Overscribing of the Lateral Groove
By Del Radomske

Although I learned a tremendous amount about log building, modern log building was still in its' infancy. There seemed to be so much more to learn and nobody back then, that I knew of, had the answers: so it was a case of trial and error. This why we have so many different building systems today - everyone was experimenting to one degree or another.

In the beginning, most builders felt their houses were great. You could stand back and marvel at your creation, the whole house looked as though the logs had grown together. Over time, you would come back to view the house. Two years pass and the house still looks great. As students, we were told that most of the shrinkage in a log house takes place in the first two years. We now believed the house would stay this shape forever. How wrong we were! The third year passes and tiny changes in the log work began to appear. The fourth year brought major changes in the log work, gaps measuring from 1/4" to 1/2", or more opened up around your once perfectly scribed round notch. Logs twisted inside their own notches and flipped open gaping holes in the lateral groove. Plate logs appeared to have twisted and rolled exposing wide lines of pink insulation. These changes can continue into the fifth year with a hot dry climate and well-exposed landscape.

When you first experience this, the big question is: "What has happened to my masterpiece?" I was terribly disturbed by the changes I saw in my houses. The amazing part about it was that the owners were still extremely happy with their log houses. I guess the old concept that log buildings are supposed to be rough and crude was shining through. I was happy for them, but it was eating away at me. I began looking at log building as being a bad joke on me and misleading to the customer. I let this get to me so badly that I had only two choices; quit, or do something about it.

I chose to do something about it. I kept thinking there had to be a way to build a log house with green logs that would maintain tight fits. I had to start somewhere, and like a lot of other builders figured I might be able to improve the fits of the notch by changing the shape of the notch. I was to learn that this is only partly true. I certainly don't regret developing the 4-point saddle, because in time it was to go hand - in - hand with overscribing of the lateral groove in order to maintain a tight fit on the sides of the notch. A future article will give the details of the Radomske 4 - point saddle notch.

I progressed with developing the 4-point saddle, but was still disappointed to see gaps form around the notch. They were usually not as large as the ones using the round notch, but nevertheless, they were gaps. Then something happened during a log building course I was teaching. I was taught as a student that there should be 75% of the weight on the notched corner and 25% of the weight on the lateral groove, and I continued to teach this to others. This makes perfect sense, because the building is structurally more sound if there is more weight bearing on the corners. The corner notches lock a building together, and help to prevent the logs from twisting.

As I explained this to my class, I suddenly realized I was contradicting myself by teaching that the final notch scribe and lateral scribe must be exactly the same setting, which is the accepted practice. By making the two scribe settings exactly the same, the initial weight of the log is evenly distributed along every inch of the fit - so many pounds per square inch. A 47' long log with 3 notches would have, say, 88' of groove and perhaps 12' of notch, this means that 12% of the weight is on the notches and 88%on the groove. This is not enough weight on the notches to keep the logs from twisting - and this is before shrinkage transfers even more weight to the groove. It has been my experience that the weight distribution does change over time until about 100% of the weight is on the lateral groove. This is what causes your once perfectly scribe-fit building to lose its original fits. I do not know why this contradiction was overlooked before - sometimes the obvious stares you in the face. Now I had the problem pinpointed and had a starting place.

Overscribing is a technique that puts 100% of the weight onto the notched corner as you build. After shrinkage and settling (a better term for settling is compression) you hope to keep 75% of the weight on the notched corners and 25% of the weight on the lateral grooves, with the result that both the notches and laterals appear to fit tightly. The trick is to know how much to overscribe to achieve this situation after shrinkage and compression. I set out to find the answer, I knew it was going to be a slow process since it takes four to five years for a building to shrink and settle.

I began by setting my lateral scribe 1/8" wider than the notch scribe and used this for three years. I monitored progress on a house - at the end of three years I could see gaps of about 1/8" showing up in some notches. I then knew we could try ¼" overscribe. For three years I monitored the first house using ¼" overscribe and I was very pleased with the results. Absolutely no changes, but I knew it could shrink and settle up to another two years. I checked the weight on the lateral groove by pushing a knife through the seam between the two logs. It appeared to me that the lateral was bearing too much weight. Guessing that in another year or two a slight gap would form around the notches, I decided to increase to 3/8" overscribe. My guess was right, because in another year the ¼" overscribe house had some gaps opening.

During this time I happened to be looking through Ray Arcands' book, Log Building Tools. In it he describes how to make scratch scribers, The Norwegian term is drawing iron. These are the old type of scriber that scratched a line into the wood fibers of the log so the builder could accurately cut and fit the logs. The drawing iron had two sets of points; one set of points was for the notch and the other for the lateral scribe. It has been assumed by many builders, including me, that these two sets of points were set an equal distance, and also assumed the set of points for notches came more directly out from the scriber and the lateral set of points was more hooked, in order to best suit the two different applications of scribing. (See Arcand's Log Building Tools, and Allan Mackie's Building With Logs).

In Arcand's book, he mentions seeing a very old scratch scriber and notes that there was a difference between the two sets of points of about ¼". He then goes on to say he could "only speculate the reason why" and stops there, giving no idea on his own speculation.

I did an experiment using one of my homemade scratch scribers and found that it is possible to scribe both the groove and the notch with one set of points, if they are not hooked too much. So why were old scratch scribers made with two sets of points? The only reason I can see is that one set of points was larger than the other set of points, for the purpose of overscribing of the lateral groove in order to put compression on the notch.

I now believe that overscribing of the long groove has long been a practice of log building and that what I have been doing is rediscovering an old technique that to my knowledge has never been documented. How old is the technique of overscribing? Your guess is a good as mine, but I would suspect hundreds of years.

Shortly after this I was teaching a six-week log building course in New Zealand. One of my students commented to me after my lecture on overscribing that while spending some time shearing sheep in Scotland, a farmer had shown him an old drawing iron, and he too had noticed the different measurements between the two sets of points. I was now more determined to make a trip to Norway and see for myself some old drawing irons. My trip came in the fall of 1989 while I was doing a home show in Germany. That was it, I was too close to Norway to let the opportunity pass. I spent seven days touring every log museum I could find. Just seeing those magnificent old log buildings was well worth the trip, but my main goal was to find old drawing irons. I found three of them in the Maihaugen Museum in Lillehammer. Each of the drawing irons had about ¼" difference between the two sets of points.

Through my experiences with overscribing and discovering how well this technique works, I don't find it hard to believe that five hundred to eight hundred years ago the log builder practiced overscribing. I firmly believe overscribing of the long groove was a common practice of years past.

The saddle notch will prevent twisting of the logs to a higher degree. First of all, you have scarfed off two angled surfaces on the log below for the next log to be notched over, establishing a better locking shape, and creating a complete bearing surface from the lateral groove to the top of the notch, preventing any side openings from occurring. Keep in mind that to maintain the bearing weight on the sides of the saddle notch; a relief opening will have to be made at the top of the notch (which can be covered by the next log that goes over). Without the relief opening at the top of the notch too much weight will bear at the peak of the saddle scarf and resulting in not enough compression on the sides of the notch as the logs shrink.

You could probably dismantle a building constructed this way one hundred years later and actually see some of the logs twist as you lifted each round off. Imagine logs twisting and moving after being locked in one position for one hundred years.

I hear log builders talk about doing compression - fit notches and I know these builders use equal settings in scribing the notch and the groove. They should be calling this 'compression fit grooves' as it is virtually impossible tohave a compression fit notch without overscribing the lateral groove.

The word compression relays a downward movement. How can a log's notch compress if all its groove edges are firmly resting on the log below? Given time, there will be a slight compression (1/16" to 1/8") of the lateral groove into the lower log, but the shrinkage that has occurred at the notch and the great resistance to compression of the groove will allow the notch to open up. By overscribing you let the notch compress, not the groove.

Be careful when cutting the notch not to weaken the edges of the notch. You want a nice biting edge to allow for compression of the notch but you want to maintain the strength of the wood fibers near the edge. If you hook, or back cut, the edge of the notch too much the wood fibers will collapse, defeating the whole purpose of overscribing.

Many builders use log pieces as they build between doors and windows in order to straighten crooked logs, cut down on wasting materials, or to increase effective log diameter. In this situation, each log section may have only one notch. Use small wedges to hold up the un-notched end to maintain a consistent gap in the groove. Once wedged, toe - nail the upper log to the lower log so that nothing will shift out of place, then continue building.

Another way to support such logs is to scribe a 2" to 3" portion on both sides of the groove with the scribe setting that is used for the notch. This eliminates the need for the small wooden wedges, though toe - nailing is still advisable since the log had just one notch. When the final trim is cut for the door or window openings the toe - nailed areas will be cut away. Be careful when cutting doors and windows at this stage, we nail up 2x4's on both sides of the opening to be cut; thereby keeping the logs in their exact position while cutting.

As more and more weight is added by successive rounds the ¼" gap will continue to close. When wall construction is done the gaps throughout the building's grooves will have closed to 1/16" to 1/8". The plate log grooves will be open about ¼" because little weight is bearing on them, but this will begin to close once the weight of the roof is on them.

Once the notches open, then the logs have even more freedom to twist and roll, causing further gaps to appear. To correct this, you must cause compression of the notch. This is accomplished by overscribing the groove to place more weight on the notch. The fits of overscribed buildings get tighter with age and they maintain stable locked corners. I believe the gap in the groove created by overscribing closes due to approximately 75% compression and 25% shrinkage at the notch.

The term "settling factor," or just "settling" is commonly used to refer to shrinkage and settling or shrinkage and compression. You should allow 3 to 5 years for settling, depending upon the character of the wood used and local climatic conditions. In a very cold or humid climate it may take three years for settling to finish. These logs will only loose moisture to whatever level of moisture content is in the air. Thus, less shrinkage would also mean less overscribe is needed. In a very hot or dry climate, settling could continue for up to five years, sucking almost all moisture content from the logs and of course, causing more shrinkage and in turn, forcing you to overscribe more.

The Scandinavian countries have a cold climate, this could be why there is a ¼" difference between the two sets of points on the old drawing irons, and not more. Another reason might well be because the majority of the notches they used were a type of lock notch that retains more of the heartwood of the log. There is less shrinkage in the heartwood - most of the shrinkage in a log will occur in the sapwood (the outermost layers of the log).

Also consider elevation. I have one building that was built in 1985 using only 1/8" overscribe. It is located in a cold area at 5000' elevation. The exterior of this building is perfectly tight. The interior, because of higher heat and lower humidity, has opened up only some of the notches to 1/8" gap. Had I used a ¼" overscribe throughout the building both exterior and interior would have been flawless.

Don't be mislead into thinking that if a building is overscribed more on the interior part of the lateral groove, this would solve the problem of more shrinkage taking place on the interior walls due to the heating condition inside the home. With more of an overscribe on the interior part of the groove, it is most probable this extra overscribe will never close entirely, because the exterior part of the notch (mainly) and groove will prevent it. Unless the log could actually roll inwards, out of its vertical position to close the extra interior overscribe. I would not want this kind of action to take place, nor do I believe it could. An exterior corner of a building has three sides of its notch facing the exterior and only one side facing in - it would be virtually impossible for this log to roll inwards.

Take for example, the building previously mentioned. Some of the interior notches were open 1/8". If we were somehow capable of removing another 1/8" from only the interior grooves, both the interior notches and grooves would be open by 1/8" because the tight fitting exterior notches and grooves would be holding the building in that position. I believe in an equal overscribe for the exterior and interior parts of the groove. The exterior parts of the notches will remain more compressed while the interior parts of the notches (due to more shrinkage) will rebound to keep them tight.

Different species of wood also have different shrinkage rates but the majority of them shrink approximately ½" for 12" diameter green logs. (A 12" diameter log in green condition will end up 11-½" diameter in dry condition.) A log shrinks in diameter, in other words, the outer log surface moves inwards towards the pith (centre of the log).

Common practice is to allow for ¾" of settling per foot of green log. Most builders recognize and accept it, but where did it come from? Is the figure of ¾" per foot an old calculation for settling?

The settling ratio of ¾" per foot of green log is used in this way: if you had 10' of green log wall and multiply that by ¾" per foot you have a total allowance of 7-1/2" for the settling factor. In other words, you are predicting that 10' of green wall will settle 7-1/2" creating a dry wall of 9'4-1/2". If we had a 4' window in this wall we would predict 4" x 3/4"=3" settling factor. We would then mark the height of the window opening as 4' 3" making the allowance for shrinkage and compression that we predict will occur in 4' of green wall.

At one time I believed that overscribed grooves would require more than ¾" per foot for settling and I started using 1" per foot. After examining these houses four years later, I found there was still 1-1/2" to 2" of settling space remaining above windows and doors. So, now I find ¾" per foot of green log to be quite adequate.

It is obvious that log builders of old had to allow for settling, but what did they use to determine the amount of settling? If overscribing of the groove was a common practice in the past, log builders would have known that each log shrinks about ½" and each log will compress the notch about ¼" because of their overscribe. This would have made their settling allowance about ¾" per foot. Slumping was not common in the past because shallow laterals and kerfs or wedges were used to help control checking. It is possible that the figure of ¾" represents overscribing in the past.

It has been said in recent years that in the 1970's log builders needed ¾" to account for the ½" of shrinkage and ¼" or so of slumping caused by checking in the grooves. This might be true because of the way we were building, but we knew nothing about slumping of logs in 1975, so how could we predict that we needed ¼" for slumping? We were taught ¾" per foot for shrinkage and settling, not for shrinkage and slumping. The slumping nature of logs caused by cutting deep lateral "V" grooves and no kerf cut on the top of wall logs only became apparent in the late 1970's and early 1980's, when we started seeing buildings that were drastically checked in the grooves and slumped down. Luckily, most builders were using ¾" per foot of and did not have problems like stuck doors and broken windows.

I have heard some log builders say that they are now able to use ½" per foot for settling by minimizing the slumping action of logs with shallow grooves and kerf cut on top of wall logs. I find this to be perfectly logical because the standard equal-scribe technique they are using will not cause compression of the notch.

One would think that with small logs you would overscribe less and with big logs overscribe more, but I have not found this necessary. I now use the same overscribe whether the logs are big, medium, or small. In my experience I have found that big logs shrink very slowly and therefore will not shrink as much by comparison, to a small log. Big logs will have less shrinkage but more weight to compress. Small logs will have more shrinkage and less weight to compress, therefore different size logs do not require different amounts of overscribe.

One might also think it necessary to overscribe soft woods more, for example Engelmann spruce or lodgepole pine, thinking of more compression and semi- hard woods less, for example Douglas fir and larch, expecting less compression. I have also found this not to be the case. We use the same overscribe whether the wood is soft or semi-hard. This is because both will compress at about the same rate because of their differences in weight and wood fiber strength. The semi-hard wood has wood fibers stronger in compression but it is very heavy in weight. The softwood has fibers weaker in compression but is also lighter in weight.

The cap log that is notched over the ceiling beams is a prime example of extreme resistance. I have been dealing with this log by overscribing 1/8" to ¼" (instead of 3/8") depending upon the distance between notches. This technique's weakness is buildings with lofts in which floor joists may come only halfway down a wall, but forcing you to overscribe a small amount the whole length of the groove. This does not compress the far wall notch enough.

I am now trying a different method in dealing with the cap log, by using a 3/8" overscribe for the groove and a 3/16" overscribe for the ceiling beam (second-floor log joist) notches. For example, if the wall notches have a final scribe of 3" and the lateral groove scribe is 3-3/8", then the final scribe for notches over the log ceiling beams will be 3-3/16", (Figure 4 and 5). I am trying to direct more weight onto the log wall corner notches than the log ceiling beam notches. It seems to me that the ceiling beam notches are somewhere between a true corner notch and a lateral groove, and should be overscribed accordingly.

A better solution is to reduce the angle of the saddle scarfs, create a relief opening, and direct all of the bearing weight onto the sides of the saddle scarfs. An additional benefit is that less steep saddle scarfs tend to remove more sapwood than steep saddle scarfs, and this also reduces shrinkage.

I have done some dovetail work in the past, but not since I started overscribing. I intend to do some experimenting with the dovetail in the future. The dovetail notch on any given log has two angled flat surfaces, and these surfaces are in direct contact with other angled flat surfaces. Dovetails do not have sharp, strong edges to bite into the log below, so there will be far less compression with this notch. I would suspect that you could use 1/8" to ¼" overscribe, but would recommend experimenting with 1/8" overscribe first, unless the building is a tool shed in your backyard, then try 1¼".

When notching the sill logs over the half-logs I underscribe the notch. Obviously, the first round of logwork has no lateral groove; there are only two flat surfaces that should meet with one another - the bottom of the first round of logs and the floor or foundation. Still, we must allow for the compression and shrinkage of these first notches of the building. This is accomplished by underscribing these notches by 5/8" (when using the 4-point saddle notch). (Sill logs have no groove, so it would be inaccurate to say, "overscribe the sill logs," instead I call this underscribing the notch. Underscribing the notch and overscribing the groove accomplish the same task.) Compression and shrinkage eventually makes the flattened sill log meet the floor or foundation. Midway along the sill log, I support this temporary gap between the sill log and the floor (or foundation) by placing about 3 strips of rubber cut from an old inner tube. This supports the sill log, but still allows for compression. The strips of rubber are discarded when the building is dismantled and reassembled on its foundation. If the building is constructed on its permanent foundation they will have to remain.

I have never liked the appearance of the half log protruding to the end of the log extensions, so at one time I used a blind notch for the half log and sill log. But the blind notch is incompatible with underscribing the first round of logwork because there is only half of a notch. Consequently, the sill log will want to tip off the half log requiring you to wedge the underscribe on the outside of the building. This bothered me for some time until I developed what I call the bicep of the half log (Figure7).

The half-log extends 15" to 16" beyond the centreline of the wall. This extension is then rounded off giving the appearance of added strength and the look of a bicep. This way there is a whole notch to underscribe (instead of half of a notch) and no unsightly half log protruding to the ends of the building.

The next two lateral grooves above the half logs and sill logs I overscribe by 5/8". The next three grooves above this get ½" overscribe, and the remainder of the grooves above this (3 or 4 grooves depending on the size of the logs and the height of the wall) are overscribed 3/8".

I use a common underscribe and overscribe to start the building, then reduce the common overscribe as the walls get higher. This is because there is more weight on the lower courses than there is nearer the top of the wall. And the sill log underscribes should be varied to account for notch resistance in the same way that overscribes are varied and for the same purpose (see Varying Overscribes above).

The overscribe technique I use for the round notch is slightly different, because the round shape won't compress as much as the triangular shape of the 4-point saddle notch. I start with a 3/8" underscribe of the sill logs to the half logs, and continue with this amount for the next 4 or 5 grooves. The remainder of the grooves above this, I overscribe by ¼" to finish the log wall height. The majority of our buildings use the 4-point saddle, and if I built more round notch buildings I would probably refine these overscribe amounts somewhat.

I haven't had the opportunity to overscribe a full two-story house. I built a number of two story houses in the past, but this was before overscribing. Obviously, there will be much more weight bearing down on the lower story of this building in comparison to a one story house. Let me speculate how I would build, for example, a house with 16 rounds of logwork, green logs, and a 4-point saddle notch. I would start by underscribing the sill log to the half log by ¾" and the next 3 grooves above would also be overscribed by ¾". The following 4 grooves would be overscribed by 5/8". The next 4 grooves would get ½" overscribe, and the building would finish with the last 4 grooves using 3/8" overscribe.

I would be comfortable using ¾" overscribe on the bottom rounds, but any more than this I would definitely question. The saddle scarf is already an extreme angled surface, and large overscribes will move the bottom of the notch up onto this angled surface. With a 1" overscribe, there would be a possibility that the bottom of the notch could embed itself too deeply into the saddle scarfs. As it is moving down the angled surfaces by compression and shrinkage and become so embedded that the log will hang up, and not allow the overscribe to close completely.

Should you overscribe logs if they are dry? I am asked this frequently, and the answer is yes. You are still trying to establish a compression fit, and dry logs will compress. Dry logs will also shrink though not as much as green logs. Using dry logs I would overscribe between 1/8" and ½", depending upon the moisture content of the logs and the type of notch used.

For example: the logs are very dry and you are using a round notch. Start the building with ¼" overscribe and continue to about halfway up the wall. Finish the remaining rounds with 1/8" overscribe. Another example: the logs are very dry and you are using the 4-point saddle notch. Start the building with 3/8" overscribe and continue to about the halfway point of the projected wall height. Finish the remaining rounds of log work with ¼" overscribe.

The biggest problem we have had in modern log building is our tight fitting long grooves. You can change the notching system to as many shapes and names that you can dream up, but if you continue to use an equal scribe for both the notch and the long groove, you can bet the notches will eventually open up. Log builders should go back to some of their older buildings and take a good hard look at them. You will probably conclude that if you could remove ¼" to 3/8" of wood from all the edges of the long groove all the notches would close and once again you would have a house with tight notches and tight laterals.

There are three possible outcomes to an overscribed building:

1. you don't overscribe enough and some notches open up ( though they are tighter than if you had used the standard equal scribe technique);

2. you overscribe too much and some of the grooves stay open slightly;

3. you overscribe just the right amount and notches and laterals are tight throughout the building.

When you look at overscribing in this way it does not appear to be as risky as you might first have thought. The third outcome is your goal, but results "1" and "2" are also okay, since you already know that the equal-scribe technique virtually assures loose notches. Therefore, any of the three outcomes will bring an improvement to your log buildings.

My suggestion is, give overscribing a try. In order to boost your confidence, try a modest ¼" overscribe, I can guarantee the overscribe will close, using green to semi-green logs with a round notch or whatever type of saddle notch you use. Once you see the better fits you get, you will have the confidence to use larger overscribes and refine them for your purposes.

I always tell prospective customers, when looking for a company to build their log house, ask to view a home that is 4 to 5 years old, because only then will they see the true picture that is representative of that company's work. The building that is on the worksite usually looks pretty good to the untrained eye and can be quite misleading.

Right after construction, the lateral grooves of an overscribed building are somewhat loose fitting, with 100% of the weight of the building on the corners. I have been asked, "How does the customer react to this"? But when I explain the technique, most people see the logic in it. I also ask them if they want a tight fitting house for two years or two hundred! It is then up to them, though I refuse to do another building without the overscribe.