Basil Greenhill in his book the History and Archaeology of the Ship identifies four basic roots for all subsequent ship designs. These roots are:-

1) The Raft boat, (pics Greenhill, Archaeology of the Boat, p 97 et seq  ) that is any boat which derives its flotation from the use of materials which are lighter than water, such as balsa or logs rafts, or skin bladders, as distinct from all other forms of boat which derive their buoyancy from the Archimedian principle of a watertight shell which floats through displacement.

The raft is the most obvious form of buoyant craft and the one which would occur most naturally to the mythical castaway, unless, of course, he happened to be Odysseus. The log raft can be built quickly and easily without tools, merely by lashing together a bundle of logs or reeds. This may be an adequate way from escaping from a tropical desert island, but in other respects the log raft is severely limited. In particular it could not be used in cold northern climates because cold water washing over the partially submerged deck could be fatally uncomfortable for the crew. In warmer climates the log raft has a role in river and lake transport and might even have built up sides so that it looks superficially like a conventional boat. Because of its lack of protection the log raft has little apparent potential for development into a displacement type of craft but a case may be made for the raft as an original form. Greenhill argues that Chinese Junks and Sampans could have developed from rafts on the grounds that their squared off overall shape and basic hull structure incorporate some of the characteristics of the raft. In particular the use of half sawn logs arranged raft fashion to form the lower part of the hull and the insertion of substantial bulkheads rather than frames to form watertight compartments might suggest the raft as an antecedent. However a similar shape could be arrived at from a dug out base. In Northern Europe the only apparent link between log rafts and displacements boats is the two archaeological finds from Humberside, the so called C5BC Brigg raft, which has now been shown not to be a raft, and the C1500BC Ferriby boats which share common structural characteristics with the Brigg vessel but cannot be reliably proved to be developed from rafts.

Rafts made from reed bundles (pics Heyerdahl) such as those used by the ancient Egyptians, and today on lake Titicaca could have a higher freeboard and, as Heyerdahl has proved, could be sailed and taken to sea. In addition it is argued by Landstrom that the reed rafts of the ancient Egyptians provided a pattern for later Egyptian wooden displacement boats on the grounds that the reeds could be protected against rot first by cladding with tarred cloth and later by cladding with wooden planks. It would then be a short step to remove the reeds and leave a plank shelled displacement vessel. It is certainly apparent that early Egyptian planked vessels adopted the same sickle shape (163) as the reed boats but, as with the Humberside boats, there are alternative explanations for this which we will return to in due course.

2) The skin boat, which floats by displacement and consists of skins or other material sewn over a skeleton framework of wood or bone. Skin boats, like rafts, are also familiar in their various forms and may still be found in use today, for example, the Eskimo Kayak and Umiak, the Irish Curragh (1067) (1037) and the Welsh Coracle. Because of their light build ancient skin boats tend not to survive as archaeological finds, but they are an obvious and relatively easy to build form of boat which would be well suited to the needs of primitive hunting societies which either do not have access to suitable timber for wooden boats, or lack the tools necessary to cut and work large timbers. The use of skin as an outer covering requires a fairly complex internal frame which may be fabricated from saplings or from bone and tied with withies. The light build and flexible nature of the skin boat make it more suitable for use in lakes and rivers because violent wave motion could cause the skins to concertina and tear with disastrous results for the occupants. For this reason skin boats have been thought to be unsuitable for open sea passages. However contemporary skin boats such as the Umiak and the Curragh are used in the sea, especially where the boat is launched from a beach into a heavy swell where the lightness of the boat allows the oarsmen to pull away from the beach very quickly (Marcus). Professor Marstrander's reconstruction of the skin boatdepicted n the Kalnes Bronze Age rock carvings (302) (pics) and Tim Severin's hypothetical reconstruction of St Brendan's Boat demonstrated by experiment both that large skin boats were technically feasible and that they could be capable of deep water cruising.

Nevertheless the frail construction of skin boats generally means that they do not last long in use, though they would be easy to repair, nor do they survive as archaeological remains so that evidence for their use in Nordic antiquity comes from iconographic representations such as the Kalnes and Gjerpen rock carvings the interpretation of which is open to serious controversy. As with the log and reed rafts there is an unresolved debate about the role of the skin boat in the evolution of later Nordic ships because the archaeological survival of the C2BC Als (Hjortspring) wooden boat appears to imitate the shape of the earlier skin boats.

Because of the limitations imposed on it by its size and by the materials used in its construction the skin boat did not provide a secure basis for development or evolution. Such vessels were well suited to the fishing and personal transportation needs of a primitive hunting society but were of little use to more settled societies.

3) The Bark boat, (pic) which is made by stripping bark from a suitable tree to form a continuous cylinder in which the ends are gathered together by stitching and the shell of the boat is reinforced by light framing sprung or lashed into place inside the shell.

The bark boat was a blind alley in the history of ship construction and it is the least susceptible of all the root types to further development. The size of the bark boat is limited by the availability of a suitable seamless bark shell and even with internal framing the skin of the boat cannot withstand severe wave motion and is vulnerable to damage. It is therefore suited only to inland waterways and lakes. On the other hand it was very well suited to its natural environment, for example the Canadian waterways.

Superficially the Bark boat is very similar in shape and size to the skin boat but they differ fundamentally in their method of construction and difference lies at the root of a crucial distinction between the two methods of construction into which all subsequent shipbuilding may be divided. In the skin boat the strength of the hull comes from the framework to which the skins are subsequently attached. It may be said to be 'skeleton' built. In the bark boat the strength and integrity of the hull comes from the bark cylinder, which is merely reinforced by light internal framing. It may be said to be 'shell' built. The dichotomy between 'skeleton' and 'shell' building is one which will recur many times in the course of this book. It is, moreover, a dichotomy which must be used with care, and qualified by observation, as will be demonstrated in due course.

4) The Dugout, () The Dugout, (178) which is apparently the simplest of all forms of boats, consisting of a hollowed out log which may be carved or shaped at its ends to form the appearance of a conventional canoe shaped boat. Notwithstanding its simplicity of construction the logboat alone was capable of development into something more grand. All of the other three roots have fundamental structural weaknesses inherent in the materials from which they are built which prevent them from evolving into larger and more sophisticated boats. For this reason rafts, skin boats and bark canoes reach the limit of their development very rapidly and thereafter retain their essential form over millennia in the kinds of riverine and lacustrine roles for which they were originally conceived.

The dugout, logboat or monoxylon is the most prolific and the best known of all primitive boat types. Dugouts are also more likely to survive as archaeological artefacts because they are made from the most massive part of the tree and are less subject to the rapid biodegradation which destroys skin and bark boats. Radio carbon dates taken form dugouts recovered within England alone show that they were widely used between the second millenium BC and 1000 AD. They continue to be used in the Third World today right and, more surprisingly, they may be found even in industrially advanced parts of Europe where more developed methods of boatbuilding are readily available, for example Germany (528) and Poland. The use of a single piece of timber gives the log boat great longitudinal strength and because it is a seamless shell it is inherently watertight.

In its most primitive form the dugout is merely a hollowed out log which may be shaped at the ends to give the appearance of a canoe shaped stem and stern. Beyond the primitive form the log boat has evolve along two lines of development which, it is argued, provide the foundation for most European edge joined and clinker built boats. The first line of development is the expanded dugout, where the log is hollowed out through a narrow slot to form a cylinder (pics Greenhill, Archaeology of the Boat  129-152). The sides of the slot are then forced apart using fire and water giving the finished dugout greater breadth and freeboard thus improving both its carrying capacity and its buoyancy. Once the sides had been forced out they could be cut and shaved to the thickness of planks then reinforced by the insertion of frames and cross beams so that they might superficially resemble a conventional plank boat built on frames and cross beams. (Greenhill, Archaeology of the Boat  132).

The alternative development is for the dugout to be extended where the freeboard is built up by fastening planks or washboards along the top of the dugout. There are various methods of doing this, some of which also require the insertion of frames into the dugout to support the side planks. (178) isometric (143) Zwammerdam 3. Both varieties can combine to form an expanded and extended dugout. Where thus happens the dugout is clearly turning into the more conventional form of plank built boat. There are, of course, many intermediate forms between the primitive monoxylon and the fully developed clinker, or carvel, built boat but two forms are particularly important for the development of European boatbuilding.

The first of these is the plank boat built on a dugout base where the dugout has become vestigial so that it has shrunk into the form of a keel. (pic) Very often this will take the form of a flat broad bottom board carved to rise towards the stem and the stern and scarfed to accept the stem and the stern posts. In this form the vestigial dugout provides the backbone of the vessel providing it with longitudinal strength to reduce the risk of hogging and offering a secure substantial base to which the stem and stern posts, frames and planks may be attached. Both the ancient mediterranean and the Nordic forms of boat building probably originated in this way.

The second variant involved a more sophisticated conception (425) Zwammerdam 2. Here the log boat is split along its length to form two L shaped chine girders which are then used to form the right angle joint between bottom of the boat and its sides. The bottom is usually formed of planks laid edge to edge in carvel fashion and held together by cross beams or floors, alternating with frames which in turn support the sides of the boat. This methods of construction is characteristic of certain types of North West European boats which are generically described as being in a Celtic tradition, as distinct from the Mediterranean and Nordic traditions of boatbuilding. The nature of the Celtic method dictates long broadish boats which might resemble punts. However, the joint between the planking and the stem and stern posts is the most difficult and potentially the weakest part of any boat. Over time the split log chine girders were gradually abandoned and the side planking attached directly to ribs and futtocks fastened to the flat bottom of the boat as in this late eighteenth century example of a 'Celtic' boat type from Llyn Padarn in Welsh Snowdonia. Because such boats do not have a proper keel they must be substantially reinforced by heavy framing, a strong central bottom board, and heavy stem and stern posts (picture of Llyn Padarn). Well developed boats of this type were noted by Caesar during the invasion of Gaul and must even then have been the product of a long line of development.

Greenhill's four roots of wooden boatbuilding help us to distinguish certain categories of boat type, though it is clear that only the dugout, and perhaps the raft, were capable of evolution. What is important but obscure is the historical interaction between the raft, the skin boat and the bark boat, and the dugout. All of the four roots existed simultaneously in specific historical times and places. For some societies, such as the Esquimos, the skin boat was the apex of their boat building technology and because it was also best suited to their way of life and their boatbuilding resources they had no incentive to change or develop it. For other societies, perhaps the early Bronze Age Scandinavian societies the skin boat was merely a transitional phase made obsolete by new tools and new cultures with different political and economic imperatives. Beyond a certain point only the log boat has any potential for a significant increase in size. What makes that increase possible, at least in part, is the insertion of ribs and frames into the dugout base, and ribs and frames are also characteristic and distinguishing features of both skin and bark boats. How and when that interaction took place is far from clear to ship historians. What is particularly interesting is that skin boats are clearly skeleton built and take their shape from the 'active' framework to which the skins are stitched, rather than being a continuous shell, into which 'passive' framing is later inserted for additional strength.

From what has been said so far it might seem that all boats may be classified as either 'shell' built or 'skeleton' built, and, indeed, such a classification was used by earlier generations of ship historians. It is, however, too simple since we now know that there are intermediate phases between shell and skeleton construction. As will be seen in due course there is archaeological evidence from both the Mediterranean and the Nordic traditions of shipbuilding which demonstrate the coexistence of passive and active framing in the same ship, shell below the waterline and skeleton above. Also, if the purpose of the skeleton is to determine the shape of the boat it is also possible to find shell built boats which have been constructed on removable moulds or jigs, which also dictate the shape of the hull, but play no part in its subsequent integrity. The Roman quinquiremes which helped to defeat the Carthagian in the third century BC may well have been built in this way. Conversely apparently skeleton built boats may be faired by creating a temporary or phantom shell consisting of laths or ribbands nailed to key frames into which the subsequent framing is fitted. A classification system which reduces all boats to either skeleton or shell built would, therefore be inherently misleading.

An alternative hypothesis is to consider the method of attaching the placing the planks of the outer hull. There are, essentially, two methods. Firstly the Carvel (pic)technique where the planks are laid (butted) edge to edge so that the outside of the hull is smooth. Secondly the Clinker (pic) or Lapstrake method where the planks are laid so that they overlap each other at the edges. For many years boats were simply divided into carvel or clinker, and the distinction is still used by wooden boatbuilders today. The carvel/clinker distinction also fits in neatly with the skeleton/shell criterion since carvel planking is usually a characteristic of skeleton built boats and clinker planking is characteristic of shell building since the planks are first fastened to each other and the framing is put in later. That is, the framework of the boat is erected first, and the planks are fastened to the framework and not to each other, whereas shell built boats have the frames inserted later, after the shell has been completed. This neat coincidence is equally misleading because early skeleton built ships can be found with clinker planking fastened to the frames and even earlier Greek and Roman ships are invariably carvel built shells in which the planks are joined to each other with mortises and tenons to form a shell into which the frames are later inserted.

The crucial distinction, and the one which is now accepted as the starting point for a system of classification, is whether or not the planks of the hull are joined to each other, or whether they are held in position by being fasted to a a prerected framework or skeleton. Using this as a criterion it may be possible to classify all wooden boats as being either 'edge joined' or 'non edge joined', the former being general characteristic of shell building and the latter of skeleton building, with various intermediate stages where both methods may coexist.

Shell built boats, whether carvel or clinker, are generally built and faired by eye, and through the experience of the builder and without the use of plans. The planks are adjusted and aligned from strake to strake and the lie of the planking is determined by the angle which the garboard planks make where they are attached to the keel. It is possible to build small skeleton boats without plans by laying down the keel, setting the stem and stern posts and inserting key ribs amidships and at either end of the boat. Temporary ribbands may then be run round the key ribs to determine the external shape of the boat and the remaining ribs may then be cut and shaped to fit inside the ribbands. When all the ribs have been fitted and faired the ribbans can be removed and the hull planking fastened into place. This method cannot, however, be used for larger skeleton built ships because the overall shape of the hull must be conceived as a whole from the beginning and the ribs and frames must be individually designed to embody the the desired hull shape. This requires planning, a knowledge and application of the mathematics of compound curves, and a high degree of precision in the assembly of the skeleton. The advantage of skeleton building compaired with shell building, is that it enables the construction of much larger ships using less skilled labour. There are theoretical limits to the maximum size of shell built boats, though both the Romans and the Vikings came by different means very close to the absolute dimensional limits of wooden shipbuilding using shell techniques. As we shall see many of the changes which take place in shipbuilding are consequences of changing social or economic needs in the societies which produce the ships. The point of widespread transition from edge joined shell to non edge joined skeleton building is the coincidence of social, economic, technological and cultural changes which made large and capacious ships an essential instrument of economic and territorial aggrandisement for the emerging Western European nation states in the fifteenth and sixteenth centuries AD.


Updated by J.S. Illsley 13/08/1999


The references listed here identify the sources for slide used to illustrate the course lectures. Where slides are taken from published sources copyright prohibits their publication on the Web. Many of the pictures are taken from George Bass's History of Seafaring and the first edition of Basil Greenhill's Archaeology of the Boat both of which are easily available on rota in the Bangor University Library, and many may also be found in other and more recent books. Slides displayed in this Web page are taken from JSI's personal collection.

This page is under development and not all references are complete.

SLIDE 143   GALLEY   C4BC  (trireme after portion)   SOURCE   Casson L., Ships and Seamanship in the ancient world (Princeton, 1973) pp: 106

SLIDE 163   REED BOAT   3100BC  (model of egyptian)   SOURCE   Casson L., Ships and Seamanship in the ancient world (Princeton, 1973) pp: 7-8

SLIDE 178   DUGOUT      (extended edge joined)   SOURCE   Greenhill B., Archaeology of the Boat (London, 1976) pp: 139 /91

SLIDE 302   KALNES ROCK CARVINGS   BRONZE AGE  (sweden)   SOURCE   Bass, G.A History of Seafaring, (London, 1974) pp: 168 /1

SLIDE 528   DUGOUT   C19TH  (Bremerhaven C20th German)   SOURCE   JSI

SLIDE 425   ZWAMMERDAM BOAT      (isometric)   SOURCE    pp:

SLIDE 1037   CURRAGHS      (modern in Galway)   SOURCE   JSI

SLIDE 1067   CURRAGH   modern  (Galway interior)   SOURCE   JSI