Sometime in the Autumn of 2020, I came across this photo of the detail of a wooden house near Kamnik in the Tuhinj Valley in Slovenia. It had been posted on social media by Jože Kozjek who works as a forester there, and is committed to photograph and document what he sees as remarkable testimonies to the cultural heritage of his region and country.

This house and many others related are extraordinary pieces of design, moving tributes to human inventiveness as it is played out within (or around) a traditional form.

These are of course words of ‘now’, written in admiration towards a process which was conducted with such fluency and ease that it could only be derived from embodied shared knowledge. We might now tend to call it ‘tradition’ which is a somewhat deadening word to pin onto something so alive, virtuosic and exuberant.

Being both related and totally foreign to my work on violins, it was fascinating for me to try to figure out how these angled joints might be designed and cut.

The reason for these few words and images is to share this fascination.

Of the houses themselves I know little. They are known to date from the 18th to the 19th centuries. My guess is that as often seen in forest dwellings [still practised today in the ‘carbets’ of the French Guyanian rain forest], a clearing is cut, and the house is built from the timber felled. It seems to be spruce, Picea Abies. It would be interesting to submit some of the ring patterns for dendrochronological study, as it may confirm the age and also the provenance of the timber.

So to these joints: a Baroque variation on a double dovetail? And why such virtuosity?

The method for stacking, alternating the beams on either side of the corner is clear if one looks at the photograph above, but the joining method is elusive at first.

The photos below show the process I followed to try to understand the basic building blocks.

Once I found what a likely shape was, I realised something amazing (and obvious once one sees it clearly!) that the pieces are of only one type, with the under and upper faces of one given piece being mirror images of each other. Thus, for two pieces to fit together they need to be actually identical!

Cutting is one thing: I found that I got a little better at it after a few tries. A gouge, a chisel and a saw can do it, there is one area near the shoulder of the tenon which cannot be sawn, so it needs a bit of digging with a chisel, but otherwise it is doable.

That is cutting two surfaces and making them meet, on a small scale model.

Building the actual house, this would be taking place using pieces of wood which (I guess) are likely to weigh somewhere between 100kg and 200kg each, where the joints need to be fitted simultaneously on the four corners around the building. This is no mean feat, and would need to be underpinned by a rigorous marking strategy before any cutting is done, in order to minimise the need to move pieces.

I speculate that a single flat sheet of material (card, thin metal, thin wood perhaps, glue-sized paper?) could be used to mark all lines, designed in such a way that convex and concave are marked using the same drawing, therefore they have to fit. This is of course the general idea for marking dovetail joints.

This is the reason for the stencil cardboard template which I show (see photos).

Having tried to put pressures on the assembled wooden pieces I made, I was struck by two properties of this joint:

# when a vertical compression is applied to it, the angled joint becomes more resistant to bending and becomes more brace-like and apt at resisting shear and tension.

# if a load is applied which forces a distortion (bending) onto the 90 degree angle, there can be a bit of play to accommodate for it, and a small amount of movement can happen without breakage.

I think these two properties would be very useful in a timber building: one needs to imagine a roof made to resist the weight of snow, and the constant movements of wood which is an extremely hygroscopic material.

Miraculously some of these houses are structurally intact after 200 years or more.


Below: photos by Jože Kozjek of that house, and of another one from the same area with many shared characteristics

This short blog article is the beginning of what I hope is ‘work in progress’:

There is thought of a version in Slovenian, and also of an exploration more in depth of the context of these fascinating structures.

I wish to thank Jože Kozjek and Nina Katarina Štular.

Thanks to Paul Mackilligin for the stimulating conversations which lead to these investigations in wood!

[ © All rights reserved 2021 Bruno Guastalla + Jože Kozjek ]


  1. Dumbstruck. The joints are an astonishing combination of ingenious, complex, and beautiful. And I love the photos of your workings out, and so impressed by the work you put in. What a challenge – and what a well met challenge . I think you may have hit on an ideal way to fend off the lockdown blues! Brilliant.

      1. Thanks for this Bruno – I look forward to the extra photos and detail. Just so remarkable to see how a combination of necessity and simple forward planning drew out such a high level of sophistication.

  2. This is a fascinating subject. I’m a log home designer and builder for 40+ years in Oregon, USA. These notches shrink tighter as they age it appears. Notice your model has a slight “underhang” of the ends of the members but the photo of the old timbers shows a slight overhang, yet the notch is very tight towards the interior of the sturcture. Of course logs weigh a lot and consderation of putting slightly more pressure on the joint as apposed to the length of the log is important too. Great research! Please continue.

    1. Good to hear from you Greg!
      Thanks for your contribution!

      I think my model suffers from ‘beginner’s approximative cuts syndrome’ ! (although as a maker of violins I sould have known better, I was a bit timid in my relation to the scribed lines..).

      To try to add to your points about overhang on the joint itself: 1/ The joints are such that a vertical load tightens them up (which is a smart move design wise!). 2/ Shrinkage over time: as a rule longitudinal shrinkage is negligeable if compared to tangential and radial shrinkage. Therefore, I would tend to believe that the overhang one sees of the long grain on the photograph of the house is simply due to this.

      So I have a suspicion that what they did was pretty exact with good fits everywhere, and that what we see now is a construction which has aged well over 200 years +. And, thinking of it, I expect you are completely right that the corners were likely to have been given a tighter fit initially than the length of the logs, so that, with the massive compression of the weight of the house, roof and all, the joints indeed tighten and the whole structure coalesces.

  3. I’m sure you’re on the right track, Bruno. And work like this is not a “one off”. It is an evolution of thinking and building over lifetimes. Whether they used templetes or had ‘templetes of the mind’ (vast experience) is hard to know but one thing is sure….these are masterpieces. I hope Joze is doing exhaustive research on this style of construction where ever it is found and we can look forward to the who, when, how and why of these acncient styles of wood working. This is a b ig deal.

  4. Great, Greg, I very much apppreciate your encouragement, and will let you know personally when more comes out. I think you are right about templates too: it is a question of what to mark and where; it could simply be done with dividers and three notches on a set square (to mark the three cutting levels on each surface top, mid, bottom), and the ‘filling in’ done by eye and feel. One thing is for sure: there is no room for error in this. In the violin making world which I have inhabited for 45 years, there are surpringly few aspects of working as naked as this, (except purfling inlays). Shapes are arrived at ‘from above’as it were, so even if one is always using the biggest tool possible at all times, one has time to see the shape emerging and the opportunity to make adjustments.

    1. Dear Magda, Thank you for your comment. Japanese joinery is indeed amazing, extremely varied and rightly well known. It would be great if you could find videos of Japanese joinery where the type of joints and processes somehow relate to this particular way of doing things. Please do post them here!

Leave a Reply to Nick Bland Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s