Newsletter Article: My Dovetail Evolution
My dovetail journey started in 1983 and while I can’t remember all the early details I do remember finding a copy of Ernest Joyce's book “The Encyclopedia of Furniture Making." This served as a foundation for my journey into learning how to hand cut dovetails and eventually teaching thousands to do the same. I put off my first attempt for a long time out of the fear of failing. This hamstrung me until my first year at Brigham Young University (BYU) where I took a furniture making class. I built a coffee table in Red Oak. It had two drawers in the center section and I was determine to hand cut the dovetails. Oak is a tough wood to cut and chisel, so I wasn’t taking the easy road on my first attempt. I did it, tails first, with half blind dovetails on the drawer fronts and through dovetails on the backs. As proud as I was I do remember thinking I could or at least needed to get a lot better at this. I don’t recall having any help other than the book I have already mentioned. I think that table is still in existence, I gave it to a good friend in Utah before we moved and I doubt she disposed of it. Blackmail for future use!
My first experience receiving professional instruction on hand cutting dovetails was attending a half-day workshop in Atlanta with Ian Kirby, around 1985. I was gobsmacked to put it mildly! I hung on every word and I espouse many of his methods to this day. Years later, Ian was the first to show me the use of a rabbet to align the tail board over the pin board during the transfer. He used the table saw to cut the rabbet and I later adapted the use of a Lie-Nielsen skew block plane to compliment my “hand tools only” approach.
My first dovetail saw was an inexpensive gents saw with small teeth, lots of set, a blue blade, and a red painted handle. I found it in one of the shops at BYU, I used a sharpening stone on the sides to reduce the set making it track much better and produce a straight cut. To help gain muscle memory I carved a flat spot on the back of the handle to produce a flat that registered against my palm just below the thumb muscle. This was a big gain; when I picked up my saw now it fit in my hand the same way each time and gave me a “feel” for where the blade was pointed. This led me to understanding why a pistol grip is superior in developing muscle memory, the saw registers the same way every time it is held. This advantage can’t be overstated, in a dark room the handle alone tells you where the blade is pointed, not so with a round “file” handle.
Something else I learned from Alan Peters was his divider method of laying out dovetails. This was the simplest and most precise method I had seen and I still think it is the best layout method. Alan's method uses two sets of dividers allowing for accurate layout of the tails. One divider is used to establish the outside half pins, measuring in from the edge of the board about 1/4”. The other divider is adjusted to the width of a tail and an inside full pin. By stepping in each direction starting at the half-pin mark, the dividers leave a small hole that later gets used to register you pen or pencil to then draw the square mark at the top of the tail board and the angled mark drawn down the face. This method is accurate and fast regardless of the width of the joint. I asked Alan how he came up with it and as best he could remember he discovered in his fathers shop as a youngster experimenting.
My fascination with the dovetail joint was fueled as got better at cutting dovetails. Soon I was dovetailing everything in sight. It wasn’t long before I was consistently assembling tight, gap-free dovetail joints directly from my saw cuts; no chisel paring required!. My tool crib had not changed much, I was using the same red handled blue bladed gents saw, I used a fret saw to remove the waste the way Ian Kirby did. It had a cast aluminium frame, the blade on this saw was held in place with small wedge-shaped clamps that were free to pivot on their contact points as you sawed. I remember how frequently the blades would break from all the flexing, a problem needing to be solved. My marking knife was an old red Stanley I had found in a school shop cupboard, I modified the blade by stoning a long bevel to allow closer marking of the tails to the pins. I used this same knife up until 8 years ago when I dropped it and broke the blade. Coincidentally, I had just received the first prototype copy of that original knife that is now my current knife manufactured by IBC and painted red like the original.
My chisels were Sorbys and had been modified as per Alan Peters. Beveled edge chisels at that time (mid 80’s) all had bevels incorrectly placed on the top instead of the side. The term “beveled edge” describes a chisel capable of fitting into the tight corners between two tails so as not to bruise the side of either tail. Since manufacturers had seemingly lost their way on this feature, Alan would grind a half oval shape on the top side of his chisels, this was easier on the hand and would easily slip into the corners without leaving a trace on the other tail. Years later I worked with IBC to correct this problem when we designed their current chisels.
The next step of my journey was to gain proficiency, mostly by increasing my speed. This is where practice came in, each morning before classes began I would show up at the BYU shop and cut a full row of dovetails, about 4 inches wide. Good fundamentals are critical, no use in getting fast at doing a lousy job. I seemed to be getting it right as well as getting pretty fast.
The summer of 1987 found me working as the shop assistant at Anderson Ranch Arts Center at Snowmass Colorado. I remember the thrill of having Alan Peters admire my joints, he even bought one of my dovetailed wood hinge boxes that he later described as “A perfect box!" I suppose once you can produce a dovetail that has no gaps, looks pleasing to the discerning eye and can be done in a timely manner, all that is left is to be able to teach it to others. This leg of the journey began shortly after my return to BYU from the Ranch in the fall of 1987. Professor Ed Hinckley suggested I offer an evening workshop to those interested. We ordered in the necessary tools as part of the admission price, everyone got the dovetail saw, marking gauge and the fret saw. I really enjoyed teaching this class and we went on to do several more. I continued to do these half day classes after moving home to New Brunswick, Canada in 1989 and they served as a second source of income to a growing family.
10 years into my furniture business and the painful truth became glaringly obvious, the public was not interested in custom made furniture at a premium price. I could build all day long if I was willing to compete with big box stores on price while offering custom work. By the end of 1999 we were down to our last few hundred dollars when two events came together. The first that should have long ago been obvious, but I had just came to realize, the only folks that appreciated what I was building didn’t want to buy it, they want to learn how to build it. The second was an out of the blue invitation from Tom Lie-Nielsen to bring his tool business into Canada. These two events convinced me the only way to stay in the woodworking field was to become a teacher of the craft. Some might think “salesman” but in this field you had to teach your customer how to use the tool they were wanting to buy. Folks didn’t know how to sharpen their tools and most thought hand-cut dovetails were well above their skill level.
My hand-cut dovetails were a big hit at wood shows. My dovetail demonstrations drew large crowds that unfortunately irritated my vendor neighbors. My crowd would spill over into their booths and make it impossible for them to conduct business. A tough problem for both of us but not something I could help. Because of this exposure It wasn’t long before I was being hired to teach at various Woodworking schools and guilds around the country. As the appreciation for what I was doing/teaching rapidly grew along with the income from tool sales, my goals changed from wanting to be the best at making to being the best at teaching!
I can’t recall how many different venues I have demonstrated in or even how many cities but for a long time I was doing this an average of 3 times a month. Over the years I had made or modified tools to do the job of cutting dovetails better and easier. Requests for these tools started to grow and left me in a tough spot. How do you profess the ease of dovetailing when the tools your using are not available to those you were teaching? In 2000 I switched to using Lie-Nielsen's dovetail saw with no regrets. Up to that point I had accumulated 16 different saws trying to find the “best”. Everything where I lived was mail order so there was no “try before you buy”. The mailman brought disappointment 16 times. These saws were poorly built, crosscut teeth instead of rip, huge and often uneven set and generally terrible tools. The Lie Nielsen dovetail saw had rip teeth, was well built, and had minimal set. It cut right out of the box, if you knew what you were doing. However, I came to find 50% of the folks wanting to learn dovetails could not make it work accurately, it cut fast but was too hard to start.
Over the years as the age of my average student climbed, I started to focus on making tools better suited to the prospective user. Modern hand tools were most often made as direct copies of antique counter parts. Think about this and the problem becomes apparent. A saw made in 1830 would more likely be bought by a teenager heading to an apprenticeship than to a 60-year-old retiree, the difference being good eyesight vs. not so, flexibility vs arthritis and agility vs, you get the point!
My first big tool venture was my dovetail saw. I had been making other tools, but this was a game changer. In response to folks not being able to accurately and easily start their saw, I had come up with the idea of combining the advantages of a large, fast-cutting, saw-tooth with a small, smooth-starting, saw-tooth. The result was instant success for accurate saw cuts, the answer to great dovetails. Chris Schwarz declared in Popular Woodworking magazine that it was the easiest starting western saw he had ever used. With rip teeth and just .002” set per side, even in a novice’s hand it would cut laser straight. A saw that cuts straight produces a flat surface. A good dovetail joint requires two flat surfaces coming together for a gap-free joint. Once the novice woodworker learned to aim my saw, the saw will cut straight, and they would be 70% on their way to great dovetails.
The other tools I had already been making included a better dovetail marker that I learned of from Earnest Joyce. The idea was to limit the steps thus limit the mistakes. This marker allowed for the square mark across the end of the tailboard and the sloped mark down the face in one setting. My fret saw, in the hands of newbies, routinely broke blades and needed upgrading. I don’t remember where I found the one I currently have but it worked well by holding the blade tight enough to prevent bowing during the cut, even in thick hardwoods. The draw back was with the handle, too small and smooth for senior adult hands and it frequently pulled off the saw frame. We solved the first problem by taping the handle like a hockey stick. This enabled folks to better grip the saw and instantly improved their performance. Drilling and pinning the handle solved the other problem. The 3-inch throat makes for a light, balanced saw so to remove the limitation of a 3” throat, a quick 30 degree twist of the blade at the clamping points kept the frame above the end of the board during a horizontal cut.
Marking knives were a serious challenge, most had VERY sharp blades with acute bevels. These were usually rigid making it hard to follow the side of the tail being traced into the end grain of the pin board. Because it was so sharp, the blade would want to follow the grain instead of the side of the tail. My old red Stanley had been the perfect tool for this job, blade was just the right stiffness with enough flex to allow a slight bend as you pressed it against the tail. It was wide enough that you had lots of reference surface to register against the tail and be able to feel when it was parallel. I had purposely dulled the edge, so it would leave a wider and easier to see mark in the pin board. I commissioned IBC to re-make this knife and in the same red color from all the videos I had produced.
The next piece of the dovetail puzzle was the marking gauge. I started with the traditional wood gauge with a round pin protruding an 1/8” out of the beam. These had to be carefully reshaped to properly slice across the grain, left as is and it would scratch a terrible, non-distinct line across the board. This gauge put new woodworkers at a huge disadvantage, the skill required to file this small pin into a knife profile, flat on one side and oval on the opposite, was considerable. Starrett had made a round gauge several years ago, Kevin Drake introduced his version around 2001 and I sold hundreds of them to happy customers. This style blade is much easier to sharpen, easier to adjust and easier to use since you can always see where the blade is engaging the wood. A slight roll and you can advance the gauge mark precisely. About 10 years ago I started making my own to simplify the tool, one locking knob and no O-ring. Now to set for the thickness of the board, you merely lay the gauge face down on the board, move to the edge and let the cutter drop to the table. Lock the knob and thickness is set.
Finally, we had user friendly tools for all the major processes of cutting dovetails. Our success rate for a one-day class was pretty good, one third of my students would cut a great joint, one third would do a decent job, and one third needed more work. I continued to travel extensively teaching this craft all the while experiencing good results. However about four years ago after my third increase in eyeglass strength, I started struggling to see the fine knife mark and it was getting to me. I remembered the method Ernest Joyce used to transfer tails to pins. He would leave the tail waste in place after making the cuts and with the tail board in position on top of the pin board he would drag his saw through the kerf leaving a saw width mark in the end grain of the pin board.
The plus of this method was the way the kerf would not allow the saw blade to wander during marking, the minus was now having to cut beside the kerf mark without allowing the saw to fall into the mark. Also, since the kerf mark was on the pin and not in the waste, this required planing to remove, OK sometimes, not OK all the time. As I reflected on the positive aspect of his method I thought if the mark ended up on the pin proper, what was needed to move it to the waste? Solution was to move the tail board one kerf width in each direction to mark in the waste side of the pins and half pins. Move the tail board a saw kerf to the right to mark the left side of the tails and move a saw kerf to the left to mark the right side of each tail. A bit more thinking and the way to measure the saw kerf came to me. In the same way I measure the thickness of the wood I could measure the saw plate and the set of both sides. Now with the cutter of the gauge protruding the kerf thickness from the face of the gauge, it is as easy and deadly accurate to off set the tail board in either direction.
The next challenge was to make the actual marking easier. Holding a saw with your fingers wrapped around the teeth is not comfortable. I had a customer suggest a special blade for my marking knife that would be the exact same as the dovetail blade. This has proven to be very successful in helping students achieve great joints in their first try
Dissecting this joint every step has allowed me to figure out easier and more accurate ways to teach it. We had pared it down to needing just two primary skills, sawing perpendicular across the end of the tail board, (slopes don’t matter since you are creating the “template”) and sawing plumb cuts on the pin board (any slope creates a wedge). Since the saw tooth blade left the exact kerf in the top of the pin board there was no longer a need to guess/figure out where to start that cut. The only challenge left was finishing the cut to be plumb. This is a tough task, folks in the learning phase frequently would drift to one side or the other and ruin the joint.
In my quest to bring senior students up to speed cutting great dovetails this week, I have found a way to eliminate the necessity to saw plumb. Now, right after marking the side of the tail with the saw tooth blade, while still holding the tail board firmly in place, carefully saw from the top down through the existing tail board kerf. A square cut on the tail board will guide the saw to automatically make a plumb cut on the face side of the pin board. This requires a very light touch so as not to mess up the tail board kerf and to allow the same kerf to accurately guide the saw down the face of the pin board. I also discovered putting painters’ tape on the pin board gauge line makes it easier to prevent sawing past it. My final step is to raise the pin board up in the vise to make it easier to use the sloped kerf in the pin board to guide my saw through the cut. Another light touch operation ensures the kerf mark guides the saw to a perfect finish.
Am I done? Probably not, I don’t know what will be next, but I won’t stop thinking about it. I was always told working smarter was wiser than working harder! Enjoy your dovetails.
Rob