If you own enough books, it’s easy to believe almost anything and yet be certain about almost nothing.
Take dovetails. I’ve seen this joint cut with a wide variety of slopes during the last 15 years. And every person who cuts this joint has a personal or historical preference about the slope they use.
For some craftsmen, the slope varies simply because they eyeball the layout. Frank Klausz, one of the two living dovetail savants I know, says he cuts his dovetails anywhere between 10° to 15° off the vertical. Tage Frid preferred slopes of “about 10°.”
Other well-known dovetailers use marking jigs to lay out the joint, which locks them into particular angles. Rob Cosman, the other living dovetail savant, uses 10° for softwoods and 8.5° for hardwoods.
For the last 15 years I’ve been cutting dovetails, I’ve used the angles used by my first instructor: 10° for softwoods and 8.5° for hardwoods, just like Cosman. But for some reason, I’ve become dissatisfied with the way the joints look when they are visible on a piece of casework.
So I hit the library a few weeks ago, and now my head hurts from the bludgeoning. Dovetails might take their name from a bird, but reading about them is a trip down the rabbit hole.
What the Dead Guys Say
To understand how little there is to understand about dovetails, let’s take an abbreviated journey through the literature. I promise to be quick like a bunny.
Charles H. Hayward, the mid-20th century pope of hand-cut joinery, suggests three slopes: Use 12° for coarse work. Use 10° or 7° for decorative dovetails. There is no advice on hardwoods vs. softwoods.
F.E. Hoard and A.W. Marlow, the authors of the 1952 tome “The Cabinetmaker’s Treasury,” say you should use 15°. Period.
“Audel’s Carpenter’s Guide,” an early 20th century technical manual, says that 7.5° is for an exposed joint and 10° is right for “heavier work.” No advice on hardwoods vs. softwoods.
“Modern Practical Joinery” the 1902 book by George Ellis recommends 10° for all joints, as does Paul Hasluck in his 1903 “The Handyman’s Book.”
So at least among our dearly departed dovetailers, the advice is to use shallow angles for joints that show and steeper angles if your work is coarse, heavy or hidden. Or just to use one angle and be done with it.
At least in my library, the advice on softwoods and hardwoods seems to become more common with modern writing. Percy Blandford, who has been writing about woodworking for a long time, writes in his new book, “The Woodworker’s Bible,” that any angle between 7.5° and 10° is acceptable. The ideal, he says, is 8.5° for softwoods and 7.5° when joining hardwoods.
My Own Eye
One Wednesday morning I laid out and cut a bunch of these dovetails. I ignored the really shallow angles (6.5° to 8.5°) because I wanted to adopt something more angular. The 10° dovetails looked OK. The 12° dovetails looked better. The 14° tails looked better still. And the 15° looked really good as well. (The photo at the top of this entry shows a 15° dovetail with a bunch of alternatives marked on it.)
But I’ve some defect in my personality that keeps me from choosing the most extreme position, so I settled on 14°. And it’s a good thing, too, because a few days after that, the mindreaders at Lee Valley Tools released a 14° dovetail marker (I really should start wearing my tinfoil hat more). I ordered one – it seemed to be a sign.
Whatever angle you use for your joint, you can rest easy knowing that someone out there (living or dead) thinks you are doing the right thing – unless you cut something more than 15°, then you’re just nuts (or use a dovetail jig with your router).
— Christopher Schwarz
Thanks for doing all that work! I was kind of surprised that the consensus was to use a shallow angle if the joints show. I would think that you’d want the sharper angle to really scream, "This is a dovetail!"
To throw another data point, check out this thread http://www.traditionaltools.us/cms/index.php?name=Forums&file=viewtopic&t=177 where Wiley Horne demonstrates and comments on 5:1 (18º) dovetails.
Interesting stuff, Chris!
I reserve the 14° angles for half-blind, as that is what works best on the Leigh D4. I just *cannot* face cutting them by hand ;-( ; something in the genes, I believe.
[img]http://i4.photobucket.com/albums/y136/Sandal_Woods/Drawers/Finisheddrawer-wideropening.jpg[/img]
Roger,
1:5 dovetails are 11° not 18°, if my geometry skills are intact.
An 18° dovetail is a machine slope I believe, so that might be what you are thinking. (Wiley’s table is very nice.)
Chris
I’m guessing the authors who suggest different angles for soft vs. hard woods do so for structural reasons, more extreme angles yield a stronger joint. What’s most interesting to me is that last paragraph in the "What the Dead Guys Say" section, which states that the advice on soft vs. hard woods becomes "more common with modern writing."
I guess I’d expect the opposite, that dovetail angles would have played a more important engineering role before modern glues. I’d think that as glues improved the angle would become more of an aesthetic choice than an engineering one–your own experiments pivot on the _look_ of the dovetail. But I’m not familiar with the history of glues; maybe we’re just waiting for you to write that book and forever etch, I mean affix, yourself in history as the "Workbench and Glue Savant".
Narayan,
What’s really interesting is what I didn’t have room for (I don’t want to bore). The really older texts offer the fewest details on dovetailing. When discussing angles, they’ll say: Make it so it’s not so extreme. Beginners tend to use extreme angles. Oh, and avoid skinny pins.
Of course, moderns love skinny pins and extreme angles. So I guess we’re just like 19th-century beginners.
And there’s no real advice on hardwoods and softwoods. Just eyeball and go.
Chris
Mea Culpa, Chris. You are correct, I was trying a new calculator and I didn’t enter the numbers correctly. I never really paid attention to the angles in degrees. The majority of references I had used 1:8 slopes for hardwoods and 1:6 for softwoods. I have a board with those slopes marked on them for setting a sliding bevel. When I bought a dovetail marking template it had those same slopes so I never used anything else.
I just did some checking and Richard Kell’s dovetail templates use 1:8 and 1:5 slopes. Other than those and the new LV 14º template, the rest are all 1:8 and 1:6.
Thanks for making us think!
I thought that it is generally believed that the "standard" angles for dovetails are 1:8 for hardwoods and 1:6 for softwoods.
Not everybody might have a degree in geometry, so for those who don’t, here’s how you would convert:
1:8 in degrees is arctan(1/8)=7.12°
1:7 in degrees is arctan(1/8)=8.13°
1:6 in degrees is arctan(1/6)=9.46°
1:5 in degrees is arctan(1/5)=11.31°
Alexander,
The standards you quote are modern ones. The earlier original source materials shied away from those particulars. And I always wondered this: What if you joined a pine drawer side to an oak drawer front? Does the slope apply to the softwood tails or the hardwood pins?
I’ve never seen a dovetail fail due to excessive slope. If anyone here has, please chime in!
And thanks for the geometry details. Good stuff.
Chris
I cut whatever angle my LV dovetail markers tell me to cut. ; )
Chris,
Having spent the last 30 years or so studying the furniture built in this country during the period from about 1650 to 1860 (the time period prior to machine cut anything), I agree with your ‘dead guy’ statement at the end of the article. Generally, joiners (softwood workers) tended to use more slope and cabinet makers (hardwood workers) tended to use less slope. I’ve seen thousands of period drawers and cases with nearly as many variations on the slope of the dovetail.
Looking at some of the very early pieces made in this country, I’ve seen angles approaching 45 degrees. The pieces are still standing. In this day of over-analyzing everything, I think we’ve lost sight of what is truly important. If it looks good to the craftsman and it functions, who’s to say it isn’t right? It’s like the argument about tails or pins first. If you cut a tight fitting, good looking set of dovetails, who’s going to know which part you cut first, or how you achieved your angle, in a couple of hundred years? I’m glad to see someone publicly espousing the notion that woodworking isn’t merely a rigid set of unyielding rules to be followed but a chance for a craftsman to express himself. Even if it is only through the angle he chooses to cut his dovetails.
Chuck Bender
Cabinet maker, joiner and furniture maker
Chris,
A mylar helment seems to work equally well.
Many of the shaker pieces vary quite a bit in the dovetail angles. I’ve noticed the chip boxes, in particular, often have angles approaching 1:4 and that matched with equally sized pins is attractive to my eye. these pieces have been around for 160 years or so.
modern break point testing indicates to me that beauty to the eye rather than perceived strength is the determining factor of dovetail angles.
A final note for those of us that like shaker pieces. The shakers sometimes just nailed things together including drawer sides. I don’t know how that is relevant to dovetail angles, but it just popped into my head under the mylar.
Greg
Chris,
With modern glues I bet failure will occur along the grain, so angle choice becomes aesthetic as you suggest.
However one thing the dead guys never mention is timber thickness. In UK we use thin 5/16" hardwood drawer sides.
1:8 looks good for the single lap dovetails in thick 7/8" drawer front material.
The same angle used for the through dovetails at the back hardly shows any slope at all, so I choose to use 1:6.
David Charlesworth
Good job Chris,
Funny thing, I too have actually spent way too much time thinking about and researching this very topic. As much as I like to think my furniture will be around in a couple of hundred years, the reality is that dovetails have become as much of a fashion statement as ‘Fancy-Ass Jeans’. To my eye the
1:8 or 1:9 works well, although in heavier timber using larger joinery I’d go all the way up to 1:14. As for soft woods vs. hard, I don’t really think about it; I tend to use the same ratio throughout…call me crazy.??
Beauty is indeed in the eye of the beholder, and those of us who spend way too much time and energy thinking about these things are indeed…er’, um, beautiful?
Cheers!
When I took a class with Alan Peters (who I consider one of the dovetail savants…he taught Cosman), he said, use 1:7 (~8+ degrees) for everything, you get used to it and cut it faster and easier, no matter what the wood. Whom am I to argue with someone whose been doing this for 60 years? Now, I just cut them and then probably average out at about that.
Gary