Because of the pull of the 8′ strings and indirectly the 4′ strings the top of the bentside slightly caves inwards. This causes lateral compression in the soundboard. I think controlled compression of the soundboard was an important goal of the old builders. A compressed soundboard behaves in a different way, it is much more responsive, resonant and alive.
Buckling (vervormen, inzakken in Dutch) of the soundboard is the enemy of a compressed soundboard because it lets the compression leak away. So how do you keep the soundboard flat and compressed without buckling? One factor is to keep the soundboard stiff enough to counteract the compression force. It has all to do with a balance between these two factors.
The Italians, in general, had relatively flexible cases. This puts a big strain on the soundboard. To keep the soundboard flat they strengthened the soundboard with ribs. With the angle of the ribs they could regulate the flexibility of the soundboard. Ruckers made a stronger case, used no ribs, but a thicker soundboard and lessened the tension on the soundboard by using upper braces and putting them in the with a certain amount of pressure. In the early French tradition they have sometimes u-frames and sometimes upper braces like Ruckers, but use ribs as well because they liked thin soundboards and because they had a second 8′. The later French builders used the Ruckers style upper braces and mostly no ribs, so they had to make the case stronger by using more upper braces, often making them less flexible by inserting triangles in the corner between upper brace and liner, or by making vertical upper braces like Hemsch.
I think nowadays there is a tendency to make the soundboards too thin, especially between the 8′ bridge and the liner. What’s more, it is generally thought that we should string our harpsichords strong to have a strong sound. The downside of this is, that if we have a soundboard which is too thin and we put too much strain on it with a heavy stringing, then the soundboard buckles and doesn’t behave optimal any more.
In my harpsichords I now use fairly thick soundboards, combined with a not too heavy stringing. This results in instruments with a strong and flexible tone. Another factor is drying the soundboard too much prior to gluing it in. In my first harpsichords I did this too. I noticed that especially in the tenor and altus region the soundboard buckled very much. Now I only dry the treble and bass with breeding lamps. The middle region gets enough room because the case is pulled slightly in by the force of the stringband.
When I first saw the diagram of soundboard thickness in the Ruckers book by O’Brien I couldn’t believe my eyes. The soundboard didn’t get thinner from the bridge to the bentside! And it was a lot thicker than I had ever dreamed possible. This could not work! But it does, and since then I have been experimenting with my soundboards to optimise them.