Alexander Grabovetskiy, woodcarver, School of Woodcarving · Published 2018
How do you carve an acanthus claw?
Quick answer: An acanthus claw is the curled point where a leaf ends, and carving one means dividing that point into three, then hollowing beneath every part. The centre goes furthest down. A small round eye sits in the widest section, and undercutting at the finish supplies the shadow.
A claw is the hooked tip of an acanthus leaf, and it is carved as three divisions with a hollow under each. Separate them with a V tool, sink the middle one deepest, put an eye and a pipe inside the largest, then undercut until the whole hook throws a shadow.
The word acanthus itself comes from the bear.
the plant itself carries the common name bear's claw, and I mention it on the recording because it explains the shape better than a drawing does: the tip curves over, comes to a point and hooks. I sit at a bench calling that same tip a claw for the same reason, and workshop 16 of Carving Hashemi Flowers is an hour spent on one of them.
What a claw is made of
- Three main divisions.
- Something hollowed out beneath every one of them.
- One eye and pipe in the widest.
- A scooped channel down each flank of that raised line.
- Undercutting.
Why three?
Three gets marked out on camera, and three is what the plant gives you. A centre with something either side of it is a balance the eye settles into without being asked, and it is the arrangement a classical acanthus leaf repeats the whole way down: one dominant lobe, smaller ones flanking.
I mark the divisions before anything else. Once they are cut everything afterwards is just depth.
Which cut comes first?
The V tool, every time. I take a number 16 at six millimetres and run it along each division. That groove does two jobs: it tells me where the shape is going, and it gives the gouge somewhere to put its chip later. Without that groove the gouge has to lever, and levering on a fragile panel is how hoiking ends badly.
Deep, too. On the middle division I go as far as the material allows, since that one is meant to disappear beneath the other two entirely.
How do you get underneath?
A quarter inch number 5 goes in first, then something narrower finishes it. I work in what I call a zig and a zipper, which is nothing more than alternating short slicing cuts from either side of the groove until the wood between them gives up.
Slicing matters even in a space that small. A pushed gouge inside a 6 mm hollow crushes as much as it carves, and the crush shows the next morning when the fibre springs back.
Then a small number 9, or a short bent V where nothing straight will fit, and the excavation is done.
Why does a short handle reach where a long one cannot?
Geometry, and it has nothing to do with the blade. The concave cut beside the eye gets attacked from below, travelling uphill, which means the handle has to sit low, almost against the panel. A standard handle runs into the work before the edge gets where it needs to be.
| Which tool, and how big | What it does on the claw | Its limit |
|---|---|---|
| V tool, number 16 at 6 mm | Marks all three divisions before anything else happens | Nothing, since it goes first |
| A quarter inch number 5 | Gets underneath the middle one, which goes deepest | Runs out of depth |
| A stub handled number 8 at three sixteenths | Opens the eye | Nothing worth naming on a claw this size |
| Number 7 at ten millimetres | Blends pipe to leaf | Reach, once the leaf above closes over it |
| A five millimetre spoon gouge | Reaches the inner leaf | Aim |
The Dockyard gouges I reach for here have unusually short handles, and beside an Ashley Iles of the same sweep the difference lives in the handle rather than the steel. That stub is the whole reason they are on the bench for this job. The blade is ordinary. The angle it can reach is not.
How big should the eye be?
My usual is one eighth, and on this one I went to 3/16 because the claw itself is large. An eye sized to its leaf reads as part of the design. An eye at a fixed size, dropped into every leaf regardless of scale, reads as a stamp rather than as growth, and once you have seen that on somebody else's panel you cannot unsee it.
The same rule runs through all ornament. Full accounts of what a pipe is and carving the eye and the pipe sit elsewhere in the school, and the only thing workshop 16 adds is that a big claw earns a big one.
What do the concave cuts beside the pipe do?
They make the shadow. A pipe standing on a flat surface is only a ridge, and from six feet it reads as nothing. Scoop a hollow down each side and it gains two dark lines, which is what the eye picks up from across a room.
Nowhere else on an acanthus leaf does it work as hard, because the hook is already leaning toward the viewer. Light hits the top and the shadow falls in exactly the place you have just hollowed, which is the whole of making a leaf catch the light. This is the same argument as carving shadow into relief, applied to an area the size of a coin.
What is wiggling, and why is it a compromise?
A small sideways shimmy with the edge already buried. On end grain a V tool wants to split ahead of itself rather than sever, and rocking the tool a fraction as it travels keeps the point working instead of wedging. You can hear it on the recording.
I am honest about what it costs. Wiggling does not give the cleanest cut, and I say so at the moment I do it. What it gives is a cut at all, in a place where a straight push would lift a chip out of a division I have just spent ten minutes establishing.
So it is a trade, and the trade is worth making only where the surface will be cleaned again afterwards. For a final face, sharpen the tool instead. There is more about that in hollows inside a lobe, where the same problem arrives inside a lobe.
What is a leaf inside a leaf?
A second complete claw carved inside the first. Once the outer hook is shaped there is a pocket of material behind it, and rather than clear that as waste I leave it standing and carve a smaller version of the same movement in there, so the claw becomes double layered.
Nobody sees the second layer from the front. That is fine; it was never carved for the person at the far wall. Walk to one side, or let low light fall across the panel, and a whole second leaf appears out of nowhere.
That is the argument for depth on this panel, and it is the argument for every hour spent on undercutting work nobody sees as well.
Should a claw sit above or below the leaf beside it?
Slightly above, and slightly is the operative word. On this panel I lifted the claw a fraction proud of the section next to it, not enough to read as a separate level, just enough that a shadow line appears where the two meet.
Equal heights are the mistake. Two shapes finishing at exactly the same plane look glued together, and the eye cannot tell which one is in front. A difference of a millimetre settles the question permanently, and there is a whole method for the opposite problem, keeping two sides level, when two sides genuinely must match.
Undercutting then finishes the job, and it comes last. undercutting comes last for a reason: everything above has to be right before you take away the material holding it up.
How deep is too deep?
The wall behind tells you, and you have to keep checking it. I say out loud in this workshop that I am watching the thickness, because on an openwork panel the space behind has already been opened up and there is far less material than the front face suggests.
Beyond that, it is taste. I like extreme, and I say so, and in high relief the hollows under this claw go deeper than any drawing called for. A carver who prefers restraint will stop earlier and the panel will still be correct.
When smaller does not mean tiny
People hear me recommend a smaller tool and reach for an eighth of an inch, and then the afternoon disappears. Taking a claw down with a 3 mm gouge would take forever. When I say small on a shape this size I mean ten millimetres, which is already small by the standards of the rest of the bench.
The rule is proportion, not miniature. Pick the largest tool that fits the space you are actually working in, and accept that this changes six times within one claw.
What does honing mid-carving look like?
A stop, a strop, and back to work inside a minute. Partway through this workshop the number 7 started leaving a surface I did not like and I said on camera that it needed honing rather than sharpening, which is a distinction worth keeping: sharpening reshapes an edge, honing only refreshes one that is still there.
The full method is in sharpening, and what what stropping does to an edge does to an edge is worth knowing before you lean too hard on leather. At the bench the signal is the surface in front of you, not the calendar on the wall.
Does any of it look right while you are doing it?
Never, and that is the part nobody warns you about. Halfway through a claw the area is a mess of grooves and half hollows, and I say plainly on the recording that it looks messy and that you sometimes have to believe it will come good and wait until it is done.
Over forty years at this and the ugly stage still arrives. What experience changes is not whether it happens. It stops frightening you.
Why carve something a viewer will never notice?
They notice it without knowing they have. The deep work under a claw does very little from across a gallery, and then somebody steps close and the panel suddenly has more in it than they expected. That gap between the far view and the near view is most of what separates ornament that holds attention from ornament that does not.
Look at any Grinling Gibbons drop, the carved frames in the V and A furniture collection, or anything in the Getty rooms, from twelve feet and then from twelve inches. The second look is the reward, and it is built entirely out of hollows nobody could see in the first one.
What about the grain?
A claw crosses it at every angle, which is why the tool changes so often. Going up the hook you are usually with the grain; coming over the top you are into end grain; down the far side you are against it and the surface will tear unless the chisel edge is genuinely sharp.
My habit is to carve the easy direction first on all three divisions, then come back and do the difficult one everywhere, rather than finishing a division at a time. Fewer tool changes, and the difficult passes all happen while I am already braced for them rather than halfway through congratulating myself on the easy ones.
What to try on your own bench
Draw one claw on a scrap of basswood or lime, big, about the width of your palm. Cut the three divisions with a V tool and stop. Look at it. Then hollow under the middle one only, and look again, and you will see the shape arrive from that single cut.
Everything after it is refinement, and the refinement is where the hours go rather than where the shape comes from. Nobody believes that until they have carved the divisions and looked. All 49 workshops of Carving Hashemi Flowers are inside the school, and this is workshop 16.
One thing about the name. I called the shape a beak on camera before correcting myself to claw, and both words get used at benches, along with hook.
A botanist and a carver end up reaching for the same syllables. Carve enough claws and the naming stops mattering, because your hand knows the movement before your mouth finds a word for it.
Next workshop: the small details inside this leaf, which look like nothing on a drawing and change the whole panel.
How it's done
- Mark the three. Draw the divisions before touching a tool, because everything that follows takes its position from them.
- Groove each one. Run a narrow V along every division so the gouge that follows has an open side to work toward.
- Sink the middle. Take the central division deepest, since it is the one meant to vanish beneath the other two.
- Cut the eye. Open a small round hollow inside the largest division, scaled to the claw rather than to habit.
- Hollow beside the pipe. Scoop a concave down each flank of the raised channel, which is where the shadow will sit.
- Leave the pocket. Keep the material behind the hook instead of clearing it, because a second layer can live there.
- Carve the inner claw. Repeat the same movement smaller inside that pocket so the shape becomes double layered.
- Undercut last. Take material from beneath the edge only once everything above it is settled.
Frequently asked questions
Why is acanthus called a bear's claw?
The plant carries that common name, and the hooked tip is the reason. I mention it at the bench because the picture explains the shape faster than any drawing of mine does.
How many divisions should a claw have?
Three on this panel, and three is what the leaf gives you almost everywhere in classical ornament. A centre with one shape either side settles the eye without asking anything of it.
Does a short handle really matter?
On a claw it decides what is reachable. The concave beside an eye has to be attacked travelling uphill, and a full length handle runs into the work before the edge arrives.
What to carve next