Mortise and tenon joint
Also called: mortise and tenon, mortice and tenon, mortise-and-tenon, stub tenon, through tenon, haunched mortise and tenon
One page. Other spellings and regional names stay on this URL.
A mortise and tenon joint fastens two pieces of wood together by cutting a tongue, called the tenon, on the end of one piece and a matching slot, called the mortise, into the other. When the tenon is pushed home into the mortise the two pieces lock together, usually with glue, and sometimes with a wooden peg driven through the joint as well.
The parts of the joint
The tenon is the projecting tongue, cut on the end of one piece so that a flat cheek runs down each broad face and a shoulder, the point where the tenon steps back to the piece's full width, sits at its base; Frid (1979) noted that the joint's strength depends entirely on how well the cheek and the shoulder are cut and fitted. Where the tenon and its mortise stop short of passing all the way through the second piece of timber, Audels (1923) called the result a stub tenon; where it does pass right through and shows on the far face it is a through tenon, and where it is deliberately hidden so that neither face nor edge shows any sign of the joint, Duginske (1999) calls it blind. Porter (1999) catalogued several shaped variants built for particular jobs: a haunched tenon carries a short stub, the haunch, alongside the main tongue, to fill part of a groove or resist twisting; a barefaced tenon has a shoulder cut on one face only, not both; a twin tenon splits the tongue into two separate tenons side by side in one mortise; and an unequal shouldered tenon has one shoulder set out of line with the other, rather than both shoulders meeting the cheek at the same point.
Why the joint holds
Glue bonds strongly to long grain, wood cut along the length of the fibres, but poorly to end grain, wood cut across the fibres; joining a piece end on to another piece's face is therefore a weak proposition by glue alone. A mortise and tenon gets round this: because the tenon's cheeks run with the grain and sit flat against the long grain walls of the mortise, most of the glued surface in the finished joint is long grain to long grain. Hoadley (1980) put it plainly: fastening the end grain of one piece to the side grain of another can be carried off with real success precisely because the joint moves most of the actual gluing surface away from that weak end grain contact. Hoadley (1980) also examined how the joint resists racking, the tendency of a frame to twist out of square under load: in an ordinary shouldered tenon, rotation of the joint is held back mainly by the shear strength of the glue line bonding the tenon's cheeks inside the mortise, working alongside the mechanical restraint the surrounding wood gives against compression and tension. Frid (1979) drew the practical conclusion: a mortise and tenon should be laid out and cut so as to give the largest possible area of long grain to long grain contact.
Variations across traditions and trades
The plain through or blind mortise and tenon has never been the only version. Some historic furniture manuals recorded wedged variants: Fairham (1920) named the fox wedged tenon and the wheelwright's self-wedging mortise and tenon joint among specialised forms, though without setting out how either is made. In timber framing, where the joint carries structural loads in a building's frame, Benson (1988) drew a boundary around the term: unlike a true mortise and tenon, where the tenon is enclosed in wood on every side once the joint is assembled, what he called an open mortise and tenon is only partly enclosed, and is secured instead with pegs or by the shape of the joint locking together; he described this open form used at the peak of a rafter pair and at the corners of a sill. Sobon (2002) looked at pairing tenons, or wedging a tenon through the mortise from the outside, and found that such arrangements can outperform a single simple mortise and tenon, reducing the tendency of the fixing pin to shear or the face of the mortise to split out under load. Noll (2002) traced a family relationship between several joints of increasing complexity in the same corner: an end lap develops into an open slot mortise, which develops in turn into a full mortise and tenon, and he noted that in a T shaped joint between a rail and an upright, the mortise and tenon, the bridle joint, and the slot mortised bridle joint all share a doubled glue surface and correspondingly greater resistance to racking than a plain lap. Outside the Western tradition, Japanese joinery developed its own family of named forms: Seike (1970) records a true mortise and tenon called hira hozo and a lapped goose neck mortise and tenon with stub tenons, while Sato (1987) records L shaped half blind and oblique dadoed variants used in furniture and framing.
Laying out and cutting
Audels (1923) described the first step as laying the work out to the dimensions the joint needs before any cutting begins. Porter (1999) set out the standard way to mark a through or blind mortise and tenon: a marking gauge, its stock held against the face side of the timber, the reference face all other measurements are taken from, is used to scribe the width and position of both the mortise and the tenon directly onto the wood, so the two parts are marked from the same settings and should therefore match.