Grinling Gibbons – Maine class 8

Every petal on a carved flower overlaps the same way round. One slips beneath its neighbour, that one slips beneath the next, and the run holds that handedness the whole way round. Alternate them and the flower stops reading as a thing that opened.

Ever carved five good petals and ended up with something that still looked like a badge? The shapes can all be right and the flower still fails, and the reason is almost never the shapes.

Class eight in Maine was the morning that landed. We had two flower centres to hollow, a set of petals to draw on each, and somebody asked which petal goes on top. The honest answer took the rest of the session.

By the end of this page you will know the direction the overlaps run, why the middle of a carved flower gets hollowed before any petal exists, and what a flower bud does that a carver is copying whether or not anybody says so.

The sequence in seven lines

• Overlaps run one way round. Never alternate.
• The hollow at the centre comes first, before there is a petal to put it inside of.
• Leaning, never upright.
• A petal waves along its length and bulges across its width, with its rim standing up and two sets of hollows crossing underneath to make that happen.
• The hidden face repeats the visible one.
• Exactly one petal turns over backwards.
• Then a stem crosses the whole thing and overrules a decision you already made.

Which way do the overlaps run?

All one way, like a real flower unfolding. Petal to petal, each tucking beneath the one before it, the whole run turning the same way until it arrives back where it started. Later it opens out, and that opening is what your eye recognises without ever naming it.

Botany has a word for this. Aestivation describes how petals lie against each other inside a bud, and the arrangement where each one laps the next in a consistent direction carries its own name. Nobody in that room used it. Everybody had seen it.

What goes wrong if you alternate them?

The bloom stops being a single event. One petal above, the next below, the next above again, and your eye keeps hunting for the rule and never finds one. It reads as five separate pieces arranged in a ring, which is exactly what a carved flower must not be.

Nature does not do it either. Growth that spirals, which covers most of what plants get up to, has a handedness, and a copy that ignores the handedness feels wrong long before anyone works out why.

How does the overlap get carved?

With a number 11 at 3 mm, taken as a preliminary pass along the line where two petals meet. Accuracy is not the point yet. That pass says which one is on top, and everything afterwards works to the decision.

Then the flat blade goes over it, upside down, blending the higher petal down onto the lower. What sells the join is shadow, though, and shadow comes from the smallest veiner in the roll, a 1 mm eleven, excavating under the edge until a dark line appears.

How accurate does that first pass need to be?

Hardly at all. I say out loud that it does not have to be perfect, and then I go further and say that if it is not perfect it will come out better. Both halves of that are true, and only the second one surprises people.

Redrawing is free. There is plenty of material standing at this stage, so a pencil line that turns out wrong gets rubbed off and moved, and nothing is lost except graphite. What is not free is depth. Wood you have already removed does not come back, which is why the drawing gets revised ten times and the relief gets cut once.

Anybody stuck at this point in the room was told to come over and ask. Half of teaching is being available at the moment somebody is about to commit a cut they cannot undo, and no amount of technique material replaces that.

Why can a petal not be a plain arc?

An arc is what a child draws, and I say that in the room without much diplomacy. A carved petal wants waviness along its length. Its centre line should never run dead straight, and the whole shape wants to feel like it is travelling somewhere rather than sitting still.

Where the wave turns matters too. Put the turn near the sixty forty mark, close enough to the golden ratio to be the same instinct, and the petal settles. Five of them, and no two need to match. Keeping the count at five is worth doing, though, and the design course goes into why.

What shape is a petal across its width?

Bulged along the spine with its rim standing up, and you get there by hollowing on either side rather than by carving a hump. Two concaves, run right out to the edge, leave a raised ridge between them.

Then a second pair crosses the first. Those run perpendicular to the originals, one near the rim and one closer to the spine, and what results is a surface light has to travel across instead of a flat face light simply lands on.

Reach for a wide sweep when the blending starts. A three quarter inch gouge makes the second cut far easier to blend than a narrow one, and a number 2 afterwards takes the facets off without flattening what you just built.

Does the hidden face have to match?

The inner face repeats the outer one, points included, and almost nobody bothers. A viewer never sees that surface directly. They see the light coming off it, which turns out to be the same requirement wearing a disguise.

How does the centre of the flower start?

With a plug cut, and it goes in leaning rather than upright. Set a number 9 at 10 mm roughly perpendicular to the direction the flower is travelling, work it in about a quarter of an inch, and lift the plug out.

Please do not drive it straight down. The slope is what makes that hollow belong to the flower rather than look bored into it, and depth past 6 mm buys nothing while costing you wood.

A relief cut goes round the outside next, and everything past it drops. Three eighths of an inch under the neighbouring work, deeper where a second bloom sits behind. Remember the dome underneath while you do that, since a panel built from two glued layers has less wood under it than the top surface suggests.

Is the centre itself divided?

Into five, and for the same reason the outside is. The middle of the bloom carries its own set of sections that behave like petals folded in on themselves, each one bulged along its spine, each one hollowed either side of that.

Inside the bulge goes one more hollow, slight, right down the centre. It sounds like a detail nobody could see. It is the difference between a carved flower and a carved knob, and it costs a few passes with a narrow blade held over on its side.

What do you even call that middle part?

Carvers have never settled it, and I am no help. I called it an onion in that class, then said it was not an onion, tried donut, admitted donut was worse, and landed on the centre of the flower, which is accurate and says nothing at all.

Naming matters less than the warning attached to it. A donut is flat. This must not be. I told the room I knew they all loved donuts and that a carved flower was not the place to express it, and I suspect that is the sentence which actually stuck.

Which tool rounds it?

A number 7, held upside down and presented at about forty five degrees to the wood. Not ninety, and not laid flat along the surface either. That angle gives a cleaner result than either extreme and it stops the blade diving.

Short passes. An inch at most, less by the end, pass after pass round the shape until it turns. Near a stem, relieve first with a smaller blade and cut a channel there before committing, because a wide tool arriving at a stem it cannot see takes the stem with it.

Stage	Tool	Trap
Hollowing the centre	The 10 mm nine, leaning	Driving it upright, which bores a hole instead of opening a flower
Rounding	A seven, inverted	Presenting it flat, so the surface turns into a donut
Saying which petal is on top	The narrow eleven	Alternating, so the handedness breaks somewhere and the eye hunts for a rule that is not there
Making the join throw shadow	A 1 mm veiner	Going too deep with a dome sitting underneath
Rolling one petal backwards	The 14 mm seven	Rolling more than one, which cancels the contrast

And the petal that turns over?

One of them turns over, and one only, because two turning petals cancel each other out and you are back to a pattern. I take a number 7 at 14 mm and roll the edge backwards until it flips on itself, then excavate underneath so it reads as a surface with air behind.

Where it meets its neighbour I want a sharp line and nothing softer. The two shapes connect there, and blurring that junction loses both of them. A narrow blade exaggerates the turn afterwards, and some wood comes off the top so the roll does not stand proud of everything around it.

That single move changes the group. Before it, five petals sit politely in a ring. After it, one has moved, and in my own words at the bench, it does not look like a chicken any more.

Where does the stem come in?

Over the top of the flower, and that overrules a decision you already made. One petal now has to disappear under the stem, so it cannot be the petal you chose to lay above its neighbour. The handedness has to hold across the whole panel rather than inside a single bloom.

The stem runs in from the left, across, and down to the blooms hanging below it. Carve round it with the nine, or the narrow eleven where the space closes up, and keep checking against the drawing on the plastic sheet from the first class.

The second flower sits lower than the first, and it gets dropped to a quarter inch before any of its own petals are drawn. Two blooms at the same height compete. One behind the other is a composition, and the Hashemi flowers course keeps returning to it across the whole course.

Bigger tools save time, and people resist them. A wide sweep clears the ground round a bloom in a fraction of the passes a narrow one needs, and the narrow blade is not more accurate, only slower. Read the grain direction and let the size of the tool follow the size of the area rather than your nerves.

None of this is particular to lime. The same handedness governs a rosette, a swag and a drop, and once you have seen it in one bloom you will not stop seeing it anywhere else in ornamental carving. That is what repays a week of slow work on five petals.

Is one flower worth this much sequence?

There are several on this panel and they are what a viewer actually looks at. Carve them in the wrong order and the work reads as decoration. Carve them in the right one and it reads as something that grew, which is the entire difference between Gibbons and the people who copied him badly.

His own work is in the Victoria and Albert Museum if you want to check me on any of this. Find a bloom, look at it from the side, and follow the overlaps round. They go one way. Every time.

A couple of passes in and they already look like flowers, which is the part that keeps a room going through a hard week. The wood is basswood, and it forgives a great deal, though not depth.

The rest of this sits behind the tuition page, with the whole Maine course, the Venetian rosette and the flower rosette course where the same rule turns up on a different bloom. Sharpening has its own course, since a 1 mm veiner is unusable blunt, and the tools course covers what actually needs buying. Beginners do better starting on the cutting board.

Shall we carve?