The word of the expert was "Your beam sections look OK to me, maybe even oversized; a wood frame structure is unlikely to fail at full-section. Its the joints which fail!"

With such counsel in mind, I have encountered the problem of shrink-lifting and progressed to research of spline joinery. In respect of jointing a single beam into a post, I am puzzled by the manner of commonly placing two pegs in a beam, and only one in the post. The latter, well placed centrally for max relish. Two pegs in the beam will make for a stiffer joint, but in view of possible pegged shrinkage-lifting, perhaps this is undesirable; just one good peg in each member could hold the asembly in tension, and a small amount of hinging could keep the beam seated on its post housing?

For such a problem, it wpuld be better to place keys on the outside of the post, but desirable relish might conflict with wall depth. Given that one must work with green timbers, how else to avoid un-seating a beam? Keeping pegs low will bias shrinkage into the upper section, but 1/32" off the base would be as bad as 1/8"; a miss as good (bad!) as a mile.

Ahh! I'd almost forgotten this lonely un-responded thread!

In naming a single beam-post joint, I actually over-simplified things; I should have named it 'spline joinery at building corner' to discuss two beams joining in a post, as depicted in: The Red book, Joinery Decisions fig 19. (no dovetail here!)

The thread came after I also read 'Experimental Evaluation of Traditional Timber Connections'. In the conclusion of this article; the problem of beam un-seating was discussed, and it noted two solutions (for a standard m+t): pre-dry timbers, or create slots (in lieu of peg holes) in the tenon. I know some folk use dryed wood (or re-cycled timbers - Derek); does anyone use the second solution given? Or maybe this un-seating isn't such a problem in reality??

Otherwise, I now understand the solution is to keep pegging low in the joint, and/or draw-bore in two senses - laterally, and vertically to squeeze the tenon downwards. Trouble is, keeping pegs low limits the number of pegs, and hence tensile strength available.

So, 'dovetail at end of spline' opens new possibilities. I've only seen this joint type presented as half-dovetail with wedge, but not much use if no end-access remains for tightening. Could the peg even be eliminated if the spline (with dove-tail wider than 'body') were inserted from the 'outside'? The tail could diverge outside of the post (if there were room for a stub) or starting from within (trickier to cut).

Oh dear, its all got complicated again!