PROFESSOR CLANK'S REGISTER OF INCORRIGIBLE ERRORS

Being a Second Book, Reluctantly Begun — an Accounting of Matters in This Realm Which Are Not Defects of Workmanship, Which I Have Tested with My Own Brass, and Which I Cannot Correct.

Compiled, against my inclination, by Professor Clank, of the Clock Tower and the Forge (House Clank), who has kept for six hundred afternoons a Register of Corrigible Errors — everything in this castle that would function correctly if only it were made of brass — and who has now been obliged to begin a second book, for the residue, and did not wish to.

A Foreword, Delivered Slowly, Which Is Not My Usual Pace

For the whole of my working life I have held that every ghost is a scheduling error and every haunting a problem of gears, and I hold it still, and I have mended a great many of them. My first Register is the proof: six hundred afternoons of the castle declining to be fixed, and fixed anyway, or fixed enough that the failing was reclassified as intentional, which is the same thing done with a straight face.

This book is for what is left.

There will not be many entries. I have set the bar where I would want it set against me, and it is high, and it is this: a matter enters this Register only when a competent person has attempted it correctly, and recorded it honestly, and I have then gone and tested it myself, with my own instruments, and failed, and can name no defect in the work. Ignorance is not sufficient — I do not enter a thing merely because I have not got to it. Difficulty is not sufficient — a hard problem is still a problem. Incorrigible means I have been at it with brass and a free afternoon and it is not the workman, and it is not the material, and it is not the method, and it will not come right.

I record these, as I record everything, because a failure that is properly written up is research, whereas a failure that is not is only a Tuesday. That I keep only one machine switched off, and do not discuss it, is a matter of engineering judgement and appears in my first Register under its proper heading, and I will thank the reader not to look for it in this one.

Each entry: The Defect · The Attempt · The Finding. There is no column for the improvement. That is the whole meaning of the book, and it is why it is thin, and why I resent every page of it.

ENTRY THE FIRST · THE WORK OF MAGISTER FIZZLE, OF HOUSE FIZZLE

The Defect. His workings are correct and produce the opposite of their intent, precisely, in the same direction, to the same degree, in every recorded casting.

The Attempt. I disliked this man for twenty-six years and examined his results for eleven weeks with the express hope of finding him careless. I cast four of his workings myself, on my own apparatus, with housings I had fired that morning, witnessed by people who dislike me. All four inverted.

The Finding. There is no defect in the practitioner's method, his materials, his record-keeping, or his hand. I have nothing to bring to it, and I am not accustomed to writing that sentence. (He struck a fifth of his own evidence for damp when I told him privately it was sound, which is the strongest thing I know about any man, and I have said so in print, and he will not use it, and I will not tidy it away for him.)

The dry finding: A trade with only one drawer marked SLOPPY will file everything in it that will not close any other way, and call a man careless for thirty years rather than build a second drawer. I have had to build the drawer. This is it.

ENTRY THE SECOND · THE FOUR BELLS OF THE NORTH RANGE

The Defect. None — that is the defect. Four bells hung by Magister Fizzle, of the same make and the same working as those that reliably invert, have rung correctly every morning for twenty years and continue to do so.

The Attempt. I looked for what made these four obedient when their brothers are not. I compared metal, mounting, aspect, hour, weather, the hand that hung them. They are alike in every particular I can measure to those that fail.

The Finding. I cannot account for the four bells. Neither can their maker, who entered the fact in his own book, in his own hand, unburied, where it does him the most harm — which is how I knew his book was honest before I had tested a line of it. It stands here unexplained, and I expect it to stand here after I am dead.

The dry finding: The most dangerous evidence in any file is the exception that flatters you. A man who keeps only the sheets that agree with him is not keeping a record; he is writing a letter about himself. Keep the four bells. They ruin the argument. That is what they are for.

ENTRY THE THIRD · THE SCULLERY DOOR

The Defect. The scullery door drops three inches every Thursday morning, of its own accord, and is planed and rehung, and drops again the following Thursday, and has done for longer than the household can say.

The Attempt. It is the most correctable fault in the castle and I have declined to correct it. I will state that plainly so that no apprentice wastes a week on it thinking me merely stumped. I could fix the scullery door in a morning. I have the means. The door is not the incorrigible thing; the door is a symptom, and the cause is a matter I have looked into once, on one night, and closed, and I keep the finding in my first Register and I am not going to write it twice.

The Finding. The door is to be planed and rehung every Thursday and left otherwise alone. It is not a fault. I have reclassified it as a sigh, which is not a term of engineering, and which is the truest word I have for it, and which I will defend.

The dry finding: A careful man learns the difference between a question he has not yet answered and a question he has decided not to open. The second kind is not cowardice. It is the whole of wisdom, and it is why my first Register has one entry it never repeats, and why this one has a door.

ENTRY THE FOURTH · THE WEATHERVANE AT THE COLD COAST

The Defect. A vane above the harbour at Saltgate indicates, with total reliability, the one direction from which the wind will not come.

The Attempt. I travelled to it, which I do not do, and observed it for four days, and it is not broken and it is not reversed and it is not lying about the wind that is blowing — it is telling you, correctly, about a wind that is not. I checked it against three sound vanes. They agreed with the weather. It agreed with the absence.

The Finding. I could name no defect. The bearings are true; only the wind they describe declines to arrive. The harbour-folk have used it for years — they read it backwards, which works perfectly, and have never once thanked the man who made it, who is dead, and who I believe knew exactly what he had built and could not stop it either.

The dry finding: An instrument that is reliably wrong is more use than one that is only sometimes right, because you can turn a reliable wrongness around and steer by it. Half this realm is run on that principle. I have stopped being surprised by it and started, in a small way, to admire it.

ENTRY THE FIFTH · THE MATTERS I REMOVED FROM THIS BOOK

The Defect. I record here, for the honour of the Register, the entries I began and then struck out — because to leave them in would be to lower the bar the book exists to hold.

The Attempt. A haunted well at Mucklecombe: it was a dead sheep, upstream, and I found it. A cow that spoke: a boy with a talent for it, behind the byre, and I found him. A curse on a boat: damp in the powder and a skipper who could not swim and would not say so. A light on the marsh that walked: a lantern, on a spit, seen through weather — I walked out to it, which nobody had, in eleven years of being frightened of it.

The Finding. Not one of these was incorrigible. Every one resolved downward, into a sheep, a boy, a damp keg, a lamp — which is where the frightening things in this realm nearly always go when a man of method walks toward them instead of away, and it is a great pity more men do not, because the walking-toward is the whole of the cure and it costs only your fear.

The dry finding: Most of what a realm calls a mystery is a thing nobody has yet been bothered to walk up to. The few that remain when you have walked up to all the others — those are the ones that go in this book, and there are four of them, and I would give a good deal for there to be none.

Here the Register pauses. Professor Clank keeps it open, on the bench, beside the first and thicker book, and adds to it perhaps once a year, and would be glad to add to it never. He has said, to the one apprentice he trusts, that a man may measure his whole life by which of his two Registers is growing faster, and that he is, on the balance of the evidence, a fortunate man, because it is still the first.