A Bazaar's Heartbeat: A Letter to an Online Artisan — Epoche B2
Dear Layla, Your letter arrived on Tuesday and I have been slow answering it, which you will forgive when I tell you I have been finishing a chess table for a customer in Shiraz and have hardly looked up in three weeks. I am glad the shop is running. Eleven orders in a month is not nothing, whatever you say about it. You asked where the inspiration comes from, working in an old trade in an old place, and I want to answer honestly rather than picturesquely, because I think the honest answer is more use to you. So I am going to give you reasons rather than atmosphere, and where the reason is a number I am going to do the sum on the page. You will see that most of what looks like devotion in this trade is arithmetic that has been done once and then never mentioned again. Where I actually am, and why I open at seven The workshop is off the Bazar-e Bozorg, the covered market of Isfahan, two turnings in from the Qeysarieh gate — the great portal at the northern end of Naqsh-e Jahan, the rectangular square laid out for Shah Abbas from 1598 and put on the UNESCO World Heritage list, under its other name Meidan Emam, in 1979. So the walk to work is through a monument. I mention this because it is exactly the sort of thing people expect to be the answer to your question, and it is not. I open at seven, and the first hour is the best of the day for a reason that has nothing to do with the atmosphere. The light comes into the vault through the openings at the crown, and at that hour the sun is low, so the beam meets my bench at a shallow angle. Now think about what a shallow angle does to a fault. Suppose a glued surface has a ridge of height $h$ standing proud of it, and the light arrives at an angle $\theta$ above the surface. The shadow the ridge throws has length $h/\tan\theta$. That is the whole of it, and everything follows from the behaviour of $\tan\theta$ near zero. Take a ridge a tenth of a millimetre high — far too small to see by looking. At $\theta = 10^\circ$, since $\tan 10^\circ \approx 0.176$, the shadow is $0.1/0.176 \approx 0.57$ mm: nearly six times the height of the fault, and easily visible. At $\theta = 60^\circ$, with $\tan 60^\circ \approx 1.73$, the same ridge throws $0.1/1.73 \approx 0.06$ mm, which is nothing. The shadow is ten times longer in the first case than the second. The fault has not changed. The light has magnified it, and the magnification is exactly $1/\tan\theta$. That is why after nine o'clock I can no longer see what I could see at seven, and it is why craftsmen here are sentimental about the bazaar in the evening and practical about it in the morning. In the evening the light flatters the work. At seven it does not. What khatam is, since I have never properly told you What I do is khatam , the Persian inlay work. You take rods of wood, brass wire and camel bone, each drawn down to a triangular section perhaps a millimetre across, and you bind six of them into a hexagonal bundle, and you bind those bundles into a larger bundle, and you glue the whole thing and leave it under pressure. Then you slice it crossways into sheets less than a millimetre thick, and each slice carries the same star pattern all the way through, like a stick of seaside rock. Hans Wulff's book on the traditional crafts of Persia describes the whole sequence, tools and all, if you want it set out properly. The number six is not a preference. The rods are equilateral in section, so each corner is $60^\circ$, and to close a ring of them around a point the corners must add to a full turn: $6 \times 60^\circ = 360^\circ$. Five would leave a gap of $60^\circ$ and seven would not fit. Triangles are used in the first place because they are the only sections that will build both the six-pointed star and the hexagon it sits in without any filler between them. You once asked why a tray takes four hundred rods, and I did not have a good answer. Here it is. Bundles pack hexagonally, like the cells in a honeycomb: one at the centre, then a ring of six around it, then a ring of twelve, then eighteen. The $k$-th ring holds $6k$ bundles, so the total out to the $n$-th ring is $$1 + \sum_{k=1}^{n} 6k = 1 + 6\cdot\frac{n(n+1)}{2} = 1 + 3n(n+1).$$ For four rings, $1 + 3 \times 4 \times 5 = 61$ bundles, and at six rods each that is $61 \times 6 = 366$ rods. Five rings would give $1 + 3 \times 5 \times 6 = 91$ bundles, or 546 rods. So "about four hundred" is a bundle built out to four rings, and now you can work out any size you like without asking me. The patience is not spiritual. It is arithmetic, and here is the arithmetic that matters. Errors in the sections do not stay where you put them. If every rod's apex is cut $\delta$ too wide, the six of them fail to close the ring by $6\delta$, so half a degree of sloppiness per rod opens a three-degree wedge — and that wedge is then bound inside a larger bundle, where it opens further. Worse, the error is permanent in a way that errors in most trades are not. A bundle thirty centimetres long, sliced at seven-tenths of a millimetre, yields about $300/0.7 \approx 430$ sheets. Every one of those 430 sheets carries the same mistake. Get the sections wrong at the start and every slice is wrong for ever, and no amount of care at the end will recover it. About the Rumi line everyone quotes at us Since you mentioned inspiration, I should deal with the line that gets quoted at people in our trade constantly: let the beauty we love be what we do. I have some sympathy with it and also some resistance, and I have a better reason for the resistance than I used to. The reason is that the line as it circulates is a modern English rendering rather than a translation. Franklin Lewis's study of how Rumi has been read documents this carefully: the popular English Rumi of the last few decades comes largely from versions made by poets working from earlier translations rather than from the Persian, and those versions smooth him into aphorism. What we are being handed, in other wor