakery pretended to be looking at something else.
The priory kept a rule in the bakery that every lot was to be written in the book by hand, in ink, and signed by the baker, and that the machine's record was a second record and not the first. Bernadette signed the lot in the second week of May.
It was the lot of the thirty-one hosts for the First Communions of the deanery of Saint-Hilaire. Levain had tested twelve from each of the five batches that went into it, as it had said it would. The means ran from 91 to 108. The range was 56 to 163. There was no host over 200, and none under 50. The card, which went into each parish's box with the hosts, said that they were made from wheat starch and a measured amount of wheat flour, to a mean gluten content of about 100 mg/kg, and that the amount had been checked in sixty hosts by a laboratory, and that the gluten in each host was about 0.06 mg, and that it was not a gluten-free host and was not to be given as one. At the bottom, in Bernadette's hand, it said: Valid matter. The sisters of Sainte-Scholastique.
She had asked Levain whether there was anything that was not on the card and should be.
"The new starch," it said. "Its price is lower and its certificate is better. We have saved some money and spent the savings on flour, a sieve and three weeks of Sister Marie-Paule's mornings. I would not say it is a bad trade. I would say that there is a version of this in which nobody looked at the column of identical symbols, and the hosts went out as clean as they were, and a priest said the words over what the Church had said was not bread, and nobody found out. I would like the lot book to say that the first lot was held and why."
Bernadette wrote that, too.
On the nineteenth of May, the sixth Sunday of Easter, in the church of Saint-Hilaire, thirty-one children went up the aisle in white. Étienne Roy was fourth from the front. He had been told that he could take either, and had been told twice, and he had made up his mind in March and told nobody.
The sisters were at the back, on the left, under the choir loft, where the priory had sat at every First Communion in the deanery for as long as it had been making the hosts. Bernadette had her pencil in her pocket. She could not see his hands. She could see the back of his head, and a boy's neck above a white collar that was too big, and a moment in which he stood and said nothing, and then said what he had been taught to say.
What she heard was the small sound that every host makes when it is broken. It is hardly a sound. It is the sound of a very thin wafer, which has been through an iron and a fan and a tray and a sieve and the hands of seven people and a Hungarian mill, parting along its edge.
The choir began the Communion hymn. In the pew in front of her Mathilde Roy had her hand over her mouth, and a hymn sheet that she was not reading, and was not, so far as Bernadette could see, thinking of the number.