A History of Woodworking ~ Raymond McInnis

The Evolution of the English Timber Roof

Westminster Hall Hammer-Beam Arches Click here for more on Westminster Hall, its hammer-beam arches, and its designer and head carpenter, Hugh Herland.

From the 13th- to the 16th-centuries, the timber roof is a testament of the achievements of British carpenters, achievements that demonstrate equally extraordinary attainment in both skills and in inventiveness. At the bottom of this page is a fairly comprehensive list of studies of the "open timber roof". Click here for a full-text study with numerous illustrations of the open timber roofs of the middle ages.

Until almost the end of the 14th-century, carpenters are content to follow the British stonemason. Such things in church architecture as canopies, tombs, sedilia and similar constructions, even the early chests, are inspired by similar constructions by British stonemasons.

In techniques of construction, the differences between the stonemason and the carpenter are subtle. Step-by-step, a stonemason chisels the solid blocks into the forms he requires. In contrast, out of a structure's component parts, the corresponding carpenter, step-by-step, erects or builds up the required forms. The carpenter builds a box with framed ends, front and top, cutting his framing from planks. He makes his framing, tenoning and mortising his stiles and rails, fixing in his panels, either in grooves or rebates.

In contrast, the stonemason can only make his frame and panel in one piece, from the solid stone.

And while stone offers greater resistance than wood to crushing weights, it lacks the tensile strength that makes wood such a versatile component of construction.

Think of either of these two structures: a Gothic church made from wood, or a tie-beam or hammer-beam roof made from stone. In each case, such a structure will collapse: the one from the structure's crushing weight, the other from the structure's inability to withstand the sagging strain on its structural beams.

The Meaning ot the Open Timber Roof

The timber roof may be regarded as a notable feature of English architecture. Although this country can show fine examples of stone vaulting in its cathedrals and great churches, the use of stone for roofing was nothing like so common here as on the Continent. On the other hand, nowhere outside England can be found such a series of magnificent timber roofs as those of which Westminster Hall, Hampton Court, and the angel roofs of East Anglia are examples. In its simplest form the framing of the timber roof consists of a series of pairs of rafters with their feet pegged to the wall-plates and their upper ends halved and pegged together. Each pair of rafters was known as a 'couple' ... .

Source: L F Salzman, Building in England Down to 1540: A Documentary History New York: Oxford University Press, 1952, 210.


Westminster Hall's Timber Interior,looking north

Thus as applied to churches and sacred buildings, it is with the timber roof that the early carpenters and joiners emancipate themselves from the stonemason's traditions.

And, where the timber roof is employed in secular houses, likewise, very little hiatus exists in the evolution of the timber roof in non-religious construction, that is, where the timber roof is left unceiled, with its timbers exposed.

Instead, in Britain, with the decline of the Great Hall and the advent of the Long Gallery, the custom arose of ceiling in, at comparatively moderate heights, and ornamenting the ceiling with moulded plaster.

This method has advantages. Under large roofs, for example, spaces can be subdivided into private rooms, with the partition walls extending up to ceiling height. With the open timber roof, however, such structural features not possible.

The Evolution of Woodworking Joints

(Note: Customarily these joints are referred to by the more widely used term, "scarfs", and the practice of making timber joints is known as "scarfing". Click here for an extensive account of scarfs and scarfing.)

In the 1960s, Cecil Alec Hewett, the prominent historian of medieval British timber building practices, argues that “a widely held view” prevails

“that carpenters, at any time in the past, had at their disposal the full range of timber joints from which they were free to choose a particular joint for their purpose”.

Hewett counters such claims with his 1969 book The Development of Carpentry, 1200-1700; an Essex study.

At bottom, the implication of the historical succession of timber buildings he describes in his 1969 work -- building by building -- is that joints of various categories are developed and refined over long periods.

Click the picture for a pdf of an adaptation of Hewett's account of the development of scarfs.

Similar claims exist for other forms of building, including furniture. For example, studies of the box chair over time show advances in both joints and in such building techniques as paneling.

Authorities who conduct research on the history of British furniture and woodwork agree that the research is difficult, because the crucial evidence they need to draw valid conclusions is unavailable. In part, this difficulty comes from the truth that, over of the passage of time, little survives, certainly not enough to show any reliable evolutionary development. Isolated specimens that are available are not very useful, because, logically, you cannot generalize for a era on the basis of one example. For any history to be reliable, the account of the evolution of woodworking techniques or furniture types must be a chronicle of the fashions which prevailed at a particular time. Unless we can produce other corresponding pieces of similar date and type, which establish the facts we need to make generalizations, we cannot definitely know.

A widely held view argues that carpenters, historically, have at their disposal the full range of timber joints known today, and from which they are free to choose a particular joint for their purpose. However, the implication of the historical succession of timber buildings Cecil Alec Hewett describes – in The Development of Carpentry, 1200-1700 Newton Abbott, England: David and Charles, 1969, pages 20-21 – is that medieval carpenters develop and refine a range of joints of various categories over a long period. The field-work examples are described, as far as is possible, in groupings which can be designated as “of the 14th-century”, or “of the 16th-century”.

As Hewett points out, his arguments apply to numerous joints used in building construction, including “scarfs”, for making longer timbers out of shorter timbers, lap joints, that is, dovetail joints, able to resist withdrawal from their sockets, even the common mortise-and-tenon joint.

Most interesting, perhaps, is that these different types of joints develop separately, some only being perfected at later dates. As a result of this, a building which has the ideal joints for its floor-joists may not have structural joints perfected a later dates.

To establish his hypothesis, Hewett examines, describes, and illustrates numerous historic buildings in Essex county which provide the whole range of evidence in what he thinks is their proper their chronological order in construction.

The joints carpenters incorporate Hewett describes chronologically – but in their separate categories – at the end of his book. All the buildings Hewett includes are dated. Hewett notes his indebtedness to the Frenchman, Henri Deneux an official of the French Historical Monuments Service. While Deneux's investigation is restricted to fewer sources, Hewett is confident that his survey substantially amplifies and endorses Deneux's methods.

We must also be mindful of the possibility of a later copy of an earlier original. In this sense, while oak dressers are made as late as the last quarter of the 18th-century, it only makes for confusion to argue that although made at that time such pieces are as examples of late-eighteenth-century furniture. True, they are of the period, but are really not typical of the period. Today, then, any solitary piece which survives from very early times cannot be reliably indicative of either the customs or the fashions of the time.

What makes Hewett's study plausible is that he, today, can study the construction techniques of medieval buildings in Britain that are still intact. For buildings with no dates of construction, by using carbon-dating techniques, approximate dates can be attached to indiviual buildings. For furniture, dates can be attached, yes, but in comparison, the process is more difficult, at least where makers' names or dates are not easily found, Moreover, the survival of furniture from the medieval period is much more tenuous.

Historic Examples of Joints

Hewett's claims about the evolution of joints apply, for example, to scarfs, a technique in woodworking for making longer timbers out of a collection of shorter timbers. By definition, scarfs are joints where two or more timbers are connected longitudinally into continuous pieces, the ends being halved, notched, or cut away so as to fit into each other with mutual overlapping. It also applies to dovetail joints designed to resist withdrawal from their sockets; and even the more familiar mortise-and-tenon joint, a feature of cabinetmaking in the Arts and Crafts tradition.

In his 1822, manual -- the New Practical Builder -- Peter Nicholson notes that

"In each piece of timber to be joined, the parts of the joints that come in contact are called scarfs".

peter nicholson on scarfing

It is a truth too that these different categories of joints develop separately, and at diifferent dates. As a result of this, strangely, a building which has ideal floor-joist joints may, at the same time, have scarf-joints suffering from detioration, since scarfing dates back before timber floors start to evolve.

For Hewett, the best way to establish this hypothesis is to provide the whole range of evidence arranged in what he thinks is their chronological order.

These carpenters' joints that Hewett analyzes in the book's main part are further described and illustrated at the book's end: "Table of Scarfs in Evolutionary Sequence"

The dates of each building's construction is indicated, including the results of carbon dating. It is fortunate, he says,

that the two oldest barns have dates which are beyond doubt since this makes many subsequent ascriptions more tenable. Similarly, towards the end of the medieval period there can be found a number of buildings which are firmly dated—by carved figures. It is possible to arrange the other structures in a logical series between these two limits, and further­more, a carbon date at the middle of the fourteenth century helps to stabilise this four-century arrangement of theoretical ascriptions.

Field surveys can never be final, and buildings and techniques relevant to Hewett's evidence -- upon which he bases his arguments -- may well exist, but are -- today -- unexamined.

The limited period examined, together with the equally limited area of the county of Essex from which almost every example of construction is chosen, have, however, simplified the task of formulating a coherent course of development and decline for the craft of carpentry within those limits.

If Hewett's arguments ring true, it follows that buildings which incorporate techniques or joints identical with any he describes can also be dated by such joints.

This method for dating carpentry's joints comes from theories first propounded by Henri Deneux -- of the French Historical Monuments Service -- and where Deneux's sources are, strangely, more limited than those of Hewett. For Hewett, his evidence, drawn exclusively from British sources, substantially amplifies and endorses Deneux's proposition.

Hewett's fieldwork examples are described, more or less, in groups, that is, groups which can be designated as "of the 14th-century", or "of the 16th-century". Interestingly, the 14th-century evidently produces exceedingly large and wide varieties of experimental techniques and devices. To cover these, Hewett divides the century into two equal periods, and applies carbon-14- dating to provide firm divisions. Usefully, after describing the buildings, the progress of each period of one half century or one full century is summarised.

Bibliography of Books and Articles on the British Timber Roof

The Westminster Hall Roof and Its 14th-Century Sources Author(s): Lynn T. Courtenay Source: Journal of the Society of Architectural Historians, Vol. 43, No. 4 (Dec., 1984), pp. 295- 309 Publi

Herbert Cescinsky and Ernest R. Gribble, "Westminster Hall and Its Roof", The Burlington Magazine for Connoisseurs, Vol. 40, No. 227 (Feb., 1922), pp. 76-79+82-8.4

THE KING'S CHIEF CARPENTERS By JOHN H. HARVEY E. Toby Morris, R. Gary Black and Stephen O. Tobriner, "Report on the Application of Finite Element Analysis to Historic Structures: Westminster Hall, London", Journal of the Society of Architectural Historians, Vol. 54, No. 3 (Sep., 1995), pp. 336- 347

Eugene S. Ferguson, "The Mind's Eye: Nonverbal Thought in Technology", Science 197 no. 4306 26 August 1977, pages 827-836

Lethaby, King's Craftsmen,

W. H. Harvey, in Connoisseur, xcvu, 1936, pages 333-336,

J. H. Harvey "The Medieval Carpenter And His Work As An Architect", Journal Of The Royal Institute Of British Architects 45, no 15, 13 June 1938, pages 733-743.

Percy J Waldram, "Science and Architecture: Wren and Hooke", Journal of the Royal Institute of Architects 62 1935, page 558

Gene Waddell, The Design of the Westminster Hall Roof Source: Architectural History, Vol. 42 (1999), pp. 47-67