Riving Article

Riving Wood For 17th Century 
Joint Furniture
By John Alexander


    Riving is the splitting of wood in the direction of its long fibers. When rived, wood is usually "tree wet", having a moisture content above 30%. This article focuses on riving wood for 17th century joint furniture. (1,2)

Figure 1  (circa 1690) 
Figure 2 (1996)

    I began riving to make parts for post and rung chairs. John Alexander, Make a Chair from a Tree, Video, ALP Productions, 1999. Charles Hummel of Winterthur Museum expanded my horizons. One day, he took me into the Museum's collection and, without comment, lifted the lid of a 17th century New England joined, frame-and-panel chest. I saw and will never forget that the two rear stiles were riven, not sawn. Each stile displayed the torn bundles of long fibers characteristic of riving. (3) I thought, "I'll never go to the lumberyard again."


 
Figure 3

    I became fascinated with the technology and then the style of 17th century joined furniture. Hummel, Benno Forman, Robert St. George and Robert F. Trent have been my mentors and native guides. After examining piece after piece of 17th century New England joined furniture, I found that, almost without exception, riven oak was used to make the frames and panels. Where long, large width boards were needed, for example, for chest lids and cupboard tops, sawn stock were used as well as edge joined riven oak. These conclusions have been confirmed by the extensive investigations of Robert F. Trent and Peter Follansbee, the joiner at Plimoth Plantation. The relatively few exceptions prove the rule. Applied molding, ornament and turnings employ a number of other woods including maple, hickory, walnut, cedar, cedrela, and snakewood.

    The choice of rived oak for framing joint furniture is understandable. Oak rives well but, as with other woods, results are most accurate and reliable over relatively short stock. With frame-and-panel construction joint chests, cupboards, chairs and stools can be constructed from relatively short riven stock. The frames of large tables and beds are a challenge. Broad widths require trees of large diameter and long stock requires trees of almost perfect straight long fibers and little or no wind. During the early years of New England's settlement, joiners had access to riches of first quality oak found in large, straight trees that had grown in the woods under competitive conditions. 
 

Riving was preferred because:

  • 1.Riving converts the greatest mass of wood in the least time with the least effort.
  • 2.Riven stock is stronger than sawn stock-it is an aggregation of parallel long fibers.
  • 3.Riven wood is less likely to absorb and lose moisture. No end grain is found on its broad surfaces.
  • 4. Moisture content change of wood riven on the ray plane causes predictable change in cross section and the stock will not crack, cup, warp or twist. (4)

Figure 4
  • 5.Riven ray plane oak surfaces are easy to plane and carve.
  • 6.Wood riven on the ray plane is attractive. However, it is likely that 17th century craftsmen and consumers ranked this attribute a low sixth. Joint furniture was often painted, carved, turned, ornamented, or covered with fabrics. See Victor Chinnery, Oak Furniture: The British Tradition, (Suffolk, Eng.: the Antique Collector's Club, 1979); New England Begins: The 17th Century, Jonathan F. Fairbanks and Robert F. Trent, Museum of Fine Arts Boston, 1982; and Benno Forman, American Seating Furniture 1630-1730, Norton, 1988.
  • 7.Though riving requires rejection of all but the best wood (5),
Figure 5
Figure 6


    Using stiles and rails with truncated cross sections conserves the best wood. Accordingly, stock was often riven on the ray plane on both sides. (6) See Peter Follansbee and John Alexander, "17th Century Joinery from Braintree, Massachusetts: the Savell Shop Tradition" in American Furniture, ed., Luke Beckerdite, (Hanover, N.H.: University Press of New England for the Chipstone Foundation, 1996).

  • 8.Ray plane oak stock rives so reliably that tenons were rough formed by first sawing the tenon shoulders and then riving off the tenon cheeks with a small cleaver or a broad chisel. (7,8)
Figure 7
Figure 8


John D. Alexander, The 17th Century Draw Bored Mortise and Tenon: The Heart of Joinery, Woodwork Magazine, No. 41, October, 1996; reprinted at the author's web site: www.greenwoodworking.com.

  • 9. When riven wood has moderate wind, this can be corrected, and the advantages of the rived stock retained. Wind is removed by hewing and then planing with a fore-plane or scrub plane. When wood is tree wet, it hews and planes easily. The stock will remain in truth as it dries.
Why rive today?

   The style of 17th century joint furniture is inextricably wedded to frame-and-panel construction of riven oak joined by drawbored mortises and tenons. (9,10)


Figure 9


Figure 10

    The use of rived oak makes stock preparation considerably easier. Layout, joining and reliability of the resulting stock create joint furniture that possesses the unique characteristics of original work.

Challenges associated with riving.

1. Log selection.  An excellent second grade veneer log is required. The wood inside a log is never as good as the log appears from the outside. It is usually worse. You fell it or buy it and it is yours. Trees that fall by themselves and gift trees are seldom satisfactory. You may be able to make a joint stool from your firewood pile. For larger pieces you must use a sizeable log. (11) 

Figure 11

    Firewood dealers, sawmills, tree surgeons, foresters and log dealers are sources. Felling a large size tree is an involved and risky task. Never try it alone and without an experienced person. 

2. Log size. The longer and larger the log, the better. Moving a large log can be difficult. The log can be converted into equal eighths on location and only suitable stock carried away. I call splitting a log into eighths "bulk riving."

3. Degradation.  Rived stock is "tree wet" and must be protected from splitting due to rapid moisture loss and wood destroying insects and microorganisms. Don’t fell or buy a tree until you have a specific need. Storage under water will keep the wood moist but surface discoloration and degradation will result over time. This is not critical. To retard checking, coat end grain with emulsified wax, paint or roofing cement. Wood destroying insects and microorganisms generally prefer the same temperatures as humans. Wood can be stored outdoors during cold months. Store off the ground out of direct sun in a moderately dry area. During summer, storage under water or prompt use is best. Leave the bark on the wood until you need it.

Interested? Let's rive parts for a 17th century Joint Stool.

    Red and white oak, both ring porous hardwoods, were and are the woods of choice. Start with red oak. Though lighter than white oak, red oak is more than strong enough. It is cheaper, rives easily, dries more reliably and works well.

    The parts of the tree from outside to the center are: corky bark, inner bark, cambium layer (where cell division takes place), sapwood, heartwood (the wood we will use), juvenile wood (near the center of the tree containing curved irregular long fibers reminiscent of the tree's early years) and, at the center, the pith. (12) Examining the cross section of oak more closely, it is easy to distinguish the circular growth rings and the distinctive rays that radiate out from the pith all the way through the inner bark. (13) 
  

Figure 12
Figure 13 (Red Oak)

    A joint stool requires the least wood and shortest stock of any manner of joined furniture. Though the firewood pile may suffice (except for the seat), here we will start with a tree trunk, the portion of the tree above the root swell and below the first limbs. (14)


Figure 14


    Avoid trunks that are curved, have leaned (the pith will be off center) or are not circular in cross section. Examine the bark for evidence of knots, metal, extreme wind or other damage. A long trunk of large diameter allows you to work around and discard defects. (15) There will be a considerable amount of unusable wood - firewood or mulch.


Figure 15

    Choosing the diameter of a log is another challenge. The log's ends will indicate the thickness of the bark and sapwood. Unfortunately, you can learn little about the extent of the curved and fractious juvenile wood until it is opened. For a stool (excepting the top), you will need at least 5" of clear wood lying in the ray plane on a radius between sapwood and juvenile wood.

     The bulk riving tools are mauls, wedges, wooden gluts, the ax and splitting hatchet. (16) 


Figure 16

    Initially, we are interested in dividing the tree into equal sections. The tips of metal wedges should have an enclosed angle of about 15° so that they will enter and stick in the log rather than spring back out. The ax is used to clean out cross-linked fibers within the split.  The splitting hatchet is an old ax head mounted on a short handle. It is used to score the end grain of large diameter wood prior to splitting. It can also be later used to  split smaller stock. It is struck with a wooden or metal maul or wooden club.

    To reduce the log to eight equal sections, first score a diameter with the splitting hatchet. Follow any existing split in the end grain.  Splits  create  planes of weakness that must not be included in finished stock. Scoring itself creates a plane of weakness that will guide the split. On large cross sectioned logs, start riving by placing a wedge on the outside perimeter of the log's end directly in line with the scored diameter. Drive the wedge into the sapwood at an angle of about 45° to the long axis of the log.

Cautions:

  1. Do not use a striking or cutting tool when anyone is in the plane of the swing. 
  2. Before each use, make sure that the tool handle is tight. 
  3. If you strike metal with metal, wear safety glasses. 
  4. Grind off all deformation of the heads of wedges, mauls, axes and hatchets. 
  5. Riving demands constant attention and thought. You or another person will be holding the wedge, froe or splitting hatchet. 
  6. A common and dangerous error is applying the same force throughout the riving. The longer the actual split, the less force needed. Otherwise, the wood may suddenly give way and hands, feet or tools injured. 
  7. The wood talks. Listen, think, and proceed deliberately.

    Increase the split by introducing another wedge into the end grain. Then, advance down the split on the outside of the log by leapfrogging wedges. After the split opens, gluts (wooden wedges) of white oak, hickory or dogwood are used. Metal wedges buried inside the log are hard and dangerous to retrieve. Remove all metal wedges from the log before chopping any interfering fibers with the ax. Wooden gluts can be left in place. Never reach into the split until the wood has given up and fallen open. A metal wedge can fly back out of the log and strike you. Or, your hand may be trapped. Rive the log on the rays into halves, quarters and then eighths. The rays are planes of weakness and reliability. If you do not leave an equal mass of wood on both sides of the split, the split may run to one side.

    With the log opened up, you will see exactly what you have. Saw away unusable sections.  Be critical. Though the freshly opened wood may look majestic, reject not only knots but also curves in the long fibers. The longer you practice riving and the closer you get to finish dimensions, the more obvious these defects will be. A stick marked with the needed stock lengths saves time. Always select the longest required stock from the best wood. Then remove the irregular juvenile wood and pith by splitting perpendicularly to the central ray of the eighths or, as we say, in the growth ring plane (really the plane tangent to the growth rings). Use only heartwood. Sapwood is weak and full of sugar that attracts consumers.

    Riving sapwood away from heartwood is messy. The split tends to run to the outside of the log. Split off as much as possible from both ends of the stock and hew away the remainder.

    Dimension riving is the second stage. The good news in stool making is that the stiles are 2x2s (or slightly less), aprons 1x4s and stretchers 1x2s. The wider dimensions of the aprons and rays  lie in the ray plane. You may not find a tree large enough to give you a seat board lying in the ray plane almost 11" wide, and 3/4" thick and free of pith, juvenile wood and sapwood. If you do, set it aside for seats and other projects. (17) This seat took 15 minutes to make after it was riven from the tree wet stock  pictured beneath it. Though the seat has since shrunk in width, because it was riven in the ray plane it has not warped or cupped. See Figure 2. 
  


Figure 17


    Use the froe and froe club or the splitting hatchet when dimension riving. (18,19) 

Figure 18
Figure 19


    Saw end grain straight across and keep it clean.  Marking is done with an indelible or artist's drafting pencil. I prefer a drafting pencil (Derwent, Drafting Pencil, Ivory Black, No. 67) because its mark is easy to read on both wet and waxed wood. It does not wick into wet end grain. Rulers and tapes take time and are clumsy. For layout I use bright colored plastic strips with the exact cross sections of stool parts. (20).


Figure 20


    Running the fat pointed pencil around the layout strip creates an allowance. Rive close to finished dimension. If you do not,  you can further reduce the stock with a single bevel hewing hatchet.

    Lay out the dimension riving scheme on the end grain to maximize the amount of stock and minimize your effort. (21)

Figure 21


    First rive the best and longest stock for stiles. Rive out a number of extra pieces. The good news is that both aprons and stretchers may, but do not have to, be rectangular in cross section. If you rive both sides of these rails on ray planes, the resulting cross section will be a truncated triangle that is 1" thick at the top and less at the bottom. This saves wood. Only the best ray plane surface will be cleaned up and then used as the fair face. In joinery, the the fair and true faces are the same. The interior surfaces of a joined frame do not have to fair or true. Rive out more aprons than stretchers. Failure aprons can be converted into stretchers. 
 

    When dimension riving, the stock is small in cross section. The froe and froe club or the splitting hatchet are used. The only 17th century illustration and reference to riving that I am aware of is found in Randle Holme, The Academy of Armory, 1688; Reprint, The Roxburghe Club. 1905; CD-ROM reprint from the original manuscript, The British Library, 2000. Holme illustrates a froe  (22) that he refers to as a lath axe and says:


Figure 22


     "…[a] Lath Axe, is an iron instrument like a Knife Blade, with a round thick back having an Eye, into which a thick strong handle is put, the edge standing outwards; With this, great Timber after it is cloven with the Maul and Wedges into small pieces, it afterwards cleaves them into Laths, Barrel Boards, and Pannels."

    The froe handle is held in the non-striking hand with its blade seated securely on the riving line. The blade is struck directly over the stock with the froe club. Focus on the blade lest when rearing back to strike with the club you allow the blade to wobble off the mark. A mislocated split creates a plane of weakness within the stock and must be followed through no matter where it is located. After the blade is buried within the split you enlarge the split by striking the projecting blade with the club and then levering to and froe. The froe is a remarkable tool. A principal advantage is its ability to assist in controlling the direction of rived splits. Almost all riving is done with equal mass on either side of the splitting edge. However, if a split starts to run to one side it may possibly be redirected by applying pressure to the stock's more massive side. This is accomplished by levering the handle of the froe so that blade's heel presses against the massive side of the split.

Figure 23


    My shorthand rule is, "Froe to the fat." A riving brake can be used to hold the stock during this procedure. (23) There are various types of break designed to hold the stock's far end between two bars. This allows the craftsman to have both hands free.

    Short stool stock (except the top) can be rived with the splitting hatchet. If the first blow does not complete the split, I lift the work piece and hatchet up and bonk them down onto the chopping block. I have found no traditional evidence for this practice, but it is very useful. For longer and wider stock the froe and club are the tools of choice.

    Layout and rive down from the end of the wood that was originally towards the top of the tree. After riving, the bottom end will be at least equal to or larger in size than the top. Because oak rives easily and reliably on the ray plane, lay out the wider face of your stock on the ray plane.

    Riving on the growth ring or tangential plane is often not as reliable. Red oak particularly sometimes tends to run out across the growth rings. Again, it is prudent, when possible, to have an equal mass on both sides of the split. Also, consistant with your need to maximize stock, it may help to have the stock as narrow across the growth rings as possible before riving on the growth ring plane.

    When the stock has been rived to dimension it will only need cleaning up on the best ray plane surface and two edges. Use a fore-plane (the 17th century joiner's name for a convex-bladed medium length plane) or a shorter Continental scrub plane. (24,25)

Figure 24
Figure 25


    Blades on regular planes can be ground into a convex curve. The inside surface is roughly cleaned up, if at all, with the joiner's hatchet and fore-plane. (26) Here is an interior surface of a 17th century stool that wasriven and hewed but never planed. 


Figure 26


    The exterior surfaces are scrubbed if necessary and then finished, depending on their lenght with a jointer or a smooth plane. A rived true ray plane surface planes smoothly and easily. It is functional and beautiful. (27,28) It makes the journey worthwhile.


Figure 27 (1998)


Figure 28 (1999) Uphosltery by Robert Trent

    John Alexander has been working with green wood for over 35 years. " I taught classes for 25 years. I thank all my students and teaching assistants for their friendship, challenge, spirit and ideas. I stopped teaching in 2004. I am redirecting my efforts to research and writing. I will continue to be available for speaking and demonstrations. I welcome letters, e-mail and phone calls about greenwoodworking."   His book, Make a Chair from a Tree: An Introduction to Working Green Wood, Taunton Press (1978); reprinted, Astragal Press (1994), now out of print, was the first book introducing what John calls greenwoodworking. His recent video of the same title is available here. John's research and craft have led him to maintain that the traditional crafts of both post and rung chair making and frame and panel joinery both used stock that was first riven from tree wet green wood. The above article is based upon portions of his forthcoming book, Make a Stool from a Tree: An Introduction to 17th Century New England Joinery, to be published jointly by Cambium and Astragal Presses.