Carving Roman Acanthus 12

The return is the small Roman ornament that makes an acanthus eye look alive. Romans carved a short movement running out of each eye and along the pipe, shaped to look as if it climbs up from inside the leaf rather than sitting on top. Three tiny concaves, a micro undercut, and the shade does the rest.

Ever looked at a Roman capital and sensed there was something happening at the eyes that you could not quite name? Carved an acanthus of your own that came out correct and somehow flat? I promised this detail five workshops ago and here it is, and it is smaller than you are expecting.

What the return actually does to a leaf

• Roughly 10 mm, three eighths of an inch, running out from the eye along the pipe.
• Subdivide that length into three and each part gets its own tiny concave.
• Drop the section a fraction, which leaves room for a micro undercut beneath it.
• Micro means barely visible. Your eye registers the shade, not the cut.
• A bent tool by Ashley Iles reaches where nothing straight will.
• Nobody will notice most of this. Carve it anyway.

What did Romans carve into the eye of an acanthus?

A short curl of movement travelling out along the crown of each pipe, carved so it seems to climb out of the leaf itself. Sculptors added it throughout history and the Roman version is what we are carving on this panel.

Mine runs about half an inch out, call it twelve or thirteen millimetres, then a second mark around 10 mm which is three eighths. Those two marks bracket the whole ornament. All of it happens inside.

Leaving it as one shape would be perfectly acceptable. Subdividing it into three is better, and three is what I do, because each of the three then takes a concave of its own and the group reads as something folded rather than something scooped. We will carve this on almost every eye and almost every pipe on the panel.

Small as it is, the detail is what separates a competent acanthus from one that looks as though it grew. Skip it and nothing is obviously wrong. Carve it and everything is obviously right.

Should you measure the ornament or judge it by eye?

Judge it by eye if you can bear to, and there is a cheap trick for when you cannot. My marks go on by eye and they land close enough that measuring would not improve them.

Uncomfortable with that? Take a scrap of paper, lay it against the end of the eye, pencil the length onto it, then carry that piece over to the matching eye opposite. No ruler, no arithmetic, and both sides agree exactly.

Training your eye is still worth the discomfort. Complicated pieces arrive eventually where measuring is genuinely impossible, curved surfaces and compound angles and places no straight edge reaches, where symmetry has to be felt, and your only remaining instrument is the one in your head. Build it early on easy work like this.

My method is to place the mark by feel first and check it afterwards. Feeling and then verifying teaches the eye something. Measuring first teaches nothing.

How deep should the micro undercut be?

About 3/32 of an inch is where I settle, having started at a sixteenth and decided it was mean. Deeper gives the undercut somewhere to live, and the undercut is the whole reason for lowering the section in the first place.

Work it in two stages. First the stop cut straight down on each mark, using a gouge whose curve matches the profile you drew. Mine is a number 7 at four millimetres and it fits almost exactly. Then the same tool, laid over, lowers the wood between those cuts.

Nothing about that tool choice is fixed. An 8 does it, and so does a 9. What matters is that the tool is small and its curve sits on your line, and one of the quiet pleasures of a full rack is finding the one that drops exactly into a shape you drew.

Micro is the right word for the undercut that follows. Barely noticeable, and yet the eye picks it up as a dark line the moment light crosses the panel. That contradiction is what the whole ornament runs on.

The step	What I have in my hand	Size of it	The thing that goes wrong if you hurry
Stop cut on marks	Sweep 7	Four millimetres, straight down on the line	You crush it
Lowering it	The same sweep 7, laid over	Down about 3/32	Going too shallow to undercut the thing afterwards at all
Rounding off the new ridges	A knife	Then a back bend, sweep 5, three millimetres	One too small
Three hollows	Sweep 8, an eighth of an inch	Only light cuts	Overdoing the outer two of them
The hidden channel running underneath the leaf edge	Veiner 11	One millimetre, going right down to the background behind	Breaking through

Does basswood hold detail this small?

Perfectly well, provided you stop pressing. Basswood gets a reputation for crumbling under fine work and my panel says the reputation is unearned, so long as every cut is a slice and none of them is a shove.

Harder timbers do make it easier. Cherry and walnut hold a crisp micro detail with almost no argument, which is one reason furniture makers reach for them. Basswood asks for a lighter hand and then gives you the same result.

Unnecessary cuts are what actually damage it. Every extra pass near a fine edge is another chance to bruise the fibre, so I decide what a section needs before the tool touches it, then make that many cuts and stop.

How do you carve a concave inside a concave?

With a gouge smaller than the one that made the first hollow, running down its middle. That is the trick nobody expects, and it costs almost no time at all.

Start with the main concave, carved with a number 8 at an eighth of an inch, which gives the illusion of the form rising out from underneath. The two outer subdivisions get their own smaller concaves beside it. Three hollows in a row, each one shallow.

Then take a sixteenth of an inch gouge and cut one more concave directly inside the middle of the first. Barely noticeable, which is the point. What it creates is a second line of shade sitting inside the first, and under raking light the eye reads depth it cannot account for.

Concave sides are characteristic of the whole family, incidentally. Look at Romanesque acanthus and you will find the same hollowed flanks on the leaves, cut inward exactly this way, so this is not a Roman peculiarity so much as a habit the classical orders handed forward.

Overdoing it is the one real risk. Each of these is small and the temptation is to make them decisive, and a decisive micro concave is simply a groove. Cut it lightly, stop early, and let the shadow do the announcing.

Rounding the ridges is what finishes it. A cheap knife starts that, then a back bend at sweep 5 and three millimetres takes over, because a pipe is meant to read round rather than faceted. Mine turned out slightly too small for this particular pipe and I said so at the bench. A wider one would have blended it in fewer passes.

Grain direction decides which way each of those rounding passes travels. Get it wrong on a ridge this fine and the fibre lifts instead of shearing, and there is no material left to recover from.

Why does a back bend matter for the deepest shade?

Clearance. Every straight tool eventually runs its handle into the carving standing above the hollow, and at that moment the blade stops regardless of how sharp it is.

Mine for this is a sweep 7 bent tool by Ashley Iles, marked 37, roughly 2.4 mm across the tip. With nothing fouling the handle it goes as deep as I want, right the way in, and the shade it leaves is as dark as anything on this kind of panel.

Without one you are not stuck, and I show the alternative on camera. Angle a straight gouge hard over and work it in on the slant, and you will get nearly all the way there. Nearly, and the difference lands exactly where the shade is meant to be deepest.

Own a few, in different sweeps, if you are serious. A real recommendation, not a shopping list. Even a couple would change what you can reach.

For cleaning out the debris afterwards, a spoon gouge at two millimetres in sweep 9. Tiny, and it reaches inside the excavation where nothing else will. Against the grain in places, so multiple light passes rather than one confident one, and a genuinely sharp edge throughout.

Why carve a channel nobody will ever see?

People do see it. Ask me whether it is necessary and my honest answer is that it is not, and I know carvers who never bother with anything that will not be noticed, and I understand them.

People who look at my work look closely. They get down near it, they tilt their heads, they find the places where the leaf lifts away from the ground and they look underneath. Somebody always does. And every small detail makes a big difference to what they find there.

What I am cutting is a narrow channel with a veiner, a number 11 at one millimetre, running underneath where the leaf floats clear. Visible only from close range and from below. It can go all the way down to the background if I want, and then the shadow under that leaf turns properly black.

Ask me whether I have to, and the answer is no. I like to. Those two sentences are most of what separates decorative work from ornament that people remember, and neither one is about skill.

Careful here, though. The wood on that side is thin and I cannot undercut further in that direction without coming through the face. So the direction of every cut gets chosen deliberately, and rushing would cost me the leaf.

How do you control a tool in a space this small?

Pressure is the enemy, and fingers are the answer. Create real force in a fragile hollow and you can wreck it in one movement, so the whole technique becomes about applying almost none.

Watch my hands rather than the blade. One finger presses down on the shank partway along, using it as a lever so the cutting force comes from a short controlled arc. My other hand grips the panel itself, tight, and gives that finger something solid to work against.

Pinching is the grip. Not a fist, not a mallet hold, just thumb and finger on the steel near the edge, and the tool travels a few millimetres at a time. Slow, and far more material comes off than you would guess.

Rotating a small gouge into the wood like a drill bit is the other move worth stealing. It opens a hollow without levering sideways, and levering sideways is what breaks both leaves and tools.

Stop when the piece feels wrong. A panel this fragile tells you, and the correct response is to move to a different area and come back tomorrow. Nobody ever regretted putting a carving down.

Worth saying that the spoon gouge turned out unnecessary here. I reached for it, used it, then tried the same passage with a plain three thirty second straight gouge and found it managed perfectly. So I put the spoon back. Reaching for the specialist tool out of habit, when the ordinary one would serve, is a thing every carver with a full rack does.

Then a sweep 5 at a quarter inch lowers the last standing ridges and blends them into the surfaces around them, and the ornament stops looking carved and starts looking grown.

Would you like to see the small things a photograph never shows you, carved slowly with the camera close in? Shall we carve?

This workshop sits inside the Roman acanthus series, and tuition opens it with the rest of the library. Workshop ten took the leaf underneath itself, the Greek leaf uses none of this ornament, and the design principles underneath all of it are their own course. Rather have me beside you? We teach in person, and a beginner starts here.