CARPENTER’S KNOWLEDGE IN SETTING OUT ROOFS AND BUILDINGS FROM SAXON TO THE LATE MEDIEVAL PERIOD

This study is the result of my observations and research while working on medieval timber frame buildings both before my retirement and later from studying many books, papers and publications on Anglo-Saxon and Viking buildings. The archaeology of these buildings shows floor plans with rows of postholes, especially, the excavations at Market Field at Steyning Sussex[1], Cowdery Down Hampshire[2] (Fig. 1), at the Viking sites 16-22 Coppergate in York[3], and the river front excavations in Dublin[4] and London[5]. There are many theories and reconstructions of Saxon/Viking buildings but no insight or understanding as to how the Saxon builders set them out and constructed them. Much speculation has been printed in archaeological journals and there have been many attempts at constructing them in open air museums. I realised that many Saxon buildings had double square floor plans (Fig. 1) [The length of the building is twice the width] and it became obvious that they were set out using a rod equal to the span or width of the building. (Figs. 1 and 2)

[1] Mark Gardiner Excavations of a Late Anglo-Saxon Settlement at Market Field Styning 1988-89, Sussex Archaeological Collections 131 (1993) 21-67

[2] Martin Millett and Simon James, Excavations at Cowdery’s Down Basingstoke, Hampshire, 1978-81, Archaeological Journal 140 (1983)151-279

[3] Richard Hall, Anglo-Scandinavian Occupation at 16-22 Coppergate York, Defining a Townscape, York Archaeological Trust 2014 523-814

[4] Hilary Murray Viking and Early Medieval Buildings in Dublin, BAR British Series 119 (1983) 1-215

[5] Valerie Horsman, Christine Milne and Gustav Milne Aspects of Saxo-Norman London: 1 Building and Street Development, London and Middlesex Archaeological Society (1988) 6-123, Gustav Milne Timber Building Techniques in London 900-1400, London and Middlesex Archaeological Society (1992), David Bowsher, Tony Dyson, Nick Holder and Isca Howell, The London Guildhall, Museum of London, Monograph 36, 3 volumes. 2007

The first thing the Saxon builder would have required was a rod. This was a long straight pole, which he would have taken to a growing wood to measure available trees that were suitable for the building; the trees would have needed to have the length and width to be cut into tie beams, wall plates, and rafters. If so, he would place the rod up against the growing tree and cut a notch at the top of the rod to give a standard length of the timber available. With this gauge, the builder would go to the building site and set out the building, using the rod and long pegs as follows (Fig. 2).

First the builder secures a centre setting-out peg A, then from peg A he lays the rod on the ground to mark the position of peg B. The rod is then moved back to peg A and laid on the ground to mark the position of peg C. The setting-out Peg C is sighted up with pegs A and B to make a straight base line, i.e., the length of the building. From peg A, the rod is placed on the ground at roughly a right angle to peg A to locate the position of peg D, the width of the building.

There is a wealth of evidence of Saxon and Viking craftsmanship in ship building, metal working and the production of high-quality steel to make weapons and tools. Intricate gold and silver working as seen in our museums, found as grave goods, Sutton Hoo in Suffolk being a good example and the Staffordshire hoard . The Saxon Graveney boat found in Kent , now in the care of the Greenwich Maritime Museum, and Viking ships on display in the museums in Scandinavia are amazing examples of woodwork craftsmanship. Even on our doorstep the Saxon door, still in use in Westminster Abbey, shows superb craftsmanship in timber for all to see close-up .

The evidence from my research has been demonstrated on many buildings as shown in the list of articles on this website. The technique was used not only to set out domestic buildings, barns and large halls but also churches and cathedrals. It was all based on the use of a length of string or cord, and did not require any numeric measurements or geometry.

Such a simple procedure, using large dividers and a rod, became a method of setting out which has lasted for thousands of years and was only out-dated by the introduction of scaled drawings on paper produced by architects and surveyors in the 17th and 18th century. These early professionals used tee squares, brass set-squares and protractors on a drawing board and the resulting drawings were then handed to the builder to build the building. By the end of the 18th century the length of string for measuring and setting out became obsolete.

Throughout history, cord or string has been an important part of the craftsman’s kit (Fig. 3). Whilst one immediately thinks of bricklayers and stonemasons working to a line, most building crafts need a length of string for one use or another. Apart from its importance as an aid to setting out in constructing buildings as diverse as the ancient pyramids, Stonehenge and our impressive cathedrals, the simple line has also been responsible for achieving a vertical plumb line and for drawing an ellipse.

Modern-day technology has dispensed with the many uses of the handy string line. The laser is now found in the modern workman’s toolkit to level, plumb and set out buildings. The historic use of the string line has almost been forgotten.

Medieval master craftsmen would have met regularly in their lodges or guilds to discuss methods of construction, problem-solving and how to use practical geometry, dividing up a circle and creating angles, not necessarily attributing it to Pythagoras or Euclid. This knowledge, while closely guarded by the guild, would be passed down from master to apprentice; when the apprentice was qualified, he became a master craftsman, keeping the knowledge secret in the same way to protect the craft.

Regulating measurement originated with Henry III or his son Edward I sometime between 1216 and 1307. The original yard was a length devised by one of these monarchs, based apparently on a rod of iron. However, not only is there no record of this early yard, crucially, it also seems not to have been widely used. It was only with the Tudors that two important Acts of Parliament set out standard measurements of length and area: the yard, the rod and the acre. Henry VII in 1497 produced a fixed yard called the Exchequer Standard, which in 1588 was slightly revised by Elizabeth I. Moreover, to cement acceptance of the official weights and measures, Elizabeth appointed inspectors to enforce their use. These measurements, the Exchequer Standard first established by Henry III or Edward I, were calculated as follows:

3 Dry barleycorns laid end to end = 1 Inch
12 Inches = 1 Foot
3 Feet = 1 Yard
5½ Yards (16½ft) = 1 Rod
40 Rods in length and 4 in breadth = 1 Acre

The medieval master craftsman was likely to have been a master carpenter or may also have been a master stonemason, or else they worked alongside one another. The Master would have had a basic knowledge of geometry, of how to use dividers and compasses to bisect angles and circles; and he would have had to use a straight edge and setsquare. He was capable of drawing full-size plans and elevations (there is evidence in the form of full-size drawings traced on the floors of Wells and York cathedrals) and especially of working-out proportions, having sufficient knowledge to build cathedrals and castles, structures where we know using a rod and a line was essential.

Supporting my argument, the late Arnold Pacey in his book Medieval Architectural Drawings, published in 2007, states that there are no surviving medieval working drawings produced by craftsmen. Indeed, in private correspondence with the author, Pacey has said that he finds the basic argument of my theory convincing – namely, that the medieval carpenters didn’t need working drawings to set out roofs and buildings because they had a much simpler method that served them very well.

Such practical geometry, as we could describe it, would have been used, for example, to make a right-angled triangle with sides in the ratio of 3:4:5 (as in Pythagoras’s Theorem), which would have been applied to the setting out of right angles for buildings. In this instance, they would have used the rod or a cord 12 fixed units long with a knot at 3 units, another knot at 4 units and a final knot at 5 units; these three lengths, when joined together with a total perimeter of 12 units, make a right-angled triangle.

At the Vernacular Architecture Group winter conference at Leicester in January 2017, in this instance on timber and trees, the resident master carpenter of the Weald and Downland Museum, Joe Thompson, gave a presentation in which he showed how the circumference of a tree can be measured with a piece of string, which is then folded twice to determine how much timber can be retrieved from a felled log. This ancient method was formalised by Edward Hoppus in the eighteenth century into tables measured in feet, providing easy references taken from a girth tape. 

This simple procedure by folding a length of string or cord can be applied to the task of setting out historic buildings. The string is lightly stretched the length of the rod secured with a nail, is folded four times until 8 equal units are formed from the half span. This method of setting out does not require any form of numerical measurement, a protractor for angles, or any form of geometry. The main factors controlling the design of a medieval building were the length of the available timbers required for tie-beams (giving the maximum width of the building) along with the length of the posts (giving the height of the building) and the available size of building plot, as well as what the owner could afford.

Tie-beam lengths are usually between 5 to 7m, as in the Weald and Downland Museum’s Pendean Farmhouse 1605d, where the span is 5.03m, (16ft 6in the Elizabethan rod) and in the same museum’s Bayleaf Wealden Hall House, which is 6.55m wide. Widths depended on the available sizes and quality of local timber, as stated by Rackham in his studies on Prittlewell Priory and Grundle House.

Given these natural restrictions, how was the medieval building designed? How did the medieval carpenters manage to maintain common roof pitches of 43, 48, 52, 55 and 60o? Take, for example, the 110 roof pitches used on medieval domestic roofs in Kent, examined as part of the Royal Commission survey on the Historic Monuments of England (RCHME) conducted from 1968 to 1992. The angles of these different roof pitches were consistent throughout this period.

The choice of pitch was down to the carpenter pitching the roof as well as the materials he intended to use: thatch and shingles required a steeper pitch, say 52 or 55°, than tiles or slate roofs, which were usually 48° respectively. The results of a survey in the Midlands by Nat Alcock and Dan Miles, The Medieval Peasant House in Midland England, including 134 cruck and box-frame buildings, which again all have these same five constant roof pitches.

Since the publication of Carpenter’s Knowledge in setting out Roofs and Buildings… in Vernacular Architecture in 2020 I have continued my research from domestic timber buildings to stone and timber tithe barns, churches and cathedrals with positive results. I have also extended my research to Europe and Scandinavia comparing churches, cathedrals and vernacular buildings. I have found some similarities basing the setting out of these buildings to my theory to see if there is a common link with the medieval craftsmen in England using the same method of setting out in Europe.