Category: The Intelligent Hand

The following is excerpted from “The Intelligent Hand,” by David Binnington Savage. It’s a peek into a woodworking life that’s at a level that most of us can barely imagine. The customers are wealthy and eccentric. The designs have to leap off the page. And the craftsmanship has to be utterly, utterly flawless.

---

I am jumping now to what a student experiences during the first week at Rowden. I am doing this because it all fits together. Without the doing, the making, the faffing about in the workshop, all the drawing and waking up, there is no context for when you become a designer and a maker.

It’s not good enough to sit in a nice clean design office and get your sweaty minions to make for you. Making is what you do. Remember William Morris, and how he was always fiddling with making something or other. Fail to grasp this, and the maker will always be in charge of the dialogue.

“No boss, that won’t work – you need three fixings there.”

You need sufficient understanding and knowledge to argue. You need to know enough to suggest a different fixing, and to maintain the smooth identity of your design.

So pick up that plane. It’s on its side on the benchtop. Wait – maybe first let’s have a look at what we have here. There is a proper full-sized cabinetmaker’s workbench, about 7′ long. It has a newly planed, flat top. The top is beech or maple, and about 3″ thick; the undercarriage is similarly heavy. A good bench should be solid, and not gallop about the workshop the moment you put the pedal to the metal.

Look at that benchtop. Many makers may have worked there before you, but it should be in pretty good condition – if not unmarked by their work, it should be at least a respectable surface. There have been accidents, yes, that caused the odd bit of damage – but it’s a dead-flat surface. We need flat, especially around the end vice, because that is where we work. The flatness of the bench transfers to the job; a hollow near the vice would show up in thin components planed on that bench. The bench has a front vice and an end vice. Working with these vices will be dogs. Whuff, whuff. No – these are pegs that fit in holes in the bench; they are used to secure your work.

The bench is probably the most important tool you will ever encounter as a maker. Later, you will make your bench; it will become the foundation upon which other work will be made. Right now, you’ll use one of the Rowden benches. Bench height is important. Stand alongside the bench; let your arm hang and bend your elbow just above the bench surface. Now spread the thumb and index finger of your opposite hand. Add that distance to your elbow. This is your bench height.

It’s good to work with a high bench because it protects your back. Much of your work will be done not with your arms and shoulders, but with your trunk – the core muscles – and thighs. This again will protect you from injury. There will be times when you need to get up higher and get on top of a job. Then, use a small “hop-up” – a 3″-high box that you stand upon. This lives under the bench’s bottom rail.

Then there is a bench light. It should be a decent, bright source of light that you can direct into the dark corners of your work. Without a good light you will not be at all times able to see where you are going. A quick maker will be pulling that light around as they move about the job. You need to see the line as you cut.

Then there is a box of hand tools – yours until you can sensibly choose your own. These are prepared tools; all edges have been sharpened by Jon Greenwood to a keen edge. They should all work straight out of the box. All the boxes are different; they hold the same tools but from different suppliers. Spend time trying out different chisels and gauges to help you make informed buying choices. Hand tools have to fit your hands comfortably; different brands offer different solutions at different prices.

With these three – a bench, a light and a box of basic hand tools – you can make a lot of things and need nothing more. Machine work can be done by a local shop; pay them by the hour. You might want some power tools to help out later on, but they are not for now. Most machines are about saving you effort and energy; you want to engage them for those reasons – not because a table saw is more accurate than you are at sawing a straight line.

On the bench, there is a piece of walnut, about 18″ by 2″ or 3″. This is to teach you about wood and tools at the same time. Look at the wood. It has been selected because it is mild and well-mannered. It will have a sawn surface, but you will be able to see the grain of the timber. Think of it as the fur of a cat – which way would rough it up, and which way would lay smooth, if you stroked it? That’s the way you plane it, smooth. Place it the right way around between the dogs and tighten the end vice to just nicely hold the job.

This is the bench doing its job in superb fashion. Now you are free to dance about the workshop, waving a No. 6 to your heart’s content. You don’t have to hold the job – it’s fixed in solid position on the benchtop. All actions now are down to you – how you hold and present that edge to the job. This is you, a sharp edge and a piece of timber. Listen. Attend to what that sharp, well-adjusted tool tells you about the surface in front of you.

Is the note the plane makes a nice, high WHUZZ, or is it getting lower, telling you the edge is getting dull? Watch the shaving as it emerges over the frog – is it a clean, full-length shaving or is there a slight hollow just over there? You want information about that surface; the plane is your primary tool of inquisition.

Consider your arms and shoulders during this process. The work should be coming from your trunk and thighs, not your arms. Imagine a piece of string tied around your plane handle and attached to your right nipple. If you push with your arms, it will pull on the string and hurt. Instead, use your core and legs. A long shaving then becomes like a dance step: forward step, forward step, forward step.

Plane a dead-flat surface on one of the 3″-wide faces. This is your “face side.” It is important, as this is the side from which all other faces are measured; get this wrong and you are in the poo. From day one, hour one, you are faced with quality. Screw this up, go too fast and slip it in under the bar, and it will come back later to bite you in the bum.

Flat is a terrible monster and can drive a newbie bonkers. But don’t let it get you. The sole of your plane is flat – I promise you – so use it. First, let’s look at the side-to-side movement of the plane. As you take a shaving, you can get it only about 1″ wide from the middle of the mouth of the plane – maybe less. If you are observant, you will see that the shaving is slightly thicker in its centre, thinning to nothing at the edge. This is deliberate; you want this.

Run the plane down the job, aiming to get a complete shaving for the whole length – then run another alongside it with a little overlap. If necessary to help you visualise, put a pencil mark across the job and plane it off. Suddenly, this is fun! The plane is working and you are doing the job – you are almost up to your knees in shavings. But the joy of taking long, ribbony strips of fragrant walnut is suddenly ruined by Jon Greenwood coming to your bench with a straightedge.

“Let’s see how flat it is.” What is usually the case is that you’ve been planing up onto the job and down off it, creating a crest in the middle. So, you learn to overcome this by applying more pressure to the plane’s toe as you come on to the work, and to the heel as you come off it.

Now check the surface with a straightedge, and hold the job up to the light and see what is going on. (You have a straightedge on the side of your plane; many skilled makers use that, as well as a purpose-made straightedge, for this and other jobs.)

High points can be removed with stopped shavings that address that area. As you approach flat, back the iron off to produce the finest shaving; this will allow the sole of the plane to have more effect. You might find yourself chasing a minor bump. Don’t do this; try taking a series of fine stopped shavings to the centre of the job, leaving pencil marks on at either end. Then, with three full-length passes, go right through, clearing off the pencil marks at both ends. There you are, nearly there.

Now check it for wind (that’s pronounced “whined”) by using winding sticks – one at each end, and peer across them from one end. Wind is twist. You can have a seemingly flat surface that is actually the shape of a propeller. If the winding sticks are out of parallel, they will visually tell you. This is the last check. You now have a face side to be proud of.

Mark it with a Face Mark, cannily pointing in the direction of grain – the direction the plane or machine must follow. It also points to the Face Edge which is the second surface. This is also a primary surface from which others are measured, so “heads up.”

Remember these steps. You may never do them again with a handplane but you will certainly do them with a pile of timber and a machine shop. The steps are identical: Face. Edge. Width. Thickness. End. Length. We use “FEWTEL” as a way to remember the order. You’ve just finished the first one. Now it’s on to the edge.

The objective is straight and flat on the narrower surface of the edge and, this is crucial, square to the face. Use a square to check this. (A good square can keep you safe; a rotten one, one that has been dropped, is a traitor in the camp that will undermine all your work.) Start off by sliding the stock of the square down the face side, and watch how the blade travels down the edge, looking for gaps between the wood and the blade. Use a bright light behind you to really see what’s going on. A square can easily be banged onto the job and tell you lies. Don’t let it.

Adjusting an out-of-square edge requires a well-set-up plane and some knowledge.

First, look at the plane. The blade, as you noticed earlier, is just a little curved – not straight across. This is important. This gives you the shavings you have previously enjoyed, with thickness in the centre tapering to nothing on the edges. Now look how that blade is set up in the plane body. Look down the sole from the toe, and you will see the thicker, dark shape of the blade in the centre fading out on either side.

This curve gives you the opportunity to adjust an out-of-square edge. Check which way the edge is out of square, then move the plane to the left or the right. What? Yes – look at the way the edge is out of square – it’s higher on either the right or the left. The centre of the plane blade takes a thicker shaving than the edge. So, shift the plane body to the left (or right). Run it down the edge and see a thicker shaving come off that side and nothing come off the other. Without that curved blade, you would not be able to adjust the edge in so accurate a way.

So, engage your brain and eyeballs before engaging the plane. Analyse the situation. Do you need more off here or over there? The square will help – again, register the stock on the finished face and slide the blade down the edge. Use your damn eyes. Hold the job and the square up to the light – don’t be lazy.
Once it’s square and straight, you have a Face Edge (the second letter of FEWTEL). Put a small “f” on the edge, adjacent to the face mark. You can, if you want, lean it in the direction you want the edge to be worked. Good job. Move on to “W” – width.

This is a surface parallel to your Face Edge, and also must be square and straight. Now you need another tool – a marking gauge. Popular at Rowden are simple wooden gauges that (with a little work) scribe a clean line. However, there is another type, a cutting gauge with a small wheel on the end that I find is simpler for the beginner to use. Both are fine. Toolmakers want us to have both marking gauges and cutting gauges, so that they can sell us more tools. I believe in fewer tools that work better. A single sharp gauge should be able to scribe a sharp line both with and across the grain – but you will need a few gauges because they are often left set up for stages of a job.

Take the job, register the gauge off the Face Edge, and scribe a line as close to the sawn width as you can. Make sure it goes all the way around the job. Keeping its stock against the Face Edge will be tricky at first; take gentle stokes to avoid digging into the job.

Now, before we go any further, let’s have look at that little wheel that made that scribe line (if you’re using that type of gauge). Note that it is beveled on one side and flat on the other, and in this case the flat side is facing the stock of the gauge and the bevel facing the outside.
Plane or pare the edge to approach the gauge line. Try to do this evenly all round so there is a little bit of gauge line flapping all around the surface. Then take one plane pass down to the actual line. Smack on.

Now use your gauge, registered off the flat Face, to mark the thickness all the way around the job. Plane to the line, using the same process as in Step 1. Now you have the T in FEWTEL. Jon will come around with a caliper to check the thickness, to within 0.2 of a millimetre. (It should be 0.1, but we are cutting you bit of slack here.)

I had a great student named Martin Dransfield, a former Yorkshire miner, who stayed on and worked for me for a while. A student asked Martin, “How do your joints come up so tight and clean?” Martin, being a Yorkshireman, is a man of few words. He paused, rubbed his chin, then said, “Well tha’ just cuts to the f-ing line.” Which is true. Leave it on and you are effed. Go past it and you are equally effed, but in the other direction.

Thank you, Martin. You are well remembered for your accuracy and your economy of language.

A word on marking-out tools. A cutting gauge, with its bevel on one side of the wheel, is one of a family of marking-out tools that give you a perfect dimension, not an approximate dimension. Mark with pencil and you get approximate. A pencil line has thickness. The marking knife is another tool in that precision family. My personal favourites, made by Blue Spruce Toolworks, are expensive, but you get what you pay for. The blade is long, thin and sensitive. What do I mean by sensitive? Will it take offence if I shout at it, go off in a huff and roll off the bench? No, I mean it will enable me to click that blade into an existing scribe line and I will feel that click in my fingers. That’s what you need from a good marking knife.

Marking with a pencil is fine for some jobs – but not the next one. You’re now at the “E” and “L,” the End Length, in FEWTEL. Place your square across the Face Side, with the stock on the Face Edge. Hold it firmly between the finger and thumb of your off hand. Pick up your marking knife with the bevel facing the waste and strike a nice firm line against the blade of the square. The line should be clear and clean from corner to corner.

Now you need to transfer that line to the face edge. Secure the job in your vice and put the knife in the existing line right at the edge of the job. Make sure the end of your knife pokes out a little. Now slide the square right up tight to it. When you are up and snug, remove the knife then use it to strike a line across the Face Edge. If you are spot on, the line will be a perfect continuation of the line on the Face Side, and you can carry it on around the work to meet the Face Side line. If it doesn’t meet up one of two things have gone wrong: you’ve mismarked around a corner or the job is not square.

If the marking out is wrong you have no chance of good work. Step back to find out the problem, and start again – either the planing or the marking out, as needed. Engage brain before plane.

I am making this point: Making is an intellectual and creative challenge. It demands that you engage fully with every atom of your being. Your full concentration. Otherwise, you will mess up.

The essence of this is that there are no shortcuts to quality work. There are lots of easier, quicker, lower-quality ways that suit lots of other situations, but if you want to do it right there is no shortcut. We see this penny drop with people; it may take a few days, or a few months. Tom, a student with us now, sat with me recently on a log of Western red cedar and said, “David, I have had a really, really, great life, and this has been the very best year, so far. However, if I had come for a trial week I would have run a mile. It took me weeks to see that there is no shortcut.” Thank you, Tom.

This “Doing Thing” isn’t easy. You need to listen to the tool in your hand. That blade has an edge engaging with the infinitely variable material before you: timber. That cutting edge in your hands is the closest you will ever get to understanding this material. The distance between you and it is, at this point, the shortest it will ever get. The information is traveling down that blade into your hand. But do you have the wit to receive it? Our bodies are wonderfully receptive information centres. The body feeds us with information. Do we listen well enough to the sound of the blade in the timber? Do we listen to the note the saw is making, to the feel of the chisel in our hands? Do we weigh the push the tool is requiring and look quizzically at the edge of the chisel? I hope so.

This is why sharpness is so important. With sharpness you have less shove and more sensitivity, control and information. Dull tool/dull maker.

We are asking you to use your eyes in a new and more intense way. This opens the door to learn drawing (more on that to come) for each stage of the process. The ability to draw quick bench notes is an essential skill in making things; it enables us to resolve what’s going on inside that joint.

The eight-week hand-tool initiation ritual builds a solid base upon which to learn machines and other processes. This is tiring, ache-inducing, hard-won accuracy. This is accuracy you have never achieved before. That’s control. That’s you in control. I like that.

The following is excerpted from “The Intelligent Hand,” by David Binnington Savage. The book is a peek into a woodworking life that’s at a level that most of us can barely imagine. The customers are wealthy and eccentric. The designs have to leap off the page. And the craftsmanship has to be utterly, utterly flawless.

How does one get to this point? And how do you stay there?

One answer to these questions is in this book. Yes, the furniture can be technically difficult to make. But a lot of the hard labor involves some unexpected skills. Listening. Seeing. Drawing. And looking into the mirror and practicing the expression: “And that will cost 20,000 pounds.”

As you will see, it’s a personal struggle – like the production of this book. On the day David began work on his manuscript, he received a cancer diagnosis with a grim prognosis. He wasn’t sure what the book was going to be about or if he could finish it. But David attacked the work with the fervor of a younger, healthier man.

So what is it about? On the one hand, “The Intelligent Hand” is the story of a boy with a stammer who became one of the leading furniture designers in the U.K., working for clients all over the world, including Saudi Arabia and China. It’s a story of extreme failure – bankruptcy – and how he built a new life using the debris from the broken one.

It’s a practical and iconoclastic guide to getting started in woodwork. David has always had pointed opinions about the tools and methods his students should use to get good results. And he shares – in great detail – his recommendations for tools, sharpening, cutting dovetails and building a proper workbench.

---

There are times when I wonder what is going on. That’s a pretty usual state of mind for me. I seem to get nearer and nearer to waving my stick at the seagulls and shouting at passing cars. But when it comes to “making,” this disconnection usually involves piles of expensive jigs and the spindle moulder (or shaper, as you guys call it). I see a pile of jigs made beautifully and look at the product – in this case four desk legs – and wonder why the desk legs were not made by hand, just as quickly.

I have never been able to win this argument with Daren, and my last workshop sought to become more efficient by making more jigs. We even had a room full of these useless implements. “Ah we should keep the jigs; we may make a chair leg very similar.” That’s how the argument goes, but by that time three things have happened. One, you can never find all the bits of jig for the necessary complex and clever sequence of machining; they are hidden beneath the pile of subsequent jigs. Two, the maker who invented this clever contraption has long since moved on and exactly how these bits of MDF fit together moved with him. Three, your old jig never quite does the new shape and you end up making a new jig.

Many makers love making jigs. Jigs keep us out of the firing line. Making even four identical forms by hand, can of course be done, but when there is a machine that will replicate shapes, most makers will dive for the cover that a jig provides. And jig design enables us to show how clever we are. Well, here we go.

Steve has made a box jig (so-called because it looks a bit like a box and the job fits inside it). The idea is clever and sensitive to the material and the design. (Note that in the photos above, Steve is running a test leg in magnolia, which is why it looks greenish.) In this case we have legs that curve out at the bottom. This means that the grain of the pear would be roughed up if machined in one direction. So, he made this jig, which enables him to turn the job over and machine the short curved ankle from one end and the long straight leg from the other direction. That is because he has two identical guiding surfaces, one directly above the other. You can see in the photos the red line that tells him to stop and turn over. Clever jig, this be.

The photo above shows the cutter block running above a guide ring. The jig is in contact with the guide ring, and the cutters trim the job above to the exact size.

Spindle moulders really are great tools. They do the same thing as a router inverted in a table, but they do it better. The heavy mass of the cutter block provides heft through the cut, so the block spins at lower speed than a router and gives a cleaner cut.

Don’t look so surprised, Steve (see above). Handwork was called “Workmanship of Risk” by David Pye in his wonderful book “The Nature and Art of Workmanship” – what you are doing is “Workmanship of Certainty” – this is the industrial process of replication coming into small workshops.

OK – back to the bench shop. This pair of legs with two rails is what we are after; these will be the outer frames joined by a rail at the back. Each of the two frames has two side rails – a bit of overkill, but our structure is potentially wobbly in the middle so we need stiffness at the side structures, and we need to transfer that stiffness to the centre. We will do that with the addition of the drawer carcases joined to the leg frame. We don’t know for sure that it will be stiff enough – had we made a maquette we could have checked that out – but we are pretty confident. Furniture makers don’t have structural engineers to tell us the tensions involved. We only go on what we have done before, the stuff that hasn’t yet fallen down.

First Steve has to take the machined surface off the legs. Above, he’s working on the twisted surface of the front leg. Careful shaping with a spokeshave is needed. The components are joined together with Festool Dominos – small loose tenons. They give us enough strength for the joint and eliminate a deal of semi-complex joinery.

This, at right [above], is our disc sander, which gets a lot of use. Steve sets it up square to sand the top of a leg component, as shown. He then takes the leg to the table saw and puts it in another jig to cut the foot to length. All the legs will be cut in this jig, so it is essential that the stop’s end is sanded square so we get accurate Leg lengths. We learnt this by making batches of two dozen chairs that didn’t wobble!

This is a big moment – getting the job to stand on its own feet so you can have a good look at the proportions and see what needs fixing. So far so good, but we next need the drawer compartments made and fitted.
The exhibition date is looming large, and both Daren and Steve are working weekends – my wages bill is eyewatering. But Daren has not yet booked the van to deliver the piece to the exhibition, which worries me. When he does that, I will know we are going to make it.

OK, last lap. Now it’s on to the final joinery and polishing before assembly. This is critical; you do not want to be polishing into corners if you can avoid it. Above, Steve using our standard shellac and wax finish for secondary and inside surfaces; apply it with a rubber that is kept in a glass jar. The rubber has a lambswool or wadding interior and fine cotton exterior. This holds the polish. As Steve moves the rubber and presses down, the polish is released. The shellac polish in the small bottle is very dilute, maybe less than a 1 lb.-cut (one pound weight of shellac dissolved in one gallon of alcohol). The other bottle contains mineral oil – use just a small dab to keep the rubber moving.

The aim is not a full shine, but to seal the surface and finish with a good wax polish. This is a quick and easy finish in the workshop, as shellac flashes off very fast. You wouldn’t want to polish with someone sanding on the next bench, but you can generally polish in the workshop without too many special measures. If I had a high-shine job to do, I would try to find a dust-free room, or I’d do the polish at the weekend when the dust had settled.

Let’s take a careful look at the glue-up above. There are eight cramps and only four components. Look at the angle-cramping block at the top of the far leg. Notice the support blocks between the double rails to stop them from bending under pressure. Notice also the small paper-faced cramping blocks to stop the blocks from marking the new polish. Ye Gods – the care these guys are now taking as they get near to the finish line.

Now we can see the drawer carcases fit. These are critical to the structure, providing stiffness to the front of the desk. Note how we avoid work that is not necessary; lippings on MDF components that are never seen after assembly are omitted. If, however, this area would be visible, even to a 4-year-old crawling underneath, it would get lipped, then veneered and polished. That 4-year-old could grow up to be a customer.

The following is excerpted from “The Intelligent Hand,” by David Binnington Savage – a peek into a woodworking life that’s at a level that most of us can barely imagine. The customers are wealthy and eccentric. The designs have to leap off the page. And the craftsmanship has to be utterly, utterly flawless.

How does one get to this point? And how do you stay there?

One answer to these questions is in this book. Yes, the furniture can be technically difficult to make. But a lot of the hard labor involves some unexpected skills. Listening. Seeing. Drawing. And looking into the mirror and practicing the expression: “And that will cost 20,000 pounds.”

If I were learning to create presentation drawings now, I would probably not be learning analogue skills, but instead go straight to computer-aided design (CAD). A good-quality program that will link properly to a computer-aided manufacture (CAM) program would be the tool here – not a cheap 2D version, but one that will show you the job, turning it around to show it from all angles. I am reliably told that a good two years should see you fluent in CAD. We teach the possibilities of CAD at Rowden, using Rhino, but getting fluent takes time; it’s a complex program.

However this is for Lost Art Press – and doing decent watercolour presentation drawings is getting to be archaic. This is a shame, as good watercolours send a clear message to the prospect that we have “a creative” in charge here. That’s a good thing.

Have a look at nearby printing facilities. Heavy watercolour paper is like thin card stock – it’s so thick. If you can find a printer to accept this thick paper, you can print from CAD using a fine grey outline that can be gone over with pencil and watercolour to a similar effect.

I wrote earlier about the tools and what you are attempting to achieve. Now I will be specific about how these drawings came about.

First the chair. I had the sketches; I would never go straight to watercolour without a sketch to work from. I want to improve upon it here, while sticking with what I’ve downloaded [the idea from his mind] to keep the proportions and the idea nice and clear.

The side elevation was first; this was a matter of choosing a scale to draw with. I knew my paper was going to be in landscape orientation, as I want the side elevation with the front elevation alongside so I could transfer dimensions from the first drawing to the second. If need be, I work a plan drawing in there some-where. The layout of the page is critical. So, I chose a scale of 1:5 to nicely fill the page.

I use what is called a scale rule – a triangular cross-sectional rule with several scales; 1:10, 1:7.5 and 1:5 are the three I use most. The paper is A3 (11.7″ x 16.5″) 140 lb. watercolour hot-pressed paper – really heavy and super smooth. I fix it to my cedar double-elephant-sized drawing board on the left in the bottom quarter, fixing on all four corners with translucent drafting tape. Over the years a mound of this tape has built up in this section of the board – a witness to the hundreds of drawings I’ve done on it.

I put the paper on the left to be near the ebony edge of the cedar board on which the T-square slides. I got rid of parallel motions and clever architects’ drawing aids years ago in an effort to have only simple tools that would always be accurate if I held them correctly.

Double elephant – huh?! Well, yes – it’s not a mea-sure often heard now, but it was a size of paper when paper was not measured in A-whatever or inches. In the old days we had elephants as a way to measure the size of a sheet of paper. This is the biggest drawing board I could find; it’s about 42″ x 31″. You need a big drawing board and a block of wood to support it at an angle on your bench. It’s got to be big to do perspective drawings – but that’s for later.

The first line on the page is the floor. I look at how large the chair might be, how high and how wide, then work out where the side elevation should sit on the paper at 1:5. A line for the height of the seat goes in next; here, it’s 14-1/2″; if the seat is to compress when you sit on it, I put this line slightly below the top of the seat.

Now I put two lines down for the back leg and the seat. I look hard at the sketch and how that seat slopes and how the back leg leans way back. Here, I use one of my two big French curves mentioned earlier, the one that is a big spiral, with the radius opening out as you go around it. I used this tool repeatedly on this drawing, as I wanted a family of shapes that would have the same feel, the same DNA.

I do this a lot on chairs when I’ve designed a curved back leg shape – using the curves to develop the front leg and arms, while trying at the same time to use it differently to create different shapes with the same origin.

Having put down the side elevation to the best of my ability, I think about the front elevation and the plan. Chairs are very three-dimensional; they have to look good from all around. This means thinking in three dimensions, accepting that this side elevation has a consequence on the front and plan elevation. You find yourself chasing around the image. Top. Bottom. Side. Side. Top. Bottom. It’s a subtle form of madness.

Bang in the first lines for the front elevation – the baseline or floor. Do this as an extension of the line on which the side elevation sits. Next, put down a centre line at 90°; this places the front elevation there. The front legs are next. I have an old American-made Windsor chair in the studio that I keep as a chair totem. I look at it and take measurements from it when questions come up such as: How wide should a chair be? How high should a seat be? This old chair is strong, light, comfy and good-looking all around. All I have to do is that – but totally differently.

So, the front and back leg positions are added. The front legs are 550mm apart overall, and the back legs lean out, opening the chair back up. I put lines in from the side elevation across to the front to show the top of the crest rail and the bottom back of the seat. OOOooo…this makes a nice square – or it will when I tighten it up. This is what I am doing at this point – tightening up the proportions, hiding squares and rectangles, making shapes I like, then checking them against Euclid.

You can see the seat plan view (at left) going in above the side elevation. Once I positioned the top of a leg on it at a 550mm width, I saw a problem. If that leg is to stay there, the side elevation must change; the leg must come further back. Grrrrr….

When this happens, you need look at what you have already done and accept that it can be better. On the paper, it’s easy to erase lines. I cut a sharp edge on my eraser so I can use it really accurately – just that line. Then, with a really sharp 3H pencil, I set about making the change. It’s always a tussle, and you can see some of my old lines compressed in the paper.

My drawings are nothing like it, but if you look at great drawings such as Leonardo’s “Vitruvian Man,” you’ll see they often have an air of having been carved in the paper – with many corrections and amendments.