News & journal

‹ Back

Bevelled tile edges: why cut not moulded?

 

The majority of medieval floor tiles, have the edges formed into a bevel so that the tile is widest on the upward face, and narrows towards the back. It is not hard to see why this might have been done: when the tiles are laid in a mortar base it means the mortar can squeeze up between the tiles, holding them firmly in position (critical for a floor tile) while still allowing the faces of the tiles to meet without a gap, so the patterns join up.

 

Drying tiles showing the mortar gaps created by the bevelled edge

 

So far, so straightforward. But, from the point of view of understanding the making process, the question is how the bevels were made. Or, more specifically, why the bevel edges seem to have been cut once the tile was formed, rather than created during the initial moulding process.

 

The latter would seem to be more efficient. If the mould itself has bevelled edges, then the tile emerges with a neat repeatable bevel from the outset; no additional work is required. Cutting, on the other hand, means that at some point while still soft enough to cut, each tile needs to have all four edges trimmed off at the desired angle, adding an additional stage to the tile making process. In addition, given that tiles were typically made with clay that contained inclusions such as small stones and grit, it was not necessarily easy to produce a good clean cut and there was a risk of damaging the tile edge.

 

However, in writing on medieval tiles there is a general view that edges were cut not moulded ^{eg 1}. This is based on evidence from the surface finish of the edges of tiles which in many cases seems to suggest trimming by some sort of cutting implement, though exactly what that was is not necessarily clear ². If this interpretation is correct, and medieval tilers cut the bevels rather than moulded them, then why?

 

One possibility would be that it was more complex to make an effective bevel-edge mould than a straight-sided one, or that the bevel-edge mould simply did not work very well. In terms of complexity, whilst it is marginally trickier to make a fixed frame mould with angled interior sides, it is not so very different from making one with vertical sides, even without the benefit of modern tools ³. 

 

A fixed frame, bevel-edged tile mould

 

However, with a fixed frame mould the an inbuilt bevel does mean that it is very difficult to use a block of wood to push the new tile out of the mould once it has been formed (you can only push from the narrower to the wider part of the tile, and this leaves the edges unsupported by the block and they stick to the mould sides and distort). So, it seems there would be good practical reasons not to use a fixed frame bevelled mould, and hence to need to cut the bevels later. However, even a straight-sided fixed frame mould requires a precisely fitting block which can be inserted to push the tile out of the mould without distorting it, especially with clays which are quite sticky, as directly dug clays often seem to be. 

 

Whilst it is commonly assumed that the moulds used were simple fixed frames, an alternative would be to use a mould that can be dismantled (which I will call a 'loose-frame' mould) to remove the newly formed tile. In this case making a mould bevelled is straightforward, and the problem of getting the formed tile out disappears: the mould is simply removed from around the tile. The simplest way to make such a mould is to form four identical side pieces, with angled internal edges, and fit these together on a board with fixed ’stops’ to hold the frame in place.

A loose-frame bevel-edged tile mould - held in position by fixed blocks

Using such a mould, the newly formed tile can be lifted out of the stops and off the board with the frame in place, sticking to the edges of the tile, placed on the drying board and the individual mould sides then easily removed by gently levering the corners apart. This means that tiles with bevelled edges are as straightforward to produce as those with vertical edges. In terms of efficiency, while there is an additional stage involved in reassembling the mould between tiles, this takes little time and certainly less than the additional time that would be required to cut the bevels. In addition, it is fairly easy for damage to occur in the cutting process — particularly if a stone or other object is encountered — which adds another potential cost to the approach.

 

Loose-frame tile mould taken out of the retaining blocks and being removed from the tile
 

In terms of the actual process of forming a tile, there seems to be little practical reason why tile makers would have chosen to make straight-sided tiles then cut the bevels rather than just mould them with bevels to start with. The possible exception would be if tiles were made by putting the stamp block in the bottom of the mould (pattern side up) and then pushing clay into the mould, and stamp, from above. In this case using a bevelled mould would involve having to fill clay ‘under’ the bevel (as the stamp is on the wider face of the tile), and this would be tricky to do effectively. However, it is generally held that this method was not common — a view supported by the fact that most tiles have clear signs of embedded sand on the back of the tile ¹, suggesting forming was done on a sanded board. Such sanding would have been unnecessary if the tile was moulded directly into the stamp, and then pushed out with a suitably sized wooden block.

 

So, if cutting rather than moulding bevels was preferred, there must have been advantages somewhere else in the process. Having built a number of tile moulds, and made a number of wooden tile stamps, I can think of one factor which might tip the balance towards cutting rather than moulding: it makes it easy to accommodate differences in tile size.

 

If a bevel-edged mould is used then as there is no later stage in which the tile is trimmed, the mould must be the correct size for the final finished tile. This means that a different mould has to be made for each different sized tile, and the mould needs to be very accurately sized. This is further complicated by the variation in clay shrinkage rates. The mould needs to be the right size to make a tile from wet clay that will shrink, as it dries and is fried, to exactly the size needed for the final tile. Given that clay shrinkage is typically about 8% - 12% or more, depending on the clay and firing temperature, using a different clay source might require new moulds to be made to achieve a specific tile size. However, if the edges of the tile are going to be trimmed anyway, then the mould simply has to be bigger than the largest size needed, and the only thing that has to be made specifically to the right size is the wooden stamp for the tile pattern.

 

This leads logically to a process where an oversized straight-sided mould is used; the tile is formed in the mould; a wooden stamp of the required size is used to impress the pattern into the surface while the tile is still in the mould (leaving an excess round the edges); the mould is then removed and then the edges of the tiles are trimmed to the correct size using the wooden stamp block as a template for the trimming. In this sequence, all elements of physically forming the tile are done together, producing a correctly sized, bevel-edged, pattern-stamped tile, ready for the application of white slip to fill in the impression.

 

Such a sequence minimises the number of times the tile needs to be handled, but also requires that the tile is solid enough to have the bevels cut as soon as the it comes out of the mould. This means the initial tile forming needs to be done with fairly stiff clay. A single piece of stiff clay (which can be around 2lb 6oz to make a large plain floor tile) is hard to push into the mould, so using stiffer clay favours a ‘piecemeal’ approach: i.e. pressing multiple smaller lumps of clay into the mould and consolidating these into a solid block using a mallet or something similar. Illustration of this process can be seen in an earlier journal article (The work of tile making I - clay and tile forming), and below.

 

Pressing clay pieces into a loose-frame bevel-edged tile mould

 

Using the clay in this state also means that impressing the pattern on the tile while still in the mould makes sense — the clay is already fairly stiff — but still soft enough for the stamp impression to be made with hand pressure, and possibly a couple of mallet blows. Being in the mould the tile is not distorted by the pressure of stamping, and using fairly dry clay means that the stamp comes away cleanly while the tile is held firmly by the mould.

 

Although it would probably have been done as a separate process (one can imagine a board of moulded, stamped, tiles being passed to another tile maker) the addition of a liquid white clay slip can be done immediately at this stage. The complete board of tiles could then be set to partly dry, prior to scraping back the slip to finish the tile face, and finally, when completely dry, being glazed.

 

Is this how the medieval tiler worked? Of course there are other possibilities, and not all tilers necessarily worked in the same  ways, but I think it is at least plausible. What is interesting is how the development of a method like this comes from the way in which the specific details of different stages of the process interact to determine the efficiency, or effectiveness, of others. Puzzling over an answer to the question of why tilers don’t seem to have moulded their tiles with bevels, sets in train a whole series of decisions about subsequent elements of the process, each setting the parameters for the next. This is where it feels like the process of actually making things using the tools and raw materials available at the time and considering some of the the practical constraints of production, can start to yield some real insight.  

 

Notes

 

1. Eames, E., 1980. Medieval lead-glazed earthenware tiles in the Department of Medieval and later Antiquities, British Museum; British Museum Publications, British Museum, London. p18

2. I hope to discuss this in a bit more detail elsewhere.

3. Interestingly, on p14 of English Tilers (Eames, E 1992. British Museum Press, London) there is a photograph of tilers' tools from the late 19th century which, it is suggested, are little different from those used by medieval tilers, but the form in the illustration does appear to have bevelled edges.